Signals & Control Systems

Home Page
Chapter 6: Performance of Feedback Control Systems
Title Page
& Steady-State Error
Contents

JJ II
College of Engineering
J I
Department of Electrical Engineering
Page 1 of 24

Go Back

Full Screen

Close

Quit
 Contents
1 Introduction 2

2 Block diagram 3
2.1 Block diagram connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.2 Canonical form of feedback systems . . . . . . . . . . . . . . . . . . . . . . . 4
2.3 Block diagram manipulation . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

3 Performance of second-order systems 8

4 Steady-state error 17
Home Page

Title Page

Contents

JJ II

J I

Page 1 of 24

Go Back

Full Screen

Close

Quit
 1. Introduction
• The ability to adjust the transient and steady-state responses of a feed-
back control system is a beneficial outcome of the design of control sys-
tems.
• One of the first steps in the design process is to specify the transient
response performance specifications.
• In this chapter, we introduce the common time-domain specifications
such as percent overshoot, settling time, time to peak and time to rise.
Home Page
• Standard input signals such as the step, ramp and parabolic inputs are
Title Page often used to test the response of the control system.
Contents

JJ II A At At 2
J I
2

Page 2 of 24

Test signal r(t) for t ≥ 0 R(s)

Go Back

Full Screen Step r(t) = A R(s) = A/s

Close

Quit
Ramp r(t) = At R(s) = A/s2

At2
Parabolic r(t) = R(s) = A/s3
2
 2. Block diagram
2.1. Block diagram connections
+ Y
X G1
+_
X G1 G2 Y

G2

Home Page Cascade connection Parallel connection

Title Page

Contents
+ Y +
X
_ G1 X G1 Y
JJ II +
J I
G2 G2
Page 3 of 24
Negative feedback loop Positive feedback loop
Go Back

Full Screen

Close
+ Y +
Quit
X
_ G1 X G1 Y
+

Negative Unity feedback loop Positive Unity feedback loop

 2.2. Canonical form of feedback systems
R
+ E
G C
+
_
or
B

H
G(s) ≡ forward transfer function ≡ direct transfer function
Home Page H(s) ≡ feedback transfer function

Title Page G(s)H(s) ≡ open-loop transfer function ≡ loop transfer function

Contents C(s)/R(s) ≡ closed-loop transfer function ≡ control ratio

JJ II C(s) G(s)
=
R(s) 1∓G(s) H(s)
J I

Page 4 of 24 C(s) G(s)

1. negative feedback loop: =
R(s) 1+G(s) H(s)
Go Back
C(s) G(s)
Full Screen
2. positive feedback loop: =
R(s) 1−G(s) H(s)
Close Closed-loop characteristic equation: the denominator of the closed loop
Quit
transfer function determines the characteristic equation of the closed-loop
system, which is usually defined as:

1∓G(s)H(s) = 0
 2.3. Block diagram manipulation
1) Combining blocks in series
_
X
X G1 G2 Y X G1G2 Y

2) Combining blocks in parallel

_ _
X + Y X + Y
X G1 X G1 _
+
Home Page X G1+ G 2 Y _ Y
X G1 G 2
G2 G2
Title Page

Contents
3) Eliminating a feedback loop

JJ II
+ Y + Y
X
_
G1 X G1
G1 + G1
J I Y Y
1 _ G1G2
X X
1 + G1G2
G2 G2
Page 5 of 24

Go Back
Negative Feedback Loop Positive Feedback Loop
Full Screen

+ Y + Y
Close X
_ G1 X G1
G1 + G1
X Y X Y
1 + G1 1 _ G1
Quit

Negative Unity Feedback Loop Positive Unity Feedback Loop

 Example: Reduce the following block diagram to the Canonical Form,
then find the closed-loop system transfer function.

G3
+
+ + +
R G1 G4 G2 C
_ +
Home Page
H1
Title Page

Contents
H2
JJ II

J I
Step 1: Combine all cascade blocks

Page 6 of 24
G1 G4 G1 G4
Go Back

Full Screen
Step 2: Combine all parallel blocks

Close
G3
Quit +
+ G2+G 3
G2
 Step 3: Eliminate all minor feedback loops

+ G1G4
G1G4
1-G1G4 H 1
+
H1

Step 4: The new block diagram is shown below

Home Page + G1G4

R G2+G 3 C
Title Page _ 1-G1G4 H 1

Contents

JJ II H2

J I which can be reduced to

+ G1G4 (G2+G 3)
Page 7 of 24

R C
Go Back
_ 1-G1G4 H 1
Full Screen

Close
H2
Quit

Step 5: The closed-loop transfer function is:

C(s) G1 G4 (G2 + G3 )
=
R(s) 1 − G1 G4 H1 + G1 G4 H2 (G2 + G3 )
 3. Performance of second-
order systems
Definition: A second-order system has generally the following standard
form
α
s + 2ζωn s + ωn2
2

where α is the amplitude factor, ζ is the damping ratio and ωn is the natural
frequency.
Home Page
Consider the single loop 2nd -order system shown in figure below.
Title Page
nd
2 order system
Contents
+ ω2n Y
R
JJ II _ G(s)=
s(s+2ξωn)
J I

Page 8 of 24

The closed-loop transfer function is given by

Go Back

G(s) ωn2
Full Screen Y (s) = R(s) =⇒ Y (s) = R(s)
1 + G(s) s2 + 2ζωn s + ωn2
Close

The dynamic response of the system for a unit-step input r(t) (R(s) = 1/s)
Quit
can be expressed as
ωn2 ωn2
Y (s) =
=  
s s2 + 2ζωn s + ωn2
p
s s + ζωn ± ωn ζ 2 − 1
 Performance of second-order
systems (Cont.)
From Laplace transform table, the time-domain response of a second-order
system (for 0 < ζ < 1) is expressed as
e−ζωn t
y(t) = 1 − sin (ωn βt + θ) , t ≥ 0
β
p
where β = 1 − ζ 2 and θ = cos−1 (ζ).
Home Page
The transient response of a second-order system for a step input is shown in
Title Page figure below for different values of ζ
Step Response
Contents
1.8
ξ=0.1
1.6 ξ=0.2
JJ II
1.4
ξ=0.4
1.2
J I ξ=0.7
1

Page 9 of 24 0.8 ξ=1

0.6
ξ=2
Go Back
0.4
0.2
Full Screen
0
0 2 4 6 8 10 12 14
Close ωn t

As ζ decreases, the response becomes increasingly oscillatory.

Quit
• If 0 < ζ < 1, the system is under-damped.
• If ζ = 1, the system is critically damped
• If ζ > 1, the system is over-damped
 Performance of second-order
systems (Cont.)

Home Page

Title Page

Contents

JJ II

J I

Page 10 of 24

Go Back

Full Screen

Close

Quit
 Performance of second-order
systems (Cont.)
Example: Determine the under-damped natural frequency and damping
ratio of the following second order system.
Y (s) 4
= 2
R(s) s + 2s + 4

Home Page Answer: Compare the denominator of the given transfer function with that
of the standard 2nd order transfer function
Title Page
α
Contents s2 + 2ζωn s + ωn2
JJ II we can write
J I s2 + 2ζωn s + ωn2 = s2 + 2s + 4

It is clear from the above equation that
Page 11 of 24

Go Back
ωn2 = 4 =⇒ ωn = 2 rad/s

Full Screen
=⇒ 2ζωn = 2, meaning that ζ = 1/ωn = 0.5

Close Example: Determine the under-damped natural frequency and damping

ratio of the following second order system.
Quit
Y (s) 5
= 2
R(s) 4s + 5s + 14

Answer: ωn = 1.8708 rad/s and ζ = 0.334

 Performance of second-order
systems (Cont.)
For 0 < ζ < 1 and ωn > 0, the 2nd order system’s response due to a unit step
input looks like
y(t)

allowable tolerance
_ 1%~ +
(+ _ 5%)

Home Page 1

Title Page

Contents
0.5
JJ II

J I

Page 12 of 24 0
tr Time (sec)
Go Back tp
ts
Full Screen

Close • The rise time, tr , is the time required for the response to rise from 0%
to 100% of its final value. (for overdamped systems, the 10% to 90%
Quit
rise time is commonly used)
• The settling time, ts , is the time required for the response to reach and
stay within a range about 1% or 2% (to 5%) of the final steady-state
value.
 Performance of second-order
systems (Cont.)
y(t) y(tp)

1 y(∞)

Home Page

0.5
Title Page tp
Contents

JJ II
0
J I tp Time (sec)

Page 13 of 24 • The peak time, tp , is the time required for the response to reach the
first peak of the overshoot.
Go Back

• The maximum overshoot, Mp , is the relative maximum peak value of

Full Screen
the response curve measured from the final value. It is defined as
Close
y(tp ) − y(∞)
Mp =
Quit y(∞)

y(tp ) − y(∞)
O% = Mp % = × 100 (Percent overshoot)
y(∞)
 Performance of second-order
systems (Cont.)
Performance approximation of 2nd order system For a second-order
system with no zeros:
Home Page
α
T (s) = (1)
Title Page s2 + 2ζωn s + ωn2
Contents
the following approximated values (for unit step input) can be obtained (set-
JJ II tling time at 2%):

1.8 π 4 √ 2
J I
tr ≈ , tp ≈ p , ts ≈ , and O% = Mp % = 100e−πζ/ 1−ζ
ωn ωn 1 − ζ 2 ζωn
Page 14 of 24

Go Back
Percent peak overshoot versus damping ratio for a second-order system is
Full Screen presented in table below
Close
Damping ratio 0.9 0.8 0.7 0.6 0.5 0.4 0.3
Quit Percent overshoot 0.2% 1.5% 4.6% 9.5% 16.3% 25.4% 37.2%
 Performance of second-order
systems (Cont.)
Example: For the system shown in figure below, determine the values of
gain K and velocity-feedback constant Kh so that the maximum overshoot
in the unit-step closed-loop response is 0.2 and the peak time is 1s. With
these values of K and Kh , obtain the rise time and settling time. Assume
that J = 1kg.m2 and B = 1N.m/rad/sec.
Home Page

Title Page + + K 1
R(s) Y(s)
Js+B s
Contents - -

JJ II Kh

J I

Page 15 of 24 Solution: The transfer function of the closed-loop system shown in figure
above is given by
Go Back

Full Screen Y (s) K K

= 2 = 2
R(s) Js + (B + KKh )s + K s + (1 + KKh )s + K
Close

Quit The standard 2nd order transfer function is

α
s2 + 2ζωn s + ωn2
 Performance of second-order
systems (Cont.)
Comparing the transfer function obtained above with the standard 2nd order
transfer function leads to
√ 2ζωn − 1 1 + KKh
K = ωn2 (or ωn = K) and Kh = (or ζ = √ )
K 2 K
Home Page • Maximum overshoot is 0.2. • The peak time is 1 s
√ 2
Title Page −π(ζ/ 1−ζ ) π
Mp = e tp = p
ωn 1 − ζ2
Contents √
e−π(ζ/ 1−ζ 2 ) = 0.2 π
JJ II p =1
ωn 1 − ζ2
J I =⇒ ζ = 0.456

Page 16 of 24 =⇒ ωn = 3.53 rad/s

Go Back

ωn = 3.53 rad/s and K = ωn2 =⇒ K = 12.5

Full Screen

2ζωn − 1
Close ζ = 0.456 and Kh = =⇒ Kh = 0.178
K
Quit
Hence, the rise time and settling time can obtained as follows

1.8 4
tr = = 0.51 s and ts = = 2.48 s
ωn ζωn
 4. Steady-state error
• Steady-state response is a measure of system accuracy when a specific
type of input is applied to a control system.
• The steady-state error is the error after the transient response has de-
cayed, leaving only continuous response.
• If the steady-state response of the feedback system output does not agree
with the steady-state of the reference exactly, the feedback system is said
Home Page to have a steady-state error.

Title Page For an LTI system described by a transfer function, L(s), the tracking error
is defined as the difference between the input, r(t), and output, y(t).
Contents

JJ II
e(t) = r(t) − y(t)

J I
R L(s) Y
Page 17 of 24

Go Back The error is then given in Laplace domain by

Full Screen E(s) = R(s) − Y (s) = R(s) − L(s)R(s) = 1 − L(s) R(s)
Close
The steady-state error is expressed as ess = lim e(t) = e(∞), which can also
t→∞
Quit be defined using the final value theorem as ess = lim sE(s).
s→0
Hence, the steady-state error is given by


ess = lim sE(s) = lim sR(s) 1 − L(s)
s→0 s→0
 Steady-state error (Cont.)
5
Example: For an LTI system with a transfer function , deter-
s2 + 7s + 10
mine the steady-state error for unit-step input.

Solution: The error in Laplace domain is given by

 
1 5
E(s) = 1− 2
s s + 7s + 10
Home Page
Hence, the steady-state error is given by
 
Title Page 1 5
ess = lim sE(s) = lim s 1− 2 = 1 − 0.5 = 0.5
s→0 s→0 s s + 7s + 10
Contents

JJ II It can also be verified from the system step response below that

J I ess = r(∞) − y(∞) = 1 − 0.5 = 0.5

Page 18 of 24 1

0.9
Go Back
0.8
ess
Full Screen 0.7

0.6
Output

Close 0.5

0.4
Quit
0.3

0.2

0.1

0
0 0.5 1 1.5 2 2.5 3 3.5 4
Time (sec)
 Steady-state error (Cont.)
For any closed-loop system, the steady-state error is defined as
 
ess = lim sE(s) = lim sR(s) 1 − Tcl (s)
s→0 s→0

where Tcl (s) is the closed-loop transfer function.

Home Page
For negative unity-feedback systems, the steady-state error becomes
Title Page  
  sR(s)
ess = lim sE(s) = lim sR(s) 1 − Tcl (s) = lim
Contents s→0 s→0 s→0 1 + H(s)G(s)
JJ II
+
R H(s) G(s) Y
J I -

Page 19 of 24

Go Back
The steady-state error of the negative unity-feedback control system
above depends on the type of the reference signal, R(s), and the type of
Full Screen the system H(s)G(s).
Close
1. If the reference input is a step function (R(s) = A/s), then the
Quit
A 
steady-state error is ess = , where Kp = lim H(s)G(s)
1 + Kp s→0
is the position error constant. The steady-state error of the closed-
loop system is small when H(0)G(0) >> 1.
 2. For a ramp reference input (R(s) = A/s2 ), the steady-state error
A 
is ess = , where Kv = lim sH(s)G(s) is the velocity error
Kv s→0
constant.

3. For a parabolic reference input (R(s) = A/s3 ), the steady-state

A 
error is ess = , where Ka = lim s2 H(s)G(s) is the accelera-
Ka s→0
tion error constant.

Home Page
Definition: the type number of feedback control system is determined by
the number of poles of the open-loop transfer function H(s)G(s) located at
Title Page the origin. For example, for an open-loop system defined by the following
transfer function H(s)G(s), the feedback control system type is m.
Contents

JJ II
K (1 + b1 s + b2 s2 + . . . + br−1 sr−1 + br sr )
H(s)G(s) =
sm (1 + a1 s + a2 s2 + . . . + an−1 sn−1 + an sn )
J I

Page 20 of 24 The steady-state errors of linear control systems depend on the type of the
reference signal and the type of the system.
Go Back

Full Screen Type number (number of Reference input

integrations in H(s)G(s)) Step A/s Ramp A/s2 Parabolic A/s3
Close
A
0 ess = ∞ ∞
Quit 1 + Kp
A
1 0 ∞
Kv
A
2 0 0
Ka
 Steady-state error (Cont.)
Example: Consider a negative unity-feedback system that has the open-loop
transfer function as
20(s + 1)
Home Page H(s)G(s) =
s(s + 2)(s + 5)
Title Page Determine the steady state-errors for unit step, ramp, and parabolic inputs.
Contents

Solution:
JJ II

J I
Kp = lim H(s)G(s) = ∞ =⇒ ess = 0 (step input)
s→0 
Kv = lim sH(s)G(s) = 2 =⇒ ess = 0.5 (ramp input)
Page 21 of 24 s→0 
Ka = lim s2 H(s)G(s) = 0 =⇒ ess = ∞ (parabolic input)
Go Back s→0

Full Screen Since the open-loop transfer function of this system has one integrator, the
steady-state error for a step input is zero, and hence, the output of the closed-
Close
loop system can perfectly track only the step input.
Quit
 Steady-state error (Cont.)
Example: Consider a second-order negative unity-feedback system whose
open-loop transfer function is given by

(s + 3)
H(s)G(s) =
(s + 1)(s + 2)

Home Page
Calculate the steady-state error for a unit-step input.

Title Page Solution: The steady-state error for a unit-step input (i.e. R(s) = A/s =
1/s) is
Contents 1
ess = (A = 1)
JJ II
1 + Kp
 (s + 3)
J I where Kp = lim H(s)G(s) = lim = 3/2
s→0 s→0 (s + 1)(s + 2)
Page 22 of 24 1
=⇒ ess = = 0.4
Go Back
1 + 1.5

Full Screen

Close

Quit
 Steady-state error (Cont.)
The unit step response of the closed-loop control system is shown in figure
below, from which it can be clearly seen that the steady-state output is equal
to 0.6, and hence, the steady-state error is equal to 1 − 0.6 = 0.4.
1

0.9

0.8 ess

0.7
Home Page
0.6

Output
Title Page 0.5

0.4
Contents
0.3

JJ II 0.2

0.1
J I
0
0 0.5 1 1.5 2 2.5 3 3.5 4
Page 23 of 24 Time (sec)

Go Back Example: Consider the feedback control system shown below. Determine
Full Screen
the range of the gain K to achieve a steady-state error less than 5% for unit
step input.
Close
+ K s+6
R C(s)= G(s)= Y
Quit s+9 (s+2)(s+4)
-
 Steady-state error (Cont.)
Solution: For the negative unity feedback system shown above, the steady-
state error for a unit-step input (i.e. R(s) = A/s = 1/s) is defined as

1
ess = lim sE(s) =
s→0 1 + Kp

such that
Home Page  K(s + 6) 6K
Kp = lim C(s)G(s) = lim =
Title Page s→0 s→0 (s + 9)(s + 2)(s + 4) 72

Contents Hence, the steady-state error can be rewritten as

JJ II 1
ess =
6K
J I 1+
72
Page 24 of 24
It is required that the steady-state tracking error must be ess < 0.05, so
Go Back solving the inequality for the gain K, yields K > 228.
Full Screen
Example: Consider the following feedback system. For e(t) = r(t) − y(t),
Close
find the steady-state error, ess .
+
Quit R
_ G(s) Y

H(s)