Symbolic Methods and Software Overview For Lie Symmetry Analysis

x
Symboli Software
for Lie Symmetry Analysis
W. Hereman1
Department of Mathemati al and Computer S ien es
Colorado S hool of Mines
Golden, CO 80401, USA
To appear as Chapter 13
CRC Handbook of Lie Group Analysis of Dierential Equations
Volume 3: New Trends in Theoreti al Developments and Computational Methods
Edited by N.H. Ibragimov
CRC Press, Bo a Raton, Florida, 1995
November 15, 1999
MCS-94-07
August 1994
1991 Mathemati s Subje t Classi ation.

Primary 68Q40, 35-04; Se ondary 22E60, 34-04
1 Resear h Supported in Part by NSF under Grant CCR-9300978
Chapter 13
Abstra t
A survey of te hniques and symboli programs for the determination of Lie symmetry groups of
systems of dierential equations is presented. The purpose, methods and algorithms of symmetry
analysis are outlined. An exhaustive review of the literature, in luding old and modern books and
papers presenting key on epts is given. Spe ial attention is paid to methods for redu ing the determining equations into standard form, and their subsequent integration. Several examples illustrate
the use of the Lie symmetry software. Throughout the paper, new trends in the development of
symboli pa kages for Lie symmetry analysis are indi ated.
13.1.
INTRODUCTION AND LITERATURE
The Norwegian mathemati ian Marius Sophus Lie (1842-1899) and Felix Klein (1849-1925), a
German geometer, pioneered the study of transformation groups that leave systems of equations
invariant. Klein's work fo used on the role of nite groups in the study of regular bodies and algebrai equations. Lie founded the theory of ontinuous transformations groups [179, 180, 181, 182,
183, 184, 185 and Lie groups. Although Lie's starting point had been geometry, the inspirational
sour e for his group theoreti investigations was the eld of dierential equations. His goal was to
establish a theory of integration of dierential equations that would mirror Abel's theory for the
solution of algebrai equations. His work brought many diverse and ad ho integration methods
for solving spe ial lasses of dierential equations under a ommon on eptual umbrella.
Later on, the on ept of symmetry evolved into one of the most explosive developments of
mathemati s and physi s throughout the twentieth entury. The theory of Lie groups and Lie
algebras is now applied to diverse elds of mathemati s in luding dierential geometry, algebrai
topology, bifur ation theory, numeri al analysis, spe ial fun tions, to name a few; and to nearly
any area of theoreti al physi s, in parti ular lassi al, ontinuum and quantum me hani s, uid
dynami s, relativity, and parti le physi s.
Lie's innitesimal transformation method provides the most widely appli able te hnique to nd
losed form solutions of ordinary dierential equations (ODEs). Standard solution methods for
rst-order or linear ODEs an be hara terized in terms of symmetries. For nonlinear ODEs, Lie's
method, when it su eeds, provides a means of redu ing the solution to a series of quadratures and
an be implemented in symboli programs [41. Under ertain onditions, rst- and se ond-order
ODEs an be linearized via Lie group transformations [149, 151. Through the group lassi ation
of ODEs, Lie also su eeded in identifying all ODEs that an either be redu ed to lower-order ones,
or ompletely integrated via group theoreti te hniques.
Applied to partial dierential equations (PDEs), the method leads to group-invariant solutions,
onservation laws, invariant enter manifolds in bifur ation theory [56, 111, 112, 252, 299, et .
Exploiting the symmetries of PDEs, new solutions an be derived from old ones, and PDEs an
be lassied into equivalen e lasses. Spe ial, physi ally signi ant solutions arising from symmetry methods allow one to investigate the asymptoti or physi al behavior of general solutions.
The group-invariant solutions obtained via Lie's approa h provide insight into the physi al models themselves. Expli it solutions also serve as ben hmarks in the design, a ura y testing, and
omparison of numeri al algorithms.
Lie's original ideas greatly in uen ed the study of physi ally important systems of dierential
equations in lassi al me hani s, uid dynami s, elasti ity, and many other applied areas. Currently, Lie symmetry methods are applied to dieren e, dierential-dieren e equations [177, 301
and integro-dierential equations. For a good review of the present state of aairs see [60.
The appli ation of Lie groups methods to on rete physi al systems involves tedious, me hani al omputations. Even the al ulation of the ontinuous symmetry group of a modest system
of dierential equations is prone to fail, if done with pen il and paper. Programmable omputer
algebra systems (CAS) su h as Mathemati a, MACSYMA, Maple, REDUCE, AXIOM and MuPAD, are extremely useful aides in su h omputations. Symboli pa kages, written in the language
of CAS, an nd the dening (or determining) equations of the Lie symmetry group. The most
sophisti ated pa kages then redu e the determining system into an equivalent but more suitable
system, subsequently solve that system in losed form, and go on to al ulate the innitesimal
generators that span the Lie algebra of symmetries.
A large body of literature exists on the topi of Lie symmetries. We list some older books
[8, 9, 10, 26, 29, 35, 48, 51, 55 [73, 84, 130, 140, 161, 174, 196, 197, 198, 220, 222, 272, more
modern books [12, 13, 18, 21, 36, 45, 78 [80, 90, 103, 105, 106, 107, 108 [110, 112, 144, 145, 147,
148, 149, 153, 162, 178 [175, 193, 214, 221, 224, 227, 229, 244, 245, 247 [248, 252, 253, 275, 281, 285,
305, 310, 316, 326, some re ent onferen e pro eedings [7, 43, 102, 155, 166, 301, published le ture
notes [230, re ent Ph.D. theses [54, 120, 141, 146, 164, 219, 276, 278, 311, and papers presenting
key on epts [6, 11, 14, 15, 28, 109 [150, 171, 251, 264, 290, 291, 292, 303, 304, 306, 312, 313, 315.
Also of interest are the le ture notes and preprints published by The International Sophus Lie
Center in Oslo (for example, [169), and the notes of the ourses organized by ERCIM on \Partial
Dierential Equations and Group Theory" [230, 270.
Extensive tables of symmetry group generators for well-known equations from mathemati al
physi s have been gathered by Rogers and Ames [244 and Ibragimov [150, 153, 154.
Many onferen es addressed issues related to group analysis of dierential equations. We single
out the latest volumes of \Group Theoreti al Methods in Physi s" [75, 76, wherein a list of
previous olloquia may be found. As of 1994, a new journal [155 overs topi s in Lie groups and
their appli ations.
We highly re ommend the spe ial issues of A ta Appli andae Mathemati ae on \Symmetries of
Partial Dierential Equations" [304, now available in book form [305, and the other two volumes
in this series edited by Ibragimov [153, 154.
Biographies of Lie and his ontemporaries are found in [19, 134, 150, 152, 226, 326. Delightful
histori al notes about symmetry resear h in general, and Lie's work in parti ular an be found in
the new, updated edition of Olver's lassi [214. Complete referen es to Lie's original work are
also given in this book, and in [153. Translations of some of Lie's fundamental papers are available
in [154, 179, 180.
Readers interested in the dierential geometri al foundation of Lie symmetry analysis, in luding
topi s like Lie-Ba klund mappings, Cartan forms, supersymmetry, graded dierential geometry, and
gauge theories, may want to onsult [83, 96, 111, 171, 172, 193, 214, 216 [224, 279, 307, 308, 310,
328, and [123, 124 for related REDUCE algorithms.
In this review, we will sparsely address the omputation of onservation laws, whi h through
Noether's famous theorem [200 and its extensions [36, 249, is intrinsi ally onne ted with the
investigation of variational symmetries. Indeed, re all that variational symmetries admitted by
a Lagrangian system|symmetries that leave the a tion integral invariant| an be obtained by
omputing the Lie-Ba klund symmetries of the orresponding Euler-Lagrange equations. On e
the Lie-Ba klund symmetries are obtained and the variational symmetries are singled out, Boyer's
formulation of Noether's theorem an be used to al ulate the onservation laws.
In Se tion 13.2 we dis uss the purpose, methods and algorithms used in the omputation of
Lie symmetries. Apart from a detailed des ription of the methods for omputing and solving the
determining equations, we address the redu tion of systems of PDEs into standard and passive
forms. This topi , in turn, ties in with the omputation of the size of the symmetry group. In this
se tion we also address some of the newest trends in the development of symboli software for Lie
symmetries.
In Se tion 13.3 we look beyond Lie-point symmetries, addressing onta t and generalized symmetries, as well as non lassi al or onditional symmetries.
Se tion 13.4 is devoted to a detailed review of modern symboli pa kages that aid in the investigation of Lie symmetries for systems of dierential equations. The software is grouped a ording
to the underlying CAS. The review of the available ode intentionally fo uses on pa kages written
after 1985. More details about pioneering eorts and software written prior to 1985 an be found
elsewhere [138. In Se tion 13.5, three examples illustrate the results that an be obtained with
the software pa kages. Finally, in Se tion 13.6 we draw some on lusions.
Although no survey an be onsidered exhaustive, our study intended to over all the Lie
symmetry software, with the ex lusion of software for Lie group omputations, su h as LiE [300,
whi h is outside the s ope of this review.
13.2.
PURPOSE, METHODS AND ALGORITHMS
13.2.1.
PURPOSE
The lassi al \Lie symmetry group of a system of dierential equations" is a lo al group of point
transformations, meaning dieomorphisms on the spa e of independent and dependent variables,
whi h map solutions of the system to solutions. Various other types of lo al symmetries [14 and
nonlo al symmetries [6, 31, 32, 35, 36, 305 have been studied, as well as approximate symmetries [20. The software reviewed in this paper attempts to ompute the lassi al Lie-point symmetries. Several pa kages go beyond that in omputing onta t (or dynami al) and generalized
(or Lie-Ba klund) symmetries, non lassi al (or onditional) symmetries. Several programs ould
be modied to ompute Cartan's dynami al symmetries [51, 285 and hidden symmetries. Two
programs that utilize point symmetries to generalize spe ial solutions or nd similarity variables
re ently be ame available [295, 296, 297, 322. The latter programs a tually try to integrate the
hara teristi system of rst-order dierential equations.
Loosely speaking, onta t symmetries are generalized symmetries of order one; i.e., the oe ients in the ve tor eld in lude rst derivatives of the dependent variables.
The name Lie-Ba klund symmetries, ommonly used for generalized symmetries, is somewhat
misleading for its onne tion with Ba klund transformations [199. Ba klund transformations onstitute a set of ase-spe i equations that allow one to transform one solution into another, whereas
generalized symmetries are innitesimal symmetries involving higher-order derivatives of the dependent variables. Also, note that Ba klund transformations do not have group properties. Among
the several ways of omputing Lie-Ba klund symmetries, this paper fo uses on an extension of the
original methods due to Lie. For a dis ussion of the dieren e with other symmetry approa hes,
in luding other omputer algebra methods, we refer the reader to [97, 99, 117, 118, 327, 329.
Conditional symmetries are found by the \non lassi al method of group-invariant solutions,"
as introdu ed in 1969 by Bluman and Cole [34. Further generalizations of the non lassi al method
lead to the less pra ti al \weak symmetries" [217, 218, and the method of dierential onstraints
[213, 214.
Nonlo al (potential) symmetries [31, 32, 33, 36, 173, 221, 305 for a system of PDEs are omputed as follows. First, one repla es (one or more of) the PDEs in the given system by equivalent
onserved forms; se ond, one introdu es auxiliary potential variables; nally, one determines the
point symmetries of the resulting auxiliary system of PDEs. The form of an innitesimal generator then determines whether or not it denes a nonlo al symmetry. The te hnique of potential
symmetries leads to interesting linearizations involving non-invertible mappings.
\Hidden" symmetries [2, 3, 4, 128 show up when, for example, the order of an ODE is in reased
by one and the number of symmetries (of the now higher-order ODE) is in reased at least by two.
Upon subsequent redu tion, a symmetry may be lost if the transformation is done in non-normal
subgroup variables. Hidden symmetries also o ur when an ODE is redu ed in order, and when
this indu es an additional symmetry that was not a symmetry of the original ODE.
Somewhat related to this is the following: If a higher-order ODE is rewritten as an equivalent
system of rst-order equations, and one analyzes the system of rst-order equations, rather than
the single equation of higher order, the lass of symmetries (and onsequently solutions) an be
substantially enlarged [220, 221, 251. A similar situation arises with PDEs. When a PDE is
redu ed to a lower-dimensional PDE one an lose or gain symmetries. Clarkson [60 provides some
examples.
Dis rete symmetries of dierential equations and dis onne ted Lie-point groups have been onsidered by Reid, Weih and Wittkopf [240.
Pre ise denitions and a dis ussion of internal symmetries (also alled dynami al symmetries)
and external symmetries are given in [14, 15. It is beyond the s ope of this survey.
13.2.2.
METHODS FOR COMPUTING DETERMINING EQUA-
TIONS
Although Lie's theory is geometri , one has to resort to dierential-algebrai methods to ompute
Lie symmetries. Indeed, the riterion for a ve tor eld to be a generator is purely geometri ,
namely, ertain fun tions must vanish on a submanifold. But dierential geometry gives us no
tools to implement su h riteria in on rete appli ations. The use of dierential algebra imposes
some restri tions on the type of problems that an be handled. For instan e, the dierential
equations must be polynomial in their variables (so that one an work in the ring of dierential
polynomials); the dierential equations must be solvable for some derivatives, et .
There are three major methods to ompute Lie symmetries. The rst one uses prolonged ve tor
elds, the se ond utilizes dierential forms (wedge produ ts) due to Cartan [51. The third one
uses the notion of \formal symmetry" [37, 38, 195. Although restri ted to evolution systems with
two independent variables, the latter method provides a very qui k way to ompute anoni al
generalized symmetries. Due to its limited s ope we will not elaborate on that te hnique.
13.2.2.1.
PROLONGED VECTOR FIELDS
The rst method is used in the algorithm or our program SYMMGRP.MAX [52 and in most of the
other Lie symmetry pa kages. Instead of looking for the Lie group G, one looks for its Lie algebra
L, realized by the ve tor eld. From the Lie algebra of symmetry generators, one an obtain the Lie
group of point transformations upon integration of a system of rst-order hara teristi equations
(see Se tion 13.5.1 for a simple example).
For notational simpli ity, let us onsider the ase of Lie-point symmetries [35, 36, 214. We
follow the method, notations and terminology used in [214.
We start with a system of m dierential equations,
i(x; u(k)) = 0; i = 1; 2; :::; m;
(13.1)
of order
k, with p independent and q dependent (real) variables, denoted by x = (x1 ; x2 ; :::; xp )
p
2 IR ; u = (u1; u2; :::; uq ) 2 IRq : We stress that m; k; p and q are arbitrary positive integers. The
partial derivatives of ul are represented using a multi-index notation, for J = (j1 ; j2; :::; jp) 2 INp;
we denote
jJ j ul
;
(13.2)
x1 j1 x2 j2 :::xp j
where jJ j = j1 + j2 + ::: + jp. Finally, let u(k) denote a ve tor whose omponents are all the partial
derivatives of order 0 up to k of all the ul .
The group transformations have the form x~ = g (x; u); u~ =
g (x; u); where the fun tions g
and
g are to be determined. Note that the subs ript g refers to the group parameters. Instead of
looking for a Lie group G, we look for its Lie algebra L, realized by ve tor elds of the form
ulJ
p
X
i
q
X
(x; u) x + 'l (x; u) u l :
=
i
i=1
l=1
(13.3)
The problem is now redu ed to nding the oe ients i(x; u) and 'l (x; u): In essen e, the omputer
onstru ts the kth prolongation pr(k) of the ve tor eld , applies it to the system (13.1), and
requests that the resulting expression vanishes on the solution set of (13.1).
This sounds straightforward, but the method involves tedious al ulations be ause the length
and omplexity of the expressions in rease rapidly as p; q; m, and espe ially k, in rease. The number
of determining equations then also rises dramati ally as is shown in [244 on page 346. Here are
the details and the steps to be performed:
1. Constru t the kth prolongation of the ve tor eld in (13.3) by means of the formula
pr(k) = +
where the oe ients
J
l
q X
X
J
l
l=1 J
(x; u(k)) u l ; 1 jJ j k;

J
(13.4)
are dened as follows. The oe ients of the rst prolongation are

Ji
l
= Di 'l (x; u)
p
X
ul
j =1
Jj Di
(x; u);
(13.5)
where Ji is a p tuple with 1 on the ith position and zeros elsewhere, and Di is the total
derivative operator
q X
X

(13.6)
+
ulJ +J l ; 0 jJ j k:
Di =
xi l=1 J
uJ
The higher-order prolongations are dened re ursively as
i
J +Ji
l
= Di
J
l
p
X
ul
j =1
J +Jj Di
(x; u); jJ j 1:
(13.7)
2. Apply the prolonged operator pr(k) to ea h equation i (x; u(k)) and require that
pr(k) i j =0 = 0 i; j = 1; :::; m:
(13.8)
The meaning of ondition (13.8) is that pr(k) vanishes on the solution set of the originally
given system (13.1). Pre isely, this ondition assures that is an innitesimal symmetry
generator of the group transformation; x~ = g (x; u); u~ =
g (x; u); i.e., that u(x) is a
solution of (13.1) whenever u~(~x) is one.
j
3. Choose, if possible, m omponents of the ve tor u(k), say v1; :::; vm, su h that:
(a) Ea h vi is equal to a derivative of a ul (l = 1; :::; q) with respe t to at least one variable
xi (i = 1; :::; p):
(b) None of the vi is the derivative of another one in the set.
( ) The system (13.1) an be solved algebrai ally for the vi in terms of the remaining
omponents of u(k), whi h we denote by w. Hen e, vi = S i(x; w); i = 1; :::; m:
(d) The derivatives of vi, vJi = DJ S i(x; w); where DJ D1j D2j :::Dpj , an all be expressed
in terms of the omponents of w and their derivatives, without ever reintrodu ing the
v i or their derivatives.
The requirements in step 3 put some restri tions on the system (13.1), but for many systems
the hoi e of the appropriate vi is quite obvious. For example, for a system of evolution
equations
ui
(x ; :::; xp 1; t) = F i(x1 ; :::; xp 1; t; u(k)); i = 1; :::; m;
(13.9)
t 1
where u(k) involves
derivatives with respe t to the variables xi but not t, an appropriate
i
u
hoi e is vi = t :
4. Use vi = S i(x; w) to eliminate all vi and their derivatives from the expression (13.8), so that
all the remaining variables are now independent of ea h other. It is ta itly assumed that the
resulting expression is now a polynomial in the ulJ :
5. Obtain the determining equations for i(x; u) and 'l (x; u) by equating to zero the oe ients
of all fun tionally independent expressions (monomials) in the remaining derivatives ulJ .
In the above algorithm the variables xi ; ul, and ulJ are treated as independent; the dependent ones
are i and 'l :
In summary, the result of implementing (13.4) is a system of linear homogeneous PDEs for i
and 'l ; in whi h x and u are independent variables. These are the so- alled determining or dening
equations for the symmetries of the system. Solving these by hand, intera tively or automati ally
with a symboli pa kage, will give the expli it forms of the i(x; u) and 'l (x; u):
The pro edure, whi h is thoroughly dis ussed in [214, onsists of two major steps: deriving the
determining equations, and solving them expli itly.
We note that in [30, Bluman proves theorems about the forms of admitted innitesimal transformations, whi h over a large lass of s alar ODEs and PDEs. In essen e, the theorems in [36
imply that the oe ients of the ve tor eld are either free or depend linearly on the dependent
variable. Use of these theorems an signi antly simplify the tedious work involved in setting up
and solving the determining equations.
1
13.2.2.2.
DIFFERENTIAL FORMS
A dierential geometri approa h to invarian e groups and solutions of PDEs was presented by
Harrison and Estabrook [131. They showed how to derive innitesimal symmetries using Cartan's
exterior dierential al ulus. The isove tor elds, whi h are the generators of geometri transformations with suitable algebrai invarian e properties, are then used to obtain invariant solutions
of several PDEs, in luding the heat equation, the Korteweg-de Vries equation, and the va uum
Maxwell equations.
Brie y, this method to nd innitesimal symmetries pro eeds as follows. The system of dierential equations is rewritten as a Pfaan system, that is a system of one-forms. The ondition for
an isove tor is then that the ontra tion of every form in the losed ideal generated by the exterior
system with the isove tor remains in the ideal. That is ompletely equivalent to the ondition in
the jet bundle approa h. From a omputational point of view there is a disadvantage: the exterior
system usually onsists of more forms than the PDE system of equations. To redu e the number of
forms, one an opt to formulate the system of dierential equations as a losed exterior dierential
system of two-forms. The underlying manifold is prolonged by new oordinates, the so- alled prolonged variables. The exterior dierential system is then prolonged by spe ial one-forms in these
prolonged variables. The ondition on these added one-forms is that the prolonged exterior dierential systems remain losed. This ondition leads to the determining equations for the oe ients
of the manifold.
For the mathemati al formulation and several worked examples, the reader is referred to [49, 86,
125, 131, 165, 279. Olver's forth oming book [216 addresses the appli ation of various approa hes,
parti ularly those of Lie and Cartan, to equivalen e and symmetry problems. That book also
dis usses the determination of invariants, the lassi ation of dierential equations, variational
problems, and so on.
It should also be noted that routines [77, 254, 255, 309 for Cartan's exterior al ulus [132 are
in luded as standard pa kages of CAS. For instan e, CARTAN in the MACSYMA Share Library,
EXCALC in REDUCE, and DierentialForms [330, 331 in Mathemati a. For a list of other
symboli pa kages that manipulate obje ts of dierential geometry we refer the reader to [255.
13.2.3.
METHODS FOR REDUCING DETERMINING EQUATIONS
To design a reliable and powerful integration algorithm for a system of determining equations the
system needs to be brought into a anoni al form. Various related on epts appear in the literature,
were authors refer to the normal, orthonomi , involutive and passive forms [45, 174. All of these
have slightly dierent denitions, for whi h the reader should onsult the later ited literature.
Involution is a geometri on ept, independent of oordinate systems. Passive systems are dened
with respe t to a given oordinate system and a given ordering of terms. The same holds for
dierential Grobner bases (DGBs).
The original theory of involutive systems goes ba k to Cartan [51, Janet [159, 160, Kahler
[161, Kol hin [167, 168, Riquier [242, Ritt [243, Thomas [286, and Tresse [288, 289. In the
introdu tion to his le ture notes [230, Pommaret dis usses the basi ideas of the theory of dierential elimination. His des ription, whi h ontains interesting histori al remarks (see also [228),
relates the work of the Fren h to the formal ohomology approa h advo ated by D.C. Spen er and
ollaborators in the USA. A on ise survey of the theory of involutive systems with omplete proofs
is also given by Finikov [95.
Roughly speaking, the methods that redu e systems of linear homogeneous PDEs into an equivalent, but mu h simpler standard form, an be viewed as generalizations of triangulation by GaussJordan elimination, but are applied here to systems of linear PDEs. First, the original system is
appended by all its dierential onsequen es. Se ond, highest derivatives are eliminated, and, if
they o ur, integrability onditions are added to the system. The pro edure is then repeated until
the new system is in involution.
Although S hwarz did valuable pioneer work on solving systems of linear homogeneous PDEs
[259, he is by far not the only one working on innovative ways of lassifying, subsequently redu ing,
and nally solving overdetermined systems of linear homogeneous PDEs. Early eorts to implement
Cartan's exterior-forms approa h to involutive systems, and the method of Riquier and Janet, are
due to Russian teams [16, 113, 114, 287.
Kornyak, Fedorova and Fush hi h [92, 170, Bo harov [40, Bo harov and Bronstein [42, Ganzha
and Yanenko [114, Gerdt and Zharkov [119, Pankrat'ev [223 and Topunov [287, amongst others,
partially implemented algorithms to redu e systems of PDEs.
The GRBASES algorithm of Pankrat'ev and the implementation of the Riquier-Janet method
by Topunov are both in REFAL. Both programs an only redu e linear systems of PDEs, this in
ontrast with the program DIFFGROB2 of Manseld dis ussed below.
A good des ription of the relation between the Riquier-Janet theory and the modern implementations (Grobner bases) is given by Topunov [287 (see also [266). For an introdu tion of Grobner
bases, in luding Bu hberger's algorithm [46, 47, we re ommend the new books by Cox, Little and
O'Shea [74, and Be ker and Weispfenning [25. Most of this work dire tly relates to Lie symmetry
omputations, with the ex eption of Pankrat'ev, who oers general algorithms for the omputation
of Grobner bases in dierential and dieren e modules.
In the West, Reid and ollaborators [234, 235, 237, 238, 241, S hwarz [259, 266, 267, and
Wolf and Brand [320, 321, 324, 323, 325, partially implemented algorithms to redu e systems of
PDEs. Their work led to sophisti ated symboli ode in MACSYMA, Maple, and REDUCE for
that purpose.
In [266, 267, S hwarz des ribes the algorithm InvolutionSystem, based on the theory of dierential equations due to Riquier [242 and Janet [159, 160, to transform a linear system of PDEs
into involutive form. In modern language: the involutive form is a DGB with respe t to the sele ted term ordering. If all onsistent orderings for the terms in the system of PDEs are known,
the algorithm InvolutionSystem may be applied repeatedly to determine a universal Grobner base
[267. S hwarz designed his algorithm InvolutionSystem for a spe i purpose, namely to determine the size of the Lie symmetry group of a given system of PDEs without having to integrate
the determining equations. We devote Se tion 13.2.5 to this important topi , where details about
S hwarz's program SYMSIZE an be found.
As dened by Reid [235, a standard form of a system is obtained by repeating the following
steps: (i) write ea h equation in solved form with respe t to its highest order derivative, (ii) repla e
these highest order derivatives throughout the rest of the system, (iii) add any new equations arising
from integrability onditions.
As said, standard-form algorithms have their origins in the work of Riquier and Janet for passive forms. Note that the Riquier-Janet form is not a standard form, but an be fairly simply
transformed into a standard form. Also, in method, design of algorithm, and in a tual implementation, modern standard form algorithms are quite dierent from the original ine ient methods
and algorithms proposed by Riquier and Janet.
A rst, but brief a ount of Reid's algorithm standard form [235, whi h also has it roots in the
lassi al Riquier-Janet theory, appeared in [233 and [234. The algorithm was rst implemented
in MACSYMA, and later on a MAPLE version be ame available [241.
The algorithm standard form redu es systems of PDEs to a simplied standard form. Again,
the pro edure an be viewed as a generalization to linear dierential equations of the Gaussian
redu tion method for matri es or linear systems. The algorithm now takes as input the system of
PDEs and a matrix whi h spe ies a omplete ordering on the derivatives appearing in the system.
It then redu es the system of PDEs to an equivalent simplied ordered triangular system with
all integrability onditions in luded and all redundan ies (dierential and algebrai ) eliminated.
Reid's algorithm implements an equivalen e lass approa h to the problem of bringing a system of
PDEs into a standard form. For that purpose, Reid developed a new ompletion method based on
a free dire tion index (rather than the monomials of the Riquier-Janet theory).
Within standard form, Reid uses an \update strategy" based on updating lists of equations: one
term equations, easy equations, hard (or yet-to-be lassied) equations; with spe ial user-dened
tuning knobs (parameters), so that the user an ontrol the ow of equations between the various
lists. Thus standard form works on easier parts of the system rst, a strategy that be omes ru ial
when dealing with large systems.
Further details about Reid's algorithm and examples of its use an be found in [237, where it
is shown how dire ted graphs representing the dependen ies amongst the system's variables an be
used to simplify or numeri ally integrate the system. On e the system is in standard form, one an
ontinue with the automati determination of a Taylor series solution of the system to a spe ied
nite degree.
Reid and Wittkopf's pa kage [241 fa ilitates automated interfa ing with major symmetry pa kages su h as DIMSYM [277, LIESYMM [49, and SYMMGRP.MAX [52, and also with the differential Grobner basis pa kage DIFFGROB2 [189. A TEX interfa e between standard form and
Hi kman's program [142, 143 that uses physi al variable notation has been provided by Lisle. Full
details and many illustrative examples of the pa kage, whi h, besides the fun tion standard form,
in ludes other powerful algorithms for symmetry analysis of PDEs, are given in [241.
Reid and M Kinnon developed a re ursive algorithm alled Rsolve Pdesys [239 that builds on
Reid's standard form [235 and on algorithms of Abramov and Kvashenko [5 and Bronstein [44.
Reid and M Kinnon's algorithm RSolve Pdesys now nds parti ular solutions of linear systems of
PDEs using only ODE solution te hniques. Applied to symmetry problems, their algorithm will
nd all polynomial/rational solutions of the determining equations provided the symmetry group
is nite-dimensional.
Several other approa hes, and onsequently implementations, are possible to omplete a given
system of PDEs to an involutive system. S hu, Seiler and Calmet [256 present an algorithm
in AXIOM to perform this task. Their method is highly geometri al and their implementation is
based in part on the Cartan-Kuranishi theorem [45, whi h assures that the integrability onditions
for the determining equations an be found in a nite number of steps. A detailed des ription of
their programs, alled JET, is given in Seiler's thesis [274. JET an be viewed as an environment
for omputations within the geometri theory of PDEs based on the jet bundle formalism. Some
standard tasks are put into AXIOM pa kages. One su h pa kage is alled CartanKuranishi, whi h
ompletes a given system to an involutive one. Another pa kage, in development at the time
of writing, ontains a pro edure to set up determining equations for lassi al and non- lassi al
symmetries.
Hartley and Tu ker [132 implemented an algorithm (in REDUCE) to analyse involutive systems
of exterior forms, based on the Cartan-Kahler theory. Later they extended their program [133,
originally in REDUCE, now partly rewritten in Maple, to non-involutive systems. For su h systems
their ompletion pro edure onstru ts the needed integrability onditions. Their work orresponds
that of S hu, Seiler and Calmet on involutive systems, however, Hartley and Tu ker use exterior
systems.
A ording to my sour es, a student of Fa kerell (Sydney) has implemented the Vessiot approa h
to involution. Vessiot's method an be viewed as a dual version to the exterior system approa h.
In the full omputer implementation of \triangulation" algorithms, one takes advantage of a

\dierential" generalization of Bu hberger's algorithm for Grobner bases. Bu hberger's algorithm,
whi h is in luded as a standard pa kage with modern CAS [47, 136, 194 (also see Appendix C in
[74), provides a te hnique for anoni ally simplifying polynomial nonlinear systems of algebrai
equations. The \dierential" generalization of that algorithm allows one to redu e systems of
nonlinear (and, onsequently, also linear) PDEs into standard form. In the \dierential" analogue
of Bu hberger's algorithm, one has to repla e ross-multipli ation by ross-dierentiation, algebrai
redu tion by dierential redu tion, et .
Carra-Ferro [50 was the rst to dene DGBs for PDE systems. She gave a method based
on dierential redu tion for the attempted onstru tion of su h DGBs, but showed that they
may be innite (unlike the ase for polynomial algebrai equations and linear systems of PDEs).
Subsequently, Ollivier [212 gave a method whi h ould, in a nite number of steps, onstru t
a DGB up to a given order of derivation (even when the DGB was innite). But riteria were
not given for determining when the DGB had been onstru ted up to the given order. Re ently
Manseld [188 has given an algorithm (and a omputer-algebra implementation) that uses pseudoredu tion instead of redu tion to attempt to onstru t DGBs. The advantage of her te hnique is
that it always terminates; however, her algorithm may not always terminate with a DGB. In short,
the dierential ideal membership problem remains unsolved; and there are even disagreements on
the denition of DGBs.
The Maple program DIFFGROB [188, 192 and its se ond version DIFFGROB2 by Manseld
[189, are designed to ompute the DGB of a nitely generated ideal of PDEs with polynomial terms.
With respe t to this basis, every member of the ideal pseudo-redu es to zero. In pseudo-redu tion
one is allowed to multiply expressions by dierential oe ients of the highest derivative terms that
o ur in the system to be redu ed. This is needed for nonlinear systems where otherwise standard
redu tion algorithms would not ne essarily terminate. DIFFGROB2 allows one to al ulate in a
systemati way: the elimination ideals, integrability onditions, and ompatibility onditions of a
system of nonlinear PDEs of polynomial type, up to ertain te hni al onstraints fully explained
in [189, 192.
The fundamental tools in Manseld's pa kage are the Kol hin-Ritt algorithm [94, whi h is a
dierential analogue of Bu hberger's algorithm, with pseudo-redu tion instead of redu tion, and
the digbasis algorithm, whi h takes into a ount algebrai as well as dierential onsequen es
of nonlinear systems. These two algorithms allow one to ompute the DGB for a wide range of
systems of PDEs. A detailed dis ussion of these algorithms is beyond the s ope of this survey
paper. For more information about DIFFGROB2 and illustrations of its use, the reader should
onsult [68, 71, 192, 190, 191, and in parti ular the manual [189.
For fairly simple examples, su h as the Boussinesq equation, DIFFGROB2 is able to automati ally redu e the nonlinear determining equations ( orresponding to non lassi al symmetries)
in standard form. For more ompli ated ases, DIFFGROB2 may need to be used intera tively.
Nevertheless, the pa kage has proven to be an ee tive tool [60, 68, 69, 71, 190, 191 in solving
overdetermined systems of linear and nonlinear PDEs arising in the study of lassi al and non lassi al symmetries. Needless to say, DIFFGROB2 an be used in appli ations other than symmetry
analysis. Su h appli ations in lude nding the ompatibility onditions for inhomogeneous systems, testing the onsisten y of systems of PDEs, and nding the least amount of ne essary data
for a formal power series solution of a linear system (the \initial data" problem). Obviously, DIFFGROB2 an also be used to bring the input equations into involutive form. For some examples
this is indeed ne essary to be able to ompute non lassi al symmetries. In passing, within DIFF-
GROB2 a pa kage alled DIRMETH is available to ompute the determining equations related to
the symmetry redu tion of PDEs via the dire t method of Clarkson and Kruskal [65.
The program CRACK by Wolf and Brand [323, 324, 325 also arries out a Grobner Basis
analysis but in slightly modied form. First, the algorithm is enri hed by the integration of PDEs
whenever possible, but only in su h a way, that the new integrated PDEs are still polynomial
in the Grobner basis. In other words, the ` riti al pair ompletion steps' of the Grobner basis
algorithm and the integrations used within CRACK are onsistent. Sele tive integration an redu e
the omplexity and aid in solving the determining equations, in parti ular for systems for whi h
pure Grobner basis methods would be unfeasible. Se ond, for e ien y reasons, only a restri ted
ompletion algorithm is used, although it is the authors intention to extend it to a omplete Grobner
basis algorithm in the future.
A ording to a re ent paper [240, Wittkopf is also developing an algorithm, alled di redu e,
whi h attempts to redu e polynomially nonlinear systems of PDEs to the form of a DGB. In
essen e, the algorithm is a dierential analogue of Bu hberger's elimination algorithm for polynomial equations. Wittkopf's algorithm uses redu tion rather than pseudo-redu tion, and in orporates strategies for e ient memory management.
Finally, Oaku [211 is designing and implementing software in the omputer algebra system
Risa/asir to automati ally nd the stru ture of the solution spa e of systems of linear PDEs.
Oaku's method is based on the notion of Grobner basis and the Bu hberger algorithm applied to
rings of dierential operators with polynomial oe ients (Weyl algebra).
13.2.4.
METHODS FOR SOLVING DETERMINING EQUATIONS
The most hallenging part of Lie symmetry analysis by omputer, involves the design of an \integrator" for the overdetermined systems of linear homogeneous PDEs. This topi is also of importan e
in the study of so- alled adjoint symmetries of dierential equations [250, and in many other areas
where determining equations of the same type o urs. Ideally, good integration ode should be
appli able to generi systems of linear dierential equations, whi h do not ne essarily ome from
symmetry analysis.
In the ontext of Lie symmetry analysis, one an aim at the design of faster and more powerful
algorithms that work for large systems of determining equations, typi ally a few hundred, and that
automati ally redu e systems to where they an be handled intera tively with the omputer, or by
hand.
Sin e the early developments [135, 257, 260, 283, 284, 285 of semi-heuristi methods to solve
determining equations, substantial progress has been made in understanding the mathemati s of
this problem and a new breed of algorithms is now available. These algorithms attempt to lose
the gap between solution te hniques for ODEs and PDEs ( onsult [280 for an impressive review
and large bibliography).
Two other important topi s tie in with the integration of the determining equations: (i) the
transformation of the determining equations into standard and passive forms; and (ii) the omputation of the size of the symmetry group dis ussed in Se tions 13.2.3 and 13.2.5, respe tively.
The design of algorithms and programs to bring the determining equations into standard form
were a major step forward. On e systems are redu ed into standard involutive form or de oupled,
subsequent integration is more tra table and reliable. One ould use separation of variables, standard te hniques for linear dierential equations, and spe i heuristi rules given below. The only
determining equations left for manual handling should be the \ onstraint" equations or any other
equations whose general solutions annot be written expli itly in losed form.
In order to be able to make the determination of ertain types of Lie generators into a de ision
pro edure, one needs an algorithm for solving linear homogeneous ODEs. Su h equations are
always obtained as the lowest equation of the redu ed determining system, with redu tion based
on lexi ographi al term ordering. An important step towards this goal is the fa torization as it is
applied in SPDE [264, 265. An in-depth review of issues related to the implementation of this and
other algorithms is given in [269.
After sear hing the relevant literature [42, 233, 234, 235, 239, 259, 266, 267, 287, 320, 321
[323, 324, 325, it is lear that many mathemati al questions remain open. Despite the innovative
eorts of Reid, S hwarz, Wolf and Brand, and many others, there is no general algorithm available
to integrate an arbitrary (overdetermined) system of determining equations that onsists of linear
homogeneous PDEs for the 's and the 's.
Most integration algorithms are based on a set of heuristi rules [135, 165, 257, 260, 277, 284.
In the omputer programs reviewed in Se tion 13.4, the following rules are used.
1. Integrate single term equations of the form
jI j f (x1 ; x2 ; :::; xn )
= 0;
(13.10)
x1 i x2 i :::xn i
where jI j = i1 + i2 + ::: + in ; to obtain the solution
1
f (x1 ; x2 ; :::; xn ) =
n iX
k 1
X
k=1 j =0
hkj (x1 ; x2 ; :::; xk 1 ; xk+1 ; :::; xn )(xk )j ;
thus, introdu ing fun tions hkj with fewer variables.

2. Repla e equations of type
n
X
fj (x1 ; x2 ; :::; xk 1 ; xk+1 ; :::; xn )(xk )j = 0;
j =0
(13.11)
(13.12)
by fj = 0 (j = 0; 1; :::; n): More generally, this method of splitting equations (via polynomial
de omposition) into a set of smaller equations is also allowed when fj are dierential equations
themselves, if the variable xk is missing.
3. Integrate linear dierential equations of rst and se ond order with onstant oe ients.
Integrate rst-order equations with variable oe ients via the integrating fa tor te hnique,
provided the resulting integrals an be omputed in losed form.
4. Integrate higher-order equations of type
n f (x1 ; x2 ; :::; xn )
= g(x1; x2; :::; xk 1; xk+1; :::; xn);
(13.13)
xk n
n su essive times to obtain
(xk )n g(x ; x ; :::; x ; x ; :::; x )
f (x1 ; x2 ; :::; xn ) =
(13.14)
1 2
k 1 k+1
n
n!
n
+ (nxk 1)! h(x1 ; x2 ; :::; xk 1; xk+1; :::; xn)

+ ::: + r(x1 ; x2; :::; xk 1; xk+1; :::; xn);
where h; :::; r are arbitrary fun tions.
5. Solve any simple equation (without derivatives) for a fun tion (or a derivative of a fun tion)
provided that it both (i) o urs linearly and only on e, and (ii) depends on all the variables
that o ur as arguments in the remaining terms.
6. Expli itly integrate exa t equations.
7. Substitute the solutions obtained above in all the equations.
8. Add dieren es, sums or other linear ombinations of equations (with similar terms) to the
system, provided these ombinations are shorter than the original equations.
With these simple rules, and perhaps a few more, the determining system an often be drasti ally
simplied. Amazingly, in many ases nothing beyond the above heuristi rules is needed to solve
the determining equations ompletely. If that is not possible, after simpli ation, the programs
return the remaining unsolved dierential equations for further inspe tion. In most programs, the
user an then intera tively simplify and fully solve the determining equations on the omputer,
thereby minimizing human errors.
At least for Lie-point symmetries, solving the determining equations an usually be done by
hand, using elementary results from the theory of linear PDEs. Solving them on a omputer may
be time onsuming, sin e the simplest approa h varies greatly from ase to ase. Furthermore, a
omputer program may a identally not at h the most general result and therefore may return
an in omplete symmetry group. The author is aware of this problem, whi h o urred when testing
some of the reviewed symmetry programs. Even worse, the omputer algorithm may not be able
to determine the solution of the determining equations in a nite number of steps.
13.2.5.
COMPUTATION OF THE SIZE OF THE SYMMETRY
GROUP
S hwarz [266, 267 and Reid [234, 235, 236 independently developed algorithms for determining
the size of the Lie symmetry groups of dierential equations without integrating the determining
equations expli itly.
S hwarz's algorithm SYMSIZE [266, 267 is available with the omputer algebra system REDUCE, as part of the pa kage SPDE (see Se tion 13.4.2). S hwarz also translated SYMSIZE into
the language of S rat hpad II, the prede essor of AXIOM. Use of SYMSIZE ir umvents some of
the short omings mentioned at the end of the previous se tion. Indeed, if a dierential equation
has no other than obvious symmetries or if the symmetry group is small (be ause all generators
are algebrai and of low degree), SYMSIZE will greatly help in ompletely solving the symmetry
problem.
In ontrast to the heuristi algorithms for the expli it omputation of the symmetry generators,
the size of the symmetry group an always be determined with SYMSIZE in a nite number of steps.
SYMSIZE a epts a system of PDEs as input, and allows one to ompute a priori the number of
free parameters if the group is nite and the number of unspe ied fun tions of the group is innite.
In turn, SYMSIZE allows one to test a posterior if the solution of the determining equations is
omplete. In ases where some, perhaps all, symmetries are known by inspe tion or from the
physi s of the problem, the knowledge of the size of the symmetry group an evade an expensive
sear h for more, perhaps nonexistent, symmetries.
At the heart of SYMSIZE is the pro edure InvolutionSystem, whi h transforms the determining system into an involutive system by means of a riti al pair/ ompletion algorithm. Similar
algorithms are applied in omputing Grobner bases in polynomial ideal theory (see Se tion 13.2.3).
Con urrently, yet independent of S hwarz, Reid [234, 235, 236 realized that triangularization
algorithms may allow bypassing the expli it solution of the determining equations and ompute
the size of the symmetry group and the ommutators immediately. Reid developed the program
SYMCAL [235, written originally in MACSYMA, but now onverted by Reid and Wittkopf into
Maple [241.
In [233, 236, a non-heuristi algorithm stru ture onstant is presented, based on Taylor and
standard form, whi h always determines (in a nite number of steps) the dimension and the stru ture onstants of the nite part of the Lie symmetry algebra. An extension of the algorithm
[233 also allows one to lassify dierential equations (with variable oe ients) a ording to the
stru ture of their symmetry groups. Furthermore, the approa h advo ated by Reid applies to the
determination of symmetries of Lie- onta t and Lie-Ba klund types, as well as potential symmetries.
The algorithm des ribed in [236, provides information about the dimension and ommutators
of the Lie symmetry algebra. It is based on expli it Taylor expansions of the symmetry generators,
and therefore is omputationally expensive and restri ted to nite dimensional Lie algebras. The
newest Maple algorithm [238 allows one to ompute the dimension and the ommutation relations
without Taylor expansions; hen e, it is appli able to innite-dimensional Lie algebras.
Readers interested in the problem of determining the \size" of the solution spa e for arbitrary
involutive systems should onsult a re ent paper by Seiler [273. Seiler's results an be applied to
linear and nonlinear determining systems.
Finally, we should mention that skillful use of the tools for redu ing systems of linear homogeneous PDEs, available within the pa kage CRACK [323, 324, 325, an also greatly assist in the
investigation of the size of the symmetry group.
13.3.
BEYOND LIE-POINT SYMMETRIES
The dis ussion of symmetries other than point symmetries is limited here to those for whi h
symmetry software is already available. For a general review of various types of symmetries we
refer to the new edition of Olver's book [214 and to Clarkson [60.
13.3.1.
CONTACT AND GENERALIZED SYMMETRIES
For the omputation of generalized symmetries or Lie-Ba klund symmetries [13, 214 the use of
symboli programs is even more appropriate. The pro edure to determine symmetries is essentially
the same as that for point symmetries, although the al ulations are lengthier and more time
onsuming. In a generalized ve tor eld, whi h still takes the form of (13.3), the fun tions i and
l may now depend on a nite number of derivatives of u, i.e.,
=
p
X
i
i=1
(x; u(k)) x +

i
q
X

'l (x; u(k) ) l :
u
l=1
(13.15)
If k = 1 the generalized symmetry determines a lassi al onta t symmetry and vi e versa, at least
in the ase of one dependent variable. The even simpler ase k = 0; with u(0) = u; leads to point
symmetries. Olver [214 dis usses various possibilities to simplify the al ulations, for example by
putting the symmetries in evolutionary form, or by xing the order of derivation on whi h the 's
and 's may depend.
13.3.2.
NONCLASSICAL OR CONDITIONAL SYMMETRIES
Re ently it was shown that the \non lassi al method of group-invariant solutions," originally introdu ed in [34, an determine new solutions of various physi ally signi ant nonlinear PDEs.
Examples in lude the nonlinear S hrodinger (NLS) equation [58, and its ylindri al version
[62; the Boussinesq equation [176; the Kadomtsev-Petviashvili equation [72 and other members
of its hierar hy [314; the Burgers equation [88, 232; the telegraph equation [207; the FitzhughNagumo equation [87, 88, 210, and other rea tion-diusion equations [66; the Helmholtz [209 and
shallow water wave equations [69, 70; and a lass of nonlinear heat equations [67, 68. Levi and
Winternitz re ently showed how onditional symmetries an be determined for the 2D-Toda latti e,
a dierential-dieren e equation [177. For a well-do umented perspe tive on the omputation of
non lassi al symmetries we re ommend [71 and [59, 60.
An example is the Boussinesq equation where the new redu tions that follow from appli ation
of the non lassi al method were dis overed earlier with a dire t method [58, 57, 65. The dire t
method was also applied to the Zabolotskaya-Khokhlov equation [61 and the Davey Stewartson
system [63, 64. Re ently, Olver [215 proved that both methods are equivalent in the ase of ber
preserving transformations, whi h means that the new independent variables depend only on the
old independent variables, not on the original dependent variables. Arrigo, Broadbridge and Hill
[17 expli itly derive riteria for whi h the dire t method and the non lassi al symmetry method
lead to the same results; they use the Burgers and Boussinesq equations as illustrative examples.
For shallow water wave equations, Clarkson and Manseld [69 have shown that the non lassi al
symmetry method an lead to parti ular solutions whi h an not be obtained via the singular
manifold method. Estevez [87, 88 and Pu i [232 pointed out some interesting onne tions between
the dire t method of Clarkson and Kruskal [65. Estevez [87, 88 and Estevez and Gordoa [89 also
ompare the non lassi al symmetry method with the singular manifold approa h from Painleve
analysis. Nu i [206 shows how Ba klund transformations an be obtained via the investigation
of non lassi al symmetries.
In ontrast to Lie-point symmetries, for example, the transformations orresponding to non lassi al (or onditional) symmetries neither leave the dierential equation invariant, nor transform all
the solutions into other solutions. They merely transform a subset of solutions into other solutions.
A ounting for \non lassi al symmetries," the program should automati ally add the q invariant
surfa e onditions [34, 213,
p
X
u
Ql (x; u(1) ) = i (x; u)
xi
i=1
l
'l (x; u) = 0; l = 1; :::; q;
(13.16)
and their dierential onsequen es, to the system (13.1). However, the in lusion of non lassi al
symmetries, and perhaps other types of symmetries as dis ussed in [14, 15, 213, requires solving systems of determining equations whi h are no longer linear. Consequently, new integration
algorithms must be designed.
It should be noted that various other types of onditional symmetries ould be onsidered. For
instan e, one ould ask under what extra onditions a lass of PDEs would admit a symmetry
hosen beforehand. Extensive work on this problem, whi h we will not address here, was done by
Fush hi h [100.
13.4.
REVIEW OF SYMBOLIC SOFTWARE
In this se tion we review the most modern Lie symmetry programs, lassied a ording to the
underlying omputer algebra system. Fo using on new trends, pa kages written before 1985 are
only brie y mentioned. Resear hers interested in more details about some of the pioneering work
ould onsult [138.
13.4.1.
COMPUTER IMPLEMENTATION
Ideally, a fully automated software pa kage for Lie symmetries should onsist of ee tive, powerful
algorithms and fast pro edures for the following tasks:
1. derivation of the determining equations for large or ompli ated systems of equations;
2. redu tion of determining equations into so- alled standard form;
3. nding the size of the symmetry group;
4. determining any obvious symmetry generators;
5. simplifying and integrating the determining equations to ompute the generators, if not all
the generators have been found yet.
Then the program should be able to exe ute the following steps in the order relevant to the spe i
appli ation:
(a) al ulation of ommutator tables, based on the results of 1, 2 and 3;
(b) al ulation of group invariant solutions.
The program should be able to handle: al ulation of nonlo al (potential) symmetries [31,
32, 36, 221, 305; al ulation of non lassi al redu tions ( onditional symmetries) and resulting
solutions; al ulation of generalized symmetries; and al ulation of equivalent onservation laws.
Furthermore, it should be able to a ept systems with free unknown ( lassi ation) fun tions
[186, 236, 277 as input. If so, questions su h as \for whi h values of these parameters or parametri
fun tions does a given ODE or PDE have pres ribed symmetries or spe ial solutions" ould be
answered.
Other ideas ould be in orporated in the design of faster and more e ient symboli software
for Lie symmetry analysis. Let me give a ouple of examples. Lengthy al ulations should be
broken up into smaller pie es by onsistently taking advantage of the \linear algebra" stru ture
of the Lie symmetry problem. For instan e, prolongations should be applied to ve tor elds of
single equations or subsets of equations, and not to the whole system at on e. Furthermore, full
expansions of the prolonged ve tor elds should be halted until the expli it forms are a tually
needed. Avoiding lengthy, redundant expansions will make the generation of the determining
equations mu h more e ient, parti ularly for non lassi al and Lie-Ba klund symmetries.
Currently, no software handles this entire ambitious program. Many Lie symmetry programs
arry out parts of the listed tasks. Making matters worse, the available Lie symmetry programs,
with the ex eption of Lie by Head [135, work with spe i CAS, whi h has to be bought separately.
In Table 1 we list the most modern software pa kages, along with information about developers
and distributors. In Table 2 we summarize the s ope of these pa kages.
With the ex eption of DELiA and DIMSYM, all the programs listed in Table 1 are publi
domain software. Potential users an obtain the software from the developers or through the
referen ed sour es. The main ost in using these pa kages is related to the ost of the underlying
CAS. As a rule of thumb, individual opies of Mathemati a, MACSYMA, Maple, REDUCE, and
the like, ost about 10% of the pri e of the platform you buy them for.
13.4.2.
REDUCE PROGRAMS
In the early '80s, S hwarz developed his well-do umented program SPDE [257, 258, 260, 261,
262, 263, 264, The program automati ally derives and often su essfully solves the determining
equations for Lie-point symmetries with minimal intervention by the user. Sin e 1986 SPDE is
distributed together with REDUCE for various types of omputers, ranging from PCs to CRAYs.
In 1994 version 1.0 of SPDE be ame available. Although S hwarz de ided to keep the old name,
the new program is drasti ally dierent. A ording to the do umentation [268, SPDE 1.0 guarantees that all innitesimal symmetry generators with algebrai oe ients will be obtained if
the equations are nonlinear and of order higher than one. Con erning the input, the equations
must be algebrai in their arguments. There is no restri tion on the number of independent and
dependent variables, and the equations an have any number of onstant parameters (no arbitrary
fun tions). The program omputes the determining equations, then generates a Grobner basis for
the determining system in a term ordering spe ied by the user (total degree, lexi ographi , et .).
The integration of the redu ed system is arried out automati ally, the symmetry generators and
their ommutator table are displayed in LATEX (if so desired).
Based on Cartan's exterior al ulus [51, 95, 140, Edelen [79, 80, 81, 82, Gragert [120, and
Gragert, Kersten and Martini [125 used omputer algebra systems to al ulate the lassi al Lie
symmetries of dierential equations. More re ently, Gragert [121, 122 added a pa kage for more
general Lie algebra omputations, in luding ode for higher-order and super symmetries and super
prolongations. Kersten [164, 165 further perfe ted the software pa kage for the al ulation of the
Lie algebra of innitesimal symmetries (in luding Lie-Ba klund symmetries) of exterior dierential
systems.
Eliseev, Fedorova and Kornyak [85, wrote ode in REDUCE-2 to generate (but not solve)
the system of determining equations for point and onta t symmetries. Their paper dis usses
the algorithm and shows three worked examples. Fedorova and Kornyak [91, 93 generalized the
algorithm to in lude the ase of Lie-Ba klund symmetries.
The intera tive REDUCE program NUSY by Nu i [202, 203, 204, 205, in luded with this book
[208, generates determining equations for Lie-point, non- lassi al, Lie-Ba klund and approximate
symmetries and provides intera tive tools to solve them. The manual [202, 204, 208 gives a
lear des ription of the various routines with their s ope and limitations, and has several worked
examples.
The pa kage CRACK by Wolf and Brand [323, 324, 325 solves overdetermined systems of dierential equations with polynomial terms. To do this, it uses ode for de oupling, separating and simplifying PDEs. Integration of exa t PDEs and dierential fa torization are also possible. CRACK
has many appli ations that are fa ilitated via spe ial tools. For instan e, the fun tion LIEORD
an aid in the investigation of Lie symmetries of ODEs. With CRACK one an also onstru t
the Lagrangian for a given se ond-order ODE, and nd rst integrals via the integrating fa tor
method. In attempting to solve standard ODEs, the program makes use of the REDUCE pa kage
ODESOLVE written by Ma Callum [187. The fun tions and tools available within CRACK allow
simpli ation and integration of linear homogeneous PDEs, beyond those derivable via symmetry
analysis.
Upon ompletion of CRACK, Wolf went on to develop three new REDUCE programs, alled
LIEPDE, QUASILINPDE and APPLYSYM, whi h all make use of the tools of CRACK.
LIEPDE [321 nds Lie-point and onta t symmetries of PDEs by deriving and solving a few
simple determining equations, before ontinuing with the omputation of the more ompli ated
determining equations. This idea, whi h makes the program highly e ient, was used in Wolf's
FORMAC program [318, 319, 320, and is also implemented in the design of the feedba k me hanism
of SYMMGRP.MAX [52. For solving the determining equations, LIEPDE makes use of modules of
the pa kage CRACK dis ussed above. The dieren e is that within LIEPDE the steps are arried
out automati ally, without intervention by the user. This approa h is parti ularly useful when
applied to large systems of PDEs, or in the omputation of higher-order symmetries, where spa e
and memory limitations ome into play.
The aim of QUASILINPDE [322 is to nd the solutions of quasi-linear PDEs. These solutions
are then used by APPLYSYM [322, whi h applies the symmetry to lower the order of ODEs, to
al ulate similarity variables for PDEs, to ee tively redu e the number of independent variables of
a system of PDEs, and to generalize spe ial known solutions of ODEs and PDEs. To our knowledge,
APPLYSYM is one of the rst symboli programs that truly applies point symmetries that an
be al ulated with the program LIEPDE. The program APPLYSYM is automati but an also be
used intera tively. Thus far, APPLYSYM is only appli able to point symmetries for whi h the
generators are at worst rational. The a tual problem solving is done in all these programs through
a all to the pa kage CRACK for solving overdetermined systems of PDEs.
In [115, 116, Gerdt introdu ed the program HSYM for the expli it omputation of higher-order
symmetries for PDEs. If the given system of equations has arbitrary parameters, the ne essary
onditions for the existen e of higher order symmetries will lead to a system of algebrai equations
in the parameters. Via the program ASYS, that algebrai system is redu ed into standard form
via a Grobner basis algorithm. The fo us in Gerdt's work is on the investigation of the integrability of polynomial type nonlinear evolution equations, by verifying the existen e of higher order
symmetries and their asso iated onservation laws.
Sarlet and Vanden Bonne [250 oer spe i pro edures to assist in the omputation of adjoint
symmetries of se ond-order ODEs. This assistan e, however, is limited to the onstru tion of
determining equations for ertain lasses of adjoint symmetries, whi h are of the same nature as
determining equations for (generalized) symmetries, and relies on other pa kages su h as DIMSYM
below for solving these determining equations. In addition, pro edures have been written for testing
whether a given adjoint symmetry an give rise to a Lagrangian or a rst integral for the original
se ond-order equations.
The program DIMSYM by Sherring [277, in ollaboration with Prin e, was inspired by Head's
symmetry program LIE [135, in turn in uen ed by SPDE [260, 264, but is mu h larger and grew
independently of it during development. It is apable of nding various types of symmetries, urrently, point symmetries, Lie-Ba klund, and onditional symmetries. DIMSYM an isolate spe ial
ases, bring the determining equations in standard form for example, and aid in the solution of
group lassi ation problems. It attempts to determine the generators and allows one to he k
whether or not the generators are orre t. It allows the user to spe ify the dependen e of the symmetry ve tor eld oe ients, whi h is parti ularly pra ti al if one wants to ompute Lie-Ba klund
symmetries. DIMSYM provides the user with a lot of exibility: ansatze an be made, simpli ation routines an be alled separately, manual intervention is possible, et . Quite often su h
interventions indeed allow the user to omplete the desired omputations, whereas the DIMSYM
in auto-pilot mode may not.
DIMSYM has routines that onvert the input equations into standard form. Another attra tive
feature of the pa kage is that the integrator for the determining equations also works for systems
of linear homogeneous dierential equations not ne essarily obtained from symmetry analysis. The
overall strategy of the solver is to put the system of determining equations into standard form
based on Reid's algorithm (see Se tion 13.2.3), while solving expli itly any equations in the system
that the algorithm is apable of solving.
Finally, we mention the programs by Ito [156, 157, 158 for the determination of symmetries
and onservation laws of systems of evolution equations. Ito's program does not use any of the
algorithms dis ussed in Se tion 13.2.2, but uses innitesimal symmetries to determine the form of
onservation laws.
13.4.3.
MACSYMA PROGRAMS
One of the rst programs was written by S hwarzmeier and Rosenau [246, 271. Their program
al ulates the determining equations, simplies them a bit, but does not solve them automati ally.
The MACSYMA version of the program SYMCON [294, whi h was originally written in muMATH, tries to ompute Lie-point and Bessel-Haagen generalized symmetries (of any order) and
their onservation laws. Vafeades later produ ed PDELIE [295, 296, 297, 298, whi h is a drasti ally improved version of SYMCON [294. The pa kage PDELIE attempts to produ e similarity
solutions of ODEs, analyze PDEs with a multipli ative or additive s alar parameter, and ompute
the ommutator table and the stru ture onstants of the Lie algebra. PDELIE also allows one to
ompute the Noether onservation laws of variational systems.
PDELIE onsists of several subroutines. Let's dis uss the main ones. The fun tion PL SYM,
produ es the determining equations and the generators of the Lie group. It uses a standard
form algorithm by Reid and a set of heuristi rules to fa ilitate the integration. The fun tion
PL SOLVE tries to nd the invariants of the symmetry group. Using these invariants, it then
dimensionally redu es the given dierential equation. In ases where the redu ed equation is an
ODE, it tries to integrate expli itly, thus arriving at spe ial similarity solutions of the original
equation. The fun tions PL COMTAB performs omputations with elements in the ommutator
table of stru ture onstants of the Lie algebra. The fun tion PL CON omputes the densities of
the Noether onservation laws of systems of variational and divergen e type.
Just as PDELIE, the program SYM DE by Steinberg [282, 283, 284 was re ently added to the
out-of- ore library of MACSYMA. Steinberg's program omputes innitesimal symmetry operators
and the expli it form of the innitesimal transformations for simple systems. In ases where the
program annot automati ally nish the omputation, the user an intervene and, for instan e, ask
for innitesimal symmetries of polynomial form. The program solves some (or all) of the determining equations automati ally and, if needed, the user an (intera tively) add extra information.
Steinberg intends to extend his program so that it would in lude the al ulation of generalized
symmetries.
The program SYMMGRP.MAX written by Champagne, Hereman and Winternitz [52, 137, is
a modi ation of an earlier pa kage [53 that has been extensively used over the last de ade at the
University of Montreal and in many pla es elsewhere. It has been tested on hundreds of systems
of equations and has thus been solidly debugged.
The exibility within SYMMGRP.MAX and the possibility of using it intera tively, allows the
user to nd the symmetry group of arbitrarily large and ompli ated systems of equations on
relatively small omputers. For example, whenever the prolongation an be applied su essfully to
the omplete system, or a subset thereof, it produ es a list of determining equations. This list is
free of trivial fa tors, dupli ation and dierential redundan ies.
To make SYMMGRP.MAX work for large systems of dierential equations, the designers followed the path that would be taken in manual al ulations. That is, obtain in as simple a manner
as possible the simplest determining equations, solve them and feed the information ba k to the
omputer. Partial information an be extra ted very rapidly. For instan e, one an derive a subset
of the determining equations, su h as those that o ur as oe ients in the highest derivatives in
the independent variables. These are usually single-term equations, whi h express that the oe ients of the ve toreld are independent of some variables or depend linearly on some of the other
variables.
A feedba k me hanism fa ilitates the solution of the determining equations step by step on the
omputer; hen e, avoiding human error in the algebrai simpli ations. Typi ally, users will provide
information about the 's and ''s, as it be omes available from solving the determining equations
step by step. The amount of intera tion by the user will depend on the omplexity of the system
of dierential equations and on the apa ity of the omputer used. A worked example showing
the use of the feedba k me hanism is given in [52. Needless to say, with the feedba k me hanism,
the program SYMMGRP.MAX an also be used to verify previously al ulated solutions of the
determining equations and, hen e, dete t errors in the literature on the subje t.
Although not designed for that purpose, the program SYMMGRP.MAX an be easily adapted
to ompute the determining equations orresponding to non lassi al symmetries [60, 67, 68, 69, 71.
In [71, Clarkson and Manseld give a detailed explanation of su h an adaptation. Their proof of
orre tness of the proposed adaptation is based on the theory of Grobner bases.
13.4.4.
MAPLE PROGRAMS
In [49, Carminati, Devitt and Fee present LIESYMM for reating the determining equations via
the Harrison-Estabrook pro edure. Within LIESYMM various intera tive tools are available for
integrating the determining equations, and for working with Cartan's dierential forms. Their
program is independent of Donsig's dierential forms pa kage diorms, also available in Maple.
Khai T. Vu (Department of Mathemati s, Monash University, Clayton, Vi toria, Australia) has
translated Head's muMATH program LIE [135, dis ussed in Se tion 13.4.7, into Maple syntax.
In 1994, the Maple version of LIE, whi h omputes Lie-point symmetries, was still being tested
and, therefore, was not yet released. The -version of the program omputes the determining
equations, solves them, and gives the expli it forms of the (ve toreld) oe ients together with
the generators.
Hi kman [142, 143 oers a olle tion of Maple routines that aid in the omputation of Liepoint symmetries, non-lo al symmetries, and Wahlquist-Estabrook-type prolongations. The tools
for symmetry analysis in lude user-friendly pro edures to enter names of variables, to reate total
derivatives, to generate and prolong ve tor elds, and to derive and partially solve determining
equations. Program and do umentation are available via anonymous FTP from math. anterbury.a .nz.
Manseld has developed the pa kage DIRMETH for the omputation of symmetries via the
dire t method proposed by Clarkson and Kruskal [65. The program DIRMETH is part of DIFFGROB2, dis ussed in detail in Se tion 13.2.3. Other eorts in the design of pa kages for the dire t
method are given in the do toral thesis of Williams [311.
13.4.5.
MATHEMATICA PROGRAMS
Herod [141 developed MathSym for deriving the determining equations orresponding to Lie-point
symmetries, in luding non lassi al (or onditional) symmetries. Upon derivation of the determining
equations, the program redu es these equations via an algorithm based on the method of Riquier
and Janet. Herod's do toral thesis ontains the well-do umented ode of MathSym and appli ations
to various equations from uid dynami s.
Re ently, the pa kages Lie.m and Bae klund.m have been added by Baumann [22, 23, 24
to MathSour e, the Mathemati a Program Library. Baumann's program Lie.m [22 follows the
stru ture of our MACSYMA program SYMMGRP.MAX [52 very losely. Users familiar with
SYMMGRP.MAX will have a short learning urve with Lie.m. In ontrast to SYMMGRP.MAX,
the program Lie.m an handle trans endental fun tions in the input equations. The newest version
of Lie.m an be used to ompute point symmetries, onta t symmetries and non lassi al symmetries. Lie.m brings the determining equations in anoni al form via the pro edure of Janet and
Riquier, and goes on to solve the determining equations automati ally. A nite set of integration
rules, similar to the ones des ribed in Se tion 13.2.4, are implemented.
On e the solution of the determining equations is obtained, the program an ontinue with the
omputation of the ve tor basis, ideals, and ommutator table of the Lie algebra, its stru ture
onstants, Casimir operators, and its metri tensor.
Baumann's pa kage Bae klund.m [23, 24 ontains fun tions that attempt to ompute generalized symmetries for PDEs and ODEs and invariants of ODEs only. When applied to se ond-order
ODEs, the program attempts to verify if the omputed symmetries are of variational type. If so,
the program al ulates the orresponding invariants (integrals of motion). For the expli it al ulations to be su essful, quite often one has to spe ify that the oe ients in the ve toreld are
polynomials in the oordinates and momenta. With this \ansatz" one may not be able to obtain
all the generalized symmetries, but one may su essfully obtain expli it forms of invariants.
Berube and de Montigny [27 produ ed Lie-symmetry ode in Mathemati a. Their program
symmgroup. omputes the determining equations for Lie-point symmetries. In its syntax and
format symmgroup. losely follows the stru ture of SYMMGRP.MAX. The data for the program
may onsist of DEs with arbitrary fun tions. Trans endental fun tions in both dependent and
independent variables are also permitted. In [27, three well- hosen examples are given to illustrate
the apabilities of the program.
Finally, Coult (while at Carleton College, Northeld, Minnesota) developed a Mathemati a
program, temporarily alled symmgroup.m, for the omputation of the determining equations orresponding to Lie-point symmetries of a large lass of dierential equations (with polynomial terms).
13.4.6.
SCRATCHPAD AND AXIOM PROGRAMS
S hwarz [263 rewrote SPDE [262, 264 for use with version 1 of S rat hpad II, a symboli manipulation program developed by IBM. S rat hpad II is now superseded by AXIOM.
Seiler and o-workers [256, 274 are designing a pa kage that will ompute determining equations
for lassi al and non- lassi al symmetries. See Se tion 13.2.3 for a des ription of their program
JET for geometry omputations based on the jet bundle formalism.
13.4.7.
MUMATH PROGRAMS
The program LIE by Head [135 is based on version 4.12 of muMATH, but is self ontained and
runs on IBM ompatible PCs. As a matter of fa t, the program omes bundled with a limited
version of muMATH. Head's program al ulates and solves the determining equations (for Liepoint symmetries) automati ally for single equations and systems of dierential equations. LIE
also omputes the Lie ve tors and their ommutators. Interventions by the user are possible but
are rarely needed. The sour e ode of the program is available, in luding the heuristi routines
that attempt to solve the determining equations. Due to the limitations of muMATH, the program
LIE is bounded by the 256 KB of memory for program and workspa e. For a program of limited
size, LIE is remarkable in its a hievements.
Version 4.2 of LIE is freely available by FTP from various publi domain software ar hives su h
as SIMTEL and asso iated ar hives. By this printing, version 4.3 of LIE also will be available. In
addition to Lie-point symmetries, the new version will be able to ompute onta t and generalized
(Lie Ba klund) symmetries.
The SYMCON pa kage written by Vafeades [294 also uses muMATH to al ulate the determining equations (without solving them). The program is restri ted to point symmetries. Furthermore,
the program veries whether the symmetry group is of variational or divergen e type and omputes
the onservation laws asso iated with the symmetries. Unfortunately, these programs are onned
to the 256 KB memory a essible by muMATH and, annot presently handle very large systems of
equations. This limitation motivated Vafeades to rewrite his SYMCON program in MACSYMA
syntax [295, 296, 297, 298. The MACSYMA version of SYMCON an handle generalized symmetries and their onservation laws.
For ompleteness, Mikhailov developed software in muMATH to verify the integrability of
systems of PDEs by testing for the existen e of higher symmetries. The program omputes spe ial
symmetries, anoni al onservation laws, and arries out onformal transformations to bring PDEs
into anoni al form. With their PC program, Mikhailov, Shabat and Sokolov [195 produ ed an
exhaustive list of integrable nonlinear S hrodinger-type equations. In this ontext, integrable means
that the equations have innitely many onserved quantities and innitely many lo al symmetries.
13.4.8.
PROGRAMS FOR OTHER SYSTEMS
Kornyak and Fush hi h [101, 104, 170 developed programs in Turbo C and AMP for the omputation of Lie-Ba klund symmetries. Their programs also lassify equations with arbitrary parameters
and fun tions with respe t to su h symmetries. It is important to note that their programs redu e
the determining equations into passive form (see Se tion 13.2.3). All integrability onditions are
then expli it and, therefore, the resulting system is in involution.
We should mention their two FORMAC programs. The rst program, alled LB, was written
in the PL/1 language by Fedorova and Kornyak [91, 92. The su essor, alled LBF, was developed
by Fush hi h and Kornyak [104. Both programs reate the system of determining equations for
Lie-Ba klund symmetries and attempt to solve these equations. The program LBF, with its 1362
lines of PL/1-FORMAC ode, is ompletely automati and onsists of 37 subroutines, one of whi h
brings the determining equations in passive (Riquier-Janet) form. The program LB [92 is available
from the Computer Physi s Communi ations Program Library in Belfast. The above programs were
designed for low-memory requirements so that they ould run on PCs.
The PL/1-based FORMAC pa kage CRACKSTAR developed by Wolf [318, 319, 320 allows
one to investigate Lie symmetries of systems of PDEs, besides dealing with dynami al symmetries
of ODEs [225, and the like. A good overview of the apabilities of CRACKSTAR is given in
[319; a des ription of the routines and worked examples are in [225. For e ien y, CRACKSTAR
generates and solves rst-order determining equations early on, and then ontinues with the higherorder determining equations. The su essor of CRACKSTAR is the REDUCE pa kage CRACK
dis ussed already in Se tion 13.4.2.
Gerdt [115, 116, Gerdt and Zharkov [119 and Gerdt, Shva hka and Zharkov [117, 118 used
REDUCE and PL/1-FORMAC to investigate the integrability of nonlinear evolution equations.
Their program FORMINT ontains algorithms to al ulate Lie-Ba klund symmetries and onserved
densities, but does not use the jet bundle formalism.
The al ulation of the Lie group by omputer was also proposed by Popov [231, who used the
program SOPHUS for the al ulation of onservation laws of evolution equations.
In [42, Bo harov and Bronstein present SCoLAr, a pa kage written in standard PASCAL that
nds innitesimal symmetries and onservation laws of arbitrary systems of dierential equations.
An appli ation of SCoLAr to the Kadomtsev-Pogutse equations is given in [129.
The PC pa kage DELiA, standing for \Dierential Equations with Lie Approa h," is an outgrowth of the SCoLar proje t [42. DELiA, written in Turbo PASCAL by Bo harov and his
ollaborators [37, 38, 39, 40, is a stand-alone omputer algebra system for investigating dierential equations. It performs various tasks based on Lie's approa h, su h as the omputation of
Lie-point and Lie-Ba klund symmetries, anoni al onserved densities and generalized onservation
laws, simpli ation and partial integration of overdetermined systems of dierential equations, et .
The methods used in DELiA and many examples are well des ribed in the user guides [37, 40.
In order to be able to handle large problems, DELiA rst generates and solves rst-order
determining equations, and then ontinues to generate and solve the higher-order determining
equations. The analyzer/integrator, whi h is available as a separate tool at the user level, in ludes
a general algorithm for passivization [42, together with a set of integration rules for linear and
quasi-linear systems of PDEs. Currently, a MS Windows version of DEliA, alled MS Win DELiA,
is under development.
Using the algorithmi language REFAL, Topunov [287 developed a software pa kage for symmetry analysis that ontains subroutines to redu e determining systems in passive form.
13.5.
EXAMPLES
In this se tion we give three examples that illustrate the omputation of Lie-point symmetries
with symboli software. The rst and simplest example involves a single s alar nonlinear equation.
The se ond example illustrates how symmetries of a nonlinear omplex equation are omputed by
splitting the equation into a system of nonlinearly oupled equations for the real and imaginary
parts of the original dependent variable. The last and most ompli ated example involves a system
of ve tor equations that needs to be split into equations for its s alar omponents in order to
ompute its Lie symmetries.
13.5.1.
THE HARRY DYM EQUATION
Consider the Harry Dym equation [1,
u3 uxxx = 0:
(13.17)
Clearly, this is one equation with two independent variables and one dependent variable. The
assignments of the variables are as follows:
x 7 ! x[1 ; t 7 ! x[2 ; u 7 ! u[1:
(13.18)
This permits us to rewrite the equation (13.17) in a form a epted by the program SYMMGRP.MAX; i.e.,
ut
e1 : u[1,[0,1-u[1^3*u[1,[3,0.
For PDELIE and SYM DE the input form would be

'DIFF(U,T)-U^3*'DIFF(U,X,3).
For SPDE and LIE the program a epts

U(1,2)-U(1)^3*U(1,1,1,1).
Next, one sele ts the variable ut for elimination, e.g.

v1 : u[1,[0,1.
Then, the tprograms automati ally

ompute the determining equations for the oe ients eta[1 = x;
u
eta[2 = , and phi[1 = ' of the ve tor eld
= x

+
t + 'u :
x
t
u
(13.19)
There are only eight determining equations,
eta2
u[1
= 0;
eta2
x[1
= 0;
eta1
u[1
= 0;
2 phi1
u[12
= 0;
(13.20)
2 eta1
x[12
= 0;
3 phi1
x[13
= 0;
3 eta1
u[13
x[13
= 0;
2 phi1
u[1x[1
phi1
x[2
3 phi1
3u[13 u[1x[12 + xeta1

[2
u[13
(13.21)
3u[1 xeta1
[1 + 3 phi1 = 0:
These determining equations are easily solved expli itly, either automati ally with SPDE, LIE and
PDELIE, or with the feedba k me hanism within SYM DE and SYMMGRP.MAX. The general
solution, rewritten in the original variables, is
x = k1 + k3 x + k5 x2 ;
(13.22)
t
= k2 3k4 t ;
(13.23)
u
' = (k3 + k4 + 2k5 x) u ;
(13.24)
where k1; :::; k5 are arbitrary onstants. The ve innitesimal generators then are
G1 = x ;
(13.25)
G2 = t ;
(13.26)
G3 = xx + uu ;
(13.27)
G4 = 3tt + uu ;
(13.28)
2
G5 = x x + 2xuu :
(13.29)
Clearly, (13.17) is invariant under translations (G1 and G2 ) and s aling (G3 and G4): The ow
orresponding to ea h of the innitesimal generators an be obtained via simple integration. As an
example, let us ompute the ow orresponding to G5. This requires integration of the rst-order
system
dx~
= x~2 ;
x~(0) = x ;
u[1
eta2
x[2
d
dt~
d
t~(0)
= 0;
t;
(13.30)
du~
d
= 2~xu~ ;
u~(0) = u ;
where is the parameter of the transformation group. One readily obtains
x
x~() =
(1 x) ;
t~()
u~()
= (1
(13.31)
(13.32)
t;
u
:
x)2
(13.33)
Therefore, we on lude that for any solution u = f (x; t) of equation (13.17), the transformed
solution
x~ ~
u~(~x; t~) = (1 + x~)2 f (
(13.34)
1 + x~ ; t)
will solve u~t~ u~3u~x~x~x~ = 0:
13.5.2.

THE NONLINEAR SCHRODINGER
EQUATION
In order to ompute the Lie-point symmetries of the elebrated nonlinear S hrodinger equation [1,
iut + uxx + ujuj2 = 0;
(13.35)
one needs to repla e the single omplex equation by a oupled system. One way of doing that is
by introdu ing the real and imaginary parts v; w of the omplex variable u via u(x; t) = v(x; t) +
iw(x; t): This yields
vt + wxx + w(v 2 + w2 ) = 0;
(13.36)
2
2
wt vxx v (v + w ) = 0:
One alternative is to repla e (13.35) by a system onsisting of the equation itself and its omplex
onjugate [58 and to interpret the variables u and v = u as real,
k ut + uxx + v u2 = 0;
(13.37)
2
k vt + vxx + u v = 0:
In order to work with real quantities throughout, the imaginary unit i was temporarily repla ed
by the onstant k during the omputations. On e the determining equations are obtained, k = i
should be reintrodu ed.
Another alternative is to write u(x; t) = R(x; t) exp(i
(x; t)); thus repla ing (13.35) by a oupled
system
Rt + 2Rx
x + R
xx = 0;
(13.38)
2
3
Rxx R
t R
x + R = 0;
for the real modulus R(x; t) and real phase
(x; t).
Adhering to (13.36), SYMMGRP.MAX (or for that matter any other symmetry program)
qui kly generates the twenty determining equations for the oe ients of the ve tor eld

= x + t + 'v + 'w :
(13.39)
x
t
v
w
The rst eleven single-term determining equations are similar to (13.20), and provide information
about the dependen ies of the 0s and the 0s on x; t; v and w; and their linearity in the latter two
independent variables. The remaining nine determining equations are a bit more ompli ated, but
the entire system is readily solved.
In the original variables, the solution reads
x = k1 + 2k4 t + k5 x ;
(13.40)
t = k2 + 2k5 t ;
(13.41)
v
' = k3 w k4 xw k5 v ;
(13.42)
w
' = k3 v + k4 xv k5 w ;
(13.43)
where k1 ; :::; k5 are arbitrary onstants. As in the previous examples, the omplete symmetry
algebra is spanned by ve ve tor elds (generators):
G1
G2
G3
G4
G5
= x ;
= t ;
= wv vw ;
= 2tx x(wv vw ) ;
= xx + 2tt vv ww :
(13.44)
(13.45)
(13.46)
(13.47)
(13.48)
If we had arried out the omputations with (13.38), where u(x; t) = R(x; t) exp(i
(x; t)); we
would have found:
G1
G2
G3
G4
G5
= x ;
= t ;
=
;
= 2tx x
;
= xx + 2tt RR :
(13.49)
(13.50)
(13.51)
(13.52)
(13.53)
Either way, (13.35) is invariant under translations in spa e and time (G1 and G2 ): Generator
G3 orresponds to adding an arbitrary onstant to the phase of u: The Galilean boost is generated
by G4 : Finally, G5 indi ates invarian e of the equation under s aling (or dilation). Similarity
redu tions an then be obtained by solving the hara teristi equations,
dx
x
dv dw
= dt
=
= ;
t 'v 'w
(13.54)
or equivalently, the invariant surfa e onditions

x vx + t vt 'v
xwx + t wt 'w
= 0;
= 0:
(13.55)
The a tual redu tions an be found in [58, where a quite general nonlinear S hrodinger equation
is treated. It is well known [176 that all the redu tions of the NLS an be obtained from G1
through G5 ; in other words, non lassi al symmetries would not lead to new symmetry redu tions.
To ompute non lassi al symmetries of (13.36), it su es to repla e vt and wt from (13.55). If
t 6= 0; we set t = 1 for simpli ity. Thus,
vt
wt
=
=
x vx + 'v ;
x wx + 'w :
(13.56)
(13.57)
The ase t = 0 has to be onsidered separately. Sin e SYMMGRP.MAX allows the user to give
information about the oe ients in the ve tor eld, the omputation an now pro eed as in the
lassi al ase. For worked examples, we refer the reader to [60, 71.
13.5.3.
THE MAGNETO-HYDRO-DYNAMICS EQUATIONS
As an example of a large system of dierential equations, we take the equations for Magneto-HydroDynami s (MHD) [98 and arry out the sear h for Lie-point symmetries with SYMMGRP.MAX.
The MHD equations, with or without dissipative terms, have be ome a ben hmark for developers of Lie symmetry pa kages. Nu i [201 omputed the lassi al symmetries in 1984 and
ommented that the al ulations by hand took her a year. Mi hel Grundland undertook the same
formidable task and also nished the job in about a year. He onveyed to the author that the long
winter in Newfoundland (where he was at the time) helped.
If we negle t dissipative ee ts, and thus restri t the analysis to the ideal ase, the MHD
equations an be redu ed to

+ (~v r) + r ~v
t
~v

1 ~ 2) (H~ r)H~

+
(
~v r)~v + r(p + H
t
2
~
H
+ (~v r)H~ + H~ r ~v (H~ r)~v
t
r H~
p
p
(
)
+
(
~v r)( )

t

= 0;
= ~0 ;
(13.58)
(13.59)
= ~0 ;
= 0;
= 0;
(13.60)
(13.61)
(13.62)
~
with pressure p, mass density , oe ient of vis osity , uid velo ity ~v and magneti eld H:
Using the rst equation, we eliminate from the last equation, hen e repla ing it by
p
+ p(r ~v) + (~v r)p = 0:
t
(13.63)
If we split the ve tor equations in s alar equations for the ve tor omponents, we have a system of
nine equations, with four independent variables and eight dependent variables. For onvenien e, we
denote the omponents of the ve tor ~v by vx; vy and vz ; not to be onfused with partial derivatives
of v:
The variables to be eliminated are sele ted as follows: for the rst seven variables and the ninth
variable we pi k the partial derivatives with respe t to t of ; vx; vy ; vz ; Hx; Hy ; Hz and p. From the
eighth equation, we sele t Hx =x for elimination.
We will only onsider the ase where 6= 0: We ran this ase on a Digital VAX 4500 with 64
MB of RAM, and on an IBM Ris 6000 workstation with 32 MB of RAM. On the VAX it took 50
minutes of CPU time, on the IBM workstation 1 hour and 50 minutes, for SYMMGRP.MAX to
reate the 222 determining equations for the oe ients of the ve tor eld

=
+

+
y + z + t + ' + 'p
x
y
z
t

p

'v
+
'v
+
'v
+
'H
+
'H
vx
vy
vz
Hx
Hy
x
+ 'H

:
Hz
Using SYMMGRP.MAX intera tively, we then integrated the determining system and obtained
the solution expressed in the original variables,
x = k2 + k5 t k8 y k9 z + k11 x ;
(13.64)
= k3 + k6 t + k8 x k10 z + k11 y ;
= k4 + k7 t + k9 x + k10 y + k11 z ;
= k1 + k12 t ;
= 2 (k11 k12 k13 ) ;
= 2 k13 p ;
= k5 k8 vy k9 vz + (k11 k12 )vx ;
= k6 + k8 vx k10 vz + (k11 k12 )vy ;
= k7 + k9 vx + k10 vy + (k11 k12 )vz ;
= k13 Hx k8Hy k9Hz ;
= k13 Hy + k8Hx k10 Hz ;
= k13 Hz + k9Hx + k10Hy :
It is lear that there is a thirteen-dimensional Lie algebra spanned by the generators:
G1 = t ;
G2 = x ;
G3 = y ;
G4 = z ;
G5 = tx + v ;
G6 = ty + v ;
G7 = tz + v ;
G8 = xy yx + vx v vy v + Hx H Hy H ;
G9 = yz zy + vy v vz v + Hy H Hz H ;
G10 = zx xz + vz v vx v + Hz H Hx H ;
G11 = xx + yy + zz 2 + vx v + vy v + vz v ;
G12 = tt + 2 (vx v + vy v + vz v ) ;
G13 = 2 + 2pp + HxH + Hy H + Hz H :
y
z
t
'
'p
'v
'v
'v
'H
'H
'H
x
y
z
x
y
z
(13.65)
(13.66)
(13.67)
(13.68)
(13.69)
(13.70)
(13.71)
(13.72)
(13.73)
(13.74)
(13.75)
(13.76)
(13.77)
(13.78)
(13.79)
(13.80)
(13.81)
(13.82)
(13.83)
(13.84)
(13.85)
(13.86)
(13.87)
(13.88)
Thus, the MHD equations (13.58)-(13.62) are invariant under translations G2 through G4, Galilean
boosts G5 through G7, rotations G8 through G10 , and dilations G11 through G13: In ontrast to the
results obtained for the 1+1 and the 2+1 dimensional versions of the MHD problem, the dimension
of the Lie algebra for (13.58)-(13.62) in the full 3+1 dimensions (x; y; z and time t) is independent
of the value of the oe ient of vis osity : Our results onrm those in [98, and of those of
Grundland and Lalague [126, 127, who omputed the lassi al and some non lassi al symmetries
of the MHD, and also lassied all the subalgebras in onjuga y lasses. The MHD system and
our results have been used by other investigators [49, 241, 277 to test their symmetry programs.
13.5.4.
OTHER INTERESTING EXAMPLES
Champagne and Winternitz [53 used SYMMGRP.MAX to ompute the Lie-point symmetries of
the Korteweg de Vries equation with variable oe ients,
ut + f (x; t)uux + g (x; t)uxxx = 0;
(13.89)
illustrating that SYMMGRP.MAX an easily handle equations involving arbitrary fun tions.
Also in [53, the point symmetries of a modied Kadomtsev-Petviashvili equation
(uxxx 2u3x 4ut)x 6uxxuy + 3uyy = 0;
(13.90)
in 2+1 dimensions are omputed with SYMMGRP.MAX. This example was hosen be ause it leads
to an innite-dimensional Lie algebra involving four arbitrary fun tions of t:
A ompletely worked example of the al ulation of Lie-point symmetries of a system of PDEs is
given in [52. This example shows the use of the feedba k me hanism within SYMMGRP.MAX to
ompletely solve the determining equations. It involves the Karpman equations [163, for whi h our
symmetries were independently veried with REDUCE programs by Kersten and Gragert (private
ommuni ation), Sherring and Prin e [278, and Wolf [321.
Finally, SYMMGRP.MAX was re ently used to ompute Lie-point symmetries of two large
systems of equations representing lassi al eld theories [139. Currently, the author is adapting
SYMMGRP.MAX for the al ulation of Lie-point symmetries of dieren e-dierential equations.
13.6.
CONCLUSION
The various programs that we reviewed need very little data and are straightforward to use provided
the user has a ess to and knows the basi s of the underlying CAS, su h as MACSYMA, Maple,
Mathemati a and REDUCE.
Apart from the theoreti al study of the underlying mathemati s, there is a need for further
development and implementation of ee tive algorithms for generating, redu ing, simplifying and
fully solving the determining equations for ( lassi al and non lassi al) Lie-point symmetries and
generalized or Lie-Ba klund symmetries.
The availability of sophisti ated symboli programs ertainly will a elerate the study of symmetries of physi ally important systems of dierential equations in lassi al me hani s, uid dynami s,
elasti ity, and other applied areas.
ACKNOWLEDGEMENTS
While writing this survey I was able to onsult with many experts in this eld. I am grateful for their
suggestions, omments, orre tions and riti isms whi h greatly helped improve the manus ript.
Sin ere thanks goes to M. Ablowitz, B. Abraham-Shrauner, G. Baumann, D. Berube, G.
Bluman, A. Bo harov, F. Cantrijn, B. Champagne, P. Clarkson, R. Conte, N. Coult, B. Fu hssteiner,
V. Gerdt, K. Govinder, P. Gragert, M. Grundland, F. Gungor, A. Head, B. Herbst, S. Herod, M.
Hi kman, D. Holm, M. Kruskal, D. Levi, M. Ma Callum, F. Mahomed, E. Manseld, L. Mar hildon,
A. Mikhailov, M.C. Nu i, P. Olver, J. Ondi h, G. Prin e, D. Rand, G. Reid, M. Roelofs, W. Sarlet,
W. Seiler, F. S hwarz, J. Sherring, W.-H. Steeb, S. Steinberg, Z. Thomova, M. Torrisi, P. Vafeades,
K. Vu, T. Wil ox, P. Winternitz, T. Wolf, and C. Wulfman.
Spe ial thanks goes to S. Collart for inviting me to write a preliminary version of this survey paper for Euromath Bulletin. It is my pleasure to thank N.H. Ibragimov for giving me the
opportunity to ompletely rework that review paper and to publish it as a hapter in this book.
I am grateful to T. Hearn and H. Melenk for providing me with a free opy of REDUCE.
Resear h for this survey paper was supported in part by Grant # CCR-9300978 of the National
S ien e Foundation of the United States of Ameri a.
Table 1
List of urrent symmetry programs
Name & System
Distributor
Developer's Address
Refs.
Email & Anonymous FTP
CRACK
LIEPDE
& APPLYSYM
(REDUCE)
REDUCE
Network Library
T. Wolf & A. Brand

T. Wolf
S hool Math. S i.
Queen Mary
& Westfield College
London E1 4NS, UK
[321
[323
T.Wolfmaths.qmw.a .uk
DELiA
(Pas al)
Beaver Soft
715 O ean View Ave
Brooklyn
NY 11235, USA
A. Bo harov et al.
A. Bo harov
Wolfram Resear h
100 Trade Center Dr.
Urbana-Champaign
IL 61820-7237, USA
[37
E. Mansfield
Dept. of Maths.
Univ. of Exeter
Exeter EX4 4QE
United Kingdom
[189
J. Sherring
G. Prin e
S hool of Maths.
Latrobe University
Bundoora, VI 3083
Australia
[277
V. Eliseev et al.
V. Eliseev
Lab. Comp. Te h. Aut.
JINR, Dubna
Mos ow Region
141980 Russia
[85
Cost:
$ 300
DIFFGROB2
(Maple)
DIMSYM
(REDUCE)
LaTrobe University
S hool of Maths.
Cost:
LIE
(REDUCE)
$ 225
CPC
Program Library
Belfast
N. Ireland
Cat. No. AABS
galois.maths.qmw.a .uk
/ftp/pub/ ra k
alexeiwri. om
lizmaths.exeter.a .uk
eu lid.exeter.a .uk
pub/liz
matjslure.latrobe.edu.au
G.Prin elatrobe.edu.au
ftp.latrobe.edu.au
/ftp/pub/dimsym
Table 1 ont.
Name & System
Distributor
Developer's Address
Refs.
LIE
(muMath)
(independent)
SIMTEL
A. Head
CSIRO
Div. Mat. S i. & Te h.
Clayton, Vi toria
3168 Australia
[135
headrivett.mst. siro.au
Lie
& LieBae klund
(Mathemati a)
Wolfram
Resear h
MathSour e
0202-622
0204-680
G. Baumann
Abt. Math. Phys.
Universitat Ulm
D-7900 Ulm
Germany
[22
[23
[24
P. Gragert & P. Kersten

P. Kersten
Dept. Appl. Math.
University of Twente
7500 AE Ens hede
The Netherlands
[120
[123
J. Carminati et al.
G. Fee
Dept. Comp. S i.
University of Waterloo
Waterloo, Canada
[49
S. Herod
Program Appl. Math.
University of Colorado
Boulder, CO 80309, USA
[141
M.C. Nu i
Dept. di Mathemati a
Universita di Perugia
06100 Perugia, Italy
[202
LIEDF/INFSYM
& others
(REDUCE)
Liesymm
(Maple)
MathSym
(Mathemati a)
NUSY
(REDUCE)
Waterloo
Maple
Software
(Pa kages)
wuar hive.wustl.edu
/edu/math/msdos/..
../adv.diff.equations/lie42
bautheophys.physik.uni-ulm.de
mathsour e.wri. om
/pub/PureMath/Cal ulus
gragertmath.utwente.nl
kerstenmath.utwente.nl
wmsidaisy.uwaterloo. a
wmsidaisy.waterloo.edu
sherodnewton. olorado.edu
newton. olorado.edu
pub/mathsym
nu igauss.dipmat.unipg.it
Table 1 ont.
Name & System
Distributor
Developer's Address
Refs.
PDELIE
(MACSYMA)
MACSYMA
Out-of-Core
Library
P. Vafeades
Dept. of Eng. S i.
Trinity University
San Antonio, TX 78212, USA
[295
[296
[297
peterengr.trinity.edu
REDUCE
Program Lib.
Rand Corp.
F. S hwarz
GMD, Inst. SCAI
D-53731 Sankt Augustin
Germany
[264
[266
fritz.s hwarzgmd.de
G. Reid & A. Wittkopf

G. Reid
Math. Dept.
Univ. Brit. Columbia
Van ouver, BC
Canada V6T IZ2
[241
S. Steinberg
Dept. Math. & Stat.
Univ. New Mexi o
Albuquerque, NM 87131, USA
[284
D. B
erub
e & M. de Montigny
M. de Montigny
D. Berub
e
Centre Traitement Inform.
Univ. Laval, St.-Froy
Canada G1K 7P4
[27
B. Champagne et al.
W. Hereman
Dept. Math. Comp. S i.
Colorado S h. of Mines
Golden, CO 80401, USA
[52
SPDE
& SYMSIZE
(REDUCE)
SYMCAL
(Maple
& MACSYMA)
SYM DE
(MACSYMA)
MACSYMA
Out-of-Core
Library
symmgroup.
(Mathemati a)
SYMMGRP.MAX
(MACSYMA)
CPC
Program Lib.
Belfast
N. Ireland
Cat. No. ACBI
gumbo.engr.trinity.edu
redu e-netlibrand.org
reidmath.ub . a
math.ub . a
pub/reid
stanlymath.unm.edu
montignyphysi s.m gill. a

berubegenesis.ulaval. a
genesis.ulaval. a
/pub/Mathemati a/symgroup
wheremanlie.mines. olorado.edu
ftp.mines. olorado.edu
pub/papers/math s dept/symmetry
or onta t: p v1.am.qub.a .uk
Table 2
S ope of symmetry programs
Name
System
Developer(s)
Point
General.
Non l.
Solves Det. Eqs.
CRACK
REDUCE
Wolf & Brand
Yes
DELiA
Pas al
Bo harov et al.
Yes
Yes
No
Yes
DIFFGROB2
Maple
Mansfield
Redu tion
DIMSYM
REDUCE
Sherring
Yes
Yes
No
Yes
LIE
REDUCE
Eliseev et al.
Yes
Yes
No
No
LIE
muMath
Head
Yes
Yes
No
Yes
Lie
Mathemati a
Baumann
Yes
No
Yes
Yes
LieBae klund
Mathemati a
Baumann
No
Yes
No
Intera tive
LIEDF/INFSYM
REDUCE
Gragert & Kersten
Yes
Yes
No
Intera tive
LIEPDE
REDUCE
Wolf & Brand
Yes
Yes
No
Yes
Table 2 ont.
S ope of symmetry programs
Name
System
Developer(s)
Point
General.
Non l.
Solves Det. Eqs.
Liesymm
Maple
Carminati et al.
Yes
No
No
Intera tive
MathSym
Mathemati a
Herod
Yes
No
Yes
Redu tion
NUSY
REDUCE
Nu i
Yes
Yes
Yes
Intera tive
PDELIE
MACSYMA
Vafeades
Yes
Yes
No
Yes
SPDE
REDUCE
S hwarz
Yes
No
No
Yes
SYMCAL
Maple/MACSYMA
Reid & Wittkopf
Redu tion
SYM DE
MACSYMA
Steinberg
Yes
No
No
Partially
symgroup.
Mathemati a
Berub
e & de Montigny
Yes
No
No
No
SYMMGRP.MAX
MACSYMA
Champagne et al.
Yes
No
Yes
Intera tive
SYMSIZE
REDUCE
S hwarz
Redu tion
Bibliography
[1
[2
[3
[4
[5
[6
[7
[8
[9
[10
[11
and
, Solitons, Nonlinear Evolution Equations and Inverse
London Mathemati al So iety Le ture Notes on Mathemati s 149 (Cambridge
University Press, Cambridge, UK, 1991).
, Hidden symmetries and linearization of the modied Painleve-In e
equation, J. Math. Phys. 34 (1993) 4809-4816.
and
, Hidden symmetries of dierential equations, Contemp. Math. 160 (1994) 1-13.
and
, Hidden symmetries of nonlinear ordinary
dierential equations, in: Exploiting Symmetry in Applied and Numeri al Analysis, Pro .
AMS-SIAM Summer Seminar, Fort Collins, Colorado, 1992, Eds.: E. Allgower, K. Georg and
R. Miranda. Le tures in Applied Mathemati s 29 (Ameri an Mathemati al So iety, Providen e, Rhode Island, 1993) 1-10.
and
, Fast algorithms to sear h for the rational solutions of linear dierential equations with polynomial oe ients, in: Pro . ISSAC '91, Bonn,
Germany, 1991, Ed.: S.M. Watt (ACM Press, New York, 1991) 267-270.
and
, Nonlo al symmetries. Heuristi
approa h, J. Sov. Math. 55 (1991) 1401-1450.
and
, Eds., Exploiting Symmetry in Applied and
Numeri al Analysis, Pro . AMS-SIAM Summer Seminar, Fort Collins, Colorado, 1992, Le tures in Applied Mathemati s 29 (Ameri an Mathemati al So iety, Providen e, Rhode Island,
1993).
, Nonlinear Ordinary Dierential Equations in Transport Pro esses (A ademi
Press, New York, 1968).
, Nonlinear Partial Dierential Equations in Engineering. Vol. I (A ademi
, Nonlinear Partial Dierential Equations in Engineering. Vol. II (A ademi
, Symmetry in nonlinear me hani s, in: Nonlinear Equations in the Applied
S ien es, Eds.: W.F. Ames and C. Rogers (A ademi Press, Boston, 1992) 33-78.
M.J. Ablowitz
S attering,
P.A. Clarkson
B. Abraham-Shrauner
B. Abraham-Shrauner
B. Abraham-Shrauner
S.A. Abramov
A. Guo
P.G.L. Lea h
K.Yu. Kvashenko
I.Sh. Akhatov, R.K. Gazizov
E. Allgower, K. Georg
W.F. Ames
W.F. Ames
W.F. Ames
W.F. Ames
N.Kh. Ibragimov
R. Miranda
[12
[13
[14
[15
[16
[17
[18
[19
[20
[21
[22
[23
[24
[25
[26
and
, Eds., Nonlinear Equations in the Applied S ien es (A ademi
Press, Boston, 1992).
and
, Lie-Ba klund Transformations in Appli ations,
Studies in Applied Mathemati s 1 (SIAM, Philadelphia, 1979).
and
, Internal, external and generalized symmetries, Adv. in Math. 100 (1993) 53-100.
and
, Internal symmetries of dierential equations, in: Modern Group Analysis: Advan ed Analyti al and Computational Methods in
Mathemati al Physi s. Pro . Int. Workshop A ireale, Catania, Italy, 1992, Eds.: N.H. Ibragimov, M. Torrisi and A. Valenti (Kluwer A ademi Publishers, Dordre ht, The Netherlands,
1993) 7-21.
and
, Dokl. Akad. Nauk S.S.S.R. 214 (1974)
p. 296, in Russian.
and
, Non lassi al symmetry solutions and the
methods of Bluman-Cole and Clarkson-Kruskal, J. Math. Phys. 34 (1993) 4692-4703.
, Le tures on Lie Groups and Systems of Ordinary and Partial Dierential Equations (Los Alamos, New Mexi o, 1983).
, Sophus Lie, The Mathemati al Intelligen er 16, no. 1 (1994) 16-19.
and
, Perturbation methods in group analysis, J. Sov. Math. 55 (1991) 1450-1490.
, Similarity, Self-similarity and Intermediate Asymptoti s (Consultants
Bureau, New York, 1979).
, Lie Symmetries of Dierential Equations: A Mathemati a program to determine Lie symmetries. Wolfram Resear h In ., Champaign, Illinois, MathSour e 0202-622,
1992.
, Generalized Symmetries: A Mathemati a Program to Determine Lie-Ba klund
Symmetries. Wolfram Resear h In ., Champaign, Illinois, MathSour e 0204-680, 1993.
, Appli ations of the generalized symmetry method, in: Modern Group Analysis: Advan ed Analyti al and Computational Methods in Mathemati al Physi s. Pro . Int.
Workshop A ireale, Catania, Italy, 1992, Eds.: N.H. Ibragimov, M. Torrisi and A. Valenti
(Kluwer A ademi Publishers, Dordre ht, The Netherlands, 1993) 43-53.
and
, Grobner Bases (Springer Verlag, Berlin, 1993).
and
, A Survey of Lie Groups and Lie Algebras with Appli ations and Computational Methods, Classi s in Applied Mathemati s 2 (SIAM, Philadelphia,
1972 & 1989).
W.F. Ames
C. Rogers
R.L. Anderson
N.H. Ibragimov
I.M. Anderson, N. Kamran
P.J. Olver
I.M. Anderson, N. Kamran
E.A. Arais, V.P. Shapeev
P.J. Olver
N.N. Janenko
D.J. Arrigo, P. Broadbridge
J.M. Hill
R. Axford
N.A. Baas
V.A. Baikov, R.K. Gazizov
N.Kh. Ibragimov
G.I. Barenblatt
G. Baumann
G. Baumann
G. Baumann
T. Be ker
V. Weispfenning
J.G. Belinfante
B. Kolman
[27
[28
[29
[30
[31
[32
[33
[34
[35
[36
[37
[38
[39
[40
and
, A Mathemati a Program for the Cal ulation of Lie Point
Symmetries of Systems of Dierential Equations. Preprint CRM-1822 (Centre de Re her hes
Mathematiques, Montreal, Canada, 1992).
, Lie-group theory for symboli integration of rst-order dierential equations,
in: Pro . 1979 MACSYMA User's Conferen e, Washington, D.C., 1979, Ed.: V.E. Lewis
(MIT Press, Boston, 1979) 362-384.
, Lezioni sulla Teoria dei Gruppi Continui Finiti di Transformatizioni (Enri i
Spoerri, Pisa, Italy, 1918).
, Simplifying the form of Lie groups admitted by a given dierential equation,
J. Math. Anal. Appl. 145 (1990) 52-62.
, Potential symmetries and linearization, in: Appli ations of Analyti and
Geometri Methods to Nonlinear Dierential Equations. NATO ASI Series C: Math. and
Phys. S s., Ed.: P.A. Clarkson (Kluwer A ademi Publishers, Dordre ht, The Netherlands,
1993) 363-373.
, Potential symmetries and equivalent onservation laws, in: Modern Group
Analysis: Advan ed Analyti al and Computational Methods in Mathemati al Physi s. Pro .
Int. Workshop A ireale, Catania, Italy, 1992, Eds.: N.H. Ibragimov, M. Torrisi and A. Valenti
, An overview of potential symmetries, in: Exploiting Symmetry in Applied
and Numeri al Analysis, Pro . AMS-SIAM Summer Seminar, Fort Collins, Colorado, 1992,
Eds.: E. Allgower, K. Georg and R. Miranda. Le tures in Applied Mathemati s 29 (Ameri an
Mathemati al So iety, Providen e, Rhode Island, 1993) 97-109.
and
, The general similarity solution of the heat equation, J. Math.
Me h. 18 (1969) 1025-1042.
and
, Similarity Methods for Dierential Equations, Applied
Mathemati al S ien es 13 (Springer Verlag, New York, 1974).
and
, Symmetries and Dierential Equations, Applied Mathemati al S ien e 81 (Springer Verlag, New York, 1989).
, DEliA: A System of Exa t Analysis of Dierential Equations using S. Lie
Approa h. Report by Joint Venture OWIMEX Program Systems Institute of the U.S.S.R.
(A ademy of S ien es, Pereslavl-Zalessky, U.S.S.R., 1989).
, DEliA: proje t presentation, SIGSAM Bulletin 24 (1990) 37-38.
, Will DEliA grow into an expert system?, in: Design and Implementation
of Symboli Computation Systems, Pro . Int. Symposium, Capri, Italy, 1990, Ed.: A. Miola,
Le ture Notes in Computer S ien e 429 (Springer Verlag, Berlin, 1990) 266-267.
, DEliA: A System of Exa t Analysis of Dierential Equations using S. Lie
Approa h. DELiA 1.5.1 User Guide (Beaver Soft Programming Team, Brooklyn, New York,
1991).
D. B
erub
e
M. de Montigny
W.A. Beyer
L. Bian hi
G.W. Bluman
G.W. Bluman
G.W. Bluman
G.W. Bluman
G.W. Bluman
G.W. Bluman
G.W. Bluman
A.V. Bo harov
A.V. Bo harov
A.V. Bo harov
A.V. Bo harov
J.D. Cole
J.D. Cole
S. Kumei
[41
[42
[43
[44
[45
[46
[47
[48
[49
[50
[51
[52
[53
[54
[55
, Symboli Solvers for Nonlinear Dierential Equations, The Mathemati a

Journal 3 (1993) 63-69.
and
, E iently implementing two methods of the geometri al theory of dierential equations: an experien e in algorithm and software design, A ta
Appl. Math. 16 (1989) 143-166.
and
, Similarity, symmetry and solutions of nonlinear boundary
value problems, Spe ial Issue of Mathl. Comput. Modelling 18 (1993) 1-170.
, On solutions of linear ordinary dierential equations in their oe ient eld,
J. Symb. Comp. 13 (1992) 413-439.
and
,
Exterior Dierential Systems. Mathemati al S ien es Resear h Institute Publi ations 18
(Springer Verlag, New York, 1991).
, Grobner bases: an algorithmi method in polynomial ideal theory, Chapter 6, in: Multidimensional Systems Theory, Ed.: N.K. Bose (D. Reidel, Dordre ht, The
Netherlands, 1985) 184-232.
, An Implementation of Grobner Bases in Mathemati a, Wolfram Resear h
In ., Champaign, Illinois, MathSour e 0205-300, 1992.
, Introdu tory Treatise on Lie's Theory of Finite Continuous Transformation
Groups (Chelsea Publishing Company, New York, 1903, 1966).
and
, Isogroups of dierential equations using algebrai
omputing, J. Sym. Comp. 14 (1992) 103-120.
, Grobner bases and dierential ideals, in: Applied Algebra, Algebrai
Algorithms, and Error-Corre ting Codes. Pro . 5th Int. Conf. AAECC-5, Eds.: L. Huguet
and A. Poli. Le ture Notes in Computer S ien e 356 (Springer Verlag, Berlin, 1989) 129-140.
, Les Systemes Dierentiels Exterieur et leurs Appli ations Geometrique (Hermann, Paris, 1946).
and
, The omputer al ulation of Lie point
symmetries of large systems of dierential equations, Comp. Phys. Comm. 66 (1991) 319-340.
and
, A MACSYMA Program for Cal ulating the Symmetry
Group of a System of Dierential Equations. Preprint CRM-1278 (Centre de Re her hes
Mathematiques, Montreal, Canada, 1985).
, Algorithms Using Lie Transformation Groups to Solve First Order Ordinary
Dierential Equations Algebrai ally. Ph.D Thesis, Department of Computer S ien e (University of California, Berkeley, California, 1980).
, Theory of Lie Groups. Vol 1 (Prin eton University Press, Prin eton, New
Jersey, 1946).
A.V. Bo harov
A.V. Bo harov
M.L. Bronstein
P. Broadbridge
J.M. Hill
M. Bronstein
R.L. Bryant, S.S. Chern, R.B. Gardner, H.L. Golds hmidt
B. Bu hberger
B. Bu hberger
J.E. Campbell
J. Carminati, J.S. Devitt
G.J. Fee
-Ferro
G. Carra
E. Cartan
B. Champagne, W. Hereman
B. Champagne
B.W. Char
C.C. Chevalley
P. Winternitz
P. Winternitz
P.A. Griffiths
[56
[57
[58
[59
[60
[61
[62
[63
[64
[65
[66
[67
[68
[69
[70
and
, Lie-point symmetries in bifur ation problems, Ann. Inst. Henri
Poin are, Physique Theorique 56 (1992) 375-414.
, New exa t solutions of the Boussinesq Equation. Euro. J. Appl. Math. 1
(1990) 279-300.
, Dimensional redu tions and exa t solutions of a generalized nonlinear
S hrodinger equation, Nonlinearity 5 (1992) 453-472.
, Non lassi al symmetry redu tions of nonlinear partial dierential equations,
Mathl. Comput. Modelling 18 (1993) 45-68.
, Non lassi al symmetry redu tions of the Boussinesq equation. Chaos, Solitons and Fra tals 4 (1994), in press.
and
, Non lassi al symmetry redu tions and exa t solutions of the
Zabolotskaya-Khokhlov equation, Euro. J. Appl. Math. 3 (1992) 381-415.
and
, Symmetry redu tions of a generalized, ylindri al nonlinear
S hrodinger equation, J. Phys. A: Math. Gen. 26 (1993) 133-150.
and
, New symmetry redu tions and exa t solutions of the Davey
Stewartson system. II. Redu tions to partial dierential equations, Preprint No. 170 Program
in Applied Mathemati s (University of Colorado at Boulder, Boulder, Colorado, USA, 1993).
and
, New symmetry redu tions and exa t solutions of the Davey
Stewartson system. I. Redu tions to ordinary dierential equations, J. Math. Phys. 35 (1994)
255-283.
and
, New similarity solutions of the Boussinesq equation,
J. Math. Phys. 30 (1989) 2201-2213.
and
, Non lassi al symmetry redu tions and exa t solutions of nonlinear rea tion-diusion equations, in: Appli ations of Analyti and Geometri
Methods to Nonlinear Dierential Equations. NATO ASI Series C: Math. and Phys. S s., Ed.:
P.A. Clarkson (Kluwer A ademi Publishers, Dordre ht, The Netherlands, 1993) 375-389.
and
, Symmetries of the nonlinear heat equation, in:
Modern Group Analysis: Advan ed Analyti al and Computational Methods in Mathemati al
Physi s. Pro . Int. Workshop A ireale, Catania, Italy, 1992, Eds.: N.H. Ibragimov, M. Torrisi
and A. Valenti (Kluwer A ademi Publishers, Dordre ht, The Netherlands, 1993) 155-171.
and
, Symmetry redu tions and exa t solutions of a lass
of nonlinear heat equations, Physi a D 70 (1993) 250-288.
and
, On a shallow water wave equation, Nonlinearity 7
(1994) 975-1000.
and
, Redu tions and exa t solutions of a shallow water
wave equation, in: Pro . 14th IMACS World Congress on Computational and Applied Mathemati s, Atlanta, Georgia, 1994, Ed.: W.F. Ames, vol. 1 (IMACS, New Brunswi k, New Jersey,
1994) 97-101.
G. Ci ogna
G. Gaeta
P.A. Clarkson
P.A. Clarkson
P.A. Clarkson
P.A. Clarkson
P.A. Clarkson
S. Hood
P.A. Clarkson
S. Hood
P.A. Clarkson
S. Hood
P.A. Clarkson
S. Hood
P.A. Clarkson
M.D. Kruskal
P.A. Clarkson
E.L. Mansfield
P.A. Clarkson
E.L. Mansfield
P.A. Clarkson
E.L. Mansfield
P.A. Clarkson
E.L. Mansfield
P.A. Clarkson
E.L. Mansfield
[71
[72
[73
[74
[75
[76
[77
[78
[79
[80
[81
[82
[83
[84
[85
and
, Algorithms for the non lassi al method of symmetry
redu tions, SIAM J. Appl. Math (1994) in press.
and
, Non lassi al symmetry redu tions for the KadomtsevPetviashvili equation, Physi a D 49 (1991) 257-272.
, An Introdu tion to the Lie Theory of One-Parameter Groups (Ste hert and Co.,
New York, 1931).
and
, Ideals, Varieties and Algorithms. An Introdu tion
to Computational Algebrai Geometry and Commutative Algebra. Undergraduate Texts in
Mathemati s (Springer Verlag, Berlin, 1992).
and
, Eds. Group Theoreti al
Methods in Physi s. Pro . XIX International Colloquium, Salaman a, Spain, 1992. Anales
de Fsi a. Monografas 1 (EDITORIAL CIEMAT, Real So iedad Expa~nola de Fsi a, Spain,
1993).
and
, Eds. Group Theoreti al Methods in Physi s. Pro . XVIII
International Colloquium Group Theoreti al Methods in Physi s, Mos ow, U.S.S.R., 1990.
Le ture Notes in Physi s 382 (Springer Verlag, Berlin, 1991).
, The Dierential Forms Pa kage, available in Maple 4.2. (The Symboli Computation Group, University of Waterloo, Ontario, Canada, 1990).
, Similarity Solutions of Nonlinear Partial Dierential Equations. Resear h
Notes in Mathemati s 88 (Pitman, Boston, Massa husetts, 1983).
, Programs for omputer implementation of the exterior al ulus with appli ations to isove tor al ulations, Comp. & Maths. with Appls. 6 (1980) 415-424.
, Isove tor Methods for Equations of Balan e (Sijtho & Noordho, Alphen
aan de Rijn, The Netherlands, 1980).
, Applied Exterior Cal ulus (Wiley-Inters ien e, New York, 1985).
, Exterior al ulus for ve tors and forms, The Mathemati a Journal 1, no.
3 (1991) 92-95.
and
, Transformation Methods for Nonlinear Partial Differential Equations (World S ienti , Singapore, 1992).
, Continuous Groups of Transformations (Prin eton University Press,
Prin eton, New Jersey, 1933).
and
, A REDUCE program for determining
point and onta t Lie symmetries of dierential equations, Comp. Phys. Comm. 36 (1985)
383-389.
P.A. Clarkson
E.L. Mansfield
P.A. Clarkson
P. Winternitz
A. Cohen
D. Cox, J. Little
D. O'Shea
M.A del Olmo, M. Santander
V.V. Dodonov
J. Mateos Guilarte
V.I. Man'ko
A. Donsig
L. Dresner
D.G.B. Edelen
D.G.B. Edelen
D.G.B. Edelen
D.G.B. Edelen
D.G.B. Edelen
L.P.
Jian-hua Wang
Eisenhart
V.P. Eliseev, R.N. Fedorova
V.V. Kornyak
[86
[87
[88
[89
[90
[91
[92
[93
[94
[95
[96
[97
[98
[99
, Dierential geometry te hniques for sets of nonlinear partial dierential

equations, in: Partially Integrable Nonlinear Evolution Equations and their Physi al Appli ations. NATO ASI Series C: Math. and Phys. S s. 310, Eds.: R. Conte and N. Bo ara
, The dire t method and the singular manifold method for the Fitzhugh-Naguma
equation, Phys. Lett. A 171 (1992) 259-261.
, Non- lassi al symmetries and the singular manifold method: the Burgers
equation and the Burgers-Huxley equation, J. Phys. A.: Math. Gen. 27 (1994) 2113-2127.
and
, Non Classi al Symmetries and the Singular Manifold
Method: Theory and Six Examples, Preprint Departmento di Fisi a Teori a (Universidad de
Salaman a, Salaman a, Spain, 1993).
and
, Continuous Symmetries, Lie Algebras and Dierential Equations (Bibliographis hes Institut Wissens haftsverlag, Mannheim, 1993).
and
, Applying omputer algebra systems for determination
of the Lie and Lie-Ba klund symmetries of dierential equations, in: Pro . Int. Conf. on
Computer Algebra and its Appli ations in Theoreti al Physi s, Dubna, U.S.S.R., 1985 (Joint
Institute for Nu lear Resear h, Dubna, U.S.S.R., 1985) 248-261, in Russian.
and
, Determination of Lie-Ba klund symmetries of dierential equations using FORMAC, Comp. Phys. Comm. 39 (1986) 93-103.
and
, A REDUCE Program for Computing Determining
Equations of Lie-Ba klund Symmetries of Dierential Equations. Report R11-87-19 (Joint
Institute for Nu lear Resear h, Dubna, U.S.S.R., 1987) in Russian.
and
, Grobner Bases, Ritt's algorithms and de ision pro edures for
algebrai theories, in: Applied Algebra, Algebrai Algorithms, and Error-Corre ting Codes.
Pro . 5th Int. Conf. AAECC-5, Eds.: L. Huguet and A. Poli. Le ture Notes in Computer
S ien e 356 (Springer Verlag, Berlin, 1989) 230-237.
, Cartan's Exterior Forms Method (Goste hizdat, Mos ow-Leningrad, U.S.S.R.,
1948) p. 432, in Russian.
and
, Prolongations to higher jets of Estabrook-Wahlquist
overings for PDEs, A ta Appl. Math. 32 (1993) 197-225.
, Symmetries and integrability, Stud. in Appl. Math. 44 (1987) 253-299.
, Symmetry groups and similarity solutions of MHD equations, J. Math. Phys.
32 (1991) 1703-1708.
and
, Computer-algebra methods for investigation of hereditary operators of higher order soliton equations, Comp. Phys. Comm. 44
(1987) 47-55.
F. Estabrook
P.G. Est
evez
P.G. Est
evez
P.G. Est
evez
N. Euler
P.R. Gordoa
W.-H. Steeb
R.N. Fedorova
V.V. Kornyak
R.N. Fedorova
V.V. Kornyak
R.N. Fedorova
A. Ferro
V.V. Kornyak
G. Gallo
S.P. Finikov
J.D. Finley III
J.K. M Iver
A.S. Fokas
J.C. Fu hs
B. Fu hssteiner, W. Oevel
W. Wiwianka
[100
, Conditional symmetries of the equations of mathemati al physi s, in: Modern Group Analysis: Advan ed Analyti al and Computational Methods in Mathemati al
[101
, Symmetry in the problems of mathemati al physi s, in: Algebrai Theoreti al Studies in Mathemati al Physi s (Kiev Inst. of Math., Kiev, U.S.S.R., 1981)
6-28, in Russian.
[102
, Ed., Symmetry Analysis of Equations of Mathemati al Physi s (Inst.
Math. A ad. Ukraine, Kiev, Ukraine, 1992).
[103
and
, Subgroup Analysis of the Galilei
and Poin are Groups and Redu tion of Nonlinear Equations (Naukova Dumka, Kiev, Ukraine,
1991) in Russian.
[104
and
, Computer algebra appli ation for determining Lie
and Lie-Ba klund symmetries of dierential equations, J. Symb. Comp. 7 (1989) 611-619.
[105
and
, Symmetries of Maxwell's Equations (D. Reidel, Dordre ht, The Netherlands, 1987).
[106
and
, Symmetry in Equations of Quantum Me hani s
(Nauka, Mos ow, U.S.S.R., 1990) in Russian.
[107
and
, Symmetry Analysis and Exa t Solutions of Nonlinear Equations of Mathemati al Physi s (Naukova Dumka, Kiev, U.S.S.R, 1989)
in Russian.
[108
and
, Symmetry Analysis and Exa t Solutions of Equations of Nonlinear Mathemati al Physi s, Mathemati s and Its Appli ations 246
(Kluwer, Dordre ht, The Netherlands, 1993).
[109
and
, Symmetry and exa t solutions of nonlinear spinor
equations, Phys. Rep. 172 (1989) 123-174.
[110
and
, Nonlinear Spinor Equations: Symmetry and Exa t
Solutions, (Naukova Dumka, Kiev, Ukraine, 1992) in Russian.
[111
, Redu tion and equivariant bran hing lemma: dynami al systems, evolution
PDEs, and Gauge theories, A ta Appl. Math. 28 (1992) 43-68.
[112
, Geometri al Symmetries of Nonlinear Equations and Physi s (1994) in preparation.
[113
and
,
Daukl. Akad. Nauk S.S.S.R. 261 (1981) p. 1044.
[114
and
, Compatibility Analysis for Systems of Dierential
Equations on a Computer. ITPM Preprint # 20-82 (ITPM, Novosibirsk, U.S.S.R., 1982) in
Russian.
W. Fush hi h
W.I.
Fush hi h
W.I. Fush hi h
W.I. Fush hi h, L.F. Barannik
W.I. Fush hi h
V.V. Kornyak
W.I. Fush hi h
A.G. Nikitin
W.I.
Fush hi h
A.G.
A.F. Barannik
Nikitin
W.I. Fush hi h, V.M. Shtelen
N.I. Serov
W.I. Fush hi h, V.M. Shtelen
N.I. Serov
W.I. Fush hi h
R.Z. Zhdanov
W.I. Fush hi h
R.Z. Zhdanov
G. Gaeta
G. Gaeta
V.G. Ganzha, S.V. Meleshko, F.A. Murzin, V.P. Shapeev
V.G. Ganzha
N.N. Yanenko
N.N. Janenko
[115
, Computer algebra tools for higher symmetry analysis of nonlinear evolution equations, in: Programming Environments for High-Level Problem Solving, Eds.: P.W.
Ganey and E.N. Houstis (Elsevier S ien e Publishers, Amsterdam, The Netherlands, 1993)
107-115.
[116
, Computer algebra, symmetry analysis and integrability of nonlinear evolution
equations, Int. J. Mod. Phys. C 4 (1993) 279-286.
[117
and
, FORMINT - A Program for Classi ation of Integrable Nonlinear Evolution Equations, Comp. Phys. Comm. 34 (1985) 303-311.
[118
and
, Computer Algebra Appli ation for
Classi ation of Integrable Non-linear Evolution Equations, J. Symb. Comp. 1 (1985) 101-107.
[119
and
, Computer generation of ne essary integrability onditions for polynomial-nonlinear evolution systems, in: Pro . ISSAC '90, Tokyo, Japan, 1990,
Eds.: S. Watanabe and M. Nagata (ACM Press, New York, 1990) 250-254.
[120
, Symboli Computations in Prolongation Theory. Ph.D. Thesis, Department of Mathemati s (Twente University of Te hnology, Ens hede, The Netherlands, 1981).
[121
, Lie algebra omputations, A ta Math. Appl. 16 (1989) 231-242.
[122
, Prolongation algebras of nonlinear PDEs, in: Pro . ERCIM Advan ed
Course on Partial Dierential Equations and Group Theory, Part II, Bonn, 1992, Ed.: J.F.
Pommaret (Gesells haft fur Mathematik und Datenverarbeitung, Sankt Augustin, Germany,
1992) 1-20.
[123
and
, Implementation of dierential geometry obje ts
and fun tions with an appli ation to extended Maxwell equations, in: Pro . EUROCAM '82,
Marseille, Fran e, 1982, Ed.: J. Calmet. Le ture Notes in Computer S ien e 144 (Springer
Verlag, New York, 1982) 181-187.
[124
and
, Graded dierential geometry in REDUCE: Supersymmetri stru tures of the modied KdV equation, A ta Appl. Math. 24 (1991) 211-231.
[125
and
, Symboli omputations in applied
dierential geometry, A ta Appl. Math. 1 (1983) 43-77.
[126
and
, Lie subgroups of the symmetry group of the equations
des ribing a nonstationary and isotropi ow I. Subgroupes de Lie des groupes de symmetry
pour les equations de la dynamique de uid I, Can. J. Phys. (1994) in press.
[127
and
, Lie subgroups of symmetry groups of the uid dynami s and the magnetohydrodynami s equations II, Can. J. Phys. (1994) in press.
[128
and
, Hidden symmetries of energy- onserving dierential
equations, IMA J. Appl. Math. 51 (1993) 147-153.
V.P. Gerdt
V.P. Gerdt
V.P. Gerdt, A.B. Shva hka
A.Yu. Zharkov
V.P. Gerdt, A.B. Shva hka
V.P. Gerdt
A.Yu. Zharkov
A.Yu. Zharkov
P.K.H. Gragert
P.K.H. Gragert
P.K.H. Gragert
P.K.H. Gragert
P.H.M. Kersten
P.K.H. Gragert
P.H.M. Kersten
P.K.H. Gragert, P.H.M. Kersten
A.M. Grundland
A.M. Grundland
A. Guo
L. Lalague
L. Lalague
B. Abraham-Shrauner
A. Martini
[129
and
, Symmetries and Conservation Laws of Kadomtsev-Pogutse Equations, A ta
Appl. Math. 15 (1989) 23-64.
[130
, Similarity Analyses of Boundary Value Problems in Engineering (Prenti e
Hall, Englewood Clis, 1964).
[131
and
, Geometri approa h to invarian e groups and solution of partial dierential equations, J. Math. Phys. 12 (1971) 653-666.
[132
and
, A onstru tive implementation of the Cartan-Kahler
theory of exterior dierential systems, J. Symb. Comp. 12 (1991) 655-667.
[133
and
, Constrained dynami s and exterior differential systems, J. Phys. A 24 (1991) 5253-5265.
[134
, The birth of Lie's theory of groups, The Mathemati al Intelligen er 16, no. 2
(1994) 6-17.
[135
, LIE: A PC Program for Lie Analysis of Dierential Equations, Comp. Phys.
Comm. 77 (1993) 241-248.
[136
Grobner Basis Pa kage, Wolfram Resear h In ., Champaign, Illinois, MathSour e 0206-794, 1994.
[137
, SYMMGRP.MAX: A symboli program for symmetry analysis of partial
dierential equations, in: Exploiting Symmetry in Applied and Numeri al Analysis, Pro .
AMS-SIAM Summer Seminar, Fort Collins, Colorado, 1992, Eds.: E. Allgower, K. Georg and
R. Miranda. Le tures in Applied Mathemati s 29 (Ameri an Mathemati al So iety, Providen e, Rhode Island, 1993) 241-257.
[138
, Review of symboli software for the omputation of Lie symmetries of differential equations, Euromath Bulletin 1, no. 2 (1994) 45-82.
[139
and
, Lie point symmetries of lassi al
eld theories, in: Group Theoreti al Methods in Physi s, Pro . XIX International Colloquium,
Salaman a, Spain, 1992, Eds.: M.A del Olmo, M. Santander, and J. Mateos Guilarte. Anales
de Fsi a. Monografas 1 (CIEMAT, Real So iedad Expa~nola de Fsi a, Spain, 1993) 402-405.
[140
, Dierential Geometry and the Cal ulus of Variations (A ademi Press, New
York, 1968).
[141
, Computer Assisted Determination of Lie Point Symmetries with Appli ation to
Fluid Dynami s. Ph.D Thesis, Program in Applied Mathemati s (The University of Colorado,
Boulder, Colorado, 1994).
[142
, The Use of Maple in the Sear h for Symmetries, Resear h Report no. 77,
Department of Mathemati s (University of Canterbury, Christ hur h, New Zealand, 1993).
V.N. Gusyatnikova, A.V. Samokhin, V.S. Titov, A.M. Vinogradov
Yumaguzhin
A.G. Hansen
B.K. Harrison
D. Hartley
F.B. Estabrook
R.W. Tu ker
D. Hartley, R.W. Tu ker
P.A. Tu key
T. Hawkins
A.K. Head
G. Helzer
W. Hereman
W. Hereman
W. Hereman, L. Mar hildon
R. Hermann
S. Herod
M. Hi kman
M. Grundland
V.A.
[143
, The use of Maple in the sear h for symmetries, in: Pro . 14th IMACS World
Congress on Computational and Applied Mathemati s, Atlanta, Georgia, 1994, Ed.: W.F.
Ames, vol. 1 (IMACS, New Brunswi k, New Jersey, 1994) 226-229.
[144
, Solution of Dierential Equations by Means of One-Parameter Groups, Resear h
Notes in Mathemati s 63 (Pitman, Boston, Massa husetts, 1982).
[145
, Dierential Equations and Group Methods for S ientists and Engineers (CRC
Press, Bo a Raton, Florida, 1992).
[146
, Non lassi al Symmetry Redu tions and Exa t Solutions of Nonlinear Partial Differential Equations. Ph.D Thesis, Department of Mathemati s (University of Exeter, Exeter,
United Kingdom, 1993).
[147
, Transformation Groups Applied to Mathemati al Physi s (D. Reidel,
Dordre ht, The Netherlands, 1985).
[148
, Primer on Group Analysis (Znanie, Mos ow, U.S.S.R., 1989) in Russian.
[149
, Essay on Group Analysis of Ordinary Dierential Equations (Znanie,
Mos ow, U.S.S.R., 1991) in Russian.
[150
, Group analysis of ordinary dierential equations and the invarian e prin iple in mathemati al physi s (for the 150th anniversary of Sophus Lie, Russian Math. Surveys
47 (1992) 89-156.
[151
, Linearization of ordinary dierential equations, in: Dierential Equations
with Appli ations to Mathemati al Physi s. Eds.: W.F. Ames, E.M. Harrell II and J.V. Herod.
Mathemati s in S ien e and Engineering 192 (A ademi Press, New York, 1993) 177-188.
[152
, Sophus Lie and harmony in mathemati al physi s, on the 150th anniversary of his birth, The Mathemati al Intelligen er 16, no. 1 (1994) 20-28.
[153
, Ed., CRC Handbook of Lie Group Analysis of Dierential Equations.
Volume 1: Symmetries, Exa t Solutions, and Conservation Laws (CRC Press, Bo a Raton,
Florida, 1994).
[154
, Ed., CRC Handbook of Lie Group Analysis of Dierential Equations.
Volume 2: Appli ations in Engineering and Physi al S ien es (CRC Press, Bo a Raton,
Florida, 1994) in press.
[155
and
, Eds., Modern Group Analysis: Theory and Appli ation in Mathemati al Modelling. Pro eedings 5th Int. Conf. on Group Analysis and Dierential Equations, University of the Witwatersrand, Johannesburg, South Afri a, January 16-22,
1994. Will appear in the new international journal on `Lie Groups and their Appli ations' 1
(1994) no. 1.
[156
, A REDUCE program for nding symmetries of nonlinear evolution equations with
uniform rank, Comp. Phys. Comm. 42 (1986) 351-357.
M. Hi kman
J.M. Hill
J.M. Hill
S. Hood
N.H. Ibragimov
N.H. Ibragimov
N.H. Ibragimov
N.H. Ibragimov
N.H. Ibragimov
N.H. Ibragimov
N.H. Ibragimov
N.H. Ibragimov
N.H. Ibragimov
M. Ito
F.M. Mahomed
[157
SYMCD{a REDUCE program for nding symmetries and onserved densities of

systems of nonlinear evolution equations, Comp. Phys. Comm. 79 (1994) 547-554.
M. Ito
[158
and
A REDUCE program for nding onserved densities of partial differential equations with uniform rank, Comp. Phys. Comm. 38 (1985) 415-419.
M. Ito
F. Kako
[159
, Les systemes d'equations aus derivees partielles, J. Math. Pure et Appl. 3 (1920)
65-151.
[160
, Le ons sur les Systemes d'Equations aux Derivees (Gauthier-Villars, Paris,
1929); Equations aux Derivees Partielles (CNRS, Paris, 1956).
[161
, Einfuhrung in die Theorie der Systeme von Dierentialglei hungen (B.G. Teubner, Leipzig, 1934).
[162
, Separation of Variables for Riemannian Spa es of Constant Curvature
(Longman, Essex, 1986).
[163
, On the equations des ribing modulational instabilities due to nonlinear
intera tion between high and low frequen y waves, Phys. Lett. A 136 (1989) 216-220.
[164
, Innitesimal Symmetries: A Computational Approa h. Ph.D. Thesis.
Department of Mathemati s. (Twente University of Te hnology, Ens hede, The Netherlands,
1985); CWI Tra t 34 (Center for Mathemati s and Computer S ien e, Amsterdam, 1987).
[165
, Software to Compute Innitesimal Symmetries of Exterior Dierential
Systems, with Appli ations, A ta Appl. Math. 16 (1989) 207-229.
[166
and
, Eds. Algebrai and Geometri Stru tures
in Dierential Equations, Pro . Meeting, Twente University of Te hnology, Ens hede, The
Netherlands, 1993, to appear as spe ial issue of A ta Appl. Math. (1994).
[167
, Dierential Algebra and Algebrai Groups (A ademi Press, New York,
1973).
[168
, Dierential algebrai groups (A ademi Press, New York, 1984).
[169
and
, Symmetries and Integrals. Preprint 15 (International
Sophus Lie Center & Matematisk Institutt, Universitetet i Oslo, Norway, 1993).
[170
and
, A program for symmetry analysis of dierential
equations, in: Pro . ISSAC '91, Bonn, Germany, 1991, Ed.: S.M. Watt (ACM Press, New
York, 1991) 315-316.
[171
, Graded manifolds, graded Lie theory and prequantization, in: Pro . Symposium on Dierential Geometri al Methods in Mathemati al Physi s, Bonn, Germany, 1975,
Eds.: K. Bleuler and A. Reetz. Le ture Notes in Mathemati s 570 (Springer Verlag, New
York, 1977) 177-306.
[172
and
, Geometry of Jet Spa es
and Nonlinear Partial Dierential Equations. Adv. Stud. Contemp. Math. 1 (Gordon &
Brea h, New York, 1986).
M. Janet
M. Janet
hler
E. Ka
E.G. Kalnins
V.I. Karpman
P.H.M. Kersten
P.H.M. Kersten
P.H.M. Kersten
I.S. Krasil'sh hik
E.R. Kol hin
E.R. Kol hin
B. Komrakov
V.V. Kornyak
V. Ly hagin
W.I. Fush hi h
B. Kostant
I.S. Krasil'sh hik, V.V. Ly hagin
A.M. Vinogradov
[173
and
Nonlo al trends in the geometry of DEs:

symmetries, onservation laws and Ba klund transformations, A ta Appl. Math. 15 (1989)
I.S. Krasil'sh hik
A.M. Vinogradov
161-209.
[174
, Le tures on Involutive Systems of Partial Dierential Equations (Publ. So .
Math., Sao Paulo, Brasil, 1967).
[175
and
, Eds., Symmetries and Nonlinear Phenomena (World S ienti , Singapore, 1988).
[176
and
, Non- lassi al symmetry redu tions: example of the Boussinesq equation, J. Phys. A: Math. Gen. 22 (1989) 2915-2924.
[177
and
, Symmetries and onditional symmetries of dierentialdieren e equations, J. Math. Phys. 34 (1993) 3713-3730.
[178
and
Group Methods in Integrating Nonlinear Dynami al
Systems (Nauka, Mos ow, U.S.S.R., 1985) in Russian.
[179
, Sophus Lie's 1880 Transformation Group Paper. Translated by M. A kerman, omments by R. Hermann, Lie Groups: History, Frontiers on Appli ations 1 (Mathemati al
S ien es Press, Brookline, Massa husetts, 1975).
[180
, Sophus Lie's 1884 Dierential Invariants Paper. Translated by M. A kerman, omments by R. Hermann, Lie Groups: History, Frontiers on Appli ations 3 (Mathemati al
S ien es Press, Brookline, Massa husetts, 1976).
[181
, Theorie der Transformationsgruppen. Volumes I, II & III (B.G. Teubner, Leipzig,
1888, 1890, 1893).
[182
, Vorlesungen uber Dierentialglei hungen mit Bekannten Innitesimalen Transformationen (B.G. Teubner, Leipzig, 1891, 1912; Chelsea Publishing Company, New York, 1967).
[183
, Gesammelte Abhandlungen. Vol. 1-7 (B.G. Teubner, Leipzig, 1922-1929).
[184
, Vorlesungen uber Continuierli he Gruppen, Abteilung III (Chelsea Publishing Company, New York, 1971).
[185
and
, Theorie der Transformationsgruppen II (B.G. Teubner, Leipzig,
1890).
[186
, Equivalen e Transformations for Classes of Dierential Equations. Ph.D. Thesis. Department of Mathemati s (University of British Columbia, Van ouver, Canada, 1992).
[187
, An ordinary dierential equation solver for REDUCE, in: Pro . ISSAC '88, Rome, Italy, 1988, Ed.: P. Gianni. Le ture Notes in Computer S ien e 358 (Springer
Verlag, Berlin, 1988) 196-205.
[188
, Dierential Grobner Bases, Ph.D Thesis, Department of Computer S ien e (University of Sydney, Sydney, Australia, 1991).
M. Kuranishi
D. Levi
P. Winternitz
D. Levi
P. Winternitz
D. Levi
P. Winternitz
A.N. Leznov
M.V. Savel
ev,
S. Lie
S. Lie
S. Lie
S. Lie
S. Lie
S. Lie
S. Lie
F. Engel
I.G. Lisle
M.A.H. Ma Callum
E.L. Mansfield
[189
, DIFFGROB2: A Symboli Algebra Pa kage for Analyzing Systems of
PDE using Maple, User's Manual for Release 2, Preprint, Department of Mathemati s (University of Exeter, Exeter, United Kingdom, 1993).
[190
and
, Appli ations of the dierential algebra pa kage
Digrob2 to redu tions of PDE, in: Pro . 14th IMACS World Congress on Computational
and Applied Mathemati s, Atlanta, Georgia, 1994, Ed.: W.F. Ames, vol. 1 (IMACS, New
Brunswi k, New Jersey, 1994) 336-339.
[191
and
, Appli ations of the Dierential Algebra Pa kage
Digrob2 to Classi al Symmetries of Dierential Equations, Preprint M94/18, Department of
Mathemati s (University of Exeter, Exeter, UK, 1994).
[192
and
, Dierential Grobner Bases, Preprint 92-10, S hool
of Mathemati s, Physi s, Computer S ien e, and Ele troni s (Ma quarie University, Sydney,
Australia, 1992).
[193
, Ed., Geometri Approa hes to Dierential Equations, Le ture Notes in Mathemati s 810 (Springer Verlag, New York, 1979).
[194
and
, GROEBNER: A Grobner basis pa kage, in:
REDUCE User's Manual, Version 3.4, Ed.: A.C. Hearn (Rand Corporation, Santa Moni a,
California, 1991).
[195
and
, The symmetry approa h to lassi ation of integrable equations, in: What is Integrability?, Springer Le ture Notes in Nonlinear
Dynami s, Ed.: V.I. Zakharov (Springer Verlag, New York, 1990) 115-184.
[196
, Lie Theory and Spe ial Fun tions (A ademi Press, New York, 1968).
[197
, Symmetry Groups and Their Appli ations, Pure and Applied Mathemati s
50 (A ademi Press, New York, 1972).
[198
, Symmetry and Separation of Variables (Addison-Wesley, Reading, Massa husetts, 1977).
[199
, Ed., Ba klund Transformations, the Inverse S attering Method, Solitons, and
Their Appli ations. Le ture Notes in Mathemati s 515 (Springer Verlag, New York, 1976).
[200
, Invariante Variationsprobleme, Na hr. Konig. Gesell. Wissen. Gottingen,
Math.-Phys. Kl. (1918) 235-257; English translation: Transport Theory and Stat. Phys. 1
(1971) 186-207.
[201
, Group analysis of MHD equations, Atti. Sem. Mat. Fis. Univ. Modena 33
(1984) 21-34.
[202
, Intera tive REDUCE programs for al ulating, lassi al, non- lassi al, and
Lie-Ba klund symmetries of dierential equations, Manual of the Program, Preprint GT
Math:062090-051, S hool of Mathemati s (Georgia Institute of Te hnology, Atlanta, Georgia, 1990).
E.L. Mansfield
E.L. Mansfield
E.L. Mansfield
E.L. Mansfield
P.A. Clarkson
P.A. Clarkson
E.D. Fa kerell
R. Martini
ller
H. Melenk, H.M. Mo
A.V. Mikhailov, A.B. Shabat
W. Miller, Jr.
W. Miller, Jr.
W.
Miller
R.M. Miura
E. Noether
M.C. Nu i
M.C. Nu i
W. Neun
V.V. Sokolov
[203
approximate symmetries of dierential equations, in: Pro . 13th IMACS World Congress,
Dublin, Ireland, 1991. Eds.: R. Vi hnevetsky and J.J.H. Miller (Criterion Press, Dublin,
Ireland, 1991) 349-350.
[204
Lie-Ba klund symmetries of dierential equations, in: Computational and Applied Mathemati s, II: Dierential Equations. Sele ted and revised papers from the IMACS 13th World
Congress, Dublin, Ireland, 1991. Eds: W.F. Ames and P.J. Van der Houwen (Elsevier S ien e
Publishers, Amsterdam, The Netherlands, 1992) 345-350.
[205
, Symmetries and symboli omputation, in: Dierential Equations with Appli ations to Mathemati al Physi s. Mathemati s in S ien e and Engineering 192. Eds.: W.F.
Ames, E.M. Harrell II and J.V. Herod (A ademi Press, New York, 1993) 249-260.
[206
, Non lassi al symmetries and Ba klund transformations, J. Math. Anal. Appl.
178 (1993) 294-300.
[207
, Iterating the Non lassi al Symmetry Method, Te h. Rep. 1993-14, Departimento de Matemati a (Universita di Perugia, Perugia, Italy, 1993).
[208
, Intera tive REDUCE programs for al ulating Lie-point, non- lassi al, LieBa klund, and approximate symmetries of dierential equations: Manual and oppy disk, in:
CRC Handbook of Lie Group Analysis of Dierential Equations. Volume 3: New Trends (CRC
Press, Bo a Raton, Florida, 1995), Chapter 14.
[209
and
, Classi al and non lassi al symmetries for the Helmholtz
equation, J. Math. Anal. Appl. 178 (1993) 584-591.
[210
and
, The non lassi al method is more general than the dire t
method for symmetry redu tions. An example of the Fitzhugh-Nagumo equation, Phys. Lett.
A 164 (1992) 49-56.
[211
, Algorithms for nding the stru ture of solutions of a system of linear partial
dierential equations, in: Pro . ISSAC '94, Oxford, UK, 1994, Eds.: J. von zur Gathen and
M. Giesbre ht (ACM Press, New York, 1994) 216-223.
[212
, Standard bases of dierential ideals, in: Applied Algebra, Algebrai Algorithms, and Error-Corre ting Codes. Pro . 8th Int. Conf. AAECC-8, Eds.: L. Huguet and A.
Poli. Le ture Notes in Computer S ien e 508 (Springer Verlag, Berlin, 1991) 304-321.
[213
, Symmetry and expli it solutions of partial dierential equations, Appl. Num.
Math. 10 (1992) 307-324.
[214
, Appli ations of Lie Groups to Dierential Equations. Graduate Texts in Mathemati s 107 (Springer Verlag, New York, 1986 & 1993{2nd edition).
[215
, Dire t redu tion and dierential onstraints, Pro . R. So . London A. 444
(1994) 509-523.
M.C. Nu i
M.C. Nu i
M.C. Nu i
M.C. Nu i
M.C. Nu i
M.C. Nu i
M.C. Nu i
M.C. Nu i
T. Oaku
F. Ollivier
P.J. Olver
P.J. Olver
P.J. Olver
W.F. Ames
P.A. Clarkson
[216
, Equivalen e, Invariants and Symmetry. (Cambridge University Press, Cambridge, UK, 1994)
[217
and
, The onstru tion of spe ial solutions to partial dierential
equations, Phys. Lett. A 114 (1986) 107-112.
[218
and
, Group-invariant solutions of dierential equations, SIAM J.
Appl. Math. 47 (1987) 263-278.
[219
, Partially Invariant Solutions of Dierential Equations. Ph.D Thesis, Department of Mathemati s (University of Minnesota, Minneapolis, Minnesota, 1989).
[220
, Group Properties of Dierential Equations (A ademy of S ien es,
Siberian Bran h, U.S.S.R., Novosibirsk, 1962); English translation by G.W. Bluman (1967)
unpublished.
[221
, Group Analysis of Dierential Equations (Nauka, Mos ow, U.S.S.R.,
1972); English translation: W.F. Ames Ed. (A ademi Press, New York, 1982).
[222
, A Global Formulation of the Lie Theory of Transformation Groups, Memoirs of
the Ameri an Mathemati al So iety 22 (Ameri an Mathemati al So iety, Providen e, Rhode
Island, 1957 & 1989).
[223
, Computations in dierential and dieren e modules, A ta Appl. Math.
16 (1989) 167-189.
[224
and
, Lo al Jet Bundle Formulation
of Ba klund Transformations. Mathemati al Physi s Studies 1 (D. Reidel, Dordre ht, The
Netherlands, 1979).
[225
and
, Automati determination of dynami al symmetries of ordinary
dierential equations, Computing 41 (1989) 297-307.
[226
, Sophus Lie (Nauka, Leningrad, 1983).
[227
, Systems of Partial Dierential Equations and Lie Pseudogroups (Gordon
& Brea h, New York, 1978); translated by A.M. Vinogradov for MIR (1983).
[228
, Dierential Galois Theory (Gordon & Brea h, New York, 1983).
[229
, Lie Pseudogroups and Me hani s (Gordon & Brea h, New York, 1988).
[230
, Le ture Notes of `Advan ed Course on Partial Dierential Equations and
Group Theory'. New Perspe tives in Computer Algebra and Appli ations to Engineering S ien es. Pro . ERCIM Advan ed Course, Gesells haft fur Mathematik und Datenverarbeitung,
Bonn, Germany, April 6-10, 1992.
[231
, Automized omputing of dening equations for a Lie group, Izvestiya AN
Belor. SSR, Ser. Phys.-Math. 2 (1985) 33-37, in Russian
[232
, Similarity redu tions of partial dierential equations, J. Phys. A: Math. Gen. 25
(1992) 2631-2640.
P.J. Olver
P.J. Olver
P.J. Olver
P. Rosenau
P. Rosenau
J.R. Ondi h
L.V. Ovsiannikov
L.V. Ovsiannikov
R.S. Palais
E.V. Pankrat'ev
F.A.E. Pirani, D.C. Robinson
H.J. Pohle
T. Wolf
E.M. Polish huk

J.F. Pommaret
J.F. Pommaret
J.F. Pommaret
J.F. Pommaret
M.D. Popov
E. Pu i
W.F. Shadwi k
[233
, A triangularization algorithm whi h determines the Lie symmetry algebra of any

system of PDEs, J. Phys. A: Math. Gen. 23 (1990) L853-L859.
[234
, Algorithmi determination of Lie symmetry algebras of dierential equations, in:
Lie Theory, Dierential Equations and Representation Theory, Pro . Annual Seminar of the
Canadian Math. So ., Montreal, Canada, 1989, Ed.: V. Hussin. Les Publi ations de Centre
de Re her hes Mathematiques (Montreal, Canada, 1990) 363-372.
[235
, Algorithms for redu ing a system of PDEs to standard form, determining the
dimension of its solution spa e and al ulating its Taylor series solution, Euro. J. Appl. Math.
2 (1991) 293-318.
[236
, Finding abstra t Lie symmetry algebras of dierential equations without integrating determining equations, Euro. J. Appl. Math. 2 (1991) 319-340.
[237
and
, Redu tion of systems of dierential equations to standard
form and their integration using dire ted graphs, in: Pro . ISSAC '91, Bonn, Germany, 1991,
Ed.: S.M. Watt (ACM Press, New York, 1991) 308-312.
[238
and
, Algorithmi determination
of ommutation relations for Lie symmetry algebras of PDEs, in: Pro . ISSAC '92, Berkeley,
California, 1992, Eds.: B. Trager and D. Lazard (ACM Press, New York, 1992) 63-68.
[239
and
, Solving systems of linear PDEs in their oe ient eld by
re ursively de oupling and solving ODEs. Preprint, Department of Mathemati s (University
of British Columbia, Van ouver, Canada, 1993).
[240
and
, A point symmetry group of a dierential
equation whi h annot be found using innitesimal methods, in: Modern Group Analysis:
Advan ed Analyti al and Computational Methods in Mathemati al Physi s. Pro . Int. Workshop A ireale, Catania, Italy, 1992, Eds.: N.H. Ibragimov, M. Torrisi, and A. Valenti (Kluwer
A ademi Publishers, Dordre ht, The Netherlands, 1993) 311-316.
[241
and
, Long Guide to the Standard Form Pa kage, Department
of Mathemati s (University of British Columbia, Van ouver, Canada, 1993).
[242
, Les Systemes d'Equations aux Derivees Partielles (Gauthier-Villars, Paris,
1910).
[243
, Dierential Algebra (Dover, New York, 1950, 1966).
[244
and
, Nonlinear Boundary Value Problems in S ien e and Engineering. Mathemati s in S ien e and Engineering 183 (A ademi Press, New York, 1989).
[245
and
, Ba klund Transformations and Their Appli ations (A ademi Press, New York, 1982).
[246
and
, Similarity Solutions of Systems of Partial Dierential Equations using MACSYMA. Report COO-3077-160 MF-94, Magneto-Fluid Dynami s
Division. Courant Institute of Mathemati al S ien es (New York University, New York, New
York, 1979).
G.J. Reid
G.J. Reid
G.J. Reid
G.J. Reid
G.J. Reid
A. Boulton
G.J. Reid, I.G. Lisle, A. Boulton
G.J. Reid
D.K. M Kinnon
G.J. Reid, D.T. Weih
G.J. Reid
A.D. Wittkopf
A.D. Wittkopf
C. Riquier
J.F. Ritt
C. Rogers
C. Rogers
P. Rosenau
A.D. Wittkopf
W.F. Ames
W.A. Shadwi k
J.L. S hwarzmeier
[247
and
Systems of Quasilinear Equations and Their
Appli ations to Gas Dynami s, 2nd Edition, Nauka, Mos ow, 1978; Translations of Mathemati al Monographs 55 (AMS, Providen e, Rhode Island, 1983).
[248
and
, Eds., Lie Methods in Opti s, Pro . CIFMOCIO Workshop, Leon, Guanajuato, Mexi o, 1985. Le ture Notes in Physi s 250 (Springer
Verlag, New York, 1986).
[249
and
, Generalizations of Noether's theorem in lassi al me hani s,
SIAM Review 23 (1981) 467-494.
[250
and
, REDUCE-pro edures for the study of adjoint symmetries of se ond-order dierential equations, J. Sym. Comp. 13 (1992) 683-693.
[251
, Group analysis of some partial dierential equations arising in appli ations,
Contemp. Math. 54 (1986) 35-44.
[252
, Group Theoreti Methods in Bifur ation Theory, Le ture Notes in Mathemati s 762 (Springer Verlag, New York, 1979).
[253
and
, Lie Groups and Algebras with Appli ations to Physi s,
Geometry, and Me hani s, Applied Mathemati al S ien es 61 (Springer Verlag, New York,
1986).
[254
, EXCALC, a system for doing al ulations in the al ulus of modern dierential geometry{User's Manual, distributed with REDUCE 3.3 (1987).
[255
, EXCALC - A pa kage for al ulations in modern dierential geometry, in:
Computer Algebra in Physi al Resear h. Pro . IV Int. Conf. Dubna, U.S.S.R., 1990, Eds.:
D.V. Shirkov, V.A. Rostovtsev and V.P. Gerdt (World S ienti , Singapore, 1990) 71-80.
[256
and
, Algorithmi methods for Lie pseudogroups, in:
Modern Group Analysis: Advan ed Analyti al and Computational Methods in Mathemati al
[257
, A REDUCE pa kage for determining Lie symmetries of ordinary and partial
dierential equations, Comp. Phys. Comm. 27 (1982) 179-186.
[258
, Automati ally determining symmetries of ordinary dierential equations, in:
Pro . EUROCAL '83, London, England, 1983, Ed.: J.A. van Hulzen. Le ture Notes in Computer S ien e 162 (Springer Verlag, New York, 1983) 45-54.
[259
, The Riquier-Janet theory and its appli ation to non-linear evolution equations,
Physi a D 11 (1984) 243-251.
[260
, Automati ally determining symmetries of partial dierential equations, Computing 34 (1985) 91-106; Addendum: Computing 36 (1986) 279-280.
B.L. Ro
zdestvensk
i
N.N. Janenko
n hez Mondrago
n
J. Sa
K.B. Wolf
W. Sarlet
F. Cantrijn
W. Sarlet
J. Vanden Bonne
C.C.A. Sastri
D.H. Sattinger
D.H. Sattinger
O.L. Weaver
fer
E. S hru
fer
E. S hru
, W.M. Seiler
J. S hu
F. S hwarz
F. S hwarz
F. S hwarz
F. S hwarz
J. Calmet
[261
, Symmetries and involution systems: some experiments in omputer algebra,

in: Topi s in Soliton Theory and Exa tly Solvable Nonlinear Equations. Pro . Meeting on Nonlinear Evolution Equations, Oberwolfa h, Germany, 1986, Eds.: M. Ablowitz, B. Fu hssteiner
and M. Kruskal (World S ienti , Singapore, 1987) 290-299.
[262
, The Pa kage SPDE for Determining Symmetries of Partial Dierential Equations. User's Manual. Distributed with REDUCE 3.3 (Rand Corporation, Santa Moni a,
California, 1987).
[263
, Programming with abstra t data types: The symmetry pa kage SPDE in
SCRATCHPAD, in: Trends in Computer Algebra, Pro . Int. Symposium, Bad Neuenahr,
Germany, 1987, Ed.: R. Janen. Le ture Notes in Computer S ien e 296 (Springer Verlag,
New York, 1988) 167-176.
[264
, Symmetries of dierential equations from Sophus Lie to omputer algebra,
SIAM Review 30 (1988) 450-481.
[265
, A fa torization algorithm for linear ordinary dierential equations, in: Pro .
ACM-SIGSAM 1989 Int. Symp. on Symboli and Algebrai Computation, ISSAC '89, Portland, Oregon, 1989, Eds.: G. Gonnet (ACM Press, New York, 1989) 17-25.
[266
, An Algorithm for Determining the Size of Symmetry Groups, Computing 49
(1992) 95-115.
[267
, Redu tion and ompletion algorithms for partial dierential equations, in:
Pro . ACM-SIGSAM 1992 Int. Symp. on Symboli and Algebrai Computation, ISSAC '92,
Berkeley, California, 1992, Eds.: B. Trager and D. Lazard (ACM Press, New York, 1992)
49-56.
[268
, SPDE: Symmetries of Partial Dierential Equations, Guidelines for the use
of SPDE 1.0, Institut SCAI (GMD, Postfa h 1316, D-53731 Sankt Augustin, Germany, 1994)
4 pages.
[269
, Computer Algebra Software for S ienti Appli ations, in: Pro . 1993 CISM
Advan ed S hool on Computerized Symboli Manipulation in Me hani s, Udine, Italy, 1993,
Ed.: E. Kreuzer (Springer Verlag, Berlin, 1994) to appear.
[270
and
, Dire tors, Partial Dierential Equations and Group
Theory, `Advan ed Course on Partial Dierential Equations and Group Theory'. New Perspe tives in Computer Algebra and Appli ations to Engineering S ien es. ERCIM Course,
Gesells haft fur Mathematik und Datenverarbeitung, Sankt-Augustin, Germany, O tober 2428, 1994.
[271
and
, Using MACSYMA to Cal ulate Similarity Transformations of Partial Dierential Equations. Report LA-UR 88-4157, Los Alamos National
Laboratory (Los Alamos, New Mexi o, 1988).
[272
, Similarity and Dimensional Methods in Me hani s (A ademi Press, New York,
1959).
F. S hwarz
F. S hwarz
F. S hwarz
F. S hwarz
F. S hwarz
F. S hwarz
F. S hwarz
F. S hwarz
F. S hwarz
F. S hwarz
J.F. Pommaret
J.L. S hwarzmeier
L.I. Sedov
P. Rosenau
[273
, On the arbitrariness of the general solution of an involutive partial dierential

equation, J. Math. Phys. 35 (1994) 486-498.
[274
, Analysis and Appli ation of the Formal Theory of Partial Dierential Equations Ph.D Thesis, Department of Physi s (University of Lan aster, Lan aster, UK, 1994).
[275
and
, Group Invarian e in Engineering Boundary Value Problems
(Springer Verlag, New York, 1985).
[276
, The Ba klund Problem, Symmetries and Conservation Laws for Some
Nonlinear Dierential Equations. Ph.D Thesis, Department of Mathemati s (King's College,
London, 1979).
[277
, DIMSYM - Symmetry Determination and Linear Dierential Equations Pa kage. Preprint, Department of Mathemati s (LaTrobe University, Bundoora, Australia, 1993).
[278
, Symmetry and Computer Algebra Te hniques for Dierential Equations Ph.D
Thesis, Department of Mathemati a (LaTrobe University, Bundoora, Australia, 1993).
[279
and
, Geometri Aspe ts of redu tion of order, Trans. Am. Math.
So . 314 (1992) 433-453.
[280
, Formal solutions of dierential equations, J. Symb. Comp. 10 (1991) 59-94.
[281
, Invertible Point Transformations and Nonlinear Dierential Equations
(World S ienti , Singapore, 1993).
[282
, Symmetry operators, in: Pro . 1979 MACSYMA User's Conferen e, Washington, D.C., 1979, Ed.: V.E. Lewis (MIT Press, Boston, Massa husetts, 1979) 408-444.
[283
, Lie series, Lie transformations, and their appli ations, in: Lie Methods in
Opti s. Pro . CIFMO-CIO Workshop, Leon, Mexi o, 1985, Eds.: J. San hez Mondragon and
K.B. Wolf. Le ture Notes in Physi s 250 (Springer Verlag, New York, 1986) 45-103.
[284
, Symmetries of dierential equations, MACSYMA Newsletter 7 (1990) 3-7.
[285
, Dierential Equations: Their Solution using Symmetries (Cambridge University Press, Cambridge, UK, 1989).
[286
, Dierential Systems. Colloquium Publi ations 21 (Ameri an Mathemati al
So iety, New York, New York, 1937).
[287
, Redu ing systems of linear dierential equations to a passive form, A ta
Appl. Math. 16 (1989) 191-206.
[288
, Sur les invariants dierentiels des groupes de transformations, A ta Math.
18 (1894) 1-88.
[289
, Determination des Invariants Pon tuel de l'Equation Dierentielle Ordinaire
du Se ond Ordre y 00 = ! (x; y; y 0) (S. Hirzel, Leipzig, 1896).
W.M. Seiler
W.M. Seiler
R. Sesahdri
T.Y. Na
W.A. Shadwi k
J. Sherring
J. Sherring
J. Sherring
M.F. Singer
W.-H.
Steeb
S. Steinberg
S. Steinberg
S. Steinberg
H. Stephani
J. Thomas
V.L. Topunov
A.M. Tresse
A.M. Tresse
G. Prin e
[290
, On variation bi omplexes of dierential equations, Osaka J. Math. 19 (1982)
T. Tsujishita
311-363.
[291
, Formal geometry of systems of dierential equations, in: Sugaku Expositions
3 (AMS, Providen e, Rhode Island, 1990) 25-73; translation of Sugaku 35 (1983) 332-357, in
Japanese.
[292
, Di. Geom. Appl. 1 (1991) 3-34.
[293
, Computer algebrai determination of symmetries and onservation laws of
PDEs, in: Pro . ISMM Int. Symposium on Computer Appli ations in Design, Simulation and
Analysis, New Orleans, Louisiana, 1990, Ed.: E.K. Park (A ta Press, Anaheim, California,
1990) 310-312.
[294
, SYMCON: A MACSYMA Pa kage for the Determination of Symmetries and
Conservation Laws of PDEs. Preprint, Department of Engineering S ien es (Trinity University, San Antonio, Texas, 1990)
[295
, PDELIE: A partial dierential equation solver, MACSYMA Newsletter 9,
no. 1 (1992) 1-13.
[296
, PDELIE: A partial dierential equation solver II, MACSYMA Newsletter 9,
no. 2-4 (1992) 5-20.
[297
, PDELIE: A partial dierential equation solver III, MACSYMA Newsletter
11 (1994) to appear.
[298
, PDELIE: Symboli Software for the Analysis of Partial Dierential Equations
by Lie Group Methods, User's Manual, Preprint, Department of Engineering S ien es (Trinity
University, San Antonio, Texas, USA, 1994).
[299
, Lo al Bifur ation and Symmetry (Pitman, London, 1982).
[300
, LiE, A software pa kage for Lie group omputations, Euromath
Bulletin 1, no. 2 (1994) 83-94.
[301
and
, Eds. Symmetries and Integrability of Dieren e
Equations, Pro . Workshop, Esterel, Quebe , Canada, May 22-29, 1994. CRM Pro eedings
and Le ture Notes (Ameri an Mathemati al So iety, Providen e, Rhode Island, 1994).
[302
, Lo al symmetries and onservation laws, A ta Appl. Math. 2 (1984)
21-78.
[303
, Symmetries and onservation laws of partial dierential equations:
Basi notations and results, A ta Appl. Math. 16 (1989) 3-21.
[304
, Ed., Symmetries of Partial Dierential Equations, Part I, A ta Appl.
Math. 15, nos. 1 & 2 (1989); Part II, ibid. 16, no. 1 (1989); Part III, ibid. 16, no. 2 (1989).
[305
, Ed., Symmetries of Partial Dierential Equations. Conservation laws
- Appli ations - Algorithms (Kluwer A ademi Publi ations, Dordre ht, The Netherlands,
1989).
T. Tsujishita
T. Tsujishita
P. Vafeades
P. Vafeades
P. Vafeades
P. Vafeades
P. Vafeades
P. Vafeades
A. Vanderbauwhede
M.A.A. van Leeuwen
L. Vinet, D. Levi
A.M. Vinogradov
A.M. Vinogradov
A.M. Vinogradov
A.M. Vinogradov
P. Winternitz
[306
and
, Nonlo al symmetries and the theory of overings, A ta Appl. Math. 2 (1984) 79-96.
[307
, Redu tion and quotient equations for dierential equations with symmetries, A ta Appl. Math. 23 (1991) 1-24.
[308
, Symmetries of ompatibility onditions for systems of dierential equations, A ta Appl. Math. 26 (1992) 61-86.
[309
, Dierential form analysis using MACSYMA, in: Pro . 1977 MACSYMA
User's Conferen e, Berkeley, California, 1977, Ed.: E.M. Andersen (NASA CP-2012, Washington, D.C., 1977) 71-74.
[310
, Foundations of Dierential Manifolds and Lie Groups (Springer Verlag,
New York, 1983).
[311
, Similarity Methods with Referen e to a High-order Nonlinear Diusion
Equation. Ph.D Thesis, Mathemati al Institute (University of Oxford, Oxford, United Kingdom, 1989).
[312
, Lie groups and solutions of nonlinear dierential equations, in: Nonlinear
Phenomena, Ed.: K.B. Wolf. Le ture Notes in Physi s 189 (Springer Verlag, New York, 1983)
263-331.
[313
, Group theory and exa t solutions of partially integrable dierential systems,
in: Partially Integrable Nonlinear Evolution Equations and their Physi al Appli ations. NATO
ASI Series C: Math. and Phys. S s. 310, Eds.: R. Conte and N. Bo ara (Kluwer A ademi
Publishers, Dordre ht, The Netherlands, 1990) 515-567.
[314
, Conditional symmetries and onditional integrability for nonlinear systems,
in: Group Theoreti al Methods in Physi s. Pro . XVIII International Colloquium, Mos ow,
U.S.S.R., 1990, Eds.: V.V. Dodonov and V.I. Man'ko. Le ture Notes in Physi s 382 (Springer
Verlag, Berlin, 1991) 263-271.
[315
, Lie group and solutions of nonlinear partial dierential equations, in: Integrable Systems, Quantum Groups, and Quantum Field Theories. NATO ASI Series C: Math.
and Phys. S s. 409, Eds.: L.A. Ibort and M.A. Rodriguez (Kluwer A ademi Publishers,
Dordre ht, The Netherlands, 1993) 429-495.
[316
, Ed., Lie Methods in Opti s II, Pro . CIFMA-CIO Workshop, Co oyo , Mexi o,
1988. Le ture Notes in Physi s 352 (Springer Verlag, New York, 1990).
[317
, An analyti algorithm for de oupling and integrating systems of nonlinear partial
dierential equations, J. Comp. Phys. 60 (1985) 437-446.
[318
, Zur analytis hen Untersu hing und exakten Losung von Dierentialglei hungen
mit Computeralgebrasystemen. Dissertation B (Friedri h S hiller Universitat, Jena, Germany,
1989).
A.M. Vinogradov
E.M. Vorob'ev
E.M. Vorob'ev
H.D. Wahlquist
F.W. Warner
S.M. Williams
P. Winternitz
P. Winternitz
P. Winternitz
P. Winternitz
K.B. Wolf
T. Wolf
T. Wolf
I.S. Krasil'sh hik
[319
, A pa kage for the analyti investigation and exa t solutions of dierential equations, in: Pro . EUROCAL '87, Leipzig, GDR, 1987, Ed.: J.H. Davenport, Le ture Notes in
Computer S ien e 378 (Springer Verlag, Berlin, 1989) 479-491.
[320
, The symboli integration of exa t PDEs, Preprint S hool of Mathemati al S ien es, Queen Mary and Westeld College (University of London, London, 1992).
[321
, An e ien y improved program LIEPDE for determining Lie-symmetries of
PDEs, in: Modern Group Analysis: Advan ed Analyti al and Computational Methods in
Mathemati al Physi s. Pro . Int. Workshop A ireale, Catania, Italy, 1992, Eds.: N.H. Ibragimov, M. Torrisi, and A. Valenti (Kluwer A ademi Publishers, Dordre ht, The Netherlands,
1993) 377-385.
[322
, A Program for Applying Symmetries of PDEs, Presentation at CATHODE Workshop on Computer Algebra Tools for Handling ODEs, London, UK, 21-24 Mar h 1994.
[323
and
, The omputer algebra pa kage CRACK for investigating PDEs,
in: Pro . ERCIM Advan ed Course on Partial Dierential Equations and Group Theory, Bonn,
1992, Ed.: J.F. Pommaret (Gesells haft fur Mathematik und Datenverarbeitung, Sankt Augustin, Germany, 1992); Also: Manual for CRACK added to the REDUCE Network Library,
S hool of Mathemati al S ien es, Queen Mary and Westeld College (University of London,
London, 1992) 1-19.
[324
and
, Heuristi s for overdetermined systems of PDEs, in: Pro . Meeting on Arti ial Intelligen e in Mathemati s, Glasgow, United Kingdom, 1991, Eds.: J.H.
Johnson, S. M Kee and A. Vella (Clarendon Press, Oxford, 1993) in press.
[325
and
, Investigating symmetries and other analyti al properties of ODEs
with the omputer algebra pa kage CRACK, Preprint S hool of Mathemati al S ien es, Queen
Mary and Westeld College (University of London, London, 1994).
[326
, Felix Klein and Sophus Lie (Birkhauser, Basel, 1988).
[327
and
, Symmetries, onserved quantities,
and hierar hies for some latti e systems with soliton stru ture, J. Math. Phys. 32 (1991)
1908-1918.
[328
, Le ture Notes on Geometri al Aspe ts of Partial Dierential Equations.
Series on Soviet and East European Mathemati s 9 (World S ienti , Singapore, 1992).
[329
, Computer lassi ation of the integrable oupled KdV-like systems with
unit main matrix, J. Symb. Comp. 15 (1993) 85-90.
[330
, Algebrai programming and dierential forms, The Mathemati a Journal 2, no.
1 (1992) 88-90.
[331
, Dierential Forms Pa kage, Wolfram Resear h In ., Champaign, Illinois, MathSour e 0205-221, 1994.
T. Wolf
T. Wolf
T. Wolf
T. Wolf
T. Wolf
A. Brand
T. Wolf
A. Brand
T. Wolf
A. Brand
I.M. Yaglom
H. Zhang, G. Tu, W. Oevel
V.V. Zharinov
A. Yu. Zharkov
F. Zizza
F. Zizza
B. Fu hssteiner

Symbolic Methods and Software Overview For Lie Symmetry Analysis

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Symbolic Methods and Software Overview For Lie Symmetry Analysis

Caricato da

Copyright:

Formati disponibili

x

1991 Mathemati s Subje t Classi ation.

1 Resear h Supported in Part by NSF under Grant CCR-9300978

INTRODUCTION AND LITERATURE

PURPOSE, METHODS AND ALGORITHMS

METHODS FOR COMPUTING DETERMINING EQUA-

PROLONGED VECTOR FIELDS

(x; u(k)) u l ; 1  jJ j  k;

are de ned as follows. The oe ients of the rst prolongation are

METHODS FOR REDUCING DETERMINING EQUATIONS

In the full omputer implementation of \triangulation" algorithms, one takes advantage of a

METHODS FOR SOLVING DETERMINING EQUATIONS

hkj (x1 ; x2 ; :::; xk 1 ; xk+1 ; :::; xn )(xk )j ;

thus, introdu ing fun tions hkj with fewer variables.

+ (nxk 1)! h(x1 ; x2 ; :::; xk 1; xk+1; :::; xn)

COMPUTATION OF THE SIZE OF THE SYMMETRY

BEYOND LIE-POINT SYMMETRIES

CONTACT AND GENERALIZED SYMMETRIES

(x; u(k)) x +

NONCLASSICAL OR CONDITIONAL SYMMETRIES

'l (x; u) = 0; l = 1; :::; q;

REVIEW OF SYMBOLIC SOFTWARE

SCRATCHPAD AND AXIOM PROGRAMS

PROGRAMS FOR OTHER SYSTEMS

THE HARRY DYM EQUATION

Consider the Harry Dym equation [1,

For PDELIE and SYM DE the input form would be

For SPDE and LIE the program a epts

Next, one sele ts the variable ut for elimination, e.g.

Then, the tprograms automati ally

There are only eight determining equations,

3u[13 u[1x[12 + xeta1

or equivalently, the invariant surfa e onditions

THE MAGNETO-HYDRO-DYNAMICS EQUATIONS

OTHER INTERESTING EXAMPLES

List of urrent symmetry programs

Name & System

Email & Anonymous FTP

T. Wolf & A. Brand

List of urrent symmetry programs

Name & System

Email & Anonymous FTP

P. Gragert & P. Kersten

List of urrent symmetry programs

Name & System

Email & Anonymous FTP

G. Reid & A. Wittkopf

montignyphysi s.m gill. a

S ope of symmetry programs

Solves Det. Eqs.

Wolf & Brand

Gragert & Kersten

Wolf & Brand

S ope of symmetry programs

Solves Det. Eqs.

Reid & Wittkopf

I.Sh. Akhatov, R.K. Gazizov

I.M. Anderson, N. Kamran

I.M. Anderson, N. Kamran

E.A. Arais, V.P. Shapeev

D.J. Arrigo, P. Broadbridge

V.A. Baikov, R.K. Gazizov

, Symboli Solvers for Nonlinear Di erential Equations, The Mathemati a

(x; u(k)) u l ; 1 jJ j k;

are dened as follows. The oe ients of the rst prolongation are

(x; u(k)) x +

3u[13 u[1x[12 + xeta1

montignyphysi s.m gill. a

, Symboli Solvers for Nonlinear Dierential Equations, The Mathemati a

, Dierential geometry te hniques for sets of nonlinear partial dierential

, On the arbitrariness of the general solution of an involutive partial dierential

, On variation bi omplexes of dierential equations, Osaka J. Math. 19 (1982)