Sei sulla pagina 1di 111

Wavelet collocation for fourth order problems

Silvia Bertoluzza
IMATI Enrico Magenes - CNR, Pavia

Joint work with Val erie Perrier, Lab. Jean Kuntzmann, INP Grenoble

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

1 / 24

Wavelet Collocation

Introduced O. Vasilyev & S. Paolucci [1995] and by S.B. & G. Naldi [1996]

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

2 / 24

Wavelet Collocation

Introduced O. Vasilyev & S. Paolucci [1995] and by S.B. & G. Naldi [1996] Two ingredients

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

2 / 24

Wavelet Collocation

Introduced O. Vasilyev & S. Paolucci [1995] and by S.B. & G. Naldi [1996] Two ingredients
Use of Deslaurier-Dubuc interpolating functions

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

2 / 24

Wavelet Collocation

Introduced O. Vasilyev & S. Paolucci [1995] and by S.B. & G. Naldi [1996] Two ingredients
Use of Deslaurier-Dubuc interpolating functions Use of a collocation approach

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

2 / 24

Wavelet Collocation

Introduced O. Vasilyev & S. Paolucci [1995] and by S.B. & G. Naldi [1996] Two ingredients
Use of Deslaurier-Dubuc interpolating functions Use of a collocation approach

Pointwise nonlinerities are easily treated (no integral computation)

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

2 / 24

Wavelet Collocation

Introduced O. Vasilyev & S. Paolucci [1995] and by S.B. & G. Naldi [1996] Two ingredients
Use of Deslaurier-Dubuc interpolating functions Use of a collocation approach

Pointwise nonlinerities are easily treated (no integral computation) Very simple yet eective rening/derening strategy

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

2 / 24

Wavelet Collocation

Introduced O. Vasilyev & S. Paolucci [1995] and by S.B. & G. Naldi [1996] Two ingredients
Use of Deslaurier-Dubuc interpolating functions Use of a collocation approach

Pointwise nonlinerities are easily treated (no integral computation) Very simple yet eective rening/derening strategy Applied to a wide class of problems

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

2 / 24

Wavelet Collocation

Introduced O. Vasilyev & S. Paolucci [1995] and by S.B. & G. Naldi [1996] Two ingredients
Use of Deslaurier-Dubuc interpolating functions Use of a collocation approach

Pointwise nonlinerities are easily treated (no integral computation) Very simple yet eective rening/derening strategy Applied to a wide class of problems
Fluid dynamics

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

2 / 24

Wavelet Collocation

Introduced O. Vasilyev & S. Paolucci [1995] and by S.B. & G. Naldi [1996] Two ingredients
Use of Deslaurier-Dubuc interpolating functions Use of a collocation approach

Pointwise nonlinerities are easily treated (no integral computation) Very simple yet eective rening/derening strategy Applied to a wide class of problems
Fluid dynamics Elasticity

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

2 / 24

Wavelet Collocation

Introduced O. Vasilyev & S. Paolucci [1995] and by S.B. & G. Naldi [1996] Two ingredients
Use of Deslaurier-Dubuc interpolating functions Use of a collocation approach

Pointwise nonlinerities are easily treated (no integral computation) Very simple yet eective rening/derening strategy Applied to a wide class of problems
Fluid dynamics Elasticity Fluid structure interaction

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

2 / 24

Wavelet Collocation

Introduced O. Vasilyev & S. Paolucci [1995] and by S.B. & G. Naldi [1996] Two ingredients
Use of Deslaurier-Dubuc interpolating functions Use of a collocation approach

Pointwise nonlinerities are easily treated (no integral computation) Very simple yet eective rening/derening strategy Applied to a wide class of problems
Fluid dynamics Elasticity Fluid structure interaction Semiconductors

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

2 / 24

Wavelet Collocation

Introduced O. Vasilyev & S. Paolucci [1995] and by S.B. & G. Naldi [1996] Two ingredients
Use of Deslaurier-Dubuc interpolating functions Use of a collocation approach

Pointwise nonlinerities are easily treated (no integral computation) Very simple yet eective rening/derening strategy Applied to a wide class of problems
Fluid dynamics Elasticity Fluid structure interaction Semiconductors

Almost no theory
Silvia Bertoluzza (IMATI-CNR) Wavelet collocation for fourth order problems 2 / 24

Wavelet Collocation

The Schauder basis


Vj : p.w. linears on uniform grid, step h = 2j
T f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f E

Basis: translated of
f f f f

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

3 / 24

Wavelet Collocation

The Schauder basis


Vj : p.w. linears on uniform grid, step h = 2j
T f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f E

Basis: translated of
f f f f

coarser space Vj 1
T t t t t  t  t  t  t  t t t t      t  t  t  t t t t t 
Silvia Bertoluzza (IMATI-CNR)

Basis: translated of

t  t  t t 
3 / 24

Wavelet collocation for fourth order problems

Wavelet Collocation

What is the dierence


T t t t t      t  t  t  t  t t t t t t t t  t  t  t  t t t t t t t t t 

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

4 / 24

Wavelet Collocation

What is the dierence


T t f  t f  t f  t f   t f t f t f t f t f t f t f t f t f t f t f t f  t f  t f  t f  t f t ff t ff t ff t ff t t t t    

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

4 / 24

Wavelet Collocation

What is the dierence


T t t t t      t  t  t  t  t t t t t t t t  t  t  t  t t t t t t t t t 

Dierence
T f f f f f ff f f f f f ff f f f f f ff f f f f f ff
4 / 24

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

Wavelet Collocation

Deslaurier-Dubuc [DD] Interpolating functions


Same structure but with more regular functions
1 0.8 0.6 0.4 0.2 0 1 1.2 1 0.8 0.6 0.4 0.2 0 0 1 0.2 5 0 5 1.2 1 0.8 0.6 0.4 0.2 0 0.2 5 0 5

1.2 1 0.8 0.6 0.4 0.2 0 0.2 10 0 10

1.2 1 0.8 0.6 0.4 0.2 0 0.2 10 0 10

1.2 1 0.8 0.6 0.4 0.2 0 0.2 20 0 20

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

5 / 24

Wavelet Collocation

Deslaurier-Dubuc [DD] Interpolating functions


Two ways of constructing : limit of a polynomial interpolatory subdivision scheme:

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

6 / 24

Wavelet Collocation

Deslaurier-Dubuc [DD] Interpolating functions


Two ways of constructing : limit of a polynomial interpolatory subdivision scheme:
(k ) = k 0 , k Z

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

6 / 24

Wavelet Collocation

Deslaurier-Dubuc [DD] Interpolating functions


Two ways of constructing : limit of a polynomial interpolatory subdivision scheme:
(k ) = k 0 , k Z {(k /2j )} {(k /2j +1 )} ((2n 1)/2j +1 ) = jn ((2n 1)/2j +1 ) with jn polynomial of degree 2L + 1 dened as jn (k ) = (k /2j ), k = n L, ., n + L

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

6 / 24

Wavelet Collocation

Deslaurier-Dubuc [DD] Interpolating functions


Two ways of constructing : limit of a polynomial interpolatory subdivision scheme:
(k ) = k 0 , k Z {(k /2j )} {(k /2j +1 )} ((2n 1)/2j +1 ) = jn ((2n 1)/2j +1 ) with jn polynomial of degree 2L + 1 dened as jn (k ) = (k /2j ), k = n L, ., n + L

is the autocorrelation function of the Daubechies scaling function with L 1 zero moments

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

6 / 24

Wavelet Collocation

Deslaurier-Dubuc [DD] Interpolating functions


Two ways of constructing : limit of a polynomial interpolatory subdivision scheme:
(k ) = k 0 , k Z {(k /2j )} {(k /2j +1 )} ((2n 1)/2j +1 ) = jn ((2n 1)/2j +1 ) with jn polynomial of degree 2L + 1 dened as jn (k ) = (k /2j ), k = n L, ., n + L

is the autocorrelation function of the Daubechies scaling function with L 1 zero moments is interpolatory ((k ) = k 0 )
Silvia Bertoluzza (IMATI-CNR) Wavelet collocation for fourth order problems 6 / 24

Wavelet Collocation

DD interpolating wavelets
Vj : span of {jk , k N}

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

7 / 24

Wavelet Collocation

DD interpolating wavelets
Vj : span of {jk , k N} Ij : C 0 (R) Vj Langrangian interpolation operator

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

7 / 24

Wavelet Collocation

DD interpolating wavelets
Vj : span of {jk , k N} Ij : C 0 (R) Vj Langrangian interpolation operator Vj Vj +1

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

7 / 24

Wavelet Collocation

DD interpolating wavelets
Vj : span of {jk , k N} Ij : C 0 (R) Vj Langrangian interpolation operator Vj Vj +1 detail space: Wj = (Ij +1 Ij )Vj +1

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

7 / 24

Wavelet Collocation

DD interpolating wavelets
Vj : span of {jk , k N} Ij : C 0 (R) Vj Langrangian interpolation operator Vj Vj +1 detail space: Wj = (Ij +1 Ij )Vj +1 Wj : span of {j ,k = (2j x k ) = (2(2j x k ) 1), k N}

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

7 / 24

Wavelet Collocation

DD interpolating wavelets
Vj : span of {jk , k N} Ij : C 0 (R) Vj Langrangian interpolation operator Vj Vj +1 detail space: Wj = (Ij +1 Ij )Vj +1 Wj : span of {j ,k = (2j x k ) = (2(2j x k ) 1), k N} Well dened correspondence between dyadic points and basis functions:

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

7 / 24

Wavelet Collocation

DD interpolating wavelets
Vj : span of {jk , k N} Ij : C 0 (R) Vj Langrangian interpolation operator Vj Vj +1 detail space: Wj = (Ij +1 Ij )Vj +1 Wj : span of {j ,k = (2j x k ) = (2(2j x k ) 1), k N} Well dened correspondence between dyadic points and basis functions:
at xed coarse level j0 x = k 2j0 j0 ,k

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

7 / 24

Wavelet Collocation

DD interpolating wavelets
Vj : span of {jk , k N} Ij : C 0 (R) Vj Langrangian interpolation operator Vj Vj +1 detail space: Wj = (Ij +1 Ij )Vj +1 Wj : span of {j ,k = (2j x k ) = (2(2j x k ) 1), k N} Well dened correspondence between dyadic points and basis functions:
at xed coarse level j0 x = k 2j0 j0 ,k for j > j0 and odd k = 2n + 1 x = k 2j j 1,n

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

7 / 24

Wavelet Collocation

DD interpolating wavelets
Vj : span of {jk , k N} Ij : C 0 (R) Vj Langrangian interpolation operator Vj Vj +1 detail space: Wj = (Ij +1 Ij )Vj +1 Wj : span of {j ,k = (2j x k ) = (2(2j x k ) 1), k N} Well dened correspondence between dyadic points and basis functions:
at xed coarse level j0 x = k 2j0 j0 ,k for j > j0 and odd k = 2n + 1 x = k 2j j 1,n

bases for Rd by tensor product.


Silvia Bertoluzza (IMATI-CNR) Wavelet collocation for fourth order problems 7 / 24

Wavelet Collocation

DD interpolating wavelets on [0,1]


Basis on [0,1] built by polynomial extrapolation
[Donoho, 92]

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

8 / 24

Wavelet Collocation

DD interpolating wavelets on [0,1]


Basis on [0,1] built by polynomial extrapolation
[Donoho, 92]

given values at xj ,k = k 2j , k = 0, . . . , 2j you want to uniquely select one function in Vj

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

8 / 24

Wavelet Collocation

DD interpolating wavelets on [0,1]


Basis on [0,1] built by polynomial extrapolation
[Donoho, 92]

given values at xj ,k = k 2j , k = 0, . . . , 2j you want to uniquely select one function in Vj obtain values at xj ,k outside [0, 1] by extrapolating values in [0, 1]

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

8 / 24

Wavelet Collocation

DD interpolating wavelets on [0,1]


Basis on [0,1] built by polynomial extrapolation
[Donoho, 92]

given values at xj ,k = k 2j , k = 0, . . . , 2j you want to uniquely select one function in Vj obtain values at xj ,k outside [0, 1] by extrapolating values in [0, 1] this yields subspace Ej Vj

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

8 / 24

Wavelet Collocation

DD interpolating wavelets on [0,1]


Basis on [0,1] built by polynomial extrapolation
[Donoho, 92]

given values at xj ,k = k 2j , k = 0, . . . , 2j you want to uniquely select one function in Vj obtain values at xj ,k outside [0, 1] by extrapolating values in [0, 1] this yields subspace Ej Vj Vj [0, 1] = Ej |[0,1]

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

8 / 24

Wavelet Collocation

DD interpolating wavelets on [0,1]


Basis on [0,1] built by polynomial extrapolation
[Donoho, 92]

given values at xj ,k = k 2j , k = 0, . . . , 2j you want to uniquely select one function in Vj obtain values at xj ,k outside [0, 1] by extrapolating values in [0, 1] this yields subspace Ej Vj Vj [0, 1] = Ej |[0,1]

Basis of Vj [0, 1]: {j ,k , k = 0, . . . , 2j } with j ,k (xj ,n ) = n,k , n = 0, . . . , 2j

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

8 / 24

Wavelet Collocation

DD interpolating wavelets on [0,1]


Basis on [0,1] built by polynomial extrapolation
[Donoho, 92]

given values at xj ,k = k 2j , k = 0, . . . , 2j you want to uniquely select one function in Vj obtain values at xj ,k outside [0, 1] by extrapolating values in [0, 1] this yields subspace Ej Vj Vj [0, 1] = Ej |[0,1]

Basis of Vj [0, 1]: {j ,k , k = 0, . . . , 2j } with j ,k (xj ,n ) = n,k , n = 0, . . . , 2j

Vj [0, 1] Vj +1 [0, 1]

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

8 / 24

Wavelet Collocation

DD interpolating wavelets on [0,1]


Basis on [0,1] built by polynomial extrapolation
[Donoho, 92]

given values at xj ,k = k 2j , k = 0, . . . , 2j you want to uniquely select one function in Vj obtain values at xj ,k outside [0, 1] by extrapolating values in [0, 1] this yields subspace Ej Vj Vj [0, 1] = Ej |[0,1]

Basis of Vj [0, 1]: {j ,k , k = 0, . . . , 2j } with j ,k (xj ,n ) = n,k , n = 0, . . . , 2j

Vj [0, 1] Vj +1 [0, 1] detail space and its basis built as for R


Silvia Bertoluzza (IMATI-CNR) Wavelet collocation for fourth order problems 8 / 24

Wavelet Collocation

Collocation method on a uniform grid


Second order problem Au = f in ]0, 1[d , B u = g on ]0, 1[d

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

9 / 24

Wavelet Collocation

Collocation method on a uniform grid


Second order problem Au = f in ]0, 1[d , B u = g on ]0, 1[d

Gj : grid of dyadic points of [0, 1] Gj = {k /2j : k Zd [0, 1]d }

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

9 / 24

Wavelet Collocation

Collocation method on a uniform grid


Second order problem Au = f in ]0, 1[d , B u = g on ]0, 1[d

Gj : grid of dyadic points of [0, 1] Gj = {k /2j : k Zd [0, 1]d }

Problem Look for uh Vj [0, 1]d such that Auh () = f () Gj ]0, 1[,
Silvia Bertoluzza (IMATI-CNR)

B u () = g () Gj ]0, 1[d
9 / 24

Wavelet collocation for fourth order problems

Wavelet Collocation

Collocation method on a non-uniform grid


detail grids: Jj = Gj +1 \ Gj

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

10 / 24

Wavelet Collocation

Collocation method on a non-uniform grid


detail grids: Jj = Gj +1 \ Gj = Gj0 j j0 Jj : set of all dyadic points

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

10 / 24

Wavelet Collocation

Collocation method on a non-uniform grid


detail grids: Jj = Gj +1 \ Gj = Gj0 j j0 Jj : set of all dyadic points interpolating wavelet

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

10 / 24

Wavelet Collocation

Collocation method on a non-uniform grid


detail grids: Jj = Gj +1 \ Gj = Gj0 j j0 Jj : set of all dyadic points interpolating wavelet h Vh Vh = span < , h >

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

10 / 24

Wavelet Collocation

Collocation method on a non-uniform grid


detail grids: Jj = Gj +1 \ Gj = Gj0 j j0 Jj : set of all dyadic points interpolating wavelet h Vh Vh = span < , h >

Problem Look for uh Vh such that Auh () = f () h ]0, 1[, B u () = f () h ]0, 1[d

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

10 / 24

Wavelet Collocation

Adaptivity
Simple adaptive strategy: look at coecients of tentative solution to design the next grid

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

11 / 24

Wavelet Collocation

Adaptivity
Simple adaptive strategy: look at coecients of tentative solution to design the next grid
if u small (|u | < r ): delete it from the grid

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

11 / 24

Wavelet Collocation

Adaptivity
Simple adaptive strategy: look at coecients of tentative solution to design the next grid
if u small (|u | < r ): delete it from the grid if u big (|u | > a > r ): add neighbouring points at higher level

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

11 / 24

Wavelet Collocation

Adaptivity
Simple adaptive strategy: look at coecients of tentative solution to design the next grid
if u small (|u | < r ): delete it from the grid if u big (|u | > a > r ): add neighbouring points at higher level

solve a sequence of problems with greed designed by looking at the solution at the previous step

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

11 / 24

Wavelet Collocation

Adaptivity
Simple adaptive strategy: look at coecients of tentative solution to design the next grid
if u small (|u | < r ): delete it from the grid if u big (|u | > a > r ): add neighbouring points at higher level

solve a sequence of problems with greed designed by looking at the solution at the previous step other strategies might be applied but no theory available

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

11 / 24

Wavelet Collocation

Adaptivity
Simple adaptive strategy: look at coecients of tentative solution to design the next grid
if u small (|u | < r ): delete it from the grid if u big (|u | > a > r ): add neighbouring points at higher level

solve a sequence of problems with greed designed by looking at the solution at the previous step other strategies might be applied but no theory available strategy eectively applied to dierent equations (Burgers equation, Convection-Diusion problems, Euler-Poisson system for semiconductors,. . . )

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

11 / 24

Wavelet Collocation

Fourth order problems


Problem Find u : [0, 1]d R such that Au = f in ]0, 1[d , where A: fourth order operator B takes values in R2 (two boundary conditions) Examples Structural mechanics (beams, plates, . . . ) Fluid dynamics (ice formations, uids in the lungs) Nanotechnologies Denoising
Silvia Bertoluzza (IMATI-CNR) Wavelet collocation for fourth order problems 12 / 24

B u = g on ]0, 1[d

Wavelet Collocation

Fourth order problems on [0, 1]


Extension to 4th order problems: two things need to be faced two boundary condition at each extrema, only one degree of freedom: if we impose the equation at all interior points we have two more equations than d.o.f.

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

13 / 24

Wavelet Collocation

Fourth order problems on [0, 1]


Extension to 4th order problems: two things need to be faced two boundary condition at each extrema, only one degree of freedom: if we impose the equation at all interior points we have two more equations than d.o.f. possible solutions

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

13 / 24

Wavelet Collocation

Fourth order problems on [0, 1]


Extension to 4th order problems: two things need to be faced two boundary condition at each extrema, only one degree of freedom: if we impose the equation at all interior points we have two more equations than d.o.f. possible solutions
do not impose equation at all interior points

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

13 / 24

Wavelet Collocation

Fourth order problems on [0, 1]


Extension to 4th order problems: two things need to be faced two boundary condition at each extrema, only one degree of freedom: if we impose the equation at all interior points we have two more equations than d.o.f. possible solutions
do not impose equation at all interior points add one d.o.f. from a higher level

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

13 / 24

Wavelet Collocation

Fourth order problems on [0, 1]


Extension to 4th order problems: two things need to be faced two boundary condition at each extrema, only one degree of freedom: if we impose the equation at all interior points we have two more equations than d.o.f. possible solutions
do not impose equation at all interior points add one d.o.f. from a higher level modify the construction of DD wavelets on [0, 1] so that we have two d.o.f. at each extrema

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

13 / 24

Wavelet Collocation

Fourth order problems on [0, 1]


Extension to 4th order problems: two things need to be faced two boundary condition at each extrema, only one degree of freedom: if we impose the equation at all interior points we have two more equations than d.o.f. possible solutions
do not impose equation at all interior points add one d.o.f. from a higher level modify the construction of DD wavelets on [0, 1] so that we have two d.o.f. at each extrema

conditioning of the matrices is so bad that round-o is killer

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

13 / 24

Wavelet Collocation

Fourth order problems on [0, 1]


Extension to 4th order problems: two things need to be faced two boundary condition at each extrema, only one degree of freedom: if we impose the equation at all interior points we have two more equations than d.o.f. possible solutions
do not impose equation at all interior points add one d.o.f. from a higher level modify the construction of DD wavelets on [0, 1] so that we have two d.o.f. at each extrema

conditioning of the matrices is so bad that round-o is killer solution: increase the precision of (selected) computations

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

13 / 24

Wavelet Collocation

Hermite boundary functions


New basis on [0,1] built by modied polynomial extrapolation

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

14 / 24

Wavelet Collocation

Hermite boundary functions


New basis on [0,1] built by modied polynomial extrapolation Lagrangian at interior points / Hermite at boundary points

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

14 / 24

Wavelet Collocation

Hermite boundary functions


New basis on [0,1] built by modied polynomial extrapolation Lagrangian at interior points / Hermite at boundary points Basis of Vj [0, 1]: {j ,k , k = 1, 0, . . . , 2j , 2j + 1, } with

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

14 / 24

Wavelet Collocation

Hermite boundary functions


New basis on [0,1] built by modied polynomial extrapolation Lagrangian at interior points / Hermite at boundary points Basis of Vj [0, 1]: {j ,k , k = 1, 0, . . . , 2j , 2j + 1, } with j ,k (xj ,n ) = n,k , k = 1, . . . , 2j + 1, n = 0, . . . , 2j

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

14 / 24

Wavelet Collocation

Hermite boundary functions


New basis on [0,1] built by modied polynomial extrapolation Lagrangian at interior points / Hermite at boundary points Basis of Vj [0, 1]: {j ,k , k = 1, 0, . . . , 2j , 2j + 1, } with j ,k (xj ,n ) = n,k , k = 1, . . . , 2j + 1, n = 0, . . . , 2j j ,k (0) = j ,k (1) = 0, k = 0, . . . , 2j

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

14 / 24

Wavelet Collocation

Hermite boundary functions


New basis on [0,1] built by modied polynomial extrapolation Lagrangian at interior points / Hermite at boundary points Basis of Vj [0, 1]: {j ,k , k = 1, 0, . . . , 2j , 2j + 1, } with j ,k (xj ,n ) = n,k , k = 1, . . . , 2j + 1, n = 0, . . . , 2j j ,k (0) = j ,k (1) = 0, k = 0, . . . , 2j j ,1 (0) = j ,2j +1 (1) = 1, j ,1 (1) = j ,2j +1 (0) = 0

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

14 / 24

Wavelet Collocation

Hermite boundary functions


New basis on [0,1] built by modied polynomial extrapolation Lagrangian at interior points / Hermite at boundary points Basis of Vj [0, 1]: {j ,k , k = 1, 0, . . . , 2j , 2j + 1, } with j ,k (xj ,n ) = n,k , k = 1, . . . , 2j + 1, n = 0, . . . , 2j j ,k (0) = j ,k (1) = 0, k = 0, . . . , 2j j ,1 (0) = j ,2j +1 (1) = 1, j ,1 (1) = j ,2j +1 (0) = 0

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

14 / 24

Wavelet Collocation

Multiresolution analysis
Vj [0, 1] dened as the span of such functions

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

15 / 24

Wavelet Collocation

Multiresolution analysis
Vj [0, 1] dened as the span of such functions Interpolation Ij : C 1 (0, 1) Vj [0, 1]
2j

Ij f = f (0)j ,1 +
k =0

f (xk )j ,k + f (1)j ,2j +1

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

15 / 24

Wavelet Collocation

Multiresolution analysis
Vj [0, 1] dened as the span of such functions Interpolation Ij : C 1 (0, 1) Vj [0, 1]
2j

Ij f = f (0)j ,1 +
k =0

f (xk )j ,k + f (1)j ,2j +1

Vj [0, 1] Vj +1 [0, 1]

[S.B., Perrier, 2012]

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

15 / 24

Wavelet Collocation

Multiresolution analysis
Vj [0, 1] dened as the span of such functions Interpolation Ij : C 1 (0, 1) Vj [0, 1]
2j

Ij f = f (0)j ,1 +
k =0

f (xk )j ,k + f (1)j ,2j +1

Vj [0, 1] Vj +1 [0, 1]

[S.B., Perrier, 2012]

Wj [0, 1] = (Ij +1 Ij )Vj +1 [0, 1]

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

15 / 24

Wavelet Collocation

Multiresolution analysis
Vj [0, 1] dened as the span of such functions Interpolation Ij : C 1 (0, 1) Vj [0, 1]
2j

Ij f = f (0)j ,1 +
k =0

f (xk )j ,k + f (1)j ,2j +1

Vj [0, 1] Vj +1 [0, 1]

[S.B., Perrier, 2012]

Wj [0, 1] = (Ij +1 Ij )Vj +1 [0, 1] Wj [0, 1] span of {j ,k = j +1,2k 1 , k = 1, . . . , 2j }

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

15 / 24

Wavelet Collocation

Smoothness charachterization

Theorem Let f =

2j0 +1

2j

fk j0 ,k +
k =1 j k =1 ,p Bq

dj ,k 2j /2 j ,k if and only if |dj ,k |p )q/p < +,


k =1

and let 1 + 1/p < < R. Then f


2j0 +1

2j

(
k =1

|fk | )

p q /p

+
j

jsq /p

with s = + 1/2 1/p.

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

16 / 24

Wavelet Collocation

Killer round-o: example


test problem: u (iv ) = 0, u (0) = u (1) = 1, u (0) = u (1) = 0

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

17 / 24

Wavelet Collocation

Killer round-o: example


test problem: u (iv ) = 0, u (0) = u (1) = 1, u (0) = u (1) = 0 errors for solution in Vj [0, 1] (scaling function basis) h .0625 L2 error 9.696e-06 L error 1.5677e-05

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

17 / 24

Wavelet Collocation

Killer round-o: example


test problem: u (iv ) = 0, u (0) = u (1) = 1, u (0) = u (1) = 0 errors for solution in Vj [0, 1] (scaling function basis) h .0625 .0313 L2 error 9.696e-06 .00015748 L error 1.5677e-05 .00025088

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

17 / 24

Wavelet Collocation

Killer round-o: example


test problem: u (iv ) = 0, u (0) = u (1) = 1, u (0) = u (1) = 0 errors for solution in Vj [0, 1] (scaling function basis) h .0625 .0313 .0156 L2 error 9.696e-06 .00015748 .0025464 L error 1.5677e-05 .00025088 .0040258

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

17 / 24

Wavelet Collocation

Killer round-o: example


test problem: u (iv ) = 0, u (0) = u (1) = 1, u (0) = u (1) = 0 errors for solution in Vj [0, 1] (scaling function basis) h .0625 .0313 .0156 .0078 L2 error 9.696e-06 .00015748 .0025464 .042886 L error 1.5677e-05 .00025088 .0040258 .067578

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

17 / 24

Wavelet Collocation

Killer round-o: example


test problem: u (iv ) = 0, u (0) = u (1) = 1, u (0) = u (1) = 0 errors for solution in Vj [0, 1] (scaling function basis) h .0625 .0313 .0156 .0078 .0039 L2 error 9.696e-06 .00015748 .0025464 .042886 3.0851 L error 1.5677e-05 .00025088 .0040258 .067578 4.8962

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

17 / 24

Wavelet Collocation

Killer round-o: example


test problem: u (iv ) = 0, u (0) = u (1) = 1, u (0) = u (1) = 0 errors for solution in Vj [0, 1] (scaling function basis) h .0625 .0313 .0156 .0078 .0039 L2 error 9.696e-06 .00015748 .0025464 .042886 3.0851 L error 1.5677e-05 .00025088 .0040258 .067578 4.8962

try with the multiscale basis

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

17 / 24

Wavelet Collocation

Killer round-o: example


test problem: u (iv ) = 0, u (0) = u (1) = 1, u (0) = u (1) = 0 errors for solution in Vj [0, 1] (scaling function basis) h .0625 .0313 .0156 .0078 .0039 L2 error 9.696e-06 .00015748 .0025464 .042886 3.0851 L error 1.5677e-05 .00025088 .0040258 .067578 4.8962 L2 error 9.696e-06 9.8408e-06 9.9162e-06 9.9545e-06 9.9734e-06 L error 1.5677e-05 1.5677e-05 1.5677e-05 1.5677e-05 1.5677e-05

try with the multiscale basis

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

17 / 24

Wavelet Collocation

Killer round-o: example


test problem: u (iv ) = 0, u (0) = u (1) = 1, u (0) = u (1) = 0 errors for solution in Vj [0, 1] (scaling function basis) h .0625 .0313 .0156 .0078 .0039 L2 error 9.696e-06 .00015748 .0025464 .042886 3.0851 L error 1.5677e-05 .00025088 .0040258 .067578 4.8962 L2 error 9.696e-06 9.8408e-06 9.9162e-06 9.9545e-06 9.9734e-06 L error 1.5677e-05 1.5677e-05 1.5677e-05 1.5677e-05 1.5677e-05

try with the multiscale basis basis reconstructs exactely polynomials of order 11, results not satisfactory at all!!
Silvia Bertoluzza (IMATI-CNR) Wavelet collocation for fourth order problems 17 / 24

Wavelet Collocation

Computation with 30 exact digits


Matlab Multiprecision toolbox
[Advanpix]

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

18 / 24

Wavelet Collocation

Computation with 30 exact digits


Matlab Multiprecision toolbox
[Advanpix]

Same computation with 30 exact digits

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

18 / 24

Wavelet Collocation

Computation with 30 exact digits


Matlab Multiprecision toolbox errors h .0625 .0313 .0156 .0078 .0039 L2 error 2.3871e-25 3.8769e-24 6.2506e-23 1.004e-21 1.6095e-20 L error 3.8769e-25 6.1761e-24 9.8818e-23 1.5811e-21 2.5297e-20
[Advanpix]

Same computation with 30 exact digits

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

18 / 24

Wavelet Collocation

Computation with 30 exact digits


Matlab Multiprecision toolbox errors h .0625 .0313 .0156 .0078 .0039 L2 error 2.3871e-25 3.8769e-24 6.2506e-23 1.004e-21 1.6095e-20 L error 3.8769e-25 6.1761e-24 9.8818e-23 1.5811e-21 2.5297e-20
[Advanpix]

Same computation with 30 exact digits

good! but very expensive solve linear system in double precision

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

18 / 24

Wavelet Collocation

Computation with 30 exact digits


Matlab Multiprecision toolbox errors h .0625 .0313 .0156 .0078 .0039 L2 error 2.3871e-25 3.8769e-24 6.2506e-23 1.004e-21 1.6095e-20 L error 3.8769e-25 6.1761e-24 9.8818e-23 1.5811e-21 2.5297e-20 L2 error 3.3959e-13 3.1068e-13 6.1219e-13 1.6684e-13 1.0915e-13 L error 9.952e-13 4.6829e-13 8.5487e-13 5.0271e-13 2.4181e-13
[Advanpix]

Same computation with 30 exact digits

good! but very expensive solve linear system in double precision

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

18 / 24

Wavelet Collocation

Computation with 30 exact digits


Matlab Multiprecision toolbox errors h .0625 .0313 .0156 .0078 .0039 L2 error 2.3871e-25 3.8769e-24 6.2506e-23 1.004e-21 1.6095e-20 L error 3.8769e-25 6.1761e-24 9.8818e-23 1.5811e-21 2.5297e-20 L2 error 3.3959e-13 3.1068e-13 6.1219e-13 1.6684e-13 1.0915e-13 L error 9.952e-13 4.6829e-13 8.5487e-13 5.0271e-13 2.4181e-13
[Advanpix]

Same computation with 30 exact digits

good! but very expensive solve linear system in double precision much more satisfactory!!
Silvia Bertoluzza (IMATI-CNR) Wavelet collocation for fourth order problems 18 / 24

Wavelet Collocation

Multiple precision computation

What exactly is computed in multiple precision:

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

19 / 24

Wavelet Collocation

Multiple precision computation

What exactly is computed in multiple precision: initialization is performed in multiple precision

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

19 / 24

Wavelet Collocation

Multiple precision computation

What exactly is computed in multiple precision: initialization is performed in multiple precision


eigenvalues/eigenvectors needed for computing derivatives of at integers

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

19 / 24

Wavelet Collocation

Multiple precision computation

What exactly is computed in multiple precision: initialization is performed in multiple precision


eigenvalues/eigenvectors needed for computing derivatives of at integers renement to compute values of at dyadic points

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

19 / 24

Wavelet Collocation

Multiple precision computation

What exactly is computed in multiple precision: initialization is performed in multiple precision


eigenvalues/eigenvectors needed for computing derivatives of at integers renement to compute values of at dyadic points extrapolation to compute basis for [0, 1]

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

19 / 24

Wavelet Collocation

Multiple precision computation

What exactly is computed in multiple precision: initialization is performed in multiple precision


eigenvalues/eigenvectors needed for computing derivatives of at integers renement to compute values of at dyadic points extrapolation to compute basis for [0, 1]

matrix and right hand side assembled in double precision

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

19 / 24

Wavelet Collocation

Multiple precision computation

What exactly is computed in multiple precision: initialization is performed in multiple precision


eigenvalues/eigenvectors needed for computing derivatives of at integers renement to compute values of at dyadic points extrapolation to compute basis for [0, 1]

matrix and right hand side assembled in double precision linear system solved in double precision

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

19 / 24

Wavelet Collocation

Test 1. [Twizell & Tirmizi, 86]

Equation: w (iv ) + 4w = 1, Boundary conditions w (1) = w (1) = 0, w (1) = w (1) = Exact solution w= 1 2(sin(1)sinh(1)sin(x )sinh(x ) + cos (1)cosh(1)cos (x )cosh(x ) 4(cosh(2) + cos (2)) sinh(2) sin(2) 4(cosh(2) + cos (2)) in (1, 1)

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

20 / 24

Wavelet Collocation

Errors h .0625 .0313 .0156 .0078 .0039 L2 error 4.8859e-10 2.087e-12 1.2167e-14 3.2334e-17 4.5559e-17 L error 8.071e-10 4.6301e-12 1.9113e-14 9.2613e-17 9.395e-17

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

21 / 24

Wavelet Collocation

Errors h .0625 .0313 .0156 .0078 .0039 L2 error 4.8859e-10 2.087e-12 1.2167e-14 3.2334e-17 4.5559e-17 L error 8.071e-10 4.6301e-12 1.9113e-14 9.2613e-17 9.395e-17 L2 error 4.8859e-10 2.8972e-12 5.7798e-09 4.4425e-09 3.1474e-08 L error 8.0717e-10 4.6153e-12 2.0393e-08 2.0393e-08 2.0393e-08

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

21 / 24

Wavelet Collocation

Test 2. Clamped-Clamped beam, half-loaded

Equation: EIw (iv ) = q Boundary conditions f , w (0) = w (2L) = w (0) = w (2L) = 0 Exact solution is known
Silvia Bertoluzza (IMATI-CNR) Wavelet collocation for fourth order problems

h .0625 .0313 .0156 .0078 .0039

L2 error 0.0173 0.0039 9.8449 e-04 2.4704 e-04 6.1876 e-05


22 / 24

Wavelet Collocation

Test 3. Cantilever beam, tip loaded

Equation: EIw (iv ) = 0 Boundary conditions w (0) = w (0) = w (L) = 0, EIw (L) = P

h .0625 .0313 .0156 .0078 .0039

L2 error 8.5334e-10 1.6134e-12 6.3875e-11 7.0735e-11 8.205e-11


23 / 24

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

Wavelet Collocation

Conclusions
Modied interpolating wavelets on [0, 1] well suited for 4th order problems

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

24 / 24

Wavelet Collocation

Conclusions
Modied interpolating wavelets on [0, 1] well suited for 4th order problems Computations must be (partially) carried out in multiple precisions

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

24 / 24

Wavelet Collocation

Conclusions
Modied interpolating wavelets on [0, 1] well suited for 4th order problems Computations must be (partially) carried out in multiple precisions First tests show that the method behaves as expected

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

24 / 24

Wavelet Collocation

Conclusions
Modied interpolating wavelets on [0, 1] well suited for 4th order problems Computations must be (partially) carried out in multiple precisions First tests show that the method behaves as expected Plenty of open problems!

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

24 / 24

Wavelet Collocation

Conclusions
Modied interpolating wavelets on [0, 1] well suited for 4th order problems Computations must be (partially) carried out in multiple precisions First tests show that the method behaves as expected Plenty of open problems! Adaptivity needs to be studied

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

24 / 24

Wavelet Collocation

Conclusions
Modied interpolating wavelets on [0, 1] well suited for 4th order problems Computations must be (partially) carried out in multiple precisions First tests show that the method behaves as expected Plenty of open problems! Adaptivity needs to be studied Preconditioning is of paramount importance

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

24 / 24

Wavelet Collocation

Conclusions
Modied interpolating wavelets on [0, 1] well suited for 4th order problems Computations must be (partially) carried out in multiple precisions First tests show that the method behaves as expected Plenty of open problems! Adaptivity needs to be studied Preconditioning is of paramount importance No theory!

Silvia Bertoluzza (IMATI-CNR)

Wavelet collocation for fourth order problems

24 / 24

Wavelet Collocation

Conclusions
Modied interpolating wavelets on [0, 1] well suited for 4th order problems Computations must be (partially) carried out in multiple precisions First tests show that the method behaves as expected Plenty of open problems! Adaptivity needs to be studied Preconditioning is of paramount importance No theory! For whom might be interested: the Advanpix Multiple Precision Toolbox for Matlab works beautifully!! http://www.advanpix.com
Silvia Bertoluzza (IMATI-CNR) Wavelet collocation for fourth order problems 24 / 24

Potrebbero piacerti anche