Introduction to Control Engineering

Course Code: ECEg4165

1
 CHAPTER 6
DESIGN OF CONTROLLER
6.1. Introduction
➢ If the system cannot achieve the asked performance, we have to design a
controller or change the control law.

❖ The goal/objectives of controller design are:

➢ To ensure safe operation of the process/system.

➢ To ensure stability.

➢ To suppress or reject the effect of disturbance.

➢ To optimize production.
2
6.2.Types of controllers
➢ There are many ways by which a control unit can react to an error
and supply an output for correcting elements. Therefore, Industrial controllers may
be classified according to their control actions as:

• Two-position or On-off controller

• Proportional controller (P-controller)

• Integral controller (I-controller)

• Derivative controller (D-controller)

• Proportional plus Derivative controller (PD-controller)

• Proportional plus Integral controller (PI-controller)

• Proportional plus integral plus derivative controller (PID-controller) 3

 Cont. …
❖ Two-position or On-off Controller:

➢ It is simple and most common type of control. The control signal 𝑢(𝑡) has only two
states either on or off (high or low) and depending on whether the actuating error
signal is positive or negative, so that
𝑢1 𝑓𝑜𝑟 𝑒(𝑡) > 0
𝑢 𝑡 =ቊ
𝑢2 𝑓𝑜𝑟 𝑒(𝑡) < 0

Fig.6.1: (a) Block diagram of an on-off controller; (b) block diagram of an on-off controller with differential gap.
4
 Cont. …

➢ Such controllers are relatively simple and inexpensive and, for this reason,

widely used in both industrial and domestic control systems.

➢ The range through which the actuating error signal must move before the

switching occurs is called the differential gap.

➢ However, reducing output oscillation by decreasing the differential gap,

increases the number of on-off switching per minute and reduces the useful

life of the component.

5
 Cont. …
❖ Proportional (P) controller:
➢ It generate a control signal proportional to the error between the desired output
and actual output.

U(s)
u t = kpe t & Gc s = = kp
E(s)

where, k p − proportional gain, 𝑟(𝑡) − set point, e(𝑡) −error signal, 𝑢(𝑡) −control
signal and 𝑦(𝑡) −actual output. 6
 Cont. …
➢ Proportional controller is essentially an amplifier with an adjustable gain. A
summing operational amplifier with an inverter can be used as a proportional
controller.

❖ Performance analysis of proportional controller:

➢ Let consider the closed loop proportional controller for a first order system given
k
by Gp s = . Where τ − time constant and k −system gain.
τs+1

7
 Cont. …
➢ The closed-loop transfer function is given by

Y(s) kpk k p kΤ(1 + k p k)

Gcl s = = = τ
R(s) τs + 1 + k p k s+1
1 + kpk

➢ Proportional controller modifies the time constant of the system. The time constant
is inversely proportional to k p , as k p increases the time constant decreases so that
the system response become faster.

➢ The steady state output for unit step input is given by

k p kΤ(1 + k p k) 1 kpk
yss = lim sY(s) = lim s τ =
s→0 s→0 s + 1 s 1 + kpk
1 + kpk
8
 Cont. …
➢ The steady state error for unit step input is given by

1
ess = lim sE(s) = lim s R s − Y(s) = lim sR(s) − lim sY s = 1 − yss =
→0 s→0 →0 →0 1+kp k

➢ Therefore, proportional controller reduces the steady state error to a certain

level but it can’t avoid completely.

➢ The sign of 𝑘𝑝 can be chosed to make the controller output increase (or
decrease) as the error signal increases.

➢ As the proportional gain is increased the system responds faster but

increases overshoot. Then the system becomes progressively underdamped
and eventually unstable.
9
 Cont. …
Exercise 6.1: Consider the following proportional controlled system.

a) Determine the proportional gain so that overshoot is 35%.

0.5
b) If a disturbance of 𝐷 𝑠 = interfere just at the output (i.e. after a plant) of
𝑠

the system determine the steady state disturbance.

10
 Cont. …
❖ Integral controller (I-controller):
➢ This produces a control action that is proportional to the integral of the error
with time. Therefore, a constant error signal will produce an increasing
correcting signal. The correction continues to increase as long as the error
persists.
t

u t = k i න e t d𝑡
0

where, k i −integral gain (adjustable constant).

U(s) k i
Gc s = =
E(s) s
11
 Cont. …
❖ Performance analysis of Integral controller:

➢ Let consider the closed-loop Integral controller for the process transfer function
k
given by Gp s = . Where τ − time constant and k −system gain.
τs+1

➢ The closed-loop transfer function is

Y(s) kik
Gcl s = = 2
R(s) τs + s + k i k
12
 Cont. …
➢ Integral controller increase the order of the system. The steady state output
for unit step input is given by
kik 1
yss = lim sY(s) = lim s 2 =1
s→0 s→0 τs + s + k i k s

➢ The steady state error of the system is

yss = lim sE(s) = lim s R s − Y(s) = lim sR(s) − yss = 1 − 1 = 0

s→0 s→0 s→0

➢ Therefore integral controller completely eliminate the steady state error.

➢ An inherent disadvantage of integral control action is a phenomenon known

as reset windup or integral windup. The integral mode causes the controller
output to change as long as the error is not zero.
13
 Cont. …
➢ When a sustained error occurs, the integral term becomes quite large and
the controller output eventually saturates and unable to bring output to
setpoint.

➢ Further buildup of the integral term while the controller is saturated is

referred to as reset windup or integral windup.

Exercise 6.2:
a) Design a unity feedback integral controller for the process transfer function
10
𝐺𝑝 𝑠 = to have underdamped response for unit step input.
4𝑠+1
12
b) Given the plant transfer function 𝐺𝑝 𝑠 = , design a unity feedback integral
9𝑠+1
controller to achieve 0.35 overshoot.
14
 Cont. …
❖ Derivative controller (D-controller):
➢ This produces a control action that is proportional to the rate at which the error is
changing.

➢ When there is a sudden change in the error signal the controller gives a large
correcting signal. When there is a gradual change only a small correcting signal is
produced.
de(t) U(s)
u t = kd ⇒ Gc s = = kds
dt E(s)
where, Td −is called the derivative time and k d −is derivative gain.

➢ A derivative control will have the effect of increasing the stability of the system,
reducing the overshoot, and improving the transient response.
15
 Cont. …
➢ It gives rapid output, reduces the time that is required to return the process
variable to setpoint in slow process.

➢ In practical case derivative controller is not used alone because of the following
reason:
𝑑𝑒(𝑡)
a) If the error is constant 𝑢 𝑡 = 𝑘𝑑 =0
𝑑𝑡

➢ This implies the controller doesn’t take any action to compensate steady state
error. Consequently, it always used in conjunction with proportional or
proportional-integral control.

b) It amplifies high frequency noise, if the time constant is very small.

➢ To avoid this case derivative controller can be designed with low pass filter.
16
 Cont. …
❖ Proportional-plus-Derivative controller (PD-controller):
➢ The control action of a proportional plus derivative controller is defined by
de(t) de(t) de(t)
u t = k p e t + k p Td = k p e t + Td = kpe t + kd
dt dt dt

and the transfer function of the controller is

U(s)
Gc s = = k p + k p Td s = k p + k d s
E(s)
where, Td − is derivative time.
17
 Cont. …
Exercise 6.3:

a) A second order process has the following transfer function

0.4
𝐺𝑝 (𝑠) =
𝑠(0.2𝑠 + 0.8)

Design a unity feedback PD-controller to achieve 24% overshoot and

5.15 𝑠𝑒𝑐 settling time within five percent.

1
b) For the plant transfer function given by 𝐺 𝑠 = , design a unity feedback
𝑠2 +𝑠

PD-controller to have desired system response such that undamped natural

frequency is 4 𝑟𝑎𝑑 Τ𝑠𝑒𝑐 and damping ratio is 0.2
18
 Cont. …
❖ Proportional-plus-Integral controller (PI-controller):
➢ The control action of a proportional-plus-integral controller is defined by
t t t
kp 1
u t = k p e t + න e t dt = k p e t + න e t dt = k p e t + k i න e t dt
Ti Ti
0 0 0

and the transfer function of the controller is

U(s) 1 ki
Gc s = = kp 1 + = kp +
E(s) Ti s s

where, Ti − is derivative time.

➢ The response becomes more oscillatory and needs longer to settle, the error
disappears.
19
 Cont. …
Exercise 6.4:

1) Consider the following proportional-plus-Integral controlled system shown

below. Design PI-controller to have desired system response such that
damping ratio is 0.7 and settling time within two percent criteria is less than
1𝑠𝑒𝑐 and the steady state error is zero for step response.

20
 Cont. …
2) Consider the plant transfer function given by

1
G s = 2
s +s

a) Design a unity feedback PI-controller to achieve 2 𝑠𝑒𝑐 settling time within

two percent criteria and a peak time of 1𝑠𝑒𝑐.

b) What is the steady state error of this system for a unit ramp reference?

21
 Cont. …
❖ Proportional-plus-Integral-plus-Derivative (PID) controller:
➢ It is the combination of proportional control action, integral control action, and derivative
control action.

➢ This combined action has the advantages of each of the three individual control actions.
The equation of a controller is given by
t
kp de(t)
u t = k p e t + න e t dt + k p Td
Ti dt
0
➢ The transfer function of the controller is
U(s) 1 ki
Gc s = = kp 1 + + Td s = k p + + k d s
E(s) Ti s s
kp
Where, k i = and k d = k p Td
Ti

➢ All design specifications can be reached. 22

 Cont. …
❖ For designing a PID controller:

1) Obtain an open-loop response and determine what is needed 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.

5) Adjust each of 𝑘𝑝 , 𝑘𝑖 and 𝑘𝑑 until you obtain a desired response.

Note: please keep in mind that you don’t need to implement all the three
controllers into a single system if not necessary. Example if PI controller gives you
a good enough response, then you don’t need to implement derivative controller
to the system. 23
 Cont. …
❖ Controller Comparison
➢ Proportional (P) controller:
i. Simplest controller to tune (k p −only one parameter).
ii. Offset with sustained disturbance or setpoint change.
➢ Proportional –plus-Integral (PI) controller:
i. More complicated to tune (k p & Ti −two parameters).
ii. Better performance than proportional only.
iii. No offset.
iv. Most popular feedback controller.
➢ Proportional-plus-Integral-plus-Derivative (PID) controller:
i. Most complicated to tune (k p , Ti & Td −three parameters).
ii. Better performance than proportional-plus-Integral.
iii. No offset.
iv. Derivative action may be affected by noise 24
 Cont. …
Table 6.1: The Characteristics of P, I, and D controllers

Controller Rise time Overshoot Settling Steady state

time error

Proportional Decrease Increase Small change Decrease

(k p )

Integral (k i ) Decrease Increase Increase Eliminate

Derivative (k d ) Shall change Decrease Decrease Small change

25
 Cont. …
❖ Controller parameter tuning:

➢ The process of selecting the controller parameters to meet given performance

specifications is known as controller tuning.

➢ Tuning parameters essentially determine:

1) How much the correction should be made is determined by the proportional mode
of the controller.

2) How long the correction should be applied is determined by the integral mode of
the controller.

3) How fast should the correction be applied is determined by the derivative mode of
the controller. 26
6.3. PID Tuning Rules (Ziegler Nichols Procedures)
➢ Consider the block diagram shown below in which PID controller used to
control a plant.

➢ The transfer function of PID controller is given as

U s ki 1
Gc s = = kp + + kds = kp 1 + + Td s
E s s Ti s
kp kd
where, Ti = and Td =
ki kp
27
 Cont. …
➢ Ziegler and Nichols suggested rules for tuning PID controllers
experimentally which are useful when mathematical models of plants are
not known.

➢ These rules can, of course, be applied to the design of system with known
mathematical models.

➢ Such rules suggest a set of values of 𝑘𝑝 , 𝑇𝑖 and 𝑇𝑑 that will give a stable
operation of the system.

➢ However, the resulting system may exhibit a large maximum overshoot in

the step response, which is unacceptable.
28
 Cont. …
➢ In such a case we need series of tunings until an acceptable result is
obtained.

➢ In fact, the Ziegler-Nichols tuning rules give an educated guess for the
parameter values and provide a starting point for fine tuning, rather than
giving the final settings for 𝑘𝑝 , 𝑇𝑖 and 𝑇𝑑 in a single shot.

➢ Ziegler and Nichols proposed rules for determining values of the 𝑘𝑝 , 𝑇𝑖 and
𝑇𝑑 based on the transient response characteristics of a given plant.

➢ Such determination of the parameters of PID controllers or tuning of PID

controllers can be made by engineers on-site by experiments on the plant.
29
 Cont. …
➢ There are two methods called Ziegler-Nichols tuning rules:

a) First method (open-loop method)

b) Second method (closed-loop method)

❖ Ziegler-Nichol’s first method

➢ In the first method, we obtain experimentally the response of the plant to a

unit step input.

Unit step response of a plant. 30

 Cont. …
➢ If the plant involves neither integrator(s) nor dominant complex-conjugate
poles, then such a unit-step response curve may look S-shaped.

S-shaped response curve. 31

 Cont. …
➢ This method applies if the response to a step input exhibits an S-shaped curve.

➢ Such step-response curves may be generated experimentally or from a dynamic

simulation of the plant.

Table 6.2
Types of 𝐤𝐩 𝐓𝐢 𝐓𝐝
controller
P T ∞ 0
L
PI T L 0
0.9
L 0.3

PID T 2L 0.5L
1.2
L
32
 Cont. …
❖ Ziegler-Nichol’s second method
➢ In the second method, we first set Ti = ∞ and Td = 0.

➢ Using the proportional control action only, increase k p from 0 to critical value k cr at
which the output first exhibits sustained oscillations.

➢ If the output does not exhibit sustained oscillations for whatever value k p may take,
then this method does not apply.

← sustained oscillation with period pcr

33
 Cont. …
➢ Thus, the critical gain and the corresponding period (pcr ) are experimentally
determined.

➢ Therefore, set the values of the parameters k p , Ti and Td according to the formula
shown in Table 6.3.

Table 6.3
Types of kp Ti Td
controller
P 0.5k cr ∞ 0

PI 0.45k cr 1 0
p
1.2 cr
PID 0.6k cr 0.5pcr 0.125pcr

34
 Cont. …
❖ Note: If the system has a known mathematical model (such as the transfer
function), then we can use the root-locus method to find the critical gain k cr
2π
and the frequency of the sustained oscillations ωcr , where pcr = .
ωcr

➢ These values can be found from the crossing points of the root-locus
branches with the jω −axis. (Obviously, if the root-locus branches do not
cross the jω −axis, this method does not apply).

Example: Consider the control system shown in Figure 6.2 in which a PID
controller is used to control the system. Apply a Ziegler-Nichols tuning rule
for the determination of the values of parameters k p , Ti and Td .
35
 Cont. …

Fig.6.2: PID controlled system.

Solution:

➢ Since the plant has an integrator, we use the second method of Ziegler-Nichols
tuning rules. By setting Ti = ∞ and Td = 0, the closed-loop transfer function is

Y(s) kp
= 3
R(s) s + 6s 2 + 5s + k p 36
 Cont. …
➢ The characteristic equation is
s 3 + 6s 2 + 5s + k p = 0

➢ The Routh array becomes as follows:

30−𝑘𝑝
For stability, > 0 and 𝑘𝑝 > 0
6

∴ 0 < k p < 30

➢ Thus sustained oscillation will occur if k p = 30 (i.e., at which root-locus cross

jω −axis). So, the critical gain k cr = 30.

➢ With gain k p set equal to k cr (30), the characteristic equation becomes

37
 Cont. …
s3 + 6s2 + 5s + 30 = 0

➢ To find the frequency of the sustained oscillation, we substitute s = jω into this

characteristic equation
3 2
𝑗𝜔 + 6 𝑗𝜔 + j5ω + 30 = 0 ⇒ 6 5 − 𝜔2 + 𝑗𝜔 5 − 𝜔2 = 0 ⇒ 𝜔 = 𝜔𝑐𝑟 = 5

➢ Hence, the period of sustained oscillation is

2π 2𝜋
pcr = = = 2.8099
ωcr 5

➢ Therefore,

k p = 0.5k cr = 0.5 × 30 = 6, Ti = 0.5pcr = 0.5 × 2.8099 = 1.405 and Td = 0.125pcr =

0.125 × 2.8099 = 1.35124 38
End

39