You are on page 1of 94

Digital Control

Objective
Control System Terminology
Computer Based Control Control Theory Classical Approach to Analog Controller Design

Control System Terminology


Control System Interconnection of components to provide a desired function Plant- The portion of the system to be controlled Controller The portion of the system that does the controlling Digital Control System- Uses digital hardware (digital computer) Analog Control System- Electronic controller made of resistors , capacitors and operational amplifiers. Signals (i) Continuous time signals (defined for all time) (ii) Discrete time signals (defined at discrete instant of time )
3

Advantage of Digital Control


Reconfiguration, Flexibility (Controlled by changing software ) Wide selection of Control Algorithms

Integrated Control of Industrial System- (Production planning, scheduling, optimization, operations control) Future Generation Control System (AI)

Command input

Disturbance input

Controlled output

Computer
A/D
D/A Final Control Element Plant

Clock

Sensor

Basic structure of a computer controlled system

Radar Tracking

Footer Text

3/9/2014

Servomechanism for Steering of Antenna


Servomotor Tachogenerator

r
Computer D/A Power amplifier

Actual shaft position

Velocity signal Shaft-angle encoder

Gear

Azimuthal servomechanism for steering of antenna


7

Functional block diagram of azimuthal servomechanism

e r

+
-

+
Power amplifier Motor + Load

Tachogenerator Rate signal

Variable Speed DC Drive


Load Speed reference + Actual speed Tachogenerator

A/D ac Full bridge rectifier circuit

Motor

Digital Computer SCR trigger control circuit

Over current Protection

SCR
9

Liquid Level Control System


Expansion slot of PC

Step Motor Controller Card

Driver Circuit

Step Motor
V1 Pump

V2

Sump

A/D conversion card

Bridge and amplifier circuit

10

Day 2
An Overview of classical approach to analog controller design Basic digital control scheme Principle of signal conversion Basic discrete time signals

11

Computer Based Control


1950 Digital Computer (main frame) 1962 Digital Computer, Direct Digital Control ( no analog controller) 1970 Small, faster, more reliable and cheap computer- Minicomputer 1975 Microcomputer DCCS Distributed Computer Control System for control of large and complex process SCADA- Supervisory Control And Data Acquisition (i) Data Acquisition and Communication (ii) Event and alarm reporting (iii) Data processing (iv) Partial process control CIPS Computer Integrated Process System

12

Machine tool numerical control- hard- wired function was replaced by software- CNC CNC- Computerized Numerical Control CIMS Computer Integrated Manufacturing System PLC Programmable Logic Controller

13

Control Theory
1940 -1950 Classical control theory- Routh-Hurwitz , Root Locus, Nyquist, Bode, Nichols use transfer function in complex frequency (Laplace Variable s)domain Limited to SISO system and linear time invariant system 1950- 1960 Modern Control System model in time domain- MIMO Lyapunov Stability criterion, pole placement by state feedback, state observers, optimal control Model based control Knowledge based controlIntelligent control using AI techniques ( Fuzzy Logic and Neural Network)
14

An Overview of classical approach to analog controller design

Y(t) Controlled variable of the system m(t) Manipulated variable Yr(t) Desired value of controlled variable Gp(s) TF of controller system H(s) TF of feedback element w(t) - Disturbance b(t) Feedback signal e(t) = yr y(t) System Error A(s) - TF of reference input element r(t) Reference input compatible with b(t) e(t) Actuating error signal D(s) TF of controller u(t) Control signal( Has knowledge about the desired control action) GA(s) TF of actuator element (develop enough torque, pressure or heat )

15

The O/P equation Y(s) is given by --------- (1)

16

----(2)

----(3)

17

Footer Text

3/9/2014

18

------(4) ------(5) Sub (4) in (5) we obtain (6) ------(6) From (6) we obtain the Reference Transfer Function M(s) ------(7)

19

----(8) ----(9) Sub (8) in (9) and solving y(s)/W(s) will give the disturbance Transfer Function Mw(s)

----(10)

20

The response to the simultaneous application of R(s) and W(s) is given by -----(11)

From equation (7) & (10) M(s) and Mw(s) are closed loop Transfer Functions

Roots of the characteristic equation poles of the system

are closed loop

21

22

23

24

Footer Text

3/9/2014

25

26

Footer Text

3/9/2014

27

28

29

30

Footer Text

3/9/2014

31

Footer Text

3/9/2014

32

Day 3
Time Domain Model for Discrete- Time System
State Variable Model Difference Equation Model Impulse Response Model Transfer Function Model
33

34

35

Time Domain Model for Discrete- Time System


Discrete time system is defined mathematically as a transformation, or an operator that maps an input sequence r(t) into an output sequence y(k).

State Variable Model


The state equation and output equation of the system together give the state variable model of the system.

36

MATLAB Program

37

State Equation

Output Equation

38

The state variable formulation as block diagram

39

40

41

Difference Equation Model

42

Impulse Response Model


------------ 1

43

44

Transfer Function Model


Analytical study of a system is to set up mathematical equation to describe the system. Let us take a linear time invariant discrete time system that is initially relaxed at k=0

45

Difference eq.

46

--------1
Z Transform Shifting theorem Z[y(k+n)]=zn Y(z)- zn Y(0)- zn-1Y(1)-. z2y(n-2)-zy(n-1)

--------2

In equation 1

--------3a

--------3b
47

Substituting 3a & 3b in 2 we get

48

Transfer Function of Unit Delayer

49

50

51

Day 4
Stability on the Z-Plane BIBO Stability Zero-input Stability Jury Stability Criterion

3/9/2014

52

Stability on the Z-Plane & Jury Stability Criterion

53

54

55

BIBO Stability

56

57

58

59

60

61

62

63

64

Zero-input Stability

65

66

Jury Stability

67

68

69

70

71

72

73

74

75

76

77

78

79

Practical aspects of the choice of Sampling Rate


Long sampling interval reduces computational load, need for rapid A/D conversion and hardware cost of the project. It result in degrading effects Limiting Factor for Choice of Sampling Rate Information loss due to sampling (i) Real signals are not band limited (ii) Ideal lowpass filters are not physically realizable (iii) ZOH introduce reconstruction errors. Information loss due to Disturbance (i) High frequency noise appear as low frequency signals due to aliasing effect causing loss of information (ii) Cut off frequency of anti aliasing filter must be higher than system bandwidth
80

Destabilizing Effects Due to conversion times and computation times digital algorithm contain dead- time . Algorithm- accuracy Effects Discretization process (ie) transformation of an algorithm from continuous time to discrete time form introduces error Word-length Effects

81

Empirical rule for selection of sampling rate


Practical experience and simulation results have produced useful rules for specification of minimum sampling rates.
Type of Variable Flow Level Pressure Temperature Sampling time(s) 1-3 5 - 10 1-5 10 - 20

(i) Table gives values from the experience of process industries (ii) Fast acting electromechanical system require shorter sampling intervals few milliseconds
82

The rule of thumb says, a sampling period needs to be selected much shorter than any of the time constants in the continuous time plant . For complex poles with imaginary part d the frequency of transient oscillation corresponding to the pole is d . Rule suggests sampling at the rate of 6 to 10 times per cycle. Sampling rates can be based on the bandwidth of the closedloop system. Reasonable sampling rates are 10 to 30 times the bandwidth. For closed loop performance the sampling interval T should be equal to or less than, one-tenth of the desired settling time.

83

Routh Stability criterion on the r - plane

84

The outside of the unit circle in the z-plane is transformed to the right half of the new complex plane. The boundary of the unit circle in the z-plane is transformed into the imaginary axis of the new complex plane. The inside of the unit circle in the z-plane is transformed into the left half of the new complex plane.

85

In the stability analysis using the bilinear transformation coupled with Routh stability criterion substitute (w+1)/(w-1) for z in the characteristic equation (z)=0 , where w=+j , to obtain the characteristic equation (w)=0

Problem: Consider the following Characteristic equation P(z) = z3-1.3z2-0.08z+0.24=0 . Determine whether or not any of the roots of the characteristic equation lie outside the unit circle in the z plane. Use the bilinear transformation and Routh stability criterion.
86

-0.14w3+1.06w2+5.10w+1.98 = 0 w3-7.571w2-36.43w-14.14 = 0

w3 w2 w1 w0

1 -7.571 -38.30 -14.14

-36.43 -14.14 0

Since there is one sign change for the coefficient in the first column there is one root in the right half of the w plane. The system is unstable.

87

88

89

90

91

92

93

94

You might also like