Sei sulla pagina 1di 11

results obtained from the root loc

Figure 9.15.
In summary, although compensa
than method b, the addition of ideal de
3/27/14  
time in each case while keeping the pe
be seen in the settling time and peak t
across all of the cases of compensation
steady-state error, even though lag com
error of the compensated system is a
system, as seen by eð1Þ and Kp. All sy
state error is expected. The reader mu
transient response always yields an im
The time response of each case i
the compensated responses are faster
response.
Now that we have seen what id
LEAD COMPENSATOR ready to design our own ideal derivat
specification. Basically, we will evalua
and zeros to a design point that is
transient response. The difference betw
angular contribution of the compensa
the position of the zero to yield the

Example 9.3
Design 3 lead compensators for the following Ideal
systemDerivative
that Compensato
will reduce the settling time by a factor of 2 while
maintaining the 30% overshoot. Compare the system
PROBLEM: Given the
system of Figu
characteristics between the three designs.
sator to yield a 16% overshoot, with
SOLUTION: Let us firs
R(s) + E(s) K C(s)
s(s + 4)(s + 6)
compensated system ope
– for the uncompensated
overshoot is equivalent
FIGURE 9.17 Feedback control system for ratio line for an odd m
Example 9.3 second-order pair of po
Lead Compensator Design: Example 1

1  
ular contribution at the design point of this pole
contribution orof
zero
thethe
Examplealongiswith
remaining the system’s
compensator pole or zero. Let us look at an example.
!9.4
when design complete, and the ensuing transient response.
Lead this
Example
oop poles and zeros. The difference between Compensator
9.4
angle and 180
For design, Design
is the required
we arbitrarily select either a lead compensator pole or zero and find
Example
ution of the remaining compensator 9.4
pole Lead
or zero.
the Let contribution
angular us look atDesign
Compensator anthe
at example.
design point of this pole or zero along with the system’s
PROBLEM:
Lead Compensator Design three
Design lead compensators for the system of Figure 9.17 that will
The difference between this angle and 180! is the required
reduce the settling
open-loop Example
poles and zeros.9.4
time by aoffactor of 2 while maintaining
Lead Compensator Design
contribution the remaining compensator pole or30% zero. overshoot.
Let us look atCompare an example.
E1C09 11/03/2010 PROBLEM: 13:29:42 Design Page 478three lead compensators for the system of Figure 9.17 that will
479 PROBLEM: Design three lead compensators the for
system the system
characteristicsof Figure 9.17
between that willthe three designs.
Example reduce9.4
PROBLEM: the
E1C09 reduce
Design
settling
11/03/2010
three time
the settling
lead bycompensators
a factor
13:29:42
Lead time byPage
ofCompensator
a factor
for the
2 while
479 of 2Design
system of 30%
maintaining
while maintaining 30% overshoot. Compare
Figure 9.17 that Compare
overshoot. will
ζ = 0.358 j ω the system characteristics between the three designs. 3/27/14  
reduce the settling time by a factor
PROBLEM: theofsystem
2Design SOLUTION:
whilecharacteristics
three Example
maintaining lead 30% First
9.4 determine
overshoot.
between
compensators the for Compare
three the the
designs.
system characteristics
of Figure 9.17 ofthatthewill
Lead Compensator Desired ζDesign
= 0.358 thej ω system characteristics
uncompensated
SOLUTION: between the
First systemthree
determine designs.
operating at
the characteristics 30% overshootof the to
ζ =compensated reduce the settling time bydetermine
a factor of 2 while maintaining
0.358
Desired

j6 SOLUTION: Lead FirstCompensator the characteristics
Design of 30% the overshoot. Compare
ζ Desired
= 0.358
dominant pole see what
uncompensated
the thesystemuncompensated
system
characteristics of settling
operating at 30%
between timeovershoot
is. Since to 30%
EM: Design three lead
compensated
jωcompensators
compensated 478 for theuncompensated
E1C09 SOLUTION:11/03/2010
system
j6 PROBLEM:
of
First
13:29:42
Figure
Chapter 9 9.17
determine
system Design
thethat
Page 479operating
via will
Root atLocus
characteristics 30% overshoot the to the three designs.
Desired
the settling time
–2.014 by adominant
factor of j6ζ2=while
pole 0.358 maintaining 30% overshoot
Design see
overshoot. what
three is
the
lead
Compareequivalent
uncompensated
compensators to fora damping
settling
the system time
ofratio is.
Figure of
Since 0.358,
9.17 30%
that we
will
dominant
compensated
9.3 pole + j5.252
Improving Transient Response
uncompensated
j5via see
jω what
Cascade the system
Compensation
operating
uncompensated SOLUTION:
9.3
at 30%
settling
Improving First overshoot
time is. Since
determine
Transient
toResponse
30%
the479 characteristics
via Cascade of the
Compensation
reduce the settling time
theovershoot by a factor
isthe of
equivalent 2 while maintaining
to aof damping 30% ratioovershoot.
of 0.358, Compare
jω we
is search along 0:358 line forwe the uncompensated
–2.014 j6
the system j5.252 characteristics
+ Desired see what between
the uncompensated three designs.
settling z ¼ is.
time Since 30%
dominant
–2.014 pole
+ j5.252 j5overshoot equivalent to
uncompensateda damping
the system ratio
system
characteristics 0.358,
operating
between atthe 30% three overshoot
designs. to
thesearch a along thefor 0:358 line for the asuncompensated
j5
compensated dominant poles on z¼ the root locus, shown in Figure

–2.014 + j5.252 j4overshoot search j6 isalong
equivalent z¼ to 0:358damping line ratio the ofuncompensated
0.358, we
SOLUTION:
on of lead compensation designs for Exampledominant
j5 First
pole determine
ζ =TABLE 9.4
9.4 along the
0.358 the jω characteristics
Comparison see
of what
lead of thethe
compensationuncompensated designs settling
for Example time 9.4
Desired is. Since 30%
Uncompensated search j4dominant poles ¼ dominant
z9.26. onFrom
0:358 the line poles
SOLUTION:
the
for the
Uncompensated:
root locus, on
pole’s the
First
uncompensated
as root
real
shown locus,
determine
part,
in aFigure weas shown
the calculate in Figure
characteristics
pole the of un-
the
uncompensated
–2.014 j4 system
Desired operating at 30% overshoot
overshoot isto equivalent to damping ratio of 0.358,
s-planetowe
dominant pole + j5.252compensated
Uncompensated
j3dominant 9.26. j5 poles
From on 9.26.
the
compensated
the root From
Uncompensated
pole’s real
uncompensated
locus, the aspole’s
settling
part, shown
we
system
real
in
time
calculate part,
Figure
Compensation asoperating
the we
T un- ¼ a at 30%
calculate
4=1:007 overshoot
location the¼un- 3:972
Compensation b
ncompensated
j6 Uncompensated Compensation
j4
see what the a
uncompensated s-planeCompensation
settling time
j6 bsearch
is. Sinceseealong
9.3 30%
what Compensation
theuncompensated
Improving
the z ¼ 0:358 linesettling
Transient c s for the
Response time uncompensated
via Cascade
is. Since Compensation
30%
–1.007 +dominant
dominant pole
Uncompensated j2.627j3 poledominant pole j3
9.26. From the
s-plane pole’s compensated
seconds. real part,
The we settling
calculate
remaining time the as
un- T ¼ 4=1:007 ¼ 3:972
compensated settling time as T sis¼equivalent thecharacteristics
4=1:007 ¼ a3:972 ofofthe un-
s-plane s
Kovershoot
–1.007 is –2.014
equivalent
+ j2.627 + j5.252 to j4 a damping ratio dominant
of
K 0.358,
overshoot poles
we on Kðsroottoþ 5Þ locus,
damping asratioshown in Figure
0.358, we þ 4Þ
Kðs
Kj5 dominant pole+ j2.627
–1.007 = 63.21
j3Kðs þ 5Þ s-plane j2compensated
Plant seconds.
and settling
Kðs seconds.

j5 þ remaining
The compensatedtime as The T ¼remaining
4=1:007
characteristics
system
s Kðs
are ¼þ characteristics
of3:972

summarizedthe un- in of
Table the un-
9.4.
search along the
Uncompensated
K = 63.21 z ¼ 0:358 TABLEline 9.4
for Comparison
the 9.26.
uncompensated of lead
search
From θccompensation
along
the the
pole’s z ¼ designs
0:358
real line
part, for
θc 6Þðs þ 42:96Þ forExample
we
θc the calculate9.4
uncompensated the un-
s þ 4Þðs–1.007þ 6Þ+ j2.627
K = 63.21
sðs þ 4Þðs þdominant 6Þðs
j2 þ pole 42:96Þ j2
seconds. sðs j3þThe
compensated
compensator 4Þðs remaining
þ 6Þðs sðs
compensated
system þ þ
Next
4Þðs
20:09Þ
are þ 6Þsðs
characteristics system
summarized
dominant
we
compensated find þ sðsof
4Þðs
poles
the
areþþthe
in on
4Þðs
summarized
6Þðs
Table
design
settling the
þun-
timeþ 8:971Þ
9.4.
root
point. locus,
as
in as
A
T
Table
shown
twofold
sðs
4=1:007
9.4.þ 4Þðs þ
in Figure
reduc- 3:972
6Þðsσ þ 20
j4 K = 63.21 dominantj2 FIGURE poles9.25 on theofroot
Three the locus, pc as shown
s-plane
j4
Next in Figure
Required
Uncompensated we From find zc
compensated pc zc
performance
Compensation pc zcs ¼
a3 calculate
¼
Compensation b
j1compensated 1 system arethe summarized9.26. in the
Table
1
the design
2 9.4.
pole’s point.
real 2
part, Awe
3 twofold reduc-
9.26. From –1.007
infinitethe
j1
Uncompensated
+possible
j2.627
pole’sleadj1real part,
Dominant
dominant110.98°
pole
Next
poles
we
we tionfind in
calculate
!1:007
design
settling
seconds.
' the un-
j2:627
point.
time
The A
!2:014
twofold
yields
remaining ' Tj5:252

reduc-
¼ 3:972=2
characteristics ¼ 1:986 of the
!2:014 the'
un-un-
sec- j5:252
1:007 ' j2:627 !2:014 ' j5:252 K =
compensator63.21 solutionstion Next
!2:014 we
j2 in as
110.98° find thetion
' j5:252timecompensated
settling
j3 s-planein
design settling
yields point.
compensated
T !2:014time
A twofold
s ¼ 3:972=2
yields
settling
' j5:252 reduc-
¼areT
1:986time
sKðs þ 3:972=2
as
sec- T
5Þ the s ¼¼ 1:986
4=1:007 sec-
¼ 3:972
Kðs þ 4Þ
j3 s-plane compensated
j1 110.98°settling
–1.007 time
+ j2.627 T onds,
s ¼63.21 from
4=1:007 ¼Kwhich
3:972
seconds.
system
the real summarized
part of in Table
desired 9.4.
thepole
110.98° tion
K in
onds, settling
Plant and
σ
from time
which yields
onds, the Treal
from s 3:972=2
¼ which
part The
of the
1423
¼the 1:986
remaining
real sec-
part
desired characteristics
of
pole the desired of pole
698.1 un-
.21 1423 seconds. The Kσ= 63.21
remaining 698.1 σ location
characteristics sðs
of þ 4Þðs
Next
the þ345.6

un- we find sðs theþ 4Þðs
design þ 6Þðspoint.
þ 42:96Þ
The A twofoldsðs
imaginary9.4. þ 6Þðs
þ 4Þðs
reduc- þ 20
7 –6 –5 – 4 – – 3– – 2– –
30 –12 σ –1 onds,
0 fromj1compensator
which j2
intersecting
the real
location theis real
part
is #zv
compensated
axis
of#2:014.
#zv ¼the
at
nthe
n ¼0.358
#4=T
system s ¼
compensator
desired
#4=T s ¼ pole #2:014.
are#2:014.
summarized
pole The and inimaginary
zero,Tableas illustrated in Figure
6 –j25 ––74 –– 63 –52 compensated
–14 system z 0 location
are0.358 summarized is #zv 0.358
n ¼
inis #4=T
tion
Table ins ¼
9.4.settling timeThe imaginary
yields ! T ¼ 3:972=2 0.358
¼ 1:986 sec-
358 0.358 location is 110.98° 9.25.part
¼We #4=T realize v¼
vs d¼ that an
¼#2:0140.358
Next
!infinite
Thewe find
tanð110:98
number the of
imaginary design
! lead point.
5:252.A twofold
s compensators
ÞÞ¼¼5:252. could reduc-
be used to mee
–5 =
– 4Closed-loop pole
– 3 =–Closed-loop
pole 2
–1 0
pole vn part is v
Dominant #zv j1npoles
¼ #2:014part
2.813 is
!1:007 d#2:014.
tanð110:98 '#2:014
j2:627Þsettling
¼ tanð110:98
5:252.
5.625 !2:014 ' j5:252 !2:014sec- ' j5:252
Closed-loop Next we find the design d the
point. A onds,
twofold
transient ! from
tion
response which
inrequirement.
reduc- the
time real
yields part of
T s ¼ 3:972=2 the ¼5.625
desired
1:986 pole
813
ed-loop =pole
Open-loop pole5.625 pole part is5.625 vd ¼ #2:014 σ We
110.98°
tanð110:98 We continue
Þ ¼5.625
continue 5:252. byby designing
designing thethelead leadcompensator.
compensator.
j1
Open-loop pole– = Open-loop K We continue How by designing
63.21
location
do onds,
the possible the
from
islead lead
which
lead 1423 compensator.
the
¼compensators real part of
differ? The the 698.1
desired
The pole
imaginary
7 – 6 tion – 5 in – 4 settling
–3 – 2time 1 yields
–%OS (
We0(28.2) T s ¼ 3:972=2
continue ¼σ 1:986
30designing
(28)
byArbitrarily the sec- #zv 30compensator
n(30.7) #4=T s ¼ #2:014. 30differences
(28.2) are in the
n-loop
RE
26(28)
9.26pole110.98°
LeadFIGURE Lead
compensator 30design,
9.26compensator
Lead (30.7) –design,
compensator
– 7 showing
6 showing
design,
– 4 –showing
– 5evaluation 3 30
evaluation
–Arbitrarily
2 evaluation
–1
Arbitrarily
0 assume
values of a
static
part
0.358
assume
compensator
error
is assume
location 30¼ (14.5)
constants, acompensator.
isazero
#zv
#2:014 compensator
thenat
0.358 #5
¼gain#4=T
tanð110:98 on s ¼
required the zero
zero
!#2:014.
Þ ¼to at at#5
reach
5:252.The#5 on
the onthethepoint on the
imaginary
design
0.358
onds, pole
= Closed-loop from which Arbitrarily the (
Ts real real part
z
assume of athe desired
compensator
3.972 (4) v pole
dzero at 1.986#5 on
(2) thejustifying 1.986 (2)approximation
Lead
ensated
972 compensator
compensated
(4) and
of uncompensated
and compensatedσ compensated
design, showing
and evaluation
(2)compensated
dominant dominant
poles
= Closed-loop pole poles
dominant
for foraxis
poles for as real real
a possible
compensated axis axis as as
partalocus,
solution.
root possible
ais possible
dUsing
vthe ¼(1.7)
#2:014 solution.
solution.
the
difficulty tanð110:98
root
in Using
Using
locus
!
Þ ¼the the
a5:252. rootroot
second-order locus locus
0 = 1.986 location
Open-loop pole is #zv ¼ #4=T 1.986
v ¼(2)#2:014. The 2.813 We
imaginary 1.986
continue by
5.625 designing the lead compensator.
5.625
ted
.4 and compensated9.4 dominant poles for pole real
n
Tp axis
( n s as a! sum possible
program,
1.196solution.
program, (1.3)sum sum Using
theWe
the the
continue
angles
0.598
angles root bylocus
from
(0.6)
from designing
both
both the
this
this lead
zero zero compensator.
and
0.598
and thethe
(0.6)
#5theon
mple 9.4
Example = Open-loop program, when the
the design
angles is
fromcomplete,
both and
this the ensuing transient response.
196 (1.3)FIGURE 9.26
0.598part
Lead is
(0.6) v ¼ #2:014
compensator tanð110:98
design, 0.598
showing
%OS
program, sum ¼ 5:252.
Þthe
(0.6)
evaluation
angles 30 (28)
from both 0.598
this (0.7)
Arbitrarily
(
30 and
zero
assume (30.7) Arbitrarily
thea lead compensator
aeither
compensator zero at #5 30pole
(28.2)
on or assume
thezero and find azero and
compensator the zero at
FIGURE 9.26 d Lead compensator design,
Kv designing showing For design,
evaluation
2.634 we arbitrarily
6.9select 5.791
of uncompensated
of We
and continue
compensated
uncompensated and by
dominant
T
compensated ( polesthe
for
the
dominant lead
angular
poles for real
3.972 axis
compensator.
real
(4)
contribution as
axis
at a
as
thepossible
a possible
1.986
design solution.
(2)
pointsolution.
of this Using
Using
pole or the
the
zero root
root
1.986
along locus
locus
(2)
with the system’
634 6.9 5.791
s 3.21
Example 9.4 Arbitrarily
showing evaluation Example 9.4 eð1Þ
assume a compensator zero0.380
atprogram,
#5 on sum
the
program, the
0.145
sum angles
the angles from
from both
both this
this zero
0.173
zero and
and !the the
380 0.145 Tp
0.173
( open-loop poles and zeros.
1.196 (1.3)0.312 The difference
0.598 (0.6) between this angle 0.598 (0.6) required
and 180 is the
minant poles for real axis as a possible Othersolution.
poles Using!7:986
theofroot
contribution locus !43.8,
the remaining compensator
!5:134 pole or zero. Let!22:06
us look at an example
Kv 2.634 6.9 5.791
7:986 !43.8, !5:134sum the angles
program, !22:06
from both this zero and !13:3,
the !5 !1:642
Zero eð1Þ None 0.380 0.145 None
0.173
one !5 None
Comments Second-order
!2 Second-order
Other poles !7:986Example 9.4 !43.8, !5:134 Second-order
!22:06
cond-order Second-order Second-order approx. No pole-zero
Zero None OK approx.
!5 OK approx. OK
None
approx. OK approx. OK approx. OK Lead Compensator
( cancellation
Design
Comments
Simulation Second-order
results are shown in parentheses. Second-order Second-order
hown in parentheses. PROBLEM: Designapprox.three
OKlead compensators
approx. OKfor the system of Figure 9.17OK
approx. that will
•  Let’s assume ( reduce
zc = -5 results
uncompensated the settling
system’s time
polesin andby a factor of 2 while maintaining 30% overshoot.
zeros, using the design point as a test point. TheCompare
Simulation are shown parentheses.
the systemjωcharacteristics between the
" " three designs.
em’s poles and zeros, using the design point as a test
resulting angle point. The . TheDesired
is !172:69 difference between this angle and 180 is the angular
" ζ"= 0.358 jω compensated j5.252
contribution
72:69 . The difference between this angle and 180 is therequired
uncompensated angular from thepoles
system’s andSOLUTION:
compensator zeros,poleusingFirst
inthe determine
order to place
design thethe
point characteristics
as adesign ofThe
point
test point. the
on
Desired dominant pole
thecompensated
root locus. Hence, " uncompensated system operating
"
angleisand at 30%
required" overshoot
from theto
j6 an angular see contribution of this
d from the compensator pole in order to place resulting
the design angle
pointison
!172:69 . The difference between !7:31 180 is the angular
dominant pole what the uncompensated
s-plane settling time is. Since 30%
ce, an angular contribution of !7:31" compensator
is–2.014contribution
required pole.
from required
the from the compensator pole in order to place the design point on
7.31° is equivalent to a damping
+ j5.252 overshoot " ratio of 0.358, we
the root locus. Hence, an angular contribution of !7:31 lineisforrequired fromof the
j5
The geometry shown in Figure 9.27along
is used
search the zto calculate
σ
¼ 0:358 thethelocation
uncompensated the
compensator pole. –pc –2.014
compensator
shown in Figure 9.27 is used to calculate the location of the pole. From j4 the figure, dominant poles on the root locus, as shown in Figure
The geometry shown=in
Uncompensated
Closed-loop
Figure pole is used to calculate the location of the
9.27
9.26. pole
From the pole’s real part, we calculate the un-
From the figure, compensator pole.j3 Froms-plane
dominant pole 5:252
= Open-loop
the figure, "
compensated settling time as T s ¼ 4=1:007 ¼ð9:19Þ 3:972
Note:p This
! figure is¼
2:014 nottan
drawn7:31
to scale.
5:252 –1.007 + j2.627
c seconds.
5:252 The remaining characteristics of the un-
¼ tan 7:31" K = 63.21 ð9:19Þ j2 FIGURE 9.27 s-plane
compensated ¼ tan 7:31"are summarized in Table 9.4.
picture
system ð9:19Þ
pc ! 2:014 from which the compensator polepcis!found 2:014 to be
used to calculate
Nextthewelocation
find the design point. A twofold reduc-
j1
mpensator pole is found to be from which the compensator
110.98° tion
pc is
pole in
of the compensator settling
42:96pole
¼ found for yields T s ¼ 3:972=2 ¼ 1:986
to time
be sec-
ð9:20Þ
Example
σ
onds, from
9.4 p ¼ 42:96 which the real part of the desired pole
pc ¼ 42:96 pc –7 –6 –The
5 –3 –2 –1 ð9:20Þ
– 4compensated system
0 root locuslocation
is sketched
c is #zvin
n ¼Figure 9.28.
#4=T s ¼ ð9:20Þ
#2:014. The imaginary
c z
= Closed-loop pole part is v ¼ #2:014 tanð110:98! Þ ¼ 5:252.
d
ystem root locus is sketched in Figure 9.28. The compensated system root locus is sketched in Figure 9.28.
= Open-loop pole We continue by designing the lead compensator.

FIGURE 9.26 Lead compensator design, showing evaluation Arbitrarily assume a compensator zero at #5 on the
of uncompensated and compensated dominant poles for ω as a possible solution. Using the root locus
real jaxis
jω Example 9.4 program, sum the angles from both this zero and the

s-plane
s-plane
s-plane σ
–42.96 –6 –5 –4 0 σ
–42.96 –6 –5 –4 0
σ 2  
–5 –4 0

= Closed-loop pole pole


= Closed-loop
= Open-loop pole pole
= Open-loop
op pole FIGURE
FIGURE 9.28 Compensated
9.28 Compensated
domi
contribution required from the compensator pole in order to place the design point on
the root locus. Hence, an angular contribution of !7:31" is required from the
compensator pole.
The geometry shown in Figure 9.27 is used to calculate the location of the =
compensator pole. From the figure, =

5:252
3/27/14   Note:
¼ tan 7:31" ð9:19Þ FIGU
pc ! 2:014
used
from which the compensator pole is found to be of th
Exam
pc ¼ 42:96 ð9:20Þ
The compensated system root locus is sketched in Figure 9.28.
RL for the lead-compensated system

s-plane

σ
–42.96 –6 –5 –4 0

= Closed-loop pole
= Open-loop pole
FIGURE 9.28 Compensated
Note: This figure is not drawn to scale. system root locus

E1C09 11/03/2010 13:29:42 Page 479

9.3 Improving Transient Response via Cascade Compensation 479

TABLE 9.4 Comparison of lead compensation designs for Example 9.4

Uncompensated Compensation a Compensation b Compensation c

K Kðs þ 5Þ Kðs þ 4Þ Kðs þ 2Þ


Plant and
sðs þ 4Þðs þ 6Þ sðs þ 4Þðs þ 6Þðs þ 42:96Þ sðs þ 4Þðs þ 6Þðs þ 20:09Þ sðs þ 4Þðs þ 6Þðs þ 8:971Þ
compensator
Dominant poles !1:007 ' j2:627 !2:014 ' j5:252 !2:014 ' j5:252 !2:014 ' j5:252
K 63.21 1423 698.1 345.6
z 0.358 0.358 0.358 0.358
vn 2.813 5.625 5.625 5.625
%OS( 30 (28) 30 (30.7) 30 (28.2) 30 (14.5)
Ts( 3.972 (4) 1.986 (2) 1.986 (2) 1.986 (1.7)
Tp( 1.196 (1.3) 0.598 (0.6) 0.598 (0.6) 0.598 (0.7)
Kv 2.634 6.9 5.791 3.21
eð1Þ 0.380 0.145 0.173 0.312
Other poles !7:986 !43.8, !5:134 !22:06 !13:3, !1:642
Zero None !5 None !2
Comments Second-order Second-order Second-order No pole-zero
approx. OK approx. OK approx. OK cancellation
(
Simulation results are shown in parentheses.

uncompensated system’s poles and zeros, using the design point as a test point. The Desired jω
compensated j5.252
resulting angle is !172:69". The difference between this angle and 180" is the angular dominant pole
contribution required from the compensator pole in order to place the design point on
s-plane
the root locus. Hence, an angular contribution of !7:31" is required from the 7.31°
compensator pole. σ
–pc –2.014
The geometry shown in Figure 9.27 is used to calculate the location of the = Closed-loop pole
compensator pole. From the figure, = Open-loop pole
Note: This figure is not drawn to scale.
5:252
¼ tan 7:31" ð9:19Þ FIGURE 9.27 s-plane picture
pc ! 2:014
used to calculate the location
from which the compensator pole is found to be of the compensator pole for
pc ¼ 42:96 ð9:20Þ
Example 9.4 3  
The compensated system root locus is sketched in Figure 9.28.

s-plane
E1C09 11/03/2010 13:29:43 Page 481
3/27/14  

E1C09 11/03/2010 13:29:43 Page 480

Chapter 9 Design via Root Locus


9.3 Improving Transient Response via Cascade Compensation
480

In orderCompensation
to justify oura,estimates
b of percent overshoot and settling time, we
must show1.4 that the Compensation
second-order c
approximation is valid. To perform this validity
check, we search for the third and fourthUncompensated
closed-loop poles found beyond !42:96
1.2 !5 and !6 in Figure 9.28. Searching these regions for the gain equal to
and between
that of the compensated dominant pole, 1423, we find that the third and fourth
poles are1.0at !43:8 and !5:134, respectively. Since !43:8 is more than 20 times the
real part0.8
of the dominant pole, the effect of the third closed-loop pole is negligible.
c(t)

Since the closed-loop pole at !5:134 is close to the zero at !5, we have pole-zero
cancellation,
0.6 and the second-order approximation is valid.
All results for this design and two other designs, which place the compensator
0.4
zero arbitrarily at !2 and !4 and follow similar design techniques, are summarized
in Table 9.4. Each design should be verified by a simulation, which could consist of
0.2
using MATLAB (discussed at the end of this example) or the state-space model
and the step-response
0 program discussed in Appendix H.1 at www.wiley.com/ FIGURE 9.29 Uncompensated
college/nise.0 We have1performed2a simulation 3 for this 4design problem,systemand theand lead compensation
results are shown by parenthetical entries
Time next to the estimated values in responses
(seconds) the table. for Example 9.4
The only design that disagrees with the simulation is the case where the compen-
sator zero is at !2. For this case the closed-loop pole and zero do not cancel.
A sketch of the root locus, which you should generate, shows why the effect of
the zero
will is pronounced,
plot the causing
root thelocus
response of to bethe
different from that predicted. system and the
uncompensated
Placing the zero to the right of the pole at !4
percent overshoot line. You will interactively creates a portion of the root locus that
select the gain,
is between the origin and the zero. In other words, there is a closed-loop pole closer
to the origin than the dominant poles, with little chance of pole-zero cancellation characteris-
after which MATLAB will display the performance
tics
except at highof the
gain. Thus,uncompensated
a quick sketch of the root system and
locus gives plot its
us information fromstep response.
which we can make better design decisions.
Using these characteristics,you will input the For this example, we want to place the desired set-
zero on, or to the left of, the pole at !4, which gives a better chance for pole-zero
tling time and a zero value for the lead compensator.You will
cancellation and for a higher-order pole that is to the left of the dominant poles and
then interactively
subsequently faster. This is verified byselect
the fact thata our
value
results for thesecond-
show good compensator pole.
orderMATLAB
approximationswillforrespond with
the cases where a was
the zero rootplacedlocus.You can then continue
at !4 and !5. Again,
decisions about wherepole
selecting to place the zero are
values based on
until thesimple
rootrules locus
of thumb goes
and through the
must be verified by simulations at the end of the design.
desired point.MATLAB will display the lead compensator,enu-
Let us now summarize the results shown in Table 9.4. First we notice
merate its performance characteristics,and plot a step re-
differences in the following:
sponse.This
1. The exercise
position of the arbitrarily solves
selected zero Example 9.4 using MATLAB.
2. The amount of improvement in the steady-state error
3. The amount of required gain, K
4. The position of the third and fourth poles and their relative effect upon the
second-order approximation. This effect is measured by their distance from the
dominant poles or the degree of cancellation with the closed-loop zero.
Skill-Assessment Exercise 9.2
Once a simulation verifies desired performance, the choice of compensation
can be based upon the amount of gain required or the improvement in steady-state
PROBLEM: A unity feedback system with the forward transfer
error that can be obtained without a lag compensator.
function
The results of Table 9.4 are supported by simulations of the step response,
shown in Figure 9.29 for the uncompensated system
K
GðsÞ and
¼ the three lead compensa-
tion solutions. sðs þ 7Þ
Differences in Performance for Different Designs
is operating
Students who arewith
usingaMATLAB
closed-loop
shouldstep response
now run that
ch9p2 in has 15%
Appendix B. overshoot. Do the
MATLAB will be used to design a lead compensator. You will input
following:
the desired percent overshoot from the keyboard. MATLAB
a. Evaluate the settling time.
b. Design a lead compensator to decrease the settling time by three times.
Choose the compensator’s zero to be at %10.
4  
ANSWERS:
a. T s ¼ 1:143 s
s þ 10
b. Glead ðsÞ ¼ ; K ¼ 476:3
s þ 25:52
The complete solution is at www.wiley.com/college/nise.
after which
0.6 MATLAB will display the performance characteris-
tics of the uncompensated system and plot its step response.
0.4
Using these characteristics,you will input the desired set-
tling time and a zero value for the lead compensator.You will
0.2
then interactively select a value for the compensator pole.
MATLAB will
0 FIGURE 9.29 Uncompensated
respond with a root locus.You can then continue
selecting0 pole values
1 2
until 3
the root system
4 goes through
locus theand lead compensation
Timedisplay
desired point.MATLAB will responses for Example3/27/14  
(seconds) the lead compensator,enu- 9.4
merate its performance characteristics,and plot a step re-
sponse.This exercise solves Example 9.4 using MATLAB.

will plot the root locus of the uncompensated system and the
percent overshoot line. You will interactively select the gain,
after which MATLAB will display the performance characteris-
tics of the uncompensated Skill-Assessment
system and Exercise
plot9.2 its step response.
UsingA these
PROBLEM: characteristics,you
unity feedback willfunction
system with the forward transfer input the desired set-
tling time and a zero value for the lead compensator.You will
K
then interactively GðsÞselect
¼ a value for the compensator pole.
sðs þ 7Þ
MATLAB will respond with a root locus.You can then continue
is operating with a closed-loop
selecting pole values step response
until thatthe
has 15%
rootovershoot.
locusDogoesthe through the
following:
desired point.MATLAB will display the lead compensator,enu-
a. merate its
Evaluate the performance
settling time. characteristics,and plot a step re-
b. sponse.This exercise
Design a lead compensator solves
to decrease theExample 9.4
settling time using
by three MATLAB.
times.
Choose the compensator’s zero to be at %10.

ANSWERS:
a. T s ¼ 1:143 s
s þ 10
b. Glead ðsÞ ¼ ; K ¼ 476:3
s þ 25:52 Skill-Assessment Exercise 9.2
The complete solution is at www.wiley.com/college/nise.
Lead Compensator Design: Example 2
PROBLEM: A unity feedback system with the forward transfer function
K
GðsÞ ¼
sðs þ 7Þ

is operating with a closed-loop step response that has 15% overshoot. Do the
following:
a. Evaluate the settling time.
b. Design a lead compensator to decrease the settling time by three times.
Choose the compensator’s zero to be at %10.

ANSWERS:
a. T s ¼ 1:143 s
s þ 10
b. Glead ðsÞ ¼ ; K ¼ 476:3
s þ 25:52
The complete solution is at www.wiley.com/college/nise.

Answers

5  
9.4 Improving Steady-State Error and Transient Response

Lag-Lead Compensator Design


In the previous example, we serially combined the concepts of ideal derivative and
mproving Steady-State Errorideal
and integral
Transient Response to arrive at the design of a 487 3/27/14  
compensation PID controller that improved
both the transient response and the steady-state error performance. In the next
esign example, we improve both transient response and the steady-state error by using a
combined the concepts lead
of ideal derivative
compensator andand
a lag compensator rather than the ideal PID. Our compensa-
tor is called
e at the design of a PID controller thata improved
lag-lead compensator.
We first
e steady-state error performance. Indesign the lead compensator to improve the transient response. Next
the next
we evaluate
response and the steady-state errorthe
byimprovement
using a in steady-state error still required. Finally, we design
the lag compensator
sator rather than the ideal PID. Our compensa- to meet the steady-state error requirement. Later in the chapter
we show circuit designs for the passive network. The following steps summarize the
.
design procedure:
ensator to improve the transient response. Next
1. Evaluate
ady-state error still required. Finally,the
weperformance
design of the uncompensated system to determine how much
dy-state error requirement.improvement in transient response is required.
Later in the chapter
ve network. The following steps summarize
2. Design the
the lead compensator to meet the transient response specifications. The
design includes the zero location, pole location, and the loop gain.

uncompensated system to3. determine


Simulate the system to be sure all requirements have been met.
how much
se is required. 4.
LEAD-LAG COMPENSATOR
Redesign
5. Evaluate
meet the transient response
if the simulation shows that requirements have not been met.
the steady-state
specifications. The error performance for the lead-compensated system to
determine how much more improvement in steady-state error is required.
, pole location, and the loop gain.
6. Design the lag compensator to yield the required steady-state error.
ll requirements have been met.
7. Simulate the system to be sure all requirements have been met.
s that requirements have not been met.
8. Redesign if the simulation shows that requirements have not been met.
erformance for the lead-compensated system to
ovement in steady-state error is required.
ield the required steady-state error.
Example 9.6
ll requirements have been met.
s that requirements have not been met. Lag-Lead Compensator Design
PROBLEM: Design a lag-lead compensator for the system of Fig-
R(s) + E(s) K
ure 9.37 so that the system will operate with 20% overshoot and a
twofold reduction in settling time. Further, the compensated system – s(s + 6)(s +
Example 9.6
will exhibit a tenfold improvement in steady-state error for a ramp
input.
Lag-Lead Compensator Design FIGURE 9.37 Uncompensa
SOLUTION: Again, our solution follows the steps just described. Example 9.6
mpensator for the system of Fig-
erate with 20% overshootStepand1a First
R(s)we
+ evaluate
E(s) the performance
K of the uncompensated
C(s) sys-
tem. Searching alongs(sthe 20%
+ 6)(s overshoot line ðz ¼ 0:456Þ in
+ 10)
Further, the compensated system –
Figure 9.38, we find the dominant poles at $1:794 % j3:501, with a gain of 192.1.
t in steady-state error for a ramp
The performance of the uncompensated system is summarized in Table 9.6.
Step 2 Next we begin
FIGURE 9.37 the lead compensator
Uncompensated design
system for by selecting the location of the
ollows the steps just described. compensated
Example 9.6 system’s dominant poles. In order to realize a twofold reduction
in settling time, the real part of the dominant pole must be increased by a factor
rmance of the uncompensated Lead-Lag Compensator
sys- of 2, since Design:
the settling time Example
is inversely proportional to the real part. Thus,
0% overshoot line ðz ¼ 0:456Þ in
$zvn ¼ $2ð1:794Þ ¼ $3:588 ð9:29Þ
inant poles at $1:794 % j3:501, with a gain of 192.1.
compensated system is summarizedThe imaginary part of the design point is
in Table 9.6.
mpensator design by selecting the location of the vd ¼ zvn tan 117:13& ¼ 3:588 tan 117:13& ¼ 7:003 ð9:30Þ
inant poles. In order to realize a twofold reduction
of the dominant pole must be increased by a factor 6  
is inversely proportional to the real part. Thus,
vn ¼ $2ð1:794Þ ¼ $3:588 ð9:29Þ
design point is
& &
2. Design the lead compensator
4. Redesign to meet
if thechoice
simulation the transient
shows
will eliminate
that response
requirements
a zero and leave specifications.
have not been met.
the lead-compensated systemThe
with
design includes
5. the zerothelocation,
Evaluate pole
steady-state
three poles, the location,
error numberand
performance
same theuncompensated
for the
that the loop gain. system
lead-compensated system
has. to
determine how muchWe
more improvement
complete the designin
bysteady-state error isofrequired.
finding the location the compensator
3. Simulate the system
6. Designto
thebe
lagsure
pole. all requirements
Using
compensator the to
root locusthe
yield have
program, been
sum
required met.
the angles
steady-stateto the design point from
error.
the uncompensated system’s poles and zeros and the compensator zero
4. Redesign if the simulation
7. Simulate shows
the system that
to !164:65
and get requirements
be sure" all requirements have
have not
. The difference between 180" been
been and thismet.
met. quantity is the
8. Redesign if theangular
simulation shows that
contribution requirements
required from the have not been
compensator met.or !15:35".
pole,
5. Evaluate the steady-state error performance for the lead-compensated
Using the geometry shown in Figure 9.39,
system to3/27/14  
determine how much more improvement in steady-state 7:003
error is required.
¼ tan 15:35" ð9:31Þ
6. Design the lag compensator to yield the required steady-state
Example 9.6error.
pc ! 3:588
from which the location of the compensator pole, pc, is found to be !29:1.
7. Simulate the system to be sure all requirements Lag-LeadhaveCompensator
been met. Design
TABLE 9.6 Predicted characteristics of uncompensated, lead-compensated, and lag-lead-compensated systems of
09 11/03/2010 13:29:45 8. Redesign
Page 488 Example 9.6if the simulation shows that requirements have not been met.
PROBLEM: Design a lag-lead compensator for the system of Fig-
ure 9.37 Uncompensated
so that the system will operate with 20% overshoot and
Lead-compensated a R(s) +
Lag-lead-compensated
E(s) K
twofold reduction in settlingRL
time.for s(s + 6)(s + 10)
488 uncompensated
Further, the compensatedsystemKðs þ 0:04713Þ
system –
will exhibit a Ktenfold improvement in steady-state
Plant and compensator
K error for a ramp
sðs þ 6Þðs þ 10Þ sðs þ 10Þðs þ 29:1Þ sðs þ 10Þðs þ 29:1Þðs þ 0:01Þ
input.
488
Dominant poles
Chapter 9 Design
!1:794 ' j3:501
via Root Locus
Example 9.6!3:574 ' j6:976
!3:588 ' j7:003 FIGURE 9.37 Uncompensated system
K 192.1 Again, our solution follows
SOLUTION: 1977 the steps just described.
1971 Example 9.6

Chapter 9
z
Design via Root vLocus
Step 1 0.456 Lag-Lead Compensator Design
0.456
First we evaluate the performance of the uncompensated 0.456sys-
ω ¼ 0:456Þ
7.869 overshoot line j ðz
n tem. Searching along the 20%
3.934 7.838 in
20 ζ =poles
%OS
PROBLEM: Design a lag-lead
The
compensator
performance of the
for 0.456
20Figure 9.38, we find the dominant
theatsystem
uncompensated
$1:794 % j3:501,
system is
with a gain of 192.1.
of 20Fig-
j4 summarized in Table
Ts
ure 9.37 so that the
2.230
system jω
will operate
1.115
with 20% overshoot
1.119 R(s) + 9.6. E(s) K
–1.794 + j3.501 Uncompensatedand a pole
dominant
Tp Step 2 0.897
Next we begin the lead compensator
0.449 design by selecting
0.450the location of the
twofold ζ = 0.456
reduction in settling time.system’s
Further, Kthe compensated
= 192.1 j3 system – s(s + 6)(s + 10
Kv compensated
3.202 j4 dominant
6.794 poles. In order to realize
31.92a twofold reduction
s-plane
will exhibit
eð1Þ a tenfold improvement
in settling
0.312
–1.794 + j3.501
time, in
the steady-state
real part of
0.147
Uncompensated dominant pole
the error
dominant for
pole a
must
j2 ramp
be increased
0.0313 by a factor
Third pole of 2, since the settling time!31:92
is inversely proportional to the real part. Thus,
input. !12:41
K = 192.1 j3 !31:91, !0:0474
Zero None $zv
s-plane n ¼ $2ð1:794Þ ¼ $3:588
None j1 !0:04713 FIGURE ð9:29Þ
9.37 Uncompensate
Comments The imaginary
Second-order approx.j2 OKof theSecond-order
part
SOLUTION: Again, our solution follows the steps just described. design point is OK 117.13°
approx. Second-order approx. OK
Example 9.6
& &
σ
–7 v –6d ¼ –5 n tan
j1 zv –4 117:13 –2 3:588
¼ –1 tan 117:13 ¼ 7:003 ð9:30Þ
Step 1 First–11we–10 –9 –8
evaluate the performance –3
117.13°of the uncompensated sys-
0
FIGURE 9.38 Root locus for –12.41
tem. Searching
= Closed-loop pole
along the 20% overshootσ line ðz ¼ 0:456Þ in
uncompensated system of
–11 –10 –9
Example 9.6
–8 –7 –6 Figure
–5 9.38, pole
–4
= Open-loop we find the dominant poles at $1:794 % j3:501, with a gain of 192.1.
–3 –2 –1 0
or –12.41 The performance of the uncompensated system is summarized in Table 9.6.
= Closed-loop pole Step 1: Analysing the uncompensated
Now we design the lead compensator.system
Arbitrarily select a location
= Open-loop pole Step 2 Nextforwe thebegin the lead compensator
lead compensator design by
zero. For this example, weselecting
select the the location
location of of the
compensated system’s
the compensator zerodominant
coincidentpoles.
with In
theorder to realize
open-loop a twofold
pole at !6. Thisreduction
Now we design the leadchoice
compensator. Arbitrarily
will eliminate apart select
zeroof
and a location
in settling
for the lead compensator zero. For
time,
this
the real
example, we select
theleave
the
the lead-compensated
dominant
location of
system withby a factor
pole must be increased
three poles, the same number that the uncompensated system has.
of 2, since
the compensator zero coincident We
the
with settling
the time pole
open-loop
complete
is by
the design
inversely
at proportional to the real part. Thus,
!6. This
finding the location of the compensator
choice will eliminate a zero andpole.leave
Usingthethelead-compensated
root locus program, system
sum thewith
angles to the design point from
$zv n ¼ $2ð1:794Þ ¼ $3:588 ð9:29Þ
•  Ts the
three poles, the same number = 2.230
that the suncompensated
uncompensated system has.
system’s poles and zeros and the compensator zero
The and
We complete the design imaginary
by finding part
the
get !164:65 " of the
location
. The of design point is180" and this quantity is the
the compensator
difference between
• 
E1C09 11/03/2010 Tsc = Ts/2
13:29:45
pole. Using the root locus program,
=
Page
sum
2.230/2
489 = 1.115
the angles torequired
the design
è Re-part = -3.587
&point
thefrom "
angular contribution from compensator pole, & or !15:35 .
the uncompensated system’s•  Im-part
poles è
Using the v ¼
andgeometry
zeros
d zv tan
andshown
n 117:13
the compensator ¼ 3:588
in Figure 9.39, zero tan 117:13 ¼ 7:003 ð9:30Þ
and get !164:65" . The • difference between 180 "
Select compensator zero at7:003and this quantity is
s = -6 (coincidethe with OL pole)
"
angular contribution required from the compensator pole, or ¼ !15:35
tan 15:35. " ð9:31Þ
Using the geometry shown in Figure 9.39, p c ! 3:588
9.4 Improving Steady-State Error and Transient Response 489
from which the location of the compensator pole, pc, is found to be !29:1.
7:003 jω
¼ tan 15:35" pc = -29.1ð9:31Þ
3:588
TABLE 9.6 Predicted characteristicspcof!uncompensated, lead-compensated, and lag-lead-compensated systems of
Example 9.6
from which the location of the compensator pole, pc, is found to be !29:1.
Uncompensated Lead-compensated Lag-lead-compensated
j7.003
racteristics of uncompensated, lead-compensated, and lag-lead-compensated systems of
s-plane
K K Kðs þ 0:04713Þ
Plant and compensator 15.35°
sðs þ 6Þðs þ 10Þ sðs þ 10Þðs þ 29:1Þ sðs þ 10Þðs þ 29:1Þðs þ 0:01Þ
Uncompensated Lead-compensated Lag-lead-compensated
Dominant poles !1:794 ' j3:501 !3:588 ' j7:003 !3:574 ' j6:976 σ
–pc – 3.588
K K 192.1 K 1977 Kðs þ 0:04713Þ 1971
= Closed-loop pole
sðs þ 6Þðs þ 10Þ sðs þ 10Þðs þ 29:1Þ sðs þ 10Þðspole
= Open-loop þ 29:1Þðs þ 0:01Þ
z 0.456 0.456 0.456
!1:794 ' j3:501 ' j7:003
!3:588 FIGURE 9.39 Evaluating the!3:574 ' j6:976
compensator pole for Example 9.6
vn 3.934 7.869 7.838
192.1 1977 1971
%OS 20 20 20
0.456 Step
0.456 2: Identifying Design
The complete Requirement
root
0.456 & Design Lead
locus for the lead-compensated systemCompensator
is sketched
Ts 2.230 in Figure 9.40. The1.115
gain setting at the design point is 1.119
found to be 1977.
3.934 7.869 7.838
Tp 0.897 Steps 3 and 4 Check the 0.449 design with a simulation. (The 0.450 result for the lead-
20 K 20
3.202 compensated 20
system
6.794is shown in Figure 9.42 and is31.92 satisfactory.)
v
2.230eð1Þ 1.115 1.119
0.312 Step 5 Continue by designing
0.147the lag compensator to improve the steady-state
0.0313
0.897Third pole 0.449 error. Since the uncompensated
0.450 system’s open-loop transfer function is
!12:41 !31:92 !31:91, !0:0474
3.202Zero 6.794 31.92 192:1
None None GðsÞ ¼
sðs þ 6Þðs þ 10Þ
!0:04713 ð9:32Þ 7  
0.312Comments 0.147
Second-order approx. OK 0.0313
Second-order approx. OK Second-order approx. OK
!12:41 !31:92 the static error constant,
!31:91,Kv,!0:0474
which is inversely proportional to the steady-
None None state error, is 3.201. Since the open-loop transfer function of the lead-
!0:04713
compensated system is
Second-order approx. OK Second-order approx. OK Second-order approx. OK
1977
GLC ðsÞ ¼ ð9:33Þ
192:1
GðsÞ ¼ ð9:32Þ
sðs þ 6Þðs þ 10Þ

the static error constant, Kv, which is inversely proportional to the steady-
state error, is 3.201. Since the open-loop transfer function of the lead-
compensated system is
1977
GLC ðsÞ ¼ ð9:33Þ 3/27/14  
sðs þ 10Þðs þ 29:1Þ

the static error constant, Kv, which is inversely proportional to the steady-
state error, is 6.794. Thus, the addition of lead compensation has improved
the steady-state error by a factor of 2.122. Since the requirements of the
problem specified a tenfold improvement, the lag compensator must be
9 11/03/2010 13:29:45 Page 489 designed to improve the steady-state error by a factor of 4.713 ð10=2:122 ¼
4:713Þ over the lead-compensated system.
RL for lead-compensated system: jω
ζ = 0.456
3/2010 13:29:45 Page 489 s-plane
j9
9.4 Improving Steady-State–3.588
Error and Transient
+ j7.003 Response
Compensated dominant pole 489
K = 1977 j6

jω j3
117.13°
σ
9.4
–33 Improving
–30 –27 –24Steady-State
–21 –18 –15Error
–12 and
–9 Transient
–6 –3 Response
0 489
–31.91
= Closed-loop pole jω j7.003
= Open-loop pole
s-plane
FIGURE 9.40 Root locus for lead-compensated system of Example 9.6
15.35°
j7.003 σ
–pc – 3.588
s-plane
= Closed-loop pole
Step 3: Check results via simulation
= Open-loop pole
15.35°
FIGURE 9.39 Evaluating the compensator pole for Example 9.6
σ
–pc – 3.588
The complete root locus
= Closed-loop pole for the lead-compensated system is sketched
= Open-loop pole
in Figure 9.40. The gain setting at the design point is found to be 1977.
9.39 3 Evaluating
FIGURESteps and 4 Checkthe compensator
the designpole
withfora Example 9.6 (The result for the lead-
simulation.
compensated system is shown in Figure 9.42 and is satisfactory.)
Step 5 The
Continue
completeby root
designing
locus the lag lead-compensated
for the compensator to improve
systemthe steady-state
is sketched
OL transfer function of uncompensated For Type 1 system & ramp input:
error.
in Figure Since
9.40. Thethe uncompensated
gain setting at the system’s open-loop
design point transfer
is found to befunction
1977. is
system Kz1z2 …z3
Steps 3 and 4 Check the design with a simulation. 192:1(The result K for lim sG(s)
v =the lead- =
GðsÞ
compensated system is shown in Figure ¼ s→0
ð9:32Þ p1 p2 … p3
sðs þ9.42
6Þðs and
þ 10Þis satisfactory.)
Step 5 Continue by designing the lag compensator to improve1 the steady-state
error. the static
Since theerror constant, Kv, system’s
uncompensated which is inversely eproportional
open-loop ss =transfer to the=steady-
= 1 /function
3.202 0.313
is
state error, is 3.201. Since the open-loop transferKfunction v of the lead-
compensated TFsystem
of leadiscompensator192:1
GðsÞ ¼ ð9:32Þ
sðs þ 6Þðs þ1977
10Þ !(! + 6)
GLC ðsÞ ¼ !!"#$ ! = ! ð9:33Þ
(! + 29.1)
sðs þ 10Þðsproportional
þ 29:1Þ
the static error constant, Kv, which is inversely to the steady-
state error, is 3.201. Since the open-loop transfer function of the lead-
the static error constant, Kv, which is inversely proportional to the steady-
compensated system
state error, TF is the
of
is 6.794. lead-compensated
Thus, system:
the addition of lead compensation has improved
the steady-state error by a factor of 2.122. Since the requirements of the
1977
GaLCtenfold
ðsÞ ¼ improvement, the lag compensator 1
problem specified
sðs þ 10Þðs þ 29:1Þ
ess = ð9:33Þ 1 / 6.794
=must be = 0.1471
designed to improve the steady-state error by a factor of 4.713 ð10=2:122 K v ¼
4:713Þ
the static over
error the lead-compensated
constant, K , which is system.
inversely proportional to the steady-
The lead
v compensator has improved ess by 0.313/0.1471 =2.122
state error, is 6.794. Thus, the addition of lead compensation j ω has improved
the steady-state error Step 4: Design
by a factor LagSince
of 2.122. Compensator
ζ = 0.456
the requirements of the
problem specified a tenfold improvement, the lag compensator j9
must be
s-plane
designed to improve the steady-state error by a factor of 4.713 ð10=2:122 ¼
–3.588 + j7.003 Compensated dominant pole
4:713Þ over the lead-compensated system. K = 1977 j6

jω j3
ζ = 0.456 117.13° 8  
s-plane σ
j9
–33 –30 –27 –24 –21 –18 –15 –12 –9 –6 –3 0
–31.91 –3.588 + j7.003 Compensated dominant pole
K = 1977 j6
= Closed-loop pole
= Open-loop pole j3
117.13°
117.13°
σ
σ
–29.1 –10 0.04713 0
–29.1 –10 0.04713 0
3/27/14  
–31.91 –31.91 –0.0474
–0.0474 –0.01
–0.01

= Closed-loop pole pole


= Closed-loop
= Open-loop
= Open-loop pole pole
Note:figure
Note: This This figure
is notisdrawn
not drawn to scale.
to scale.
Root
FIGURE 9.41 •  locus
Need for lag-lead-compensated
tenfold reduction in esssystem è theof Example 9.6
lag compensator
FIGURE 9.41 Root locus for lag-lead-compensated system of Example 9.6
must be designed to improve the ess by
Step 6 10/2.122
We arbitrarily choose = 4.713
the lag compensator pole at 0.01, which then places
Step 6 We arbitrarily
the •  choose the
lagArbitrarily
compensator lag compensator
choose
zero the0.04713,
at pole at 0.01,
lag compensator
yielding pole which
at 0.01,then places
the lag compensator zero at 0.04713, yielding
which places the lag compensator zero at 0.04713,
ðs þ 0:04713Þ
yeilding Glag ðsÞ ¼ ð9:34Þ
ðs þðs0:04713Þ
þ 0:01Þ
Glag ðsÞ ¼ ð9:34Þ
ðs þ 0:01Þ
as the lag compensator. The lag-lead-compensated system’s open-loop
•  Therefore,
transfer function isthe lead-lag-compensated system’s OL transfer
as the lag compensator. The lag-lead-compensated system’s open-loop
function
transfer function is Kðs þ 0:04713Þ
GLLC ðsÞ ¼ ð9:35Þ
sðs þ 10Þðs þ 29:1Þðs þ 0:01Þ
Kðs þ 0:04713Þ
G LLC ðsÞ ¼
where the uncompensated
ð9:35Þ
sðssystem pole
þ 10Þðs at %6 canceled
þ 29:1Þðs þ 0:01Þthe lead compen-
sator zero at %6. By drawing the complete root locus for the lag-lead-
where compensated system and
the uncompensated by searching
system pole atalong the 0.456 damping
%6 canceled the leadratio line,
compen-
we find
sator zero at the
%6.dominant,
By drawing closed-loop poles to root
the complete be at locus & j6:976,
%3:574for with a
the lag-lead-
gain of 1971.
compensated Theand
system lag-lead-compensated
by searching along rootthe
locus is shown
0.456 in Figure
damping ratio9.41.
line,
11/03/2010 13:29:46 Page 490 A summary of our design is shown in Table 9.6. Notice that the
we find the dominant, closed-loop poles to be at %3:574 & j6:976, with a
lag-lead compensation has indeed increased the speed of the system, as
gain ofwitnessed
1971. Thebylag-lead-compensated
the settling time or therootpeaklocus
time.isThe
shown in Figureerror
steady-state 9.41.
A summary of our design is shown in Table 9.6.
for a ramp input has also decreased by about 10 times, as seen Notice that the
lag-lead compensation
from eð1Þ. has indeed increased the speed of the system, as
490 witnessed
Step by the
Chapter 9
7 The final settling
of our time
Design
proof
via Root oristhe
Locus
designs peak
shown by time. The steady-state
the simulations of Figures error
9.42
for a and
ramp 9.43. The improvement in the transient response isjωshown inas
input has also decreased by about 10 times, seen
Figure
from eð1Þ.
9.42, where we see the peak time occurring sooner in the lag-lead-
Step 7 compensated
The final system.
proof of our Improvement
designs is shown in
bythe
thesteady-state
simulations error for a ramp
of Figures 9.42
ζ = 0.456 each step of our design yields more
input is seen in Figure 9.43, where
and 9.43. The improvement–3.574 in the transient
+ j6.976 for the
response
Compensated is shown in Figure
improvement. The improvement lead-compensated system is
9.42, where we see the peak time
K = 1971 occurring
dominantsooner
pole
shown in Figure 9.43(a), and the final improvement due tos-plane
in the lag-lead-
the addition
compensated system. Improvement
of the lag is shown in Figure 9.43(b). in the steady-state error for a ramp
input is seen in Figure 9.43, where each step of our design yields more
improvement. The improvement for the lead-compensated system is
shown in Figure 9.43(a), and the final improvement due117.13° to the addition
of the lag is shown in Figure 9.43(b). σ
–29.1 –10 0.04713 0

–31.91 –0.0474 –0.01

= Closed-loop pole
= Open-loop pole
Note: This figure is not drawn to scale.
FIGURE 9.41 Root locus for lag-lead-compensated system of Example 9.6

Step 6 We arbitrarily choose the lag compensator pole at 0.01, which then places
the lag compensator zero at 0.04713, yielding
ðs þ 0:04713Þ
Glag ðsÞ ¼ ð9:34Þ
ðs þ 0:01Þ

as the lag compensator. The lag-lead-compensated system’s open-loop 9  


transfer function is
Kðs þ 0:04713Þ
GLLC ðsÞ ¼ ð9:35Þ
sðs þ 10Þðs þ 29:1Þðs þ 0:01Þ

where the uncompensated system pole at %6 canceled the lead compen-


uncompensated system of
Example 9.6 = Open-loop pole

Now we design the lead compensator. Arbitrarily select a location


for the lead compensator zero. For this example, we select the location of
the compensator zero coincident with the open-loop pole at !6. This
choice will eliminate a zero and leave the lead-compensated system with
three poles, the same number that the uncompensated system has.
We complete the design by finding the location of the compensator
pole. Using the root locus program, sum the angles to the design point from 3/27/14  
the uncompensated system’s poles and zeros and the compensator zero
and get !164:65" . The difference between 180" and this quantity is the
angular contribution required from the compensator pole, or !15:35" .
Using the geometry shown in Figure 9.39,
7:003
¼ tan 15:35" ð9:31Þ
pc ! 3:588
from which the location of the compensator pole, pc, is found to be !29:1.
TABLE 9.6 Predicted characteristics of uncompensated, lead-compensated, and lag-lead-compensated systems of
Example 9.6

Uncompensated Lead-compensated Lag-lead-compensated

K K Kðs þ 0:04713Þ
Plant and compensator
sðs þ 6Þðs þ 10Þ sðs þ 10Þðs þ 29:1Þ sðs þ 10Þðs þ 29:1Þðs þ 0:01Þ
Dominant poles !1:794 ' j3:501 !3:588 ' j7:003 !3:574 ' j6:976
K 192.1 1977 1971
z 0.456 0.456 0.456
vn 3.934 7.869 7.838
%OS 20 20 20
Ts 2.230 1.115 1.119
Tp 0.897 0.449 0.450
Kv 3.202 6.794 31.92
eð1Þ 0.312 0.147 0.0313
Third pole !12:41 !31:92 !31:91, !0:0474
Zero None None !0:04713
Comments Second-order approx. OK Second-order approx. OK Second-order approx. OK

E1C09 11/03/2010 13:29:46 Page 491

9.4 Improving Steady-State Error and Transient Response

Lead- and lag-lead-compensated


1.25
Uncompensated
1.00

0.75
c(t)

0.50

0.25
FIGURE 9.42 Improvement in
0 step response for lag-lead-
0 1 2 3 4 5
compensated system of
Time (seconds)
Example 9.6

5.0 Input
4.5
4.0 Lead-compensated
3.5
3.0
c(t)

2.5
2.0 Uncompensated
1.5
1.0
0.5
0 t 10  
0 1 2 3 4 5
Time (seconds)
(a)

5.0 Input
E1C09 11/03/2010 13:29:46 Page 491

9.4 Improving Steady-State Error and Transient Response 491

Lead- and lag-lead-compensated


1.25

1.00
Uncompensated
3/27/14  
0.75

c(t)
0.50

0.25
FIGURE 9.42 Improvement in
0 step response for lag-lead-
0 1 2 3 4 5
compensated system of
Time (seconds)
Example 9.6

5.0 Input
4.5
4.0 Lead-compensated
3.5
3.0

c(t)
2.5
2.0 Uncompensated
1.5
1.0
0.5
0 t
0 1 2 3 4 5
Time (seconds)
(a)

5.0 Input
4.5
4.0
3.5
3.0
c(t)

2.5
2.0 Lag-lead-compensated
1.5
1.0
0.5 FIGURE 9.43 Improvement
0 t in ramp response error for
0 1 2 3 4 5 the system of Example 9.6:
Time (seconds) a. lead-compensated;
(b) b. lag-lead-compensated

In the previous example, we canceled the system pole at !6 with the lead
compensator zero. The design technique is the same if you place the lead compen-
sator zero at a different location. Placing a zero at a different location and not
canceling the open-loop pole yields a system with one more pole than the example.
This increased complexity could make it more difficult to justify a second-order
approximation. In any case, simulations should be used at each step to verify
performance.

11  

Potrebbero piacerti anche