Dynamics & Control Transient Response

Transient Response
A. INTRODUCTION
One of the most important tasks in the analysis and design of control systems is mathematical
modeling of the systems. A mathematical model of a system is defined as a set of equations
that represent the dynamics of the system.
The dynamics of many systems, whether they are electrical, mechanical, thermal, economic,
biological may be described in terms of differential equations. Such differential equations may
be obtained by utilizing physical laws governing the particular system e.g. Kirchhoff's Law for
electrical system, Newton's Law for mechanical system.
Once such a model is obtained, various methods are variable for the analysis of the system
performance.
In practice, the input signal to a control system is not known in advance but is random in nature.
In analyzing and designing control systems, we must have a basis of comparison of
performance of the various control systems. This basis may be set up by specifying particular
test input signals and by comparing the responses of various systems to these input signals.
Once a control system is designed on the basis of test signals, the performance of the system
in response to actual input is generally satisfactory.

B. DEFINITIONS
1. Dynamics, Linear System and Linear Time Invariant systems
 Dynamics:
When at least one element in a physical system is capable of storing energy, such system
is known to be dynamic.
 Linear System:
A system is called linear if the principle of superposition can be applied.
 Linear Time Invariant system:
A dynamic system is liner time-invariant means that the coefficients of the differential
equations govern the system are constants or functions only of independent variable.

Page C- 1
Dynamics & Control Transient Response

2. Transfer Function - I/O relationship (review)

Laplace Transform of Output

Transfer Function 
Laplace Transform of Input All Initial Conditions  0

With the obtained Laplace-transformed equations, simple algebra manipulations can be

used to obtain the transfer function. Thus the output of a component or system can be
determined when a specific input is applied. Following figure shows a clear illustration,
assuming capacitor C has zero initial voltage.

Simple R-C circuit

R
Unit step input Vi(t) Output Vo(t) = ?

Vi(t) C Vo(t)

0, t 0   A
Unit step input : Vi (t)    Input : Vi (s)  s
A, t  0  
 1  Transforma
Laplace
  1

     RI ( s )  Cs I ( s )  Vi ( s )
Ri(t)  C  i (t )dt  Vi (t )
tion

System Model :  Model : 
  V ( s )  1 I ( s )
V (t )  1 i (t )dt
 o
C  


  o Cs

Operational Algebra
Calculus

1 Inverse Laplace
- t 1
Output : Vo (t)  1  e RC
  
Transforma tion
V o ( s) 
s(RCs  1)

Input Vi(t)
Output Vo(t)
A
A

0 Time 0 Time

Page C- 2
Dynamics & Control Transient Response

3. Transfer Function – example (cont’d, in Laplace Transform)

Consider a simple RC circuit shown in Figure C -1:

R

i(t) Input - vi(t)

vi(t) C vo(t) Output - vo(t)

Figure C-1: simple RC circuit

1
C
Apply Kirchhoff's law: v i (t )  i(t )R  v o (t ) v o (t)  i( t )dt

Laplace transform with zero initial condition:

1
Vi (s)  I(s)R  Vo (s) Vo (s)  I(s)
Cs
Vo (s) 1
Combine both equations, we have the transfer function: 
Vi (s) RCs  1
4. Typical Input Signals

Input Signal Graphical Presentation Mathematical Description Note

r(t) A t0 Unit-step: A = 1,

Step
r( t )  
0 t<0 r(t) = 1
A 1
R( s)  , A can be +ve or -ve R(s) 
0 t s s

r(t)  At t0 Unit-ramp: A = 1,

r ( t)  
Ramp 0 t=0 R(t) = t
A 1
R( s)  , A can be +ve or -ve R(s) 
0 t s2 s2
1 2
r(t)  At t0
Parabolic r(t )   2
 0 t=0
Unit-parabolic:
A=1
A
0 t R( s)  3 , A can be +ve or -ve
s

r(t)
Unit
1 t0
Impulse r (t )   or, r(t) = (t)
0 t0

0 t R(s)  1

Page C- 3
Dynamics & Control Transient Response

C. FIRST-ORDER SYSTEMS

1. A First-order System
Block diagram:

R(s) E(s) C(s) R(s) a C(s)

a
Ts Ts + a

(a) Block diagram (b) Simplified block diagram

Figure C-2 Block diagram of a first-order system

C ( s) a a/a 1
Transfer function:    (C-1)
R( s) Ts  a Ts  a  / a (T / a) s  1
Here, the numerator “1” is called the system gain, or DC gain of the system, when the input is
a step function (DC input) resulting in an eventual constant steady state output (DC output),
and the coefficient of s term in the denominator, “T/a”, is called the time constant. Generally
speaking, you can find system DC gain by setting s=0 to the closed-loop transfer function.

C ( s) 1 1 1
When a=1,    , which we will use to draw output curve below.
R( s ) T T Ts  1
s 1 s 1
a 1

2. Step Response of First-Order Systems

The system response to a step input:
A
Input: R( s) 
s
a A a/a A 1 A
Output: C ( s)   
Ts  a s T a s T s
s s 1
a a a
Taking the inverse Laplace Transform
 
t

 T 
Output: c(t )  c(0)  A1  e a  ( t0) (C-2)
 
 

Page C- 4
Dynamics & Control Transient Response

 
t

When a=1, we have c(t )  c(0)  A1  e T 
 ( t0) (C-2’)
 
If A=1 where input is unit step function, assuming zero initial condition(s) we have
t

c(t )  1  e T
( t 0) (C-2’’)
Response of the system when t > 0: (see graph 1 below)

Time t 0 T 2T 3T 4T 5T infinity

Response [c(t)-c(0)]/A 0.000 0.632 0.865 0.950 0.982 0.993 1.000

Assuming A=1 or unit step input, the slope of the tangent line at t = 0 is:

d c(t )  c(0) dc(t ) 1 1

  e t / T 
dt t 0 dt t 0 T t 0 T

The response of this first-order system is shown in Figure C-3 graph 2. Note that for y-axis,
the output c(t) shows values with respect to initial value of c(0), or actually values of c(t)-c(0).
When under zero initial conditions, c(0) = 0 and y-axis would give values of c(t), see graph 3
below.

Page C- 5
Dynamics & Control Transient Response

Graph 1 (Equation C-2) Graph 2 (Equation C-2’)

Graph 3 (Equation 3-2’’)

Figure C-3: Exponential response curve

One important characteristic of such exponential response curve c(t) under zero initial
conditions (graph 3), is that at t = T the value of c(t) is 0.632, or the response c(t) has reached
63.2% of its final output from initial value 0.

T is defined as the time constant of the system. The value of time constant indicates
how fast the system tends to reach the final output. Note that the smaller the time constant T,
the faster the system response.
1 1
 t  t
Steady-state error: ess  lim e(t )  lim [r (t )  c(t )]  lim [1  (1  e T
)]  lim e T
0
t  t  t  t 

3. Ramp Response of First-order Systems under zero initial conditions

A
The system response to a ramp input R( s ) 
s2
1 A 1 T T2 
Output: C ( s)   A
 s 2 s Ts  1 
 
Ts  1 s 2  
Taking the inverse Laplace Transform assuming zero initial condition(s)

Page C- 6
Dynamics & Control Transient Response

  t 
1

Output:  T 
c(t )  A t  T  Te  ( t0)
 
The response of a first-order system unit-ramp input is shown in Figure C-4.

r(t), c(t) Input r(t) = t

Steady State Error = AT

Output c(t)

t
Nominal 0

Figure C-4: Ramp response of the first-order system

Steady-state error (also called offset):

1 1
 t  t
ess  lim e(t )  lim [r (t )  c(t )]  lim A[t  (t  T  Te T
)]  lim AT (1  e T
)  AT
t  t  t  t 

4. Steady-State Error (Offset) For General Control Systems

Consider the system shown in Figure C-5.

R(s) E(s) C(s) r(t) e(t) c(t)

G(s) G(s)

H(s) H(s)

Figure C-5 control system in both s domain and time domain

The closed-loop transfer function:

C( s ) G( s )

R( s ) 1  G( s )H( s )
The error signal:
E(s)  R(s)  C(s)H(s)

The transfer function between the actuating error signal E(s) and the input signal R(s):
E( s ) C( s ) 1
 1 H( s ) 
R( s ) R( s ) 1  G( s )H( s )

Page C- 7
Dynamics & Control Transient Response

The error signal:

1
E( s )  R( s )
1  G( s )H( s )

Applying the final value theorem to find the steady-state error ess:
sR( s )
ess  lim e( t )  lim sE( s )  lim
t  s0 s  0 1  G( s )H( s )

Note that final value theorem applies to all signals in the control system, for example
C ( s) G( s) sG( s) R( s)
c ss  lim c(t )  lim sC( s)  lim s R( s)  lim s R( s)  lim
t  s 0 s 0 R( s ) s 0 1  G ( s ) H ( s ) s 0 1  G ( s ) H ( s )

D. SECOND-ORDER SYSTEM

1. Step Response of Second-order Systems

In general, the block diagram of a second-order system is shown as Figure C-6.

R(s) E(s) 2n

 C(s)
s(s+ 2n 

Figure C-6 Second-order system

C (s)  n2
The closed-loop transfer function:  2
R( s) s  2 n s   n2

where : the damping ratio of the system

n: the un-damped natural frequency
In addition, we can also define its system gain or DC gain, when the reference input is a step
function, i.e., R(s) = A/s.

By setting s=0 to the closed-loop transfer function above, we find system/DC gain is equal to
1 in this special case. Take note that the system gain can be different from 1, depending on the
resultant closed-loop transfer function.

Page C- 8
Dynamics & Control Transient Response

In general, however, a second order system has the closed-loop transfer function of
C (s) ks  g n2
 2 . As an exercise, try to find the system gain for this system.
R( s) s  2 n s   n2

Roots of characteristic equations and different kinds of damping

The characteristic equation of the second-order system is obtained by setting denominator of

the closed-loop transfer function to 0:
s 2  2n s  n2  0 (C-3)


Roots: s1  n  jn 1   2 let d n 1
 s1  n  jd
2
 (C-4)


s2  n  j n 1   2 let d n 1
 s 2  n  j d
2
 (C-5)

where: d  n 1   2 (d : damped natural frequency) (C-6)

Note that  d is well defined only when 0    1 so that it is a positive real number.

Depending on the values of damping ratio  , we have different values of  d , hence different

kinds of roots, s1 and s 2 namely

1)   0 ,

s1   n  j n 1   2  0 n  j n 1  0 2   j n
s 2   n  j n 1   2  0 n  j n 1  0 2   j n
where the two roots are on the imaginary axis of the complex plane, and are complex
conjugate. This is un-damped case.

2) 0    1 ,

s1   n  j n 1   2   n  j d
s 2   n  j n 1   2   n  j d
where the two roots are on the left-hand side of the complex plane, and are complex
conjugate. This is underdamped case.

Page C- 9
Dynamics & Control Transient Response

3)   1 ,

s1   n  j n 1   2  1   n  j n 1  12   n
s 2   n  j n 1   2  1   n  j n 1  12   n
where the two roots are on the negative real axis of the complex plane, and are at the
location. This is critically damped case.

4)   1 ,

s1   n  j n 1   2   n   n  2  1
s 2   n  j n 1   2   n   n  2  1
where the two roots are on the negative real axis of the complex plane, and are at
different locations. This is over-damped case.

5) If   0 the system is called negatively damped case, which we will explore more in
chapter 5 when we talk about system stability and root locus.

Step responses
A
When input is a step signal, i.e. R( s) 
s
 n2 A
Response of the system: C ( s) 
s  2 n s   n s
2 2

Taking the inverse Laplace Transform: c(t) = L-1[C(s)]

(1) Over-damped Case ( > 1)

Roots of the characteristic equation: two unequal roots at

s 1  n  n  2  1
s 2  n  n  2  1
Response:
  n2 A
c(t )  L1 c( s)  L1  2 
  A1  e
 (   2 1 )nt  (t > 0)
    2 
 s 2 n s n s 

Page C- 10
Dynamics & Control Transient Response

The response for over-damped case is an exponential curve. The speed of the
response depends on the value of damping ratio .

(2) Underdamped Case (0 <  < 1)

Roots of the characteristic equation: two complex roots at

s 1  n  jn 1   2
s 2  n  jn 1   2
Time domain response:
  
  n2 A e  nt   t  tan 1 1  
2
c(t )  L1 c( s )  L1  2   A 1  sin  ( t  0)
 s  2 n s   n s 
2
 1 2  d  
  
The response for underdamped case is damped sinusoid signal with an exponential envelope.
The frequency of transient oscillation is the damped natural frequency d and thus varies with
the damping ratio .
Figure C-7 shows response curves with different damping ratios in the range of 0    1 while
the input is a unit step function.

Figure C-7 response c(t) with various value of damping ratio 

(3) Un-damped Case ( = 0)

Roots of the characteristic equation: two imaginary roots at
s1  jn
s 2  jn

Time domain response:

  n2 A
c(t )  L c( s)  L  2
1 1
  A1  cos n t  (t > 0)
 s  2n s   n s 
2

Page C- 11
Dynamics & Control Transient Response

The response for un-damped case is a pure sinusoid signal. The frequency of oscillation is the
un-damped natural frequency n.
(4) Critically Damped Case ( = 1):
Roots of the characteristic equations: two equal real roots at
s1  s2  n

Time domain response:

  n2 A
c(t )  L1 c( s)  L1  2
    2
 n t

  A 1  e (1   n t ) (t > 0)
 s 2 n s n s 

The response of a second-order system for critically damped case is an exponential curve,
which is similar to but faster than over-damped case.

Damping ratio, roots of CE and system response

Damping Characteristic Equation Roots System Response

Ratio (Closed-loop Poles)

=0

0<<1

=1

>1

Page C- 12
Dynamics & Control Transient Response

Damping ratio, Natural frequency and System response

Unit step response of a second order system Unit step response of a second order system
for different damping ratio when natural for different natural frequency when damping
frequency remains unchanged. ratio remains unchanged.

E. DEFINITIONS OF TRANSIENT RESPONSE SPECIFICATIONS

In many special cases, the desired performance characteristics of control systems are specified
in terms of time-domain quantities. Systems with energy storage cannot respond
instantaneously and will exhibit transient responses whenever they are subjected to inputs or
disturbances.
The transient response of a practical control system often exhibits damped oscillations before
reaching steady state. In specifying the transient-response characteristics of a control system to
a unit-step input, it is common to specifying as Delay time, Rise time, Peak time, Maximum
overshoot, and Settling time.

Figure C-8 Unit-step response curve showing td , tr , tp , Mp, and ts

Page C- 13
Dynamics & Control Transient Response

Specifications:

1. Delay Time, td: The time required for the response to reach 50% of its final value at the
very first time.

2. Rise Time, tr: The time required for the response to rise from 0% to 100% of its final
value for underdamped response where 0    1 , and 10% to 90% of final value for
critically damped or over-damped response where   1 and system output can never reach
100% in finite time.
We obtain the rise time tr by letting c(tr) = 1.

Thus, the rise time: tr  (C-7)
d
Note:  must be in Radian.
j
Where  is defined as:
jd

n   cos 1  (rad)


- 0 
n

Clearly, for small value of tr, n must be large.

3. Peak Time, tp: The time required to reach the first, or maximum, peak.
We may obtain the peak time by differentiating c(t) with respect to time and letting this
dc( t )
derivative equal zero. i.e. 0
dt t  t p


The peak time: tp  (C-8)
d
The peak time tp corresponds to one-half cycle of the frequency of damped oscillation. This
definition applies to under damped response only as  d is well defined where 0    1 .

For critically and over-damped responses where   1 , peak time is not applicable.

Page C- 14
Dynamics & Control Transient Response

4. Maximum (Percent) Overshoot, MP: The amount that the waveform overshoots the
steady-state, or final, value at the peak time, expressed as a percentage of the steady-state
value. Maximum overshoot directly indicates the relative stability of the system.
c(tp )  c()
Maximum Percent Overshoot =  100% (C-9)
c()
The maximum overshoot occurs at the peak time, i.e. MP = c(tP) - 1

The Maximum Percent Overshoot: Mp  e / 1  2

 100% (C-10)

This definition applies to under-damped response where 0    1 . For critically and over-
damped responses where   1 , peak time is not applicable as there is no overshoot at all.
Conversely, if MP is known, then the system is under-damped. We can use

ln M P 2
  to find damping ratio  , where 0    1 .
ln M P 2   2

5. Settling Time, ts: The time required for the transient's damped oscillations to reach and
stay within +2% or +5% of the steady-state value.
4
The settling time: ts  (2% criterion) (C-11)
 n

3
ts  (5% criterion) (C-12)
 n

F. STEADY-STATE ERROR ANALYSIS

Any physical control system inherently suffers steady-sate error in response to certain types of
inputs. A system may have no steady-state error to a step input, but the same system may
exhibit nonzero steady-state error to a ramp input. Whether or not a given system will exhibit
steady-state error for a given type of input depends on the type of open-loop transfer function
of the system.

1. Classification of Control Systems

Control systems may be classified according to their ability to follow step inputs, ramp inputs,
parabolic inputs, and so on.

Consider the open-loop transfer function of a control system G(s)H(s):

Page C- 15
Dynamics & Control Transient Response

K (Ta s  1)(Tb s  1)... (Tm s  1)

G(s)H(s)  (C-13)
sN (T1s  1)(T2 s  1)... (Tp s  1)

It involves the term sN in the denominator, representing the number of the roots at the origin.
The present classification scheme is based on the number of integration indicated by the open-
loop transfer function.

N=0  Type 0 system

N=1  Type 1 system
N=2  Type 2 system
: : : : : : : : :
Note: classification of system type is different from classification of system order. As the type
number is increased, accuracy is improved; however, increasing the type number
aggravates the stability problem. A compromise between steady-state accuracy and
relative stability is always necessary.

2. Steady-State Error (Offset) Revisited

Consider the system shown in Figure C-9,

R(s) E(s) C(s) r(t) e(t) c(t)

G(s) G(s)

H(s) H(s)

Figure C-9 control system in both s domain and time domain

The closed-loop transfer function:

C( s ) G( s )

R( s ) 1  G( s )H( s )
The error signal:
E(s)  R(s)  C(s)H(s)
The transfer function between the actuating error signal E(s) and the input signal R(s):
E( s ) C( s ) 1
 1 H( s ) 
R( s ) R( s ) 1  G( s )H( s )
The error signal:

Page C- 16
Dynamics & Control Transient Response

1
E( s )  R( s ) (C-14)
1  G( s )H( s )
Applying the final value theorem to find the steady-state error ess:
sR( s )
ess  lim e( t )  lim sE( s )  lim (C-15)
t  s0 s0 1  G( s )H( s )

Note that final value theorem applies to all signals in the control system, for example
C ( s) G( s) sG( s) R( s)
c ss  lim c(t )  lim sC( s)  lim s R( s)  lim s R( s)  lim
t  s 0 s 0 R( s ) s 0 1  G( s) H ( s) s  0 1  G( s) H ( s)

Interested students are directed to any reference book to explore the relationship between
system type and steady-state error.

Page C- 17