Mu’tah University

Faculty of Engineering
Electrical Eng. Depart.
AUTOMATIC CONTROL
LECTURE 8
PID CONTROLLER
DESIGN & IMPLEMENTATION
Dr. Khaled Alawasa
Fall 2014 1
 Controller Types
Controller (compensator)
R(s)
+ Gc(s) Y(s)
G (s )
-

• PID {P, PI, PD, PID}

• Lead-lag Your role, as ECE Engineer, is to select
• Hysteresis 1) the appropriate control type and
• Dead-beat, 2) control parameters
• Variable Structure To achieve the desired response
• Fuzzy
• …. etc. 2
 The need for controller [PID]
Example
• Consider a simple mass, spring, damper problem.

• The dynamic model is such as:

mx  bx  kx  f
• Taking the Laplace Transform, we obtain:
ms 2 X (s)  bsX (s)  kX (s)  F (s)
• The Transfer function is then given by:
X (s) 1
 2
F ( s ) ms  bs  k
 Example ….cont’d
• Let
m  1kg , b  10N .s / m, k  20N / m, f  1N
• By plugging these values in the transfer function:
X (s) 1
 2
F ( s ) s  10s  20
• The goal of this problem is to show you how controller
parameters contribute to obtain excellent control
performance i.e.,:
1. Fast system( lower rise time),
2. Minimum overshoot,
3. No(zero) steady-state error.
 Example ….cont’d
R(s)[F(s)] 1 Y(s)[X(s)]

s 2  10s  20

• The (open) loop transfer function is given by:

X (s) 1
 2
F ( s ) s  10s  20

• The steady-state value for the output is:

X (s ) 1
e ss  lim x (t )  lim sX (s )  lim sF (s ) 
t  s 0 s 0 F (s ) 20
5
 Ex (cont’d): Open-loop step response
• 1/20=0.05 is the final
value of the output to
0.05
Step Response
an unit step input.
0.04

0.03 • This corresponds to a

Amplitude

0.02 steady-state error of

0.01 95%, quite large!
0
0 0.5 1 1.5 2 2.5
Time (sec) • The settling time is
about 1.5 sec.

What is the rise time from the figure: ??

6
 Adding [Proportional Controller]
Open Loop
R(s)[F(s)] U(s) 1 Y(s)[X(s)]
Kp s 2  10s  20
Controller

• The open loop transfer function is given by:

X (s ) Kp  Kp  Kp
 2  lim  2 
F (s ) s  10s  20 y ss
s 0 s  10s  20

  20

When Kp=20 the error become zero… that is GOOD…

but Remember!! open loop control system has many weaknesses
compared with feedback system……..previous chapter 7
 From previous Chapter
Comparison between
Open Loop Closed loop
Sensitivity To
S 1
T 1
Parameter S 
T

1  Gc ( s)G ( s)
G
G
Variations

Disturbance Signals not able to reject Ability to “ completely”

disturbance reject the disturbance
Poor ( slow) Excellent ( fast
Transient Response response) response)

Steady-state Error not able to maintain able to maintain zero

zero steady state steady state

8
 Adding [Proportional Controller] (1)
(Closed Loop)
R(s)[F(s)]
E(s) U(s) 1 Y(s)[X(s)]
+
Kp s 2  10s  20
-
Controller

• The closed loop transfer function is given by:

Kp
X ( s) s 2
 10 s  20 Kp
  2
F (s) Kp s  10s  (20  K p )
1 2
s  10s  20
Kp K n 2
 2 Find ζ ωn ???
s  10s  (20  K p ) s  2ns  n 2
2

9
 Adding [Proportional Controller] (2)
Closed-loop response w ith Kp=300
1.4

1.2
System: CL_Kp
Final Value: 0.938
1

Let Kp=300 0.8 System: CL_Kp System: CL_Kp

Amplitude
Time (sec): 0.105 Settling Time (sec): 0.772
Amplitude: 0.912
0.6

 Kp  300 0.4

e ss  lim 1  2   0.0625
s 0
 s  10s  (20  K p )  320
0.2

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
Time (sec)

• The above plot shows that the proportional

controller reduced both the rise time and the
steady-state error, increased the overshoot, and
decreased the settling time by small amount.
10
 Adding [Proportional Controller] (3)
• Approximate equation Rise time

• Adding Kp, makes the system similarity to

standard 2nd order TF.
Kp K n 2
 2
s  10s  (20  K p ) s  2ns  n 2
2

5
n  20  K p ,   , when K p : n  and    Tr 
20  K p
As Kp increase wn increased ( more oscillation) and damping ratio decrease.
From the rise time expression, When the damping ratio wn increased , rise time decreases.
For Kp=300 , rise time=0.067 sec. ( a bit different from the result on the plot
in the previous slide, why?) 11
 Adding [Proportional Controller] (4)
• Question: Calculate the damping ratio and overshoot
percent for the system with Kp=100, 200, 300.
1
100(  120) K n 2
Step Response
1.2 
1.4

s 2  10s  (20  100) s 2  2ns  n 2 kp=300

1.2
5
   0.4564  P .O %  19.95% kp=100
20  100 1

•With Kp=100  P.O% = 20% OR 0.8

Amplitude
•With Kp=200  P.O% = 32.5% 0.6

•With Kp=300  P.O% = 40% 0.4

•Overshoot= (Maximum peak- Final value)/ Final value

0.2
Peak vaule  1.0
 Kp  Kp 0

y ss Kp 100  lim  2   0.833

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

s 0 s  10s  (20  K )

 20  K p
Time (sec)
 p 

P .O .%  (1.31  0.833) / 0.833  19.95% 12

 Steady-State Errors- Modifying the TF
• We learned that steady-state error can be improved by
adding an open-loop pole at the origin in the forward
path, thus increasing the system type and driving the
associated steady-state error to zero.
• This additional pole at the origin requires an integrator
for its realization.
• Therefore by modifying the type of the system ( from
type 0 to type 1) we can eliminate the error….

13
 Adding Integral Gain
PI Controller (1)
F(s) + E(s) Ki U(s) 1 X(s)
Kp 
- s s 2  10s  20
Controller

• Improving transient Response via PI controller

• The closed loop transfer function is given by:
K p  Ki / s
X ( s) s 2
 10 s  20 K p s  Ki
  3
F (s) K p  Ki / s s  10s 2  (20  K p ) s  K i
1 2
s  10s  20
Q: Tr, Tp, Ts ….. be calculated? 14
 Adding Integral Gain
PI Controller (2)
Closed-loop response w ith Kp=30 and Ki=70
1.4

1.2
• Let kp=30, Ki=70
System: CL_Kp System: CL_Kp
Rise Time (sec): 0.684 Final Value: 1
1
System: CL_Kp
System: CL_Kp • We have reduced the
proportional gain because
Peak amplitude: 1.01
Settling Time (sec): 0.62
0.8 Overshoot (%): 1.26
Amplitude

At time (sec): 0.81

0.6
the integral controller also
reduces the rise time and
0.4
increases the overshoot as
0.2
the proportional controller
does (double effect).
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8
Time (sec)

under step input :

• The above response shows
that the integral controller
 K ps  K i  Ki eliminated the steady-state
y ss  lim  3  1
s 0 s  10s 2  (20  K )s  K
 p i  K i error.
1 1 1 1
e ss ()  ( )   0
1  lim sGc (s )G (s )  K ps  K i
 K i 1 
s 0 1  lim  2  1 15
s 0 s (s  10s  20) 0
 
 Adding Derivative Gain
PD- Controller (1)
F(s) + E(s) U(s) 1 X(s)
K p  K Ds
- s 2  10s  20
Controller

• The closed loop transfer function is given by:

K p  Kd s
X ( s) s 2
 10 s  20 K p  Kd s
  2
F (s) K p  Kd s s  (10  K d ) s  (20  K p )
1 2
s  10s  20

16
 PD Controller (2)
1.4
System: CL_Kp
Closed-loop response w ith Kp=300 and KD=10
• Let kp=300, Kd=10
• This plot shows that the
Peak amplitude: 1.08
1.2 Overshoot (%): 15.3
At time (sec): 0.17
System: CL_Kp System: CL_Kp

proportional derivative
Settling Time (sec): 0.29 Final Value: 0.938
1

System: CL_Kp
0.8 Rise Time (sec): 0.103
controller reduced both the
Amplitude

0.6

0.4
overshoot and the settling
0.2
time, and had small effect on
0
0 0.1 0.2 0.3 0.4 0.5 0.6
the rise time and no effect on
Time (sec)
the steady-state error.
under step input :
 K p  Kds  Kp 300
y ss  lim  2   1   0.9375
s 0 s  (10  K )s  (20  K ) 
 d p  20 K p 320

1 1 20
e c ( )  ( )   0.0625
1  lim sGc (s )G (s )  K  K d s  20  K p
s 0 1  lim  2 p  K p 300
s 0 s  10s  20
  17
 Three Controller ( P, I, D)
PID Controller (1)
F(s) + E(s) Ki U(s) 1 X(s)
Kp   K Ds
- s s 2  10s  20
Controller

• The closed loop transfer function is given by:

K p  K d s  Ki / s
X ( s) s 2
 10 s  20 K s 2
 K p s  Ki
  3
d

F (s) K p  K d s  K i / s s  (10  K d ) s 2  (20  K p ) s  K i

1
s 2  10s  20
18
 Three Controller ( P, I, D)
PID Controller (2)
Let Kp=350, Ki=300, Kd=50
Closed-loop response w ith PID
1

0.8

0.6
Amplitude

0.4

0.2

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Time (sec)

Now, we have obtained an optimal design for the

system with no overshoot, fast rise time, lower
settling time, and no steady-state error.
19
 PID Controllers Comparison
This Figure shows the system step-response under
four controller types.
Closed-loop response w ith different controllers
1.4

1.2

0.8
Amplitude

0.6
(P-mode)
0.4 (PD-mode)
(PI-mode)
0.2
(PID-mode)

0
0 0.5 1 1.5 2 2.5 3 3.5 20
Time (sec)
 The PID Controller
• The PID algorithm is the most popular feedback
controller algorithm used in industrial process
control.
• It is a robust easily understood algorithm that can
provide excellent control performance despite the
varied dynamic characteristics of processes.
• As the name suggests, the PID algorithm consists
of three basic modes:
– the Proportional mode, (P-mode)
– the Integral mode, (I-mode)
– the Derivative mode. (D-mode)

21
 P, PI or PID Controller
• Utilizing the PID algorithm, it is necessary to decide
which modes are to be used (P, I or D, or combination)
and then specify the parameters (or settings) for each
mode used.
• Generally, three basic algorithms are used: P, PI or PID.
• Controllers are designed to eliminate the need for
continuous operator attention. (so it’s Automatic
Control)
– Examples: Cruise control in a car and a house thermostat are common
examples of how controllers are used to automatically adjust some
variable to hold a measurement (or process variable) to a desired variable
(or set-point)

22
 PID Definitions
R(s) + E  Ki U Plant/ Y(s)
Kp  K s process
-  s d 
 
PID Controller

• In the s-domain, the PID controller may be

represented, in a transfer function, as:
 Ki 
U (s )   K p   K d s  E (s )
 s 
• In the time domain:
t de(t )
u (t )  K p e(t )  K i  e(t )dt  K d
0 dt
proportional gain integral gain derivative gain23
 PID Definitions
• In the time domain:
In the s-domain:
t de(t )
u (t )  K p e(t )  K i  e(t )dt  K d  1 
0 dt U (s )  K p 1  T d s  E (s )
 1 t de(t )   Ti s 
 K p  e(t )   e(t )dt  Td 
 Ti 0 dt 

integral time constant derivative time constant

Kp Kd
where T i  , Td 
Ki Kp derivative gain
proportional gain
integral gain
 Poles and Zero of PID
• a PID controller introduces a transfer function
with :
– One pole at the origin and
– Two zeros that can be located anywhere in the s-
plane.

 K d s 2  K ps  K i 
PID   
 s 

25
 PID Controller Functions

• Output feedback
 from Proportional action : compare output with set-point

• Eliminate steady-state offset (=error)

 from Integral action : apply constant control even when
error is zero

• Anticipation
 From Derivative action: react to rapid rate of change
before errors grows too big
 Controller Output (1)
• The variable being controlled is the output of the
controller (U) (and the input of the plant):
R(s) E U
+ Controller
Plant/ Y(s)
process
-
System to
provides excitation to the plant
be controlled

• The output of the controller will change in

response to a change in measurement or set-
point (that said a change in the tracking error, E)
U=E*Controller
27
 Controller Output (2)
R(s) + E  Ki U Plant/ Y(s)
Kp  K s process
-  s d 
 
Controller

t de(t )
u (t )  K p e(t )  K i  e(t )dt  K d
0 dt
• The signal u(t) will be sent to the plant, and a new output
y(t) will be obtained. This new output y(t) will be sent back
to the sensor again to find the new error signal e(t). The
controllers takes this new error signal and computes its
derivative and its integral gain. This process goes on and on.

U=E*Controller 28
 Controller Effects
• A proportional controller (P)
– reduces error responses to disturbances, but still
allows a steady-state error.
• When the controller includes a term proportional
to the integral of the error (I),
– then the steady state error to a constant input is
eliminated, although typically at the cost of
deterioration in the dynamic response.
• A derivative control
– typically makes the system better damped and more
stable.
29
 Closed-loop Response
Rise time Maximum Settling Steady-
overshoot time state error
Increasing Decrease Increase Small Decrease
KP change
Increasing Decrease Increase Increase Eliminate
KI
Increasing Small Decrease Decrease Small
KD change change
• Note that these correlations may not be exactly accurate,
because P, I and D gains are dependent of each other.
 1  Kp Kd
U (s )  K p 1  T d s  E (s ) Ti  , Td 
 Ti s  Ki Kp
 Effect of Proportional,
Integral & Derivative Gains on the
Dynamic Response
Effect of change Proportional Controller
Gain (Kp)
• Pure gain (or attenuation) since: the
controller input is error the controller output
is a proportional gain
R(s) E U
+ Plant/ Y(s)
Kp process
-
Controller

E ( s ) K p  U ( s )  u (t )  K p e(t )

32
 Effect of change Proportional
Controller Gain (Kp)
Closed-loop response w ith different controllers
1.5
(KP=10)
(KP=50) Increase in Kp gain:
(KP=100)
(KP=500)
Upgrade both steady-
1 state and transient responses
 Reduce steady-state
Amplitude

error
Step Response
2

0.5 1.8

1.6

1.4

1.2

Amplitude
1

0.8

Kp increases 0.6

0.4 Kp=1000
0.2

0 0
0 0.5 1 1.5 2 2.5 3 0 0.1 0.2 0.3 0.4 0.5
Time (sec)
0.6 0.7 0.8 0.9 1

Time (sec)

Rise time Maximum Settling time Steady-state error

overshoot
Increasing KP Decrease Increase Small change Decrease
33
 Effect of change Integral gain
Controller
• a Integral of error with a constant gain:
 Increase the system type by 1
 Eliminate steady-state error for a unit step input
 Amplify overshoot and oscillations
To see the effect of changing Ki, Kp is kept
constant.
F(s) + E(s) Ki U(s) 1 X(s)
Kp 
- s s 2  10s  20
Controller

t
Ki
E (s )  U (s )  u (t )  K pe (t )  K i  e (t )dt 34
s 0
 Effect of change Integral Controller
Gain (Ki)
1.8
Step Response
• Increase in ki gain:
(Ki=0)
1.6 (Ki=50)  upgrade steady-
1.4
With Kp=30 (Ki=100)
state responses
(Ki=300)

1.2
 Zero error
1
 Increase slightly
Amplitude

0.8
settling time
0.6
 Increase oscillations
0.4
Ki increases and overshoot!
0.2

0
0 0.5 1 1.5 2 2.5 3
Time (sec)
Rise time Maximum Settling time Steady-state
overshoot error
Increasing KI Decrease Increase Increase Eliminate
35
Effect of change for gain PD controller (2)
• Differentiation of error with a constant gain:
 Detect rapid change in output
 Reduce overshoot and oscillation
 Does not affect the steady-state response
To see the effect of changing KD, Kp is kept
constant.
F(s) + E(s) U(s) 1 X(s)
K p  K Ds
- s 2  10s  20
Controller

de (t )
E (s )K d s  U (s )  u (t )  K pe (t )  K d
dt 36
 Effect of change for gain PD controller (2)
Step Response
1
• Increase in gain:
0.9

0.8
 Upgrade transient
0.7
response
0.6
 small change in settling time
 Decrease the peak and rise
Amplitude

0.5 (KD=0)

0.4
(KD=50) time
(KD=100)

0.3 (KD=300)  decrease overshoot

0.2
and settling time!
0.1

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Time (sec)

Rise time Maximum Settling time Steady-state

overshoot error
Increasing KD Small change Decrease Decrease Small change
37
 Application of PID Control
• PID regulators provide reasonable control of
most industrial processes, provided that the
performance demands is not too high.
• PI control are generally adequate when
plant/process dynamics are essentially of 1st-
order.
• PID control are generally ok if dominant plant
dynamics are of 2nd-order.
• More elaborate control strategies needed if
process has long time delays, or lightly-damped
vibrational modes
 Closed-loop Response
Rise time Maximum Settling Steady-
overshoot time state error
Increasing Decrease Increase Small Decrease
KP change
Increasing Decrease Increase Increase Eliminate
KI
Increasing Small Decrease Decrease Small
KD change change
• Note that these correlations may not be exactly accurate,
because P, I and D gains are dependent of each other.
 1  Kp Kd
U (s )  K p 1  T d s  E (s ) Ti  , Td 
 Ti s  Ki Kp
 PID Tuning
• There are many methods available to determine acceptable
values of the PID gains.
• The process of determining the gains is often called PID
tuning.
– Manual PID tuning method. (Open-loop method)
– Ziegler-Nichols PID tuning method. (closed-loop method)
– Tuning programs (MATLAB, Special software's,…)
• http://www.straightlinecontrol.com/software.html
• A common approach to tuning is to use manual PID tuning
methods, whereby the PID control gains are obtained by
trial-and-error with minimal analytic analysis using step
responses obtained via simulation, or in some cases, actual
testing on the system and deciding on the gains based on
observations and experience

40
 Manual PID tuning
• One approach to manual tuning is to first set Kd=0 and Ki=0 .
• This is followed by slowly increasing the gain until the output of the
closed-loop system oscillates just on the edge of instability.
• This can be done either in simulation or on the actual system if it
cannot be taken off-line.
• Once the value of (with and) is found that brings the closed-loop
system to the edge of stability, you reduce the value of gain to
achieve what is known as the quarter amplitude decay: (i.e. reduce
the gain to 50%).
• A rule-of-thumb is to start by reducing the proportional gain KP by
one-half.
• The next step of the design process is to increase Kj (eliminate the
error) and KD manually to achieve a desired step response.
– Ki to eliminate the error
– Kd to reduce the overshoot.

41
 Manual PID tuning Example
Step#1
• Consider the closed-loop system in Figure

To begin the manual tuning process:

Step#1: set Kj = 0 and KD = 0 and
increase Kp until the closed-loop system
has sustained oscillations.
KP = 885.5 KP = 442.75. 42
 Manual PID tuning Example…con’t
Step#1
• Reduce KP = 885.5 by half as a first step to
achieving a step response with approximately
a quarter amplitude decay.

The step response is shown in Figure where we

note that the peak amplitude is reduced to
one-fourth of the maximum value in one
period, as desired.
To accomplish this reduction, we refined the
value of KP by slowly reducing the
value from KP = 442.75 to KP = 370.

43
 Manual PID tuning Example…con’t
Step#2
• The root locus for KP = 370, Ki= 0, and 0 < KD < ∞
is shown in Figure below
• In this case, the characteristic equation is

KD increases, the root locus shows that the

closed-loop complex poles move left, and in
doing so, increases the associated damping
ratio and thereby decreases the percent
overshoot.
The movement of the complex poles to the
left also increases the associated ζωn thereby
reducing the settling time.

44
 Manual PID tuning Example…con’t
Step#2

According to required response

Select the value of KD

45
 Manual PID tuning Example…con’t
Step#3
• The root locus for Kp= 370, KD = 0, and 0 < Ki < ∞ is
shown in Figure below .
• The characteristic equation is

Ki increases, the root locus shows that the

closed-loop complex pair poles move right.
This decreases the associated damping ratio
and thereby increasing the percent overshoot.
In fact, when Ki = 778.2, the system is
marginally stable

46
 Manual PID tuning Example…con’t
Step#3
• The percent overshoot and settling time as a function of K{
are shown in Figure 7.36. The trends in Figure 7.36 are
consistent with Table 7.6.
• To meet the percent overshoot and settling time
specifications, we can select KP = 370, KD = 60, and Ki =
100.
• The step response shown in Figure 7.37 indicates a Z;. = 2.4
s and P.O. = 12.8% meeting the specifications.

47
 Manual PID tuning Example…con’t
Step#3
The percent overshoot and settling time as a function of K

Percent overshoot
and settling time
with KP = 370,
KD = 0, and
50 < K, < 600.

48
 Output response
Percent overshoot
and settling time
with final design
KP = 370, KD = 60,
andK/ = 100.

49
 Ziegler-Nichols Tuning
• In 1942 Ziegler and Nichols, described simple
mathematical procedures for tuning PID
controllers.
• The method had a large impact in making PID
feedback controls acceptable to control engineers
• These procedures are now accepted as standard in
control systems practice.
• The closed-loop Zielgier-Nichols tuning method
considers the closed-loop system response to a
step input (or step disturbance) with the PID
controller in the loop.
50
 Ziegler-Nichols PID tuning method (1)
• Initially the derivative and integral gains, KD and Ki
respectively, are set to zero. The proportional gain
KP is increased (in simulation or on the actual
system) until the closed-loop system reaches the
boundary of instability.
• The gain on the border of instability, denoted by Ku,
is called the ultimate gain. The period of the
sustained oscillations, denoted by Tu, is called the
ultimate period.
• Once Ku and Tu are determined, the PID gains are
computed using the relationships in Table 7.7
according to the Ziegier-Nichols tuning method. 51
Ziegler-Nichols PID tuning method (2)
• How to find the ultimate gain and ultimate
period. By using RH-method.

52
Ziegler-Nichols PID tuning method.

For the previous example

53
 Comparison
Manual tuning method Ziegler-Nichols tuning method

54
 Ziegler-Nichols PID tuning method
Example
Consider a process with transfer function :
1
G (s ) 
(s  1)(s  3)(s  5)

+ Y(s)
PID System
-

The system is to be controlled by a PID

controller, using Ziegler-Nichols method,
determine the PID gain/parameters
55
 Ziegler-Nichols PID tuning method
Example –Solution
1. We can determine the limiting gain for
stability (before oscillations) by use of the
Routh-Hurwitz condition.
2. Obtain characteristic equation with Kp and
Set Ki=0 and KD =0
3. The characteristic equation, p(s), with
Proportional control is:
(s )  p (s )  1  K pG (s )  0
 (s  1)(s  3)(s  5)  K p
=s 3  9s 2  23s  15  K p   0 56
Ziegler-Nichols PID tuning method
Example –Solution
The Routh array is
s3 1 23
2
s 9 15+K
s1 192-K 0
0
s 15+K
From this we see that the range of K for stability is
1- 15+K>0  K>-15 and
2- 192-K >0  K< 192
At K=192 we will have row of zeros, So KU=192.
57
 Ziegler-Nichols PID tuning method
Example –Solution
When K = 192, we have imaginary roots since the s1 row is
identically 0. The corresponding auxiliary equation is
9s2+(15+192)=0 with roots at s = ±j4.8.
Since this is a quadratic factor of the characteristic polynomial
=> the sustained oscillation at the ultimate gain: is at
4.8rad/s. Thus, ultimate period. Tu = 1.309sec
1 2 2
TU     1.309
f  4.8
Now the ultimate gain and ultimate period are determined,
we can determine the PID gains:
This gives for full PID control from the table as
• Kp = 0.6Ku = 115.2;
• Ki = (1.2KU) /Tu = 176.0 ;
• Kd = (0.6KUTu)/8 = 18.85 58
Step Response for System with
Ziegler-Nichols Tuning method

Can we get a better response ??

59
 Tuning Using MATLAB
• At command window type
>> sisotool

Examine it in your free time …!!!!! 60

 Tips for Designing a PID Controller (1)
1. Obtain an open-loop response and determine what
needs to be improved
2. Add a proportional control to improve the rise time
3. Add a derivative control to improve the overshoot
4. Add an integral control to eliminate the steady-state
error
Adjust each of Kp, Ki, and Kd until you obtain a desired
overall response.

61
 Tips for Designing a PID Controller (2)
• Lastly, keep in mind that you do not need to
implement all three controllers (proportional,
derivative, and integral) into a single system, if
not necessary.
• For example, if a PI controller gives a good
enough response , then you don't need to
implement derivative controller to the system.
• Keep the controller as simple as possible.

62
 Tips for Designing a PID Controller (3)
System with added Controller:
• If the TF similar to 2nd standard transfer, the design will
be quite easily: use the developed equation to find Tr,
Tp, Ts, O.S%,
• If the TF is a 3rd order :
– Check if the added pole/zero can be cancelled, so the
response can be approximated to that by 2nd standard
transfer …. the design then will be easily.
– Otherwise, use root locus technique to select the
controller parameters (Kp, Ki, KD) to achieve the desired
response( speed, overshoot, error)
• Normally Root locus technique is used for design/find
the PID gains
• Use Ziegler-Nichols tuning method for achieve fast
and simple designs 63
 PID in our Control Lab
Process- Control Simulator Plant/ Process

Feedback
Sensor Actuator

Output

Input

PID controller
Identify the control system components 64
 PID Application
PID controller in AVR system - ABB

EATEN
Real system

Is it Digital or analogue ??

Equivalent TF.

65
 Controller Implementation (1)
• The use of the computer has played an major role in
recent advances in the design of Automatic control
systems.
• The computers may divided into TWO types:
– Analogue computers and
– Digital computers.
• Analogue computer is one in which the equation
describing the operation of the computer in analogue
to that for the actual system.
• The most commonly used analog computer is the
electronic analog system.
• The electronic analog computer is a very powerful tool
for investigation the performance of control systems
66
 Controller Implementation (2)
Process
R(s)
+ Analogue Controller (Operation to be Y(s)
(via Analogue computer) controlled)
-

Process
R(s) Digital Controller
+ (via Digital computer,
(Operation to be Y(s)
controlled)
- PC)

Some of You will study this in Digital control course

67
 Physical Realization of
Controller - Analogue computer
How to build the controller in the real life

R(s) Process
+ Controller
(Operation to be Y(s)
controlled)
-

• Active-Circuit Realization

68
 Op-amps

Inverting op-amp
-
+

inputs output

-
69
 Physical Realization of
Controller - Analogue computer
• Electronic analog computer; typically use op-amps

op-amps
At our Control Lab

70
Active realization of controllers using
an operational amplifier

71
 PID realization (implementation using Analog
computer )
Vo R
P ?????   2  K P
Vi R1

PD K p  ?
PD  K p (TD s  1)  
 TD  ?
First op-amp. Second op-amp.

PI  1 
 ( s  1) 
K K s  Ki T
PI  K p  i  p Kp  i 
s s  1
s 
 T i  72
The second op-amp acts as
K p  ?
a sign inverter as well as a gain adjuster 
 Ti  ?
 PID Second op-amp.
Second op-amp.
To invert the signal

**Changing R1, for example, to tune Ti,

also affect the Kp and Td, so careful tuning
is required in selecting the parameters.

Comparing with PID standard TF.

 1 
PID  K p 1  T d s 
 Ti s  73
 Implement the PID controller
Example
• The figure below represent a PID
implementation; derive the transfer function and
obtain the the PID controller parameters( gains)

Kp=56.42, KD=1, Ki=27.96

….Find TI and TD
74
Kp K
Ti  & Td  d
Ki Kp
 Analogue computer: Example
• AC can be used to built a complete system and
analysis it performance : 1st order, 2nd order, and Nth
other
• Using analogue computer, build a first-order shown
below:
R(s) + 1 Y(s)

- s
Summing point

1
integral with time constant ; 
R4C 1
R3 C1

R1
Input - R4
Output
-
+
R2 +

R6
Make R5  R6 to get a unity feeeback
R5
-
+ 75
 Analogue computer:
Exercise#1 and 2
• Exercise#1: Using analogue computer, built a
Second order shown below:
R(s) + a Y(s)

- s(s  2b)

• Exercise#2: Using analogue computer, built a

complete ( controller and a second order process)
closed loop system shown below:
R(s) + a Y(s)
PID
- s(s  2b)
76
 Summary of the Unit
Student should be able to:
• Define P, I, and D controllers
• Identify the effect of each controller on the
system performance
• Design and tune a PID controller
• Build P, PI, PD and PID controllers using analogue
computer and find the expression for each gain/
time constant
• Build simple systems using analogue computer

77