EEE-335

1
Control System I
Course Teacher: Mohammad Kamruzzaman Khan Prince
Assistant Professor
Dept. of EEE, SUST
2 Books to Follow

 1. Control Systems Engineering by Norman S Nise , John Wiley & Sons, New York
 2. Modern Control Systems by Richard C. Dorf and Robert H. Bishop
 3. Modern control engineering by Ogata, Katsuhiko, Prentice-Hall
3 Importance, Pre-requisite and Outcomes of this
Course
 Obviously to get your degree!
 Control systems are everywhere- toaster, battery charger, heating systems, DVD player
etc…
 Widespread applications in science and industry
 Great career opportunities
 Solid backgrounds on maths such as-
 Electrical circuit analysis
 To be able to design a control system for a physical system satisfying the control goals.
4 What is a Control System?

 Subsystems and processes (or plants) assembled for the purpose of obtaining a desired
output with desired performance, given a specified input.
5 Control Systems Around Us

 Human body
 Communication- Radar antenna
 Robotics
 Elevator
 Room temperature control
 Aero-space engineering
 And many more…
6 Configurations of Control Systems

 Open loop
 Closed loop
 The block diagrams are-
7 Open and Closed Loop Systems

 The plant can be a furnace where the input is fuel and the output is temperature.
 The open loop systems cannot compensate the disturbances.
 Ex. of an open loop system: toaster
 Closed loop system overcomes the problem of the open loop system.
 Ex. is toaster ovens.
 Complex and expensive.
8 Primary Design Objectives of a Control
System
 Producing desired transient response
 Reducing the steady-state error
 Achieving stability
 Example of the elevator.
9 Modeling in the Frequency Domain

 Mathematical modeling of a physical system is a precondition to control the system.

 The modeling can be done in the frequency or time domain.
10 Transfer Function

r(t) c(t) [c(t)]=C(s)=

Plant

 Considering the integral converges

R(s) C(s) 𝑠=𝜎 + 𝑗 𝜔

G(s)

[F(s)]=
𝐶 (𝑠)
𝐺 ( 𝑠 )=
𝑅 (𝑠)
 The transfer function is defined for a linear time
invariant (LTI) system.
{
𝑢 ( 𝑡 )= 1(𝑡 > 0)
0(𝑡 < 0)
11 Laplace Transform Table

© Nise, N. S. (2011). Control system engineering, john wiley & sons. Inc, New York.
12 Transfer Function- Electrical Network

 Kirchhoff’s voltage law or mesh analysis and differential equation.

© Nise, N. S. (2011). Control system engineering, 7th ed., john wiley & sons. Inc, New York. (Ex. 2.6)
13 Transfer Function- Electrical Network Cont’d

 Solve the same problem by using-

 Voltage division rule
 Ohm’s Law
 Kirchhoff’s current law or node analysis
 Explore other methods…
14 Transfer Function- Translational Mechanical
System
15 Linear and Nonlinear Systems

 So far, we have discussed only LTI systems.

 LTI systems possess two properties-
 1. Superposition and
 2. Homogeneity
 To derive the transfer function the nonlinear system can be linearized.
 Linearization
16  Step 1: Identify the nonlinear component and write the
nonlinear differential equation
 Step 2: Find the equilibrium point
 Step 3: Linearize the nonlinear differential equation and take the
Laplace transform
 Step 4: Separate the input and output and derive the transfer
function
 f(x) is a nonlinear function of x
 Let’s linearize f(x) at point A where
 At , f(x)=f()
 Let’s make a very small changes to and assumes the changes
occur along the slope of the curve.
 Here, x- = δx and f(x)-f() ≈ δf(x)
 The slope of the blue line is
 So we get f(x)-f() ≈ δx
 Or, f(x) ≈ f() + δx (eq. 1)
17 Linearization Cont’d

 Eq. (1) in the previous slide resembles with the Taylor series expansion in eq. (2)

eq. (2)

 Ignoring higher order terms-

𝑓 ( 𝑥 ) ≈ 𝑓 ( 𝑥0)+
𝑑𝑓 ( 𝑥 )
𝑑𝑥 |𝑥=𝑥 0
𝛿𝑥 eq. (3)

 We can use eq. 3 to linearize a nonlinear system.

 Linearization Cont’d
18

0.1 𝑣 𝑟
𝑖𝑟 =2 𝑒
𝑉 𝐿(𝑠 )
𝑓 ( 𝑥 )=5 cos 𝑥 =?
𝑉 (𝑠)
© Nise, N. S. (2011). Control system engineering, 7th ed., john wiley & sons. Inc, New York. (Ex. 2.26 and 2.28)
 19 Modeling in Time Domain - State Space
Representation
 An alternative modeling technique of an LTI system along with the frequency domain
modeling.

 Find the state-space representation of

the electrical network and current
through the resistor.

© Nise, N. S. (2011). Control system engineering, 7th ed., john wiley & sons. Inc, New York. (Ex. 3.1)
 20 State-Space Representation Cont’d

 Find the state and output equations if the output vector is

© Nise, N. S. (2011). Control system engineering, 7th ed., john wiley & sons. Inc, New York. (Ex. 3.2)
 21 Transfer Function to State-Space Conversion

ML

ML

© Nise, N. S. (2011). Control system engineering, 7th ed., john wiley & sons. Inc, New York. (Ex. 3.4 and Ex. 3.5)
22 State-Space to Transfer Function Conversion
 We have seen so far the state and output equations are

 The transfer function is

 Where I is the identity matrix.
 Find the transfer function ML
𝑇 ( 𝑠 ) =𝑌 ( 𝑠 ) /𝑈 ( 𝑠 )

© Nise, N. S. (2011). Control system engineering, 7th ed., john wiley & sons. Inc, New York. (Ex. 3.6)
23 Time Response
 Mason’s Rule
24
 A technique to reduce signal flow graphs to single transfer function.
 Loop gain: The product of branch gains found by traversing a path that starts at a node and ends at the
same node, following the direction of the signal flow, without passing through any other node more
than once.
 Mason’s Rule Cont’d
25
 Forward-path gain: The product of gains found by traversing a path from the input node to the output
node of the signal-flow graph in the direction of signal flow.
 Mason’s Rule Cont’d
26
 Nontouching Loops: Loops that do not have any nodes in common.
 G2(s)H1(s) doesnot touch loops G4(s)H2(s), G4(s)G5(s)H3(s) and G4(s)G6(s)H3(s).

 Nontouching-loop gain: The product of loop gains from nontouching loops taken two, three, four, or
more at a time.
 Mason’s Rule Cont’d
27
 Mason’s Rule Cont’d
28
 Find the transfer function C(s)/R(s).
 Mason’s Rule Cont’d
29
 Signal-Flow Graphs of State Equations
30
 Alternative Representation in State Space- Cascade Form
31
 Alternative Representation in State Space Cascade Form Cont’d
32
 Alternative Representation in State Space-Parallel Form
33
 Alternative Representation in State Space Parallel Form Cont’d
34
 MATLAB Example
35
'(ch5p2) Example 5.3' % Display label.
numg=[25]; % Define numerator of G(s).
deng=poly([0 -5]); % Define denominator of
G(s).
' G(s)' % Display label.
G=tf(numg,deng) % Create and display G(s).
' T(s)' % Display label.
T=feedback(G,1) % Find T(s).
[numt,dent]=tfdata(T,'v'); % Extract
numerator & denominator
% of T(s).
wn=sqrt(dent(3)) % Find natural frequency.
z=dent(2)/(2*wn) % Find damping ratio.
Ts=4/(z*wn) % Find settling time.
Tp=pi/(wn*sqrt(1-z^2)) % Find peak time.
pos=exp(-z*pi/sqrt(1-z^2))*100 % Find
percent overshoot.
step(T,'g') % Generate step response.
pause
 Introduction to MATLAB Simulink
36
 Definitions of Stability
37
 Three requirements in designing a control system- transient response, stability and steady-state errors.
 Stability is the most important system specification.

 Definitions based on natural response:

 A linear, time-invariant system is stable if the natural response approaches zero as time approaches infinity.
 A linear, time-invariant system is unstable if the natural response grows without bound as time approaches
infinity.
 A linear, time-invariant system is marginally stable if the natural response neither decays nor grows but remains
constant or oscillates as time approaches infinity.
 BIBO definition:
 A system is stable if every bounded input yields a bounded output.
 A system is unstable if any bounded input yields an unbounded output.
 Definitions of Stability Cont’d
38
 Stable systems have closed-loop transfer functions with poles only in the left half-plane.
 Unstable systems have closed-loop transfer functions with at least one pole in the right half-plane
and/or poles of multiplicity greater than 1 on the imaginary axis.
 Marginally stable systems have closed-loop transfer functions with only imaginary axis poles of
multiplicity 1 and poles in the left half-plane.
 Routh-Hurwitz Criterion
39
 A method that yields stability information without the need to solve for the closed-loop system poles.
 Routh-Hurwitz Criterion Cont’d
40

The number of roots of the polynomial that are in the right half-plane is equal to the number of sign
changes in the first column.
41 Routh-Hurwitz Criterion Cont’d
 Routh-Hurwitz Criterion Cont’d
42
 Two special cases can arise
 Zero in the first column and
 Entire row of zeros
 Routh-Hurwitz Criterion Cont’d
43

Zero in the first column is avoided and the result in example 6.2 is
confirmed.
The complexity is reduced considerably.
 Routh-Hurwitz Criterion Cont’d
44

The system is stable.

 Routh-Hurwitz Criterion Cont’d
45

4 poles on the imaginary axis, 2 in the right half plane

and 2 in the left half plane.
 Routh-Hurwitz Criterion Cont’d
46
 Routh-Hurwitz Criterion Cont’d
47
 Routh-Hurwitz Criterion Cont’d
48
 Stability in State Space
49
 Stability in State Space Cont’d
50
 Steady-State Errors
51
 Steady-state error is the difference between the input and the output for a prescribed test input as .
 Steady-State Errors Cont’d
52
 Steady-State Error for Unity Feedback Systems
53
 E(s)=R(s)-C(s)
 C(s)=R(s)T(s)
 E(s)=R(s)[1-T(s)]
 According to the final value theorem,

R(s) T(s)=G(s)/ C(s)

(1+G(s))
 Steady-State Error for Unity Feedback Systems Cont’d
54
 E(s)=R(s)-C(s)
 C(s)=E(s)G(s)
 E(s)=R(s)/[1+G(s)]
 According to the final value theorem,

What if the input R(s) is a step or ramp or parabola?

55
Steady-State Error for Unity Feedback Systems Cont’d
56
Static Error Constants
 Steady-state errors for a step input, u(t) The three terms in the denominator that are
taken to the limit determine the
steady-state error. We call these limits static
error constants.

Position constant, K_p

 For a ramp intput, tu(t)

Velocity constant, K_v

 For a parabolic input, 0.5t^2 u(t)

Acceleration constant, K_a
57
Static Error Constants Cont’d
58
Steady-State Error Specifications
 (a) What information is contained in the specification K_v=1000?
 (b) What information is contained in the specification K_p=1000?
 (a) Answer: (i) The system is stable with a constant error between input and output.
 (i) The test signal is a ramp input since K_v is specified.
 (ii) It is a Type-1 system (check the table).
 (iii) The SS error is 1/K_v

Find the value of K so that the SS error is 10%

59
Steady-State Error Due to Step Disturbance
60
SS Error for Non-unity Feedback Systems
61
SS Error for Non-unity Feedback Systems Cont’d
62
Sensitivity
 The degree to which changes in system parameters affect system transfer functions, and hence
performance, is called sensitivity.
 A system with zero sensitivity is ideal.
63
Sensitivity of a Closed-Loop Transfer Function
64
Root Locus Techniques
 Root locus, a graphical presentation of the closed-loop poles as a system parameter is varied.
 It is a powerful method of analysis and design for stability and transient response.
 The root locus method can be used to design a 1st , 2nd or higher order system with desired transient
response.
 The root locus gives a graphical representation of a system’s stability.
 Sketching the Root Locus
65
 Number of branches: The number of branches of the root locus equals the number of closed-loop poles.
 Symmetry: The root locus is symmetrical about the real axis.
 Real-axis segments: On the real axis, for K > 0 the root locus exists to the left of an odd number of real-
axis, finite open-loop poles and/or finite open-loop zeros.

 The root locus begins at the finite and infinite poles of G(s)H(s) and ends at the finite and infinite zeros of
G(s)H(s).
66
Sketching the Root Locus Cont’d
 Behaviour at infinity:
 If the function approaches infinity as s approaches infinity, then the function has a pole at infinity.
 If the function approaches zero as s approaches infinity, then the function has a zero at infinity.
 Examples: G(s) = s (pole at infinity) and G(s) = 1/s. (zero at infinity)
 Including infinite poles and zeros every function of s has an equal number of poles and zeros.
 The root locus approaches straight lines as asymptotes as the locus approaches infinity. Further, the
equation of the asymptotes is given by the real-axis intercept, σ_a and angle, θ_a as follows:

 Where k = 0, , …. (After a while the values of k yield similar angles)

67
Sketching the Root Locus Cont’d
68
Sketching the Root Locus Cont’d
 Sketch the root locus. Find the breakaway and break-in points.
69
Sketching the Root Locus Cont’d
 The jω-axis crossing is a point on the root locus that separates the stable operation of the system from the
unstable operation.
70
Sketching Root Locus Cont’d
71
Sketching Root Locus Cont’d
72
Sketching Root Locus Cont’d
73
Root Locus for 3rd Order System
74
Root Locus for 3rd Order System Cont’d
75
Control System Design via Root Locus
76
Design via Root Locus Cont’d
 Improving Steady-State Error via Cascade Compensation
77
78
Effect of an Ideal Integral Compensator
 Lead Compensation
79

© Modern Control Engineering by Katsuhiko Ogata

80
Lead Compensation Cont’d
81
 From the performance specifications,
determine the desired location for the
dominant closed-loop poles.
 Draw the root locus of the uncompensated
system (original system). Can the gain
adjustment yield the desired closed loop
poles?

 If not, calculate the angle deficiency. This angle must be contributed by the lead compensator.
 If static error constants are not specified, determine the location of the pole and zero of the lead
compensator so that the lead compensator will contribute the necessary angle.
 If no other requirements are imposed on the system, try to make the value of alpha as large as possible. A
larger value of alpha generally results in a larger value of K_v.
 Determine the value of K_c of the lead compensator from the magnitude condition.
82
 Once a compensator has been designed, check to see whether all performance specifications have been
met. If the compensated system does not meet the performance specifications, then repeat the design
procedure by adjusting the compensator pole and zero until all such specifications are met.
 If a large static error constant is required, cascade a lag network or alter the lead compensator to a lag–lead
compensator.
 Note that if the selected dominant closed-loop poles are not really dominant, or if the selected dominant
closed-loop poles do not yield the desired result, it will be necessary to modify the location of the pair of
such selected dominant closed-loop poles.
83
Lead Compensation Cont’d
84
Lead Compensation Cont’d

Determination of pole and zero of a lead network.

85
Lead Compensation Cont’d
86
Lead Compensation Cont’d

The pole and zero cancelled each other

87
Lead Compensation Cont’d
88
Lead Compensation Cont’d
89
Lag Compensation
 90
Lag Compensation Cont’d

 Draw the root-locus plot for the uncompensated system whose open-loop transfer function is G(s). Based on the transient-
response specifications, locate the dominant closed-loop poles on the root locus.
 Assume the transfer function of the lag compensator is:

 Then the open-loop transfer function of the compensated system becomes Gc(s)G(s).
 Evaluate the particular static error constant specified in the problem.
 Determine the amount of increase in the static error constant necessary to satisfy the specifications.
91
Lag Compensation Cont’d
 Determine the pole and zero of the lag compensator that produce the necessary increase in the particular
static error constant without appreciably altering the original root loci.
 Draw a new root-locus plot for the compensated system. Locate the desired dominant closed-loop poles on
the root locus.
 Adjust gain of the compensator from the magnitude condition so that the dominant closed-loop poles lie at
the desired location. (will be approximately 1).
92
Lag Compensation Cont’d
93
Lag Compensation Cont’d

x = open loop poles, black dot presents closed loop poles

94
Lag Compensation Cont’d

Compensated system with a lag compensator

95
Lag Compensation Cont’d
96
Lag Compensation Cont’d
97
Lag Compensation Cont’d
 Lag-Lead Compensator
98
99 Lag-Lead Compensator Cont’d
100
Lag-Lead Compensator Cont’d
101
Lag-Lead Compensator Cont’d
 Control System Analysis and Design by the Frequency Response
102
Method
 By the term frequency response, we mean the steady-state response of a system to a sinusoidal input. In
frequency-response methods, we vary the frequency of the input signal over a certain range and study the
resulting response.
 In designing a closed-loop system, we adjust the frequency-response characteristic of the open-loop
transfer function by using several design criteria in order to obtain acceptable transient-response
characteristics for the system.
103
Presentation of Sinusoidal Transfer Function
 The sinusoidal transfer function, a complex function of the frequency ω, is characterized by its
magnitude and phase angle, with frequency as the parameter. There are three commonly used
representations of sinusoidal transfer functions:
 Bode diagram or logarithmic plot
 Nyquist plot or polar plot
 Log-magnitude-versus-phase plot (Nichols plots)
 A Bode diagram consists of two graphs: One is a plot of the logarithm of the magnitude of a sinusoidal
transfer function; the other is a plot of the phase angle; both are plotted against the frequency on a
logarithmic scale.
104
Basic Factors of G(jw)H(jw)
105
Bode Plot of Sinusoidal Transfer Function
106
107