Dynamics of Linear Systems

College of Engineering Trivandrum

asokkumarsuma@gmail.com

September 27, 2017

(CET) Dynamics of Linear Systems September 27, 2017 1 / 103

Overview

1 Syllabus

2 Feedback Systems

3 Compensator

(CET) Dynamics of Linear Systems September 27, 2017 2 / 103

Syllabus: Course Objectives

To provide a strong foundation on classical and modern control

theory.
To provide an insight into the role of controllers in a system.
To design compensators using classical methods.
To design controllers in the state space domain.
To impart an in depth knowledge in observer design

(CET) Dynamics of Linear Systems September 27, 2017 3 / 103

Syllabus: Expected Outcome

Analyze a given system and assess its performance.

Design a suitable compensator to meet the required specifications.
Design and tune PID controllers for a given system.
Realize a linear system in state space domain and to evaluate
controllability and observability.
Design a controller and observer for a given system and evaluate its
performance.

(CET) Dynamics of Linear Systems September 27, 2017 4 / 103

Syllabus: References

1. A. Thomas Kailath, Linear System, Prentice Hall Inc., Eaglewood

Cliffs, NJ, 1998
2. M. Gopal, Control Systems-Principles and Design, Tata McGraw-Hill.
3. Richard C. Dorf& Robert H. Bishop, Modern Control Systems
Pearson Education, Limited, 12th Ed., 2013.
4. Gene K. Franklin & J. David Powell, Feedback Control of Dynamic
Systems, Pearson Education, 5th Edition, 2008
5. Control System Design: An Introduction to State Space Methods,
Courier Corporation, 2005

(CET) Dynamics of Linear Systems September 27, 2017 5 / 103

Syllabus: Course Plan
Module I
Design of feedback control systems- Approaches to system design
compensators performance measures- cascade compensation networks
phase lead and lag compensator design using both Root locus and Bode
plots (7 hours)

Module II
PID controllers-effect of proportional, integral and derivative gains on
system performance-PID tuning-integral windup and solutions (7 hours)

Module III
State Space Analysis and Design- Analysis of stabilization by pole
cancellation- reachability and constructability - stabilizability -
controllability - observability-grammians-Analysis of stabilization by output
feedback-Transfer function approach - state feedback and zeros of the
transfer function.Solution
(CET)
of Linear Time Varying Systems
Dynamics of Linear Systems
(10 hours) 6 / 103
September 27, 2017
Syllabus: Course Plan

Module IV
Linear state variable feedback for SISO systems, -modal controllability
formulae for feedback gain -significance of controllable Canonic form-
Ackermann’s formula feedback gains in terms of Eigen values
Mayne-Murdoch formula - non controllable realizations and stabilizability
-controllable and uncontrollable modes - regulator problems (12 hours).

Module V
Observers: Asymptotic observers for state measurement-open loop
observer-closed loop observer-formulae for observer gain implementation of
the observer - full order and reduced order observers separation principle -
combined observer -controller optimality criterion for choosing observer
poles (10 hours).

(CET) Dynamics of Linear Systems September 27, 2017 7 / 103

Syllabus: Course Plan

Module VI
Direct transfer function design procedures Design using polynomial
equations - Direct analysis of the Diophantine equation. MIMO systems:
Introduction, controllability, observability, different companion forms for
MIMO systems (10 hours).

(CET) Dynamics of Linear Systems September 27, 2017 8 / 103

Feedback Systems

What is Feedback?
The term feedback is used to refer to a situation in which two (or more)
dynamical systems are connected together such that each system
influences the other and their dynamics are thus strongly coupled. By
dynamical system, we refer to a system whose behavior changes over time,
often in response to external stimulation or forcing.

(CET) Dynamics of Linear Systems September 27, 2017 9 / 103

Feedback Systems

What is Feedback?
Figure 1.1 illustrates in block diagram form the idea of feedback. We often
use the terms open loop and closed loop when referring to such systems. A
system is said to be a closed loop system if the systems are interconnected
in a cycle, as shown in Figure 1.1a. If we break the interconnection, we
refer to the configuration as an open loop system, as shown in Figure 1.1b.

(CET) Dynamics of Linear Systems September 27, 2017 10 / 103

Feedback Systems: What is Control?

The term control has many meanings and often varies between
communities.
Control includes such examples as feedback loops in electronic
amplifiers, setpoint controllers in chemical and materials processing,
fly-by-wire systems on aircraft and even router protocols that control
traffic flow on the Internet.
Emerging applications include high confidence software systems,
autonomous vehicles and robots, real-time resource management
systems and biologically engineered systems.
At its core, control is an information science, and includes the use of
information in both analog and digital representations.

(CET) Dynamics of Linear Systems September 27, 2017 11 / 103

Feedback Systems: What is Control?

A modern controller senses the operation of a system, compares that

against the desired behavior, computes corrective actions based on a
model of the systems response to external inputs and actuates the
system to effect the desired change.
This basic feedback loop of sensing, computation and actuation is the
central concept in control.
The key issues in designing control logic are ensuring that the
dynamics of the closed loop system are stable (bounded disturbances
give bounded errors) and that they have additional desired behavior
(good disturbance rejection, fast responsiveness to changes in
operating point, etc).

(CET) Dynamics of Linear Systems September 27, 2017 12 / 103

Feedback Systems: What is Control?

These properties are established using a variety of modeling and

analysis techniques that capture the essential dynamics of the system
and permit the exploration of possible behaviors in the presence of
uncertainty, noise and component failures.

(CET) Dynamics of Linear Systems September 27, 2017 13 / 103

Compensator
What is compensator?
A compensator is an additional component or circuit that is inserted into a
control system to compensate for a deficient performance.

What is compensation?

In control systems certain performance specifications are to be met.

In many practical situations the plant may be fixed and not
modifiable. We must adjust parameters other than those in the fixed
plant. This can be done by adding poles or zeroes.
The process of addition of poles and zeroes into a system to meet
desired specifications is called compensation.
The additional device which is introduced into the system to
compensate a deficient behaviour is called compensator.
(CET) Dynamics of Linear Systems September 27, 2017 14 / 103
Compensator

Necessity of Compensation

To meet the desired performance of the system.

Compensate an unstable system to make it stable.
To improve response of already stable system.
To improve transient response specifications- like minimize overshoot ,
reduce settling time etc.
To increase the steady state accuracy of the system-reduce steady state
error.

(CET) Dynamics of Linear Systems September 27, 2017 15 / 103

Compensator

Methods of Compensation

Compensation for improved performance may be provided to the

control system by inserting a network either in series as a cascade or
series compensation or in parallel as feedback compensation.

Series compensation
1 Compensator connected in series
with the plant.
Parallel/Feedback compensation
1 Compensator connected in inner
feedback loop.

(CET) Dynamics of Linear Systems September 27, 2017 16 / 103

Compensator

Choice of compensation

The choice between series compensation and parallel compensation

depends on:
The nature of the signals
The power levels at various points
Available components
The designer’s experience
Economic considerations and so on

(CET) Dynamics of Linear Systems September 27, 2017 17 / 103

Compensator

Commonly used compensator’s

Lead compensator
Lag compensator
Lag-Lead compensator

Compensator’s are realized by:

Electronic devices such as circuits using OPAMP

Electrical networks (RC networks)
Mechanical networks (Spring, Mass, Dampers)
Pneumatic

(CET) Dynamics of Linear Systems September 27, 2017 18 / 103

Compensator

Cascade compensation Networks

The compensation network function Gc (s) is cascaded with the speceified
process G (s) in order to provide a suitable loop transfer function
L(s) = Gc (s)G (s)H(s).

The compensator Gc (s) can be chosen to alter either the shape of the root
locus or the frequency response.

In either case, the network may be chosen to have a transfer function

K M
Q
(s + zi )
Gc (s) = Qn i=1 (1)
j=1 (s + pj )

(CET) Dynamics of Linear Systems September 27, 2017 19 / 103

Compensator
First order compensator
First order compensator with a transfer function

K (s + z)
Gc (s) = (2)
(s + p)

The design problem then becomes the selection of z, p and K in order to

provide a suitable performance. When |z| < |p|, the network is called a
phase lead network.

Figure : 3. Pole-zero diagram of phase-lead network

(CET) Dynamics of Linear Systems September 27, 2017 20 / 103
Compensator

First order compensator

If the pole was negligible, that is |p| >> |z|, and zero occurred at the
origin of the s-plane, it will have a differentiator so that
K
Gc (s) = s (3)
p

Thus, a compensation network of the form of equ. 2 is a differentiator

type network. The differential network of equ. 3 has the frequency
characteristic and phase angle of +900
 
K K 0
Gc (jω) = j ω = ω e +j90 (4)
p p

(CET) Dynamics of Linear Systems September 27, 2017 21 / 103

Compensator

First order compensator

The frequency response of differentiating network of equ. 2 is

K (jω + z) (Kz/p)[j(ω/z) + 1] K1 (1 + jωατ )

Gc (jω) = = = (5)
jω + p j(ω/p) + 1 1 + jωτ

where τ = 1/p, p = αz, and K1 = K /α. The angle of the frequency

characteristic is
φ(ω) = tan−1 (αωτ ) − tan−1 (ωτ ) (6)

(CET) Dynamics of Linear Systems September 27, 2017 22 / 103

Compensator

First order compensator

Figure : 4. Bode diagram of phase-lead network

(CET) Dynamics of Linear Systems September 27, 2017 23 / 103

Compensator
From the phase-lead network, Transfer function can be obtained as

Using KCL
i1 + i2 = i
d(vi − v0 ) 1 1
C + (vi − v0 ) = v0 (7)
dt R1 R2
Taking LT
1 1 1
sCVi (s) − sCV0 (s) + Vi (s) − V0 (s) = V0 (s)
R1 R1 R2

(CET) Figure Dynamics

: 5. Phase-lead network
of Linear Systems September 27, 2017 24 / 103
Compensator

From the phase-lead network, Transfer function can be obtained as

   
1 1 1
Vi (s) sC + = V0 (s) sC + + (8)
R1 R1 R2
h i
1
V0 (s) s + R1 C
=
Vi (s) s + R11 +R
(R
R2 C
2)

h i
s + R11C
= (9)
s + h R2 1 i
R1 +R2
R1 C

S + T1
=
s + α1 T

(CET) Dynamics of Linear Systems September 27, 2017 25 / 103

Compensator

From the phase-lead network, Transfer function can be obtained as

Where
T = R1 C and α = R1R+R
2
2
<1

The maximum value of α is generally 0.5

The lead compensator has a zero at s = −1/T and pole at s = −1/αT
Since 0 < α < 1, the zero is always located to the right of the pole

Figure : 3. Pole-zero diagram of phase-lead network

(CET) Dynamics of Linear Systems September 27, 2017 26 / 103

Compensator
maximum phase-lead angle, and α
V0 (s) α(1 + jωT )
= (10)
Vi (s) 1 + (jωαT )
Magnitude
√
V0 (s) α 1 + ω2T 2
| |=M= √ (11)
Vi (s) 1 + α2 ω 2 T 2

The phase angle

φ = tan−1 ωT − tan−1 αωT (12)

Differentiating the phase angle w.r.to ω and equating with zero,

2 1 1
ωm = 2
andωm = √ (13)
αT T α
(CET) Dynamics of Linear Systems September 27, 2017 27 / 103
Compensator

maximum phase-lead angle, and α

ωm is the mean of the two corner frequencies of the compensator
which are ωc1 = 1/T and ωc2 = 1/αT

tanφ = tan tan−1 ωT − tan−1 αωT

 

ωT − αωT ωT (1 − α) (14)
= =
1 + ωT αωT 1 + ω2T 2α
√
at ω = ωm = 1/T α, and

1−α 1−α
tanφ = √ = √ (15)
α(1 + 1) 2 α

(CET) Dynamics of Linear Systems September 27, 2017 28 / 103

Compensator

maximum phase-lead angle, and α

 
1−α 1−α
sinφm = or φm = sin−1 (16)
1+α 1+α
The above equation relate α and the maximum lead angle φm .

(CET) Dynamics of Linear Systems September 27, 2017 29 / 103

Compensator

ωc1 = 1/T for a simple zero at s = −1/T and

ωc2 = 1/αT for a simple pole at s = −1/αT and K = α

Figure : 4. Bode diagram of phase-lead network

(CET) Dynamics of Linear Systems September 27, 2017 30 / 103

Compensator:Frequency Response approach

In frequency domain design is to reshape the frequency response

characteristics, so that the desired specifications are satisfied.
In bode plot, the desired performance specifications are given in terms
of frequency domain specifications and steady state error requirement.
The stability requirement is specified in terms of gain cross over
frequency, band width and resonant frequency.

(CET) Dynamics of Linear Systems September 27, 2017 31 / 103

Design of Lead Compensator by Bode plot: Procedure

Step 1
The lead compensator has a gain α at zero frequency. It provides an
attenuation for α < 1 The practical lead compensator is designed
with an amplifier having a gain Kc in series with the basic lead
network to cancel the attenuation.
Practical TF of a lead compensator is assumed as

(1 + Ts) (1 + Ts)
Gc (s) = Kc α =K (17)
1 + αTs 1 + αTs
where, K is DC gain = Kc α
Open loop transfer function for the system

K (1 + Ts)
Gc (s)G (s) = G (s) (18)
1 + αTs

(CET) Dynamics of Linear Systems September 27, 2017 32 / 103

Design of Lead Compensator by Bode plot

(1 + Ts) (1 + Ts)
Gc (s)G (s) = KG (s) = G1 (s) (19)
1 + αTs 1 + αTs
where, G1 (s) = KG (s)
This is the open loop TF of compensated network.

Step 2
The bode plot of G (jω) is drawn using the value of K . The phase
margin is computed from the Bode plot.

Step 3
PM is given for a design problem, the necessary phase lead φm
required to be added is determined using the relation φm = φ + ε,
where, ε is the margin of safety as crossover frequency may shift due
to compensation and ε = 50 − 150 .
(CET) Dynamics of Linear Systems September 27, 2017 33 / 103
Design of Lead Compensator by Bode plot

Step 4
1−α
Find the value of α from the equation sinφm = 1+α

Step 5
The frequency ωm is calculated so that
the magnitude of
uncompensated system is −10 log10 α1 dB at ω = ωm . This
1
frequency is taken as new gain crossover frequency, where ωm = T
√
α
Here, α is known. Therefore, can determine 1/T

Step 6
The two following corner frequencies of the lead compensator are
determined.
1 1
ωc1 = andωc2 =
T αT
(CET) Dynamics of Linear Systems September 27, 2017 34 / 103
Design of Lead Compensator by Bode plot

Step 7
Since K = Kc α, can obtain the value of Kc

Step 8
If the gain margin of tje compensated system is not satisfactory, the
design is repeated by modifying the pole-zero location of the
compensator till a satisfactory result is obtained.

(CET) Dynamics of Linear Systems September 27, 2017 35 / 103

Design of Lead Compensator by Bode plot

Effects of lead compensation

Since a lead compensator adds a dominant zero and a pole, the
damping of the closed-loop system is increased. The less overshoot,
less rise time and less settling time are obtained due to increase of
damping and hence there is improvement in the transient response.
it improves the phase margin of the closed-loop system.
improvement of gain and phase margin and hence relative stability
occurs due to the reduction in the slope of the magnitude plot in a
bode diagram of the forward path transfer function at the gain
crossover frequency.
Bandwidth of the closed loop system is increased and hence the
response is faster.
The steady-state error does not get affected.

(CET) Dynamics of Linear Systems September 27, 2017 36 / 103

Design of Lead Compensator by Bode plot

Example 1
A system shown in the fig. Design a lead compensation such that the
closed loop system satisfy the following specifications.
Static velocity error constant = 24s − 1
Phase margin = 550
Gain margin 1 13 dB

(CET) Dynamics of Linear Systems September 27, 2017 37 / 103

Design of Lead Compensator by Bode plot: Example 1

Step 1
(1 + Ts) (1 + Ts)
Gc (s) = Kc α =K
1 + αTs 1 + αTs
12K
G1 (s) = KG (s) =
s(s + 2)
Kv = 24 = lim sG (s)Gc (s)
s→0

12K (1 + Ts)
24 = lim s
s→0 s(s + 2)(1 + αTs)
24 = 6K , K = 4
48
G1 (s) =
s(s + 2)

(CET) Dynamics of Linear Systems September 27, 2017 38 / 103

Design of Lead Compensator by Bode plot: Example 1
Step 2

Figure : 6. Bode diagram of uncompensated system

φ1 = PM = 16.40 , ωgc = 6.8rad/sec., GM = +∞dB

(CET) Dynamics of Linear Systems September 27, 2017 39 / 103
Design of Lead Compensator by Bode plot: Example 1

Step 3

φs = 550 , φm = φs − φ1 + Er
Let, Er = 50 , φm = 550 − 16.40 + 50 = 43.60

Step 4
1−α 1−α
sinφm = ∴ sin43.60 = = 0.6896
1+α 1+α
ie.,α = 0.1837 w 0.18

(CET) Dynamics of Linear Systems September 27, 2017 40 / 103

Design of Lead Compensator by Bode plot: Example 1
Step 5
 
1
−10 log10 = −7.44 dB for α = 0.18
α
From the uncompensated bode plot, the frequency at which the gain of
the uncompensated system is -7.44 dB is ωm = 10.6 rad/sec.
We know that
1 1 √ √
ωm = √ ie., = ωm α = 10.6 × 0.18 = 4.497
T α T

Step 6
Two corner frequencies of the lead compensator are,
1 1 1
ωc1 = = 4.497 and ωc2 = = 4.497 = 24.98
T αT 0.18
(CET) Dynamics of Linear Systems September 27, 2017 41 / 103
Design of Lead Compensator by Bode plot: Example 1

Step 7
K 4
K = Kc α = 0.18 ∴ Kc = = == 22.22
α 0.18

Step 8
(1 + 0.2224s) 3.99(1 + 0.2224s)
Gc (s) = 22.22 × 0.18 =
(1 + 0.0400s) (1 + 0.0400s)
This is the designed compensator.

12 3.99(1 + 0.2224s)
∴ Gc (s)G (s) = ×
s(s + 2) (1 + 0.0400s)
47.88(1 + 0.2224s)
=
s(s + 2)(1 + 0.0400s)

(CET) Dynamics of Linear Systems September 27, 2017 42 / 103

Design of Lead Compensator by Bode plot: Example 1

Figure : 7. Bode diagram of compensated system

φ1 = PM = 54.80 w 550 , ωgc = 10.5rad/sec., GM = ∞dB

Therefore, the compensated system satisfies all the specifications.
(CET) Dynamics of Linear Systems September 27, 2017 43 / 103
Compensator

Phase-lag network

Figure : 8. Phase-lag network

Phase-lag network is shown in the figure 8, the transfer function of the

phase-lag network is
 
V0 (s) 1 s + 1/T
= (20)
Vin (s) β s + 1/βT
Generally, β is taken to be 10.
(CET) Dynamics of Linear Systems September 27, 2017 44 / 103
Compensator
The lag compensator has zero at z = −1/T and pole at p = −1/(βT ),
since, β > 1, the pole lies closed to the origin of the s-plane, as shown in
the figure 9. This type of compensating network is often called an
integrating network.
The Bode diagram shown in the figure 8, of the network is obtained from
the transfer function

Figure : 9. Pole-zero diagram of Phase-lag network

(CET) Dynamics of Linear Systems September 27, 2017 45 / 103

Compensator
maximum phase-lag angle, and β
V0 (s) (1 + jωT )
= (21)
Vi (s) 1 + (jωβT )
Magnitude
√
V0 (s) 1 + ω2T 2
| |=M= p (22)
Vi (s) 1 + β 2ω2T 2

The phase angle

φ = tan−1 ωT − tan−1 (βωT ) (23)

Differentiating the phase angle w.r.to ω and equating with zero,
s  
1 1 1
ωm = √ = (24)
T β T βT
(CET) Dynamics of Linear Systems September 27, 2017 46 / 103
Compensator

maximum phase-lag angle, and β

1
The lag compensator has the following corner frequencies: ωc1 = βT for a
1
small pole at s = − βT and
ωc2 = T1 for a small pole at s = − T1
The pole dominates over zero. The transfer function of basic lag network
is,
(1 + Ts)
T (s) = (25)
1 + βTs)

(CET) Dynamics of Linear Systems September 27, 2017 47 / 103

Compensator
The Bode diagram shown in the figure 8, of the network is obtained from
the transfer function

Figure : 8. Bode diagram of Phase-lag network

(CET) Dynamics of Linear Systems September 27, 2017 48 / 103

Design of Lag Compensator by Bode plot: Procedure

Step 1
Assume the lag compensator
!
V0 (s) 1 s + T1 (1 + Ts)
Gc (s) = = 1
= (26)
Vin (s) β s + βT (1 + βTs)

It is taken as G1 (s) = KG (s). The value of K is given from the error

constant and it must satisfy the steady-state performance.

Step 2
The bode plot of G1 (jω) is drawn using the value of K . The phase
margin (φ1 ) is computed from the Bode plot.

(CET) Dynamics of Linear Systems September 27, 2017 49 / 103

Design of Lag Compensator by Bode plot: Procedure

Step 3
Usually the PM is specified. Let the given phase margin be φs .
Therefore, φ2 = φs + ε. Here, ε compensate for the phase lag of the
lag compensator and ε is the margin of safety = 50 − 150 .

Step 4
Corresponding the phase margin of φ2 degrees, the frequency ω2 is
determined, ie., the frequency at which the phase angle of an open
loop transfer function is 1800 + φ2 . This frequency is taken as the
new gain crossover frequency.

(CET) Dynamics of Linear Systems September 27, 2017 50 / 103

Design of Lag Compensator by Bode plot: Procedure

Step 5
For ω2 as the new gain crossover frequency, the attenuation to shift
the magnitude curve up of down to 0dB is calculated and this shift is
due to the contribution of β which is 20 log10 (1/β).
∴ shift to ω2 as new gain crossover = 20 log10 (1/β) = −20 log10 β.
Down shift must be taken negative while up shift must be taken
positive.
The value of β is determined.

(CET) Dynamics of Linear Systems September 27, 2017 51 / 103

Design of Lag Compensator by Bode plot: Procedure

Step 6
The upper corner frequency 1/T is chosen which 1/2 or 1/10 below
the ω2 determined in step 4.
1 ω2 ω2
∴ ωc2 = = or,
T 2 10
This determine the value of T
1
For the lag compensator, the other corner frequency is ωc1 = βT

Step 7
Plot the compensated system Bode plot. If the gain margin of the
compensated system is not satisfactory, the design is repeated by
modifying the pole-zero location of the compensator till a satisfactory
result is obtained.
(CET) Dynamics of Linear Systems September 27, 2017 52 / 103
Design of Lag Compensator by Bode plot

Effects of lag compensation

Since a lag compensator allows high gain at low frequencies, it is
basically a low pass filter. Therefore, it improves the steady state
performance.
Since the attenuation due to a lag compensator shifts the gain
crossover frequency to a lower frequency point, the band width of the
system is reduced. Hence a slower response is occurs. Therefore, rise
time and settling time are usually longer and hence the transient
response lasts for longer time.
The system is very sensitive to parameter variation.
The system tends to make a less stable.

(CET) Dynamics of Linear Systems September 27, 2017 53 / 103

Design of Lag Compensator by Bode plot

Example 2
A system shown in the fig. Design a suitable lag compensator to
satisfy the following specifications.
Velocity error constant = 27.3s − 1
Phase margin = 550

(CET) Dynamics of Linear Systems September 27, 2017 54 / 103

Design of Lag Compensator by Bode plot: Example 2

Step 1
0.035K
G1 (s) = KG (s) =
s(1 + 0.5s)(1 + 0.04s)
Kv = 27.3 = lim sG1 (s)Gc (s)
s→0

0.035K (1 + Ts)
27.3 = lim s
s→0 s(1 + 0.5s)(1 + 0.04s) (1 + βTs)
27.3 = 0.035K , K = 780
0.035 × 780 27.3
G1 (s) = =
s(1 + 0.5s)(1 + 0.04s) s(1 + 0.5s)(1 + 0.04s)

(CET) Dynamics of Linear Systems September 27, 2017 55 / 103

Design of Lag Compensator by Bode plot: Example 2
Step 2

(CET) Dynamics of Linear Systems September 27, 2017 56 / 103

Design of Lag Compensator by Bode plot: Example 2

Step 3

φs = 450 , ε = 50
φ2 = φs + ε = 450 + 50 = 500

Step 4
The frequency ω2 at phase margin becomes 500 is given by
ω2 = 1.48rad/sec.

(CET) Dynamics of Linear Systems September 27, 2017 57 / 103

Design of Lag Compensator by Bode plot: Example 2

Step 5
Let ω2 =1.48 rad/sec. be the new gain crossover. It is necessary to bring
the magnitude curve down by 23.3 dB.

−20 log10 β = −23.3, ∴ β = 14.62

Step 6
Choose
ω2 1.48
ωc2 = = = 0.148rad/sec.
10 10
1
ωc2 = = 0.148rad/sec.
T
1 1
T = = 6.75 and ωc1 = = 0.0101
0.148 βT

(CET) Dynamics of Linear Systems September 27, 2017 58 / 103

Design of Lag Compensator by Bode plot: Example 2

Step 7
The lag compensator is

(1 + 6.75s)
Gc (s) =
(1 + 98.68s)

The transfer function for this compensated system is;

27.3(1 + 6.75s)
Gc (s)G (s) = (27)
s(1 + 0.5s)(1 + 0.04s)(1 + 98.68s)

(CET) Dynamics of Linear Systems September 27, 2017 59 / 103

Design of Lag Compensator by Bode plot: Example 2

Figure : 7. Bode diagram of compensated system

PM = 45.50 , ωpc = 6.8rad/sec., GM = 22.5dB, ωgc = 1.5rad/sec.

(CET) Dynamics of Linear Systems September 27, 2017 60 / 103
Compensator
Lag-Lead compensation network
Lag-Lead compensator is a combination of a lag compensator and a
lead compensator.
The lag-section has one real pole ane one real zero with the pole to
the right of zero.
The lead-section also has one real pole and one real zero but the zero
is to the right of the pole.
These compensator used when both transient and steady-state
performance of the system have to be reshaped.

(CET) Dynamics of Linear Systems September 27, 2017 61 / 103

Compensator

Lag-Lead compensation network

Overall transfer function of the network

V0 (s) R2 + 1/C2 s
Gc (s) = =
Vi (s) R2 + (1/C2 s) + RR11+(1/C
(1/C1 s)
s)
  1  
1
s + R1 C1 s + R21C2
=   
(28)
s 2 + R11C1 + R21C1 + R21C2 s + R1 R21C1 C2
   
s + T11 s + T12
=     , β > 1, α < 1
1 1
s + βT1 s + αT2

Lag section Lead section

(CET) Dynamics of Linear Systems September 27, 2017 62 / 103
Compensator

where, T1 = R1 C1 , T2 = R2 C2

 
(s + 1/T1 ) (s + 1/T2 ) (s + z1 ) (s + z2 )
Gc (s) = = (29)
(s + 1/βT1 ) (s + 1/αT2 ) (s + p1 ) (s + p2 )

Figure : 10. Pole-zero diagram of Phase lag-lead network

(CET) Dynamics of Linear Systems September 27, 2017 63 / 103

Compensator

Figure : 11. Bode diagram of Phase lag-lead network

(CET) Dynamics of Linear Systems September 27, 2017 64 / 103

Design of Lag-Lead Compensator by Bode plot: Example 3

Example 3
Consider the following system with transfer function
1
G (s) =
s(1 + 0.1s)(1 + 0.2s)

Design a lag-lead compensator C(s) such that the phase margin of the
compensated system is at least 450 at gain crossover frequency around 10
rad/sec and the velocity error constant Kv is 30.
 
(1 + T1 s) (1 + T2 s)
Gc (s) = K where, β > 1 and α < 1
(1 + βT1 s) (1 + αT2 s)

Kv = 30 = lim sG1 (s)Gc (s) where G1 (s) = KG (s) ∴ K = 30

s→0

(CET) Dynamics of Linear Systems September 27, 2017 65 / 103

Design of Lag-Lead Compensator by Bode plot: Example 3

30
Uncompensated system is G1 (s) =
s(1 + 0.1s)(1 + 0.2s)

(CET) Dynamics of Linear Systems September 27, 2017 66 / 103

Design of Lag-Lead Compensator by Bode plot: Example 3

From the plot, PM = 17.20 , ωgc = 9.77rad/sec.

Since the PM of the uncompensated system with K is negative, we

need a lead compensator to compensate for the negative PM and
achieve the desired phase margin.
However, we know that introduction of a lead compensator will
eventually increase the gain crossover frequency to maintain the low
frequency gain.
Thus the gain crossover frequency of the system cascaded with a lead
compensator is likely to be much above the specified one, since the
gain crossover frequency of the uncompensated system with K is
already 9.77 rad/sec.

(CET) Dynamics of Linear Systems September 27, 2017 67 / 103

Design of Lag-Lead Compensator by Bode plot: Example 3

Thus a lag-lead compensator is required to compensate for both.

We design the lead part first.
From the bode diagram, it is seen that at 10 rad/sec the phase angle
of the system is 1980 .
Since the new !g should be 10 rad/sec, the required additional phase
at ωg , to maintain the specified PM, is 45 (180 198) = 630 . With
safety margin 20 ,
1 − sin650
 
1
α= = 0.05 and 10 = √
1 + sin650 T1 α

(CET) Dynamics of Linear Systems September 27, 2017 68 / 103

Design of Lag-Lead Compensator by Bode plot: Example 3

which gives T2 = 0.45. However, introducing this compensator will

actually increase the gain crossover frequency where the phase
characteristic will be different than the designed one. This can be
seen from Figure below bode diagram

(CET) Dynamics of Linear Systems September 27, 2017 69 / 103

Design of Lag-Lead Compensator by Bode plot: Example 3

The gain crossover frequency is increased to 23.2 rad/sec. At 10

rad/sec, the phase angle is 1340 and gain is 12.6 dB. To make this as
the actual gain crossover frequency, lag part should provide an
attenuation of 12.6 dB at high frequencies.
At high frequencies the magnitude of the lag compensator part is
1/β. Thus ,
20 log10 β = 12.6,
which gives β = 4.27. Now, 1/T1 should be placed much below the
new gain crossover frequency to retain the desired PM. Let 1/T1 be
0.25. Thus T1 = 4

(CET) Dynamics of Linear Systems September 27, 2017 70 / 103

Design of Lag-Lead Compensator by Bode plot: Example 3

The overall compensator is

1 + 4s 1 + 0.45s
C1 (s) = 30 (30)
1 + 17.08s 1 + 0.0225s
The frequency response of the system after introducing the above
compensator is shown in Figure, which shows that the desired
performance criteria are met.

(CET) Dynamics of Linear Systems September 27, 2017 71 / 103

Design of Lag-Lead Compensator by Bode plot: Example 3

(CET) Dynamics of Linear Systems September 27, 2017 72 / 103

Design of Compensator

Phase Lead Design using Bode diagram: Example 4

Consider the following system,
1
G (s) = , H(s) = 1
s(s + 1)

Design a cascade lead compensator so that the PM is at least 450 and

steady state error for a unit ramp input is ≤ 0.1

(CET) Dynamics of Linear Systems September 27, 2017 73 / 103

Design of Compensator

Phase Lead Design using Bode diagram: Example 4

The lead compensator is
τs + 1
C (s) = K , where,α < 1
ατ s + 1
when s −→ 0, C (s) −→ K
Steady state error for unit ramp input is
1 1 1
= =
lims→0 sC (s)G (s) C (0) K

Thus K1 = 0.1, or K = 10 PM of the closed loop system should be 450 . Let

the gain crossover frequency of the uncompensated system with K be ωg .

(CET) Dynamics of Linear Systems September 27, 2017 74 / 103

Design of Compensator

Phase Lead Design using Bode diagram: Example 4

1
G (jω) =
jω(jω + 1)
1
Mag . = √
ω 1 + ω2
Phase = −900 − tan−1 ω
10
⇒ q =1
ωg 1 + ωg2

100
=1
ωg2 (1 + ωg2 )
⇒ ωg = 3.1

(CET) Dynamics of Linear Systems September 27, 2017 75 / 103

Design of Compensator

Phase Lead Design using Bode diagram: Example 4

Phase angle at ωg = 3.1 is −90 − tan−1 3.1 = −1620 . Thus the PM of
the uncompensated system with K is 180 .
If it was possible to add a phase without altering the magnitude, the
additional phase lead required to maintain PM = 450 is 450 − 180 = 270
at ωg = 3.1 rad/sec.
However, maintaining same low frequency gain and adding a compensator
would increase the crossover frequency. As a result of this, the actual
phase margin will deviate from the designed one. Thus it is safe to add a
safety margin of ε to the required phase lead so that if it devaites also, still
the phase requirement is met. In general ε is chosen between 50 to 150 .

(CET) Dynamics of Linear Systems September 27, 2017 76 / 103

Design of Compensator

Phase Lead Design using Bode diagram: Example 4

So the additional phase requirement is 270 + 100 = 370 , The lead part of
the compensator will provide this additional phase at ωmax .
Thus
φmax = 37
 
1 − sin(φmax )
⇒α= = 0.25
1 + sin(φmax )
The only parameter left to be designed is τ . To find τ , one should locate
the frequency at which the uncompensated system has a logarithmic
1
magnitude of −20 log10 √ .
α

(CET) Dynamics of Linear Systems September 27, 2017 77 / 103

Design of Compensator

Phase Lead Design using Bode diagram: Example 4

Select this frequency as the new gain crossover frequency since the
1
compensator provides a gain of 20 log10 √ at ωmax . Thus
α
1
ωmax = ωg new = √
τ α

In this case ωm ax = ωg new = 4.41 . Thus

1
τ= √ = 0.4535 The lead compensator is thus
4.41 α
0.4535s + 1
C (s) = 10
0.1134s + 1
With this compensator actual phase margin of the system becomes 49.60
which meets the design criteria.

(CET) Dynamics of Linear Systems September 27, 2017 78 / 103

Design of Compensator
Phase Lead Design using Bode diagram: Example 4
The corresponding Bode plot is shown in Figure.

(CET) Dynamics of Linear Systems September 27, 2017 79 / 103

Design of Compensator

Phase Lag Design using Bode diagram: Example 5

1
Consider the following system G (s) = , H(s) = 1 Design
(s + 1)(0.5s + 1)
a lag compensator so that the phase margin (PM) is at least 500 and
steady state error to a unit step input is ≤ 0.1.
τs + 1
The overall compensator is C (s) = KClag (s) = K α , where, α > 1
ατ s + 1
When s → 0, C (s) → K α
Steady state error for unit step input is
1 1 1
= =
1 + lims→0 C (s)G (s) 1 + C (0) 1 + Kα
1
Thus, = 0.1 or K α = 9
1 + Kα

(CET) Dynamics of Linear Systems September 27, 2017 80 / 103

Design of Compensator

Phase Lag Design using Bode diagram: Example 5

Now let us modify the system transfer function by introducing K with the
original system. Thus the modified system becomes
K
Gm (s) = PM of the closed loop system should be 500 .
(s + 1)(0.5s + 1)
Let the gain crossover frequency of the uncompensated system with K be
ωg .
K
Gm (jω) =
(jω + 1)(0.5jω + 1)
K
Mag . = √ √
1+ ω2 1 + 0.25ω 2
Phase = −tan−1 ω − tan−1 0.5ω

(CET) Dynamics of Linear Systems September 27, 2017 81 / 103

Design of Compensator
Phase Lag Design using Bode diagram: Example 5
Required PM is 500 . Since the PM is achieved only by selecting K , it
might be deviated from this value when the other parameters are also
designed. Thus we put a safety margin of 50 to the PM which makes the
required PM to be 550 .

⇒ 1800 − tan−1 ωg − tan−1 0.5ωg = 550

ωg + 0.5ωg
or , −tan−1 = 1250
1 − 0.5ωg2
1.5ωg
or , −tan−1 = tan1250 = −1.43
1 − 0.5ωg2
or , 0.715ωg2 − 1.5ωg − 1.43 = 0
⇒ ωg = 2.8rad/sec
(CET) Dynamics of Linear Systems September 27, 2017 82 / 103
Design of Compensator

Phase Lag Design using Bode diagram: Example 5

To make ωg = 2.8rad/sec, the gain crossover frequency of the modified
system, magnitude at ωg should be 1. Thus
K
q q =1
1 + ωg2 1 + 0.25ωg2
Putting the value of ωg in the last equation, we get K = 5.1.
9
Thus, α = = 1.76
K
The only parameter left to be designed is τ . Since the desired PM is
already achieved with gain K , We should place ω = 1/τ such that it does
not much effect the PM of the modified system with K .

(CET) Dynamics of Linear Systems September 27, 2017 83 / 103

Design of Compensator

Phase Lag Design using Bode diagram: Example 5

If we place 1/τ one decade below the gain crossover frequency, than,
1 2.8
= , or, τ = 3.57
τ 10
The overall compensator is
3.57s + 1
C (s) = 9
6.3s + 1
With this compensator actual phase margin of the system becomes 52.70 ,
as shown in Figure, which meets the design criteria.

(CET) Dynamics of Linear Systems September 27, 2017 84 / 103

Design of Compensator

Figure : Bode plot of the compensated system for Example 5

(CET) Dynamics of Linear Systems September 27, 2017 85 / 103
Compensator:Root locus Design
Transient Response Specifications requirements is specified in terms of:
Peak overshoot
Settling time
Rise time for a step input

Steady State error requirements is specified in terms of:

Error constants in standard inputs

The root loci are reshaped by introducing a compensator

The compensator will add a pole/or a zero in the open loop transfer
function
The addition of a pole to the open loop transfer function has the
effect of pulling the root locus to the right, which reduce the relative
stability of the system and increase the settling time and improve the
steady state performance.
(CET) Dynamics of Linear Systems September 27, 2017 86 / 103
Compensator:Root locus Design

The addition of zero to the open loop transfer function has the effect
of pulling the root locus to the left which make the system more
stable and reduce the settling time

The transfer function of lead compensator will have a zero and pole:
The zero can be placed in the real axis to satisfy transient response
specifications.
The introduction of zero will amplify high frequency noise which is
eliminated by pole.
If the pole is located far away from zero then it will not be effective in
suppressing the noise.
If the pole is too close to zero then it will not allow the zero to do its
job. In order to avoid this conflict, the pole is located at 3 to 10
times the value of zero locations.
(CET) Dynamics of Linear Systems September 27, 2017 87 / 103
Compensator:Root locus Design

The lag compensator is employed to getting a reduced steady state error

The zero and pole are place very close to each other, to the origin,
then the error constants will increase, which will reduce the steady
state error.

The lag-lead compensator is employed when both the transient and

steady state specifications are not satisfactory.
The lead compensator will improve the transient response and lag
compensator will reduce the steady state error.

The two main specifications which are used in the root locus
compensation are:
Damping factor and
Undamped natural frequency
(CET) Dynamics of Linear Systems September 27, 2017 88 / 103
Design of Lead Compensator by Root locus: Procedure

Step 1
The desired locations of dominant closed loop poles are found from
the given specifications. where, α < 1 and Kc is determined from the
requirement of open-loop gain.

Step 2
The lead compensator is assumed as

Kc α(1 + Ts)
Gc (s) = (31)
(1 + αTs)

where, α < 1 and Kc is determined from the requirement of

open-loop gain.

(CET) Dynamics of Linear Systems September 27, 2017 89 / 103

Design of Lead Compensator by Root locus: Procedure

Step 3
The sum of the angles at the desired location of one of the dominant
closed loop poles with the open loop poles and zeros of the original
system is to be calculated, which must be odd multiple of 1800 .
If it is not equal to 1800 , the necessary angle φ is calculated which is
to be add to get the sum as an odd multiple of 1800 . This angle φ is
contributed by the lead compensator.
If this angle φ is more than 600 , two or more networks may be used.
The is angle φ also helps to determine values of α and T .

(CET) Dynamics of Linear Systems September 27, 2017 90 / 103

Design of Lead Compensator by Root locus: Procedure
Step 4
To calculate α and T for the known angle φ, a horizontal line from
one of the dominant closed loop pole say P is drawn, the origin is
joined to P.

Figure :
(CET) Dynamics of Linear Systems September 27, 2017 91 / 103
Design of Lead Compensator by Root locus: Procedure

Step 4 (Continue)
The origin is joined P
The angle between PA and PO is bisected
The two lines PC and PD are drawn, which makes an angle ± φ2 with
the bisector PB.
The intersection of PC and PD with negative real axis will give the
required pole and zero of compensator.

Step 5
The magnitude condition at point P is applied to determine the
open-loop gain.

(CET) Dynamics of Linear Systems September 27, 2017 92 / 103

Design of Lead Compensator by Root locus: Procedure
Step 6
The compensated system is checked whether it satisfies all the
specifications or not.
If does not satisfy, the compensation pole and zero are adjusted till all
the specifications are satisfied.

Example 6
Design a suitable lead compensation for a system with unity feedback
and having open-loop transfer function
K
G (s) =
s(s + 1)(s + 2)

This has to meet the specifications: (i) Damping ratio ξ = 0.6 and
(ii) Undamped natural frequency ωn = 2rad/sec.
(CET) Dynamics of Linear Systems September 27, 2017 93 / 103
Design of Lag Compensator by Root locus

The compensated system is checked whether it satisfies all the

specifications or not.
The compensator increase the open-loop gain, but it does not change
the root locus appreciably near the dominant closed loop poles.
The angle contribution by the lag compensator is usually 50 , which is
very small.
The gain is increased, the static error constant is also increased.
The lag compensator is used for the case to satisfy static error
constant specification using root locus.

(CET) Dynamics of Linear Systems September 27, 2017 94 / 103

Design of Lag Compensator by Root locus: Procedure

Step 1
The root locus of the uncompensated system is drawn and the
dominant closed loop poles on the root locus are located.

Step 2
The lag compensator is assumed as

Kc (s + 1/T )
Gc (s) = (32)
(s + 1/βT )

where, β > 1 and Kc is determined from the requirement of

open-loop gain.

(CET) Dynamics of Linear Systems September 27, 2017 95 / 103

Design of Lag Compensator by Root locus: Procedure

Step 3
The static error constant specified in the problem is calculated

Step 4
The of increase in static error constant which is necessary to satisfy
the specifications is calculated.

Step 5
The pole and zero of the compensator do not produce appreciable
change in original root locus but give necessary increase in static error
constant.

(CET) Dynamics of Linear Systems September 27, 2017 96 / 103

Design of Lag Compensator by Root locus: Procedure

Step 6
The root locus of the compensated system is drawn. The dominant
closed-loop poles are located.

Step 7
Kc is adjusted from the magnitude condition, so as to place the
dominant closed-loop poles at the desired location.

Example 7
Design a lag compensator for a system with open-loop transfer
K
function, G (s) = s(s+1)(s+3) to satisfy the following specifications:
Damping ratio = +0.5, Velocity error constant > 5.5s −1 , and settling
time = 10 s.

(CET) Dynamics of Linear Systems September 27, 2017 97 / 103

Design of Lag-Lead Compensator by Root locus

The lead compensator increases the speed of response and improve

the stability while lag compensator improves steady state response.
The lag-lead compensator is used when the requirement of
improvement both transient and steady-state responses is desired.

Let the transfer function of the lag-lead compensator is assumed as

(s + 1/T1 ) (s + 1/T2 )
Gc (s) = Kc ×
(s + 1/βT1 ) (s + 1/αT2 )

where, β > 1 and α < 1

(CET) Dynamics of Linear Systems September 27, 2017 98 / 103

Design of Lag-Lead Compensator by Root locus:
Procedure

Step 1
The root locus of the uncompensated system is drawn and the
dominant closed loop poles on the root locus are located.

Step 2
Find the angle of deficiency φ for the dominant closed-loop poles to
be part of root locus.

Step 3
Select the zero, pole and Kc of lead portion from the angle and
magnitude criteria.

(CET) Dynamics of Linear Systems September 27, 2017 99 / 103

Design of Lag-Lead Compensator by Root locus:
Procedure

Step 4
Find β such that it satisfies the given static velocity error constant Kv
where,
Kv = lim sGc (s)G (s)
s→0

(CET) Dynamics of Linear Systems September 27, 2017 100 / 103

 The End

(CET) Dynamics of Linear Systems September 27, 2017 101 / 103