EES42042
Fundamental of Control Systems
Transient Response
Electrical Engineering Department University of Indonesia
Lecturer:
Dr. Wahidin Wahab M.Sc.
Aries Subiantoro, ST. MSc.
Transient Response - First 2
Order System
Consider simple closed loop system with an integrator in feed-forward path
+ -
R(s) C(s)
Ts 1
1 1
+ Ts
R(s) C(s)
Transient Response - First 3
Order System
Unit Step Response
1 1
+ Ts
R(s) C(s)
s T s
Ts T s s
C
s s Ts
s C s
R
1 1 1
1 ) 1
(
1 1 ) 1
( 1 ,
) (
− + + =
−
=
→
+ ⋅
=
=
Transient Response - First 4
Order System
Unit Step Response
1 1
+ Ts
R(s) C(s)
T t
e t
c
s T s s
C
−
−=
→
− +
=
1 )
(
1 1 ) 1
(
Transient Response - First 5
Order System
Unit Step Response
1 1
+ Ts
R(s) C(s)
T t
e t
c ( ) = 1 −
−1.0
Initial slope = 1/T
0.632
Transient Response - First 6
Order System
Unit Step Response
1 1
+ Ts
R(s) C(s)
Time c(t)
T 0.632
2T 0.865
3T 0.95
4T 0.982
5T 0.993
Transient Response - First 7
Order System
Unit Ramp Response
1 1
+ Ts
R(s) C(s)
1.0 t r(t)
1.0
2
) 1
( s s
R =
Transient Response - First 8
Order System
Unit Ramp Response
1 1
+ Ts
R(s) C(s)
2 2
1 1
) 1 ( ) 1 (
s s Ts
C s s R
+ ⋅
=
→
=
Transient Response - First 9
Order System
Unit Ramp Response
1 1
+ Ts
R(s) C(s)
T t
Te T
t t
c
Ts T s
T s s
C
s s Ts
C
+
−−
=
∴
+ +
−
=
→
+ ⋅
=
) (
1 ) 1
(
1 1
) 1 (
2 2
2
Transient Response - First 10
Order System
Unit Ramp Response
1 1
+ Ts
R(s) C(s)
{ } e t T
e T
t c t
t c t
r t
e
T t
=
=
⎟⎟ ⎠
⎜⎜ ⎞
⎝
⎛ −
=
−
=
−
=
∞
→
−
) ( lim
Error Tracking
State Steady
1 )
(
) ( )
( )
( as error Define
input ramp
g in trackin Error
t
Transient Response - First 11
Order System
Unit Ramp Response
1 1
+ Ts
R(s) C(s)
{ }
system.
the of
constant time
by the determined
offset small
a be always will
There ) ( lim
Error Tracking
State Steady
t
→
=
=
→∞e t T
Transient Response - First 12
Order System
Unit Ramp Response
1 1
+ Ts
R(s) C(s)
Steady State error = T
r(t)
c(t)
Transient Response - First 13
Order System
Unit Impulse Response
1 1
+ Ts
R(s) C(s)
Note that the unit impulse response of an LTI system is
simply the inverse Laplace Transform of the transfer
function
Transient Response - First 14
Order System
Unit Impulse Response
1 1
+ Ts
R(s) C(s)
T t
T e L Ts
t
c
−=
−⎭ ⎬
⎫
⎩ ⎨
⎧
= + 1
1 ) 1
(
1Transient Response - First 15
Order System
Unit step response = derivative of unit ramp response
Unit impulse response = derivative of unit
step response
16
Example from Ogata
Liquid level controller - p140 Ogata
+ -
R(s) H(s)
K
pK
v+ 1 RCs
R
K
b17
Example from Ogata
Upon block reduction
+ -
R(s) H(s)
+ 1 Ts 1/K
bK
X(s)
RC T
RK K
K
K
p v b=
=
18
Example from Ogata
Consider response to step change in reference input r(t)
+ -
R(s) H(s)
+ 1 Ts 1/K
bK
X(s)
Note that this is equivalent to step change in x(t)
19
Example from Ogata
T s K
K K s
K K
s K Ts
s K H
K Ts
K s
X s H
+ + + ⋅
− + ⋅
=
+ ⋅
= +
+
= +
1 1 1
1 1
fractions partial
into Splitting
1 ) 1
(
response Step
1 )
(
)
(
20
Example from Ogata
K T T
t K e
t K h
T s K
K K s
K s K
H
T t
= +
⎟ ≥
⎟
⎠
⎞
⎜ ⎜
⎝
⎛ −
= +
→
+ + + ⋅
− + ⋅
=
−
1
0 ,
1 1 )
(
sides both
of LT inverse
the taking
1 1 1
1 ) 1
(
1
1
21
Example from Ogata
( ) K
e K K
h K
t K e
t K h
T t
t
T t
= +
⎪⎭
⎪ ⎬
⎫
⎪⎩
⎪ ⎨
⎧
⎟ ⎟
⎠
⎞
⎜ ⎜
⎝
⎛ −
= +
∞
⎟ ≥
⎟
⎠
⎞
⎜ ⎜
⎝
⎛ −
= +
−
∞
→
−
1 1 lim 1
value state
Steady
0 ,
1 1 )
(
1 1
Note that this could also be found from the Final Value Theorem
of Laplace Transforms
22
Example from Ogata
( ) { }
K K
s K
Ts s K
s H s
h
s
s
= +
⎭ ⎬
⎫
⎩ ⎨
⎧ ⋅
+
⋅ +
=
⋅
=
∞
→
→
1
1 lim 1
) ( lim
theorem value
final the
using
0
0
23
Example from Ogata
( )
K K
K x
= +
− +
=
=
∞
1 1 1 1
offset or
error state
Steady
0 . 1 Since
1.0
Note that this offset could be eliminated by placing an
integrator (K/s) in the feed-forward path
24
Second Order Systems
General form of 2nd order transfer function
C s
R s s s
n
n n
( ) ( ) =
+ +
=
=
ω
ςω ω ς
ω
2
2 2
2
damping ratio of system
undamped natural frequency
n
Time domain performance of a 2nd order system
linked to these two quantities
25
Effect of Damping Ratio
0 1
1
2 1
1 2
2 2
2
< < −
→
= ⋅
+ +
⎧ ⎨
⎩
⎫ ⎬
⎭
= − =
−
ς
ω
ςω ω
ω ω ς
UNDERDAMPED CASE poles are complex conjugate
damped oscillatory response Response to Unit Step input
Define damped natural frequency c t L
s s s
n
n n
d n
( )
26
Effect of Damping Ratio
→ = − +
−
⎛
⎝ ⎜⎜ ⎞
⎠ ⎟⎟
= − − ⎛ + −
⎝ ⎜⎜ ⎞
⎠ ⎟⎟
= +
−
⎛
⎝ ⎜⎜ ⎞
⎠ ⎟⎟
−
− −
−
c t e t t
e t
c t
e t t
n
n
n
t
d d
t
d
t
d d
( ) cos sin
sin tan ,
( )
cos sin
1
1 1
1
1
1
2
2
1 2
2 ςω
ςω
ςω
ω ς
ς ω
ς ω ς
ς
ω ς
ς ω
Error signal = 1 -
27
Effect of Damping Ratio
Note that if the damping ratio is zero system will undergo sustained undamped
oscillations
ς = → 0 c t ( ) = − 1 cos ω
nt
Note also that if the damping ratio >1 then
response is OVERDAMPED and system will not
oscillate
28
Critical Damping
( )
( )
Critical damping = 1
System poles are equal and real L.T. of unit step response is
→
→
→
= +
→ = −
−+ ≥
ς
ω ω
ω
ω C s
s s
c t e t t
n n
t
n n
( )
( ) ,
2 2
1 1 0
29
Overdamped Case
( )( )
( ) ( )
OVERDAMPED > 1 poles are real and distinct L.T. of unit step response
Poles:
→
→
→
= + + − + − −
= + − = − −
ς
ω
ςω ω ς ςω ω ς
ς ς ω ς ς ω
C s
s s s
s s
n
n n n n
n n
( )
,
2
2 2
1
2
2
2
1 1
1 1
30
Overdamped Case
( )
( )
c t e
e
n
n
t
t
( ) = +
− + −
− − − −
− +⎛ −
⎝⎜ ⎞
⎠⎟
− −⎛ −
⎝⎜ ⎞
⎠⎟
1 1
2 1 1
1
2 1 1
2 2
1
2 2
1
2
2
ς ς ς
ς ς ς
ς ς ω
ς ς ω
Step response = sum of two decaying exponential functions
31
Overdamped Case
( )
( )
c t e
e
n
n
t
t
( ) = +
− + −
− − − −
− +⎛ −
⎝⎜ ⎞
⎠⎟
− −⎛ −
⎝⎜ ⎞
⎠⎟
1 1
2 1 1
1
2 1 1
2 2
1
2 2
1
2
2
ς ς ς
ς ς ς
ς ς ω
ς ς ω
When faster decaying exponential has died out response
is similar to 1st order system
Performance of Second Order 32
Systems
0 2 4 6 8 10 12 14
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Unit Step Response Curves of Second Order System ωn=1.0 rad/sec
c(t)
Underdamped Critically damped Overdamped
ζ = 2.0 ζ = 1.0 ζ = 0.9 ζ = 0
ζ = 0.2 ζ = 0.4
ζ = 0.7
ζ = 0.8
Performance of Second Order 33
Systems
Underdamped systems respond fastest but will oscillate about steady state value
Overdamped systems tend to be sluggish but lack oscillation.
Critically damped systems give the quickest
rise time without overshoot.
Transient Response 34
Specifications
Means of specifying the desired performance of a control system
Frequently specified in terms of response to
unit step input
Transient Response 35
Specifications
Terms
– Delay Time, t
d– Rise Time, t
r– Peak time, t
p– Maximum Overshoot, M
p– Settling Time, t
sTransient Response 36
Specifications
Delay Time
– Time to reach 50% of final value for the first time
1.0 0.5
Example - Unit Step Response
Transient Response 37
Specifications
Rise Time
– Time for system to go from 0-100% of final value or 10-90%, or 5-95%
1.0
Example - Unit Step Response
Transient Response 38
Specifications
Peak Time
– Time for system to reach first peak
1.0
Example - Unit Step Response
Transient Response 39
Specifications
Maximum Percent Overshoot
– Maximum percent value of response from steady state value
1.0
Example - Unit Step Response
M
pTransient Response 40
Specifications
Maximum Percent Overshoot
1.0
Example - Unit Step Response M
p( ) ( )
( ) − ∞ ∞ × 100
= c
c t
M
pc
pTransient Response 41
Specifications
Settling Time
– Time for response to reach and stay within a certain range of the steady state value -
typically 5% or 2%
1.0
Example - Unit Step Response
Allowable
tolerance
Transient Response 42
Specifications
Desirable Response
– fast and well damped
– damping ratio should be between 0.4 and 0.8 – too small a damping ratio leads to a poorly
damped response – too much overshoot
– too large a damping ratio leads to a sluggish response
Conflict between max. rise time and allowable overshoot
Second Order Systems and 43
Transient Response Specs.
Second Order system pole locations
2 2
1
1 ,
,
Poles
ς ω
ω ςω
σ
ω σ
−
=
−
=
±
=
n d
n
j
ds
s
Second Order Systems and 44
Transient Response Specs.
Second Order system pole locations
2 2
1
1 ,
, Poles
ς ω
ω ςω
σ
ω σ
−
=
−
=
±
=
n d
n
j
ds s j ω
σ x
x -σ=−ζω
nω
nω
dβ
β ς cos
Note =
s
1s
2=s
*1Second Order Systems and 45
Transient Response Specs.
Rise Time
large.
be
must time
rise small
a for that
Note
1 tan
1d d
d d
t
rω ω
β π
σ ω ω
= −
⎟ ⎠
⎜ ⎞
⎝
⎛
=
−−
Second Order Systems and 46
Transient Response Specs.
Peak Time
proof.
of details for
153 -
152 pp.
Ogata See
d
t
pω
= π
Second Order Systems and 47
Transient Response Specs.
Maximum Overshoot
proof.
of details for
153 p.
Ogata See
% 100
1 2
×
=
⎟⎟
⎠
⎞
⎜⎜
⎝
⎛
− −
ς π ς
e M
pNote that this ONLY depends on damping ratio ζ
Second Order Systems and 48
Transient Response Specs.
Settling Time
s
n
t ςω
ςω
%) 3 5
(
value final
its of
5%
within be
will system
constants time
3 After
constant 1 time
of value
on depends
response of
decay of
Speed
=
∴
Second Order Systems and 49
Transient Response Specs.
Settling Time
n
t
sςω
%) 4 2
(
value final
its of
2%
within be
will system
constants time
4 After
=
∴ These Settling Time formulae are very
important to remember as settling time
is a very common control system spec.
Second Order Systems and 50
Transient Response Specs.
Settling Time
– inversely proportional to product of damping ratio with natural frequency
– damping ratio is usually set by max. overshoot – therefore settling time determined by choosing
ω
nonce damping ratio has been set.
51
Example
Consider the closed loop system shown below. Determine the values of K
h& K
such that the max. overshoot to a step input is 0.2 and the peak time is 1 second. Also determine the rise time and settling time.
Assume J=1, B=1
52
Example
R(s) +
- +
- Js B
K
+ s
1
C(s)
K
hNeed to reduce this to a single loop
53
Example
R(s) +
- s ( Js + B K + KK
h) C(s)
Reduced system
( )
J K KJ
KK B
K s
KK B
Js
K R(s)
C(s)
n h
s
+ =
=
→
+ +
= +
ω
ς ,
2
T.F.
loop Closed
2
54
Example
456 .
0
61 . 1 1
2 . 0 Overshoot
Max
2 1 2
=
→
⎟ =
⎟
⎠
⎞
⎜ ⎜
⎝
⎛
→ −
=
=
⎟⎟
⎠
⎞
⎜⎜
⎝
⎛
− −
ς
ς π ς
ς π ς
e
M
p55
Example
53 . 3 1
14 . 3
1 Time Peak
2
=
= −
→
=
→
=
=
ς ω ω
ω ω
π
d n
d d
t
p56
Example
178 .
0 456
. 2 0
5 . 12 53
. 3
is frequency natural
Now
=
→ + =
=
=
→
=
=
h h
n
KJ K KK B
J K K
ς
ω
57
Example
65 . 0
10 . 1 tan
, Time Rise
1
=
→
⎟ =
⎠
⎜ ⎞
⎝
= ⎛
= −
−r
d d
r
t
t σ
β ω ω
β
π
58
Example
sec 86
. 3 1
time settling
5%
sec 48
. 4 2
time settling
2%
Time Settling
=
=
=
=
σ
σ
Impulse Response of Second 59
Order System
Simply inverse Laplace transform of transfer function
or derivative of unit step response
different expressions for max. overshoot
see Ogata pp. 158-160
60
Significance of Pole Locations
Real and negative
– impulse response is a stable exponential decay
Real and positive
– impulse response is an unstable exponential rise
Complex with negative real parts
– impulse response is oscillatory and stable
» damped oscillation
61
Significance of Pole Locations
Complex with positive real parts
– impulse response is oscillatory and unstable
» undamped oscillation
angle of complex pole w.r.t. negative real axis = cosine of damping ratio
magnitude of complex pole = natural frequency
Note that complex poles ALWAYS occur in conjugate pairs
62
Significance of Pole Locations
63
Higher Order Systems
Consider general transfer function of the form
( )
( ) ( )
∏ ∏
∏
= =
=
+ +
+
+
=
qj
r
k
k k
k j
M
i
i
s s
p s
z s
K s
G
1 1
2 2
1
2 )
(
ω ω
ς
M zeros, at least q real poles, & at most r pairs of complex
conjugate poles
64
Higher Order Systems
The response to a unit step input is determined as
( )
( ) ( )
fractions.
partial of
method by the
expanded be
can then This
2 ) 1
(
1 1
2 2
1
∏ ∏
∏
= =
=
+ +
+
+
⋅
=
qj
r
k
k k
k j
M
i
i
s s
p s
z s
K s s
C
ω ω
ς
65
Higher Order Systems
( )
responses order
second and
first of
set a
of
sum system
order higher
of Response
2
1 fractions
partial by
Expanding
1
2 2
2
1
=
→
+ +
− +
+ + + +
= ∑ ∑
=
=
r
k k k k
k k
k k
k k
q
j j
j
s s
c s
b p
s a s
C(s) a
ω ω
ς
ς ω
ω ς
Residues
66
Higher Order Systems
( ) ( )
( )
∑
∑
∑
=
−
=
−
=
−
− +
− +
+
=
r
k
k k
t k
r
k
k k
t k
q
j
t p j
t e
c
t e
b e
a a
t c
k k
k j k
1
2 1
2 1
1 sin
1 cos
then is
response domain
Time
ς ω
ς ω
ω ς
ω ς
Relative size of each component determined by system residues
which relate to system zeros.
67
Higher Order Systems
Note that components of c(t)
corresponding to poles with large negative real parts will decay rapidly
– shorter settling times
More dominant poles are those with longer settling times
– closer to the imaginary axis in the s-plane
Additional Pole in 2
Additional Pole in 2 nd nd Order Order Model
Model
Step Response Change Due to Step Response Change Due to
Extra Zero
Extra Zero
70
Dominant Poles
Relative dominance of pole determined by ratio of real parts
If ratio of real parts is greater than 5 & no
zeros are close by then the closed loop poles
closest to the imaginary axis will dominate
the system response.
71
Dominant Poles
Useful for getting an approximate idea of system behaviour.
Make sure of underlying assumptions
before using.
Transfer Functions in 72
MATLAB
Transfer functions may be implemented in Matlab using the “tf” command
s polynomial r
denominato and
numerator the
of ts coefficien the
containing arrays
are and
where )
, (
syntax has
tf.m
den num
den num
tf
sys =
Transfer Functions in 73
MATLAB
);
, (
];
10 1
[ ];
1 [
Matlab in
write could
10 we ) 1
(
function transfer
the implement to
example For
den num
tf G
den num
s s G
=
=
=
= +
Transfer Functions in 74
MATLAB
);
, (
];
10 2
1 [
];
5 1 [
Matlab in
write could
10 we 2
) 5 (
function transfer
the implement Or to
2
den num
tf G
den num
s s
s s G
=
=
=
+ +
= +
In this way Matlab has the transfer function stored as variable G
which it can use in the control design tools you are going to use.
75
A Useful Trick
( form 2 )( for example 3 ) 5 6 but might be given as
expanded in the
ly convenient be
not given will
are you
polynomial r
denominato or
numerator the
Sometimes
2