Chapter 3

Continuously Process Control

 Objectives

Understand two-position and floating-control
mode.

Understand function of proportional control
mode.

Understand function of integral-control mode.

Understand function of derivative-control
mode.

Understand composite and 3-mode controllers.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 2

 Control System Parameters

The error, or deviation of the controlled variable from

the setpoint –

e=r–b

e = error,
b = measured indication of variable,
r = setpoint (or reference) of variable.
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 3
 Error
To express error as percent of span, it is
necessary to write both the setpoint and
measurement in terms of percent of span:
r −b
ep = ×100%
bmax − bmin
where
ep = error expressed as percent of span
bmax = max of measured value
bmin = min of measured value
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 4
 Control Parameter Range
 The ouput is often expressed as a percentage where 0%
is the minimum controller output and 100% the
maximum.
 The controller output as a percent of full scale when the
output varies between specified limits is given by
u − umin
p= ×100%
umax − umin
where
p = controller output as percentage of full scale
u = value of the output
umax = max value of controlling parameter
umin = min value of controlling parameter
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 5
Example 3.1
What is the error if the setpoint is 10.5 mA, and the measured value is
13.7 mA in the standard instrumentation signal range of 4 to 20 mA.
r −b 10.5mA − 13.7 mA
ep = ×100% = × 100% = −20%
bmax − bmin 20mA − 4mA

Example 3.2
A process has a temperature range of 300 K to 440 K and a setpoint of
384 K. What is the percent of span error for the measured
temperature of 379 K?
r −b 384 − 379
ep = ×100% = ×100% = 3.6%
bmax − bmin 440 − 300

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 6

Example 3.3
A controller outputs a 4 to 20 mA signal to control the speed of a
motor from 140 to 600 rpm with linear dependence. Calculate (a)
current corresponding to speed of 310 rpm. (b) the value of (a)
expressed as the percent of control output
(a)Linear equation: S p = mI + S 0
140 = 4m + S 0 − −(1)
600 = 20m + S 0 − −(2)
Solve (2) – (1), m = 28.75 rpm/mA and then we find S0 = 25 rpm

Therefore, S = 28.75I + 25
At 310 rpm, we have 310 = 28.75I = 25. which gives I = 9.91 mA

(b) u − umin 9.91 − 4

p= ×100% = ×100% = 36.9%
umax − umin 20 − 4
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 7
 Discontinuous Controller Modes
The controller modes that show
discontinuous changes in controller output
as controlled variable error occurs.

Discontinuous controller modes include:

 Two-position mode
 Multi-position mode
 Floating-control mode

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 8

 Two-Position Mode (ON/OFF mode)
 When the measured value is less than the setpoint,
full controller output results.
 When it is more than the setpoint, the controller
output is zero.
 Example: heater control,
1. If the temperature drops below a setpoint, the
heater is turned ON.
2. If the temperature rises above the setpoint, it turns
OFF.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 9

  0% error < -∆e p
p=
100% error > + ∆e p

 In virtually any practical implementation of the two-

position controller, there is an overlap as ep increases or
decreases through zero.
 In this span, no change in controller output occurs.
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 10
Example 3.4
A level controller operates between 2 to 3 m with the signal of 4 to 20
mA linearly. A relay closes an inlet valve at 12 mA and opens it at 10
mA. Find (a) the relation between displacement level and current (b)
the neutral zone (displacement gap) in meters.

Solution
(a) Linear equation: H = KI + H 0

2 = 4 K + H 0 − −(1)
3 = 20 K + H 0 − −(2)

Solve (2) – (1), K = 0.0625 m/mA and then we find H0 = 1.75 m

Therefore, H = 0.0625I + 1.75

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 11
Solution
(b) The relay closes at 12 mA, which is a high level, HH

H H = 0.0625 × 12 + 1.75
H H = 2.5m

The lower level, HL occurs at 10 mA, which is

H L = 0.0625 ×10 + 1.75
H L = 2.375m
Thus, the neutral zone is HH – HL = 2.5 – 2.375 = 0.125m

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 12

 Two-Position Mode Application
Large-scale systems with relative slow process
rates.
Thus, in the example of either a room heating or
air-conditioning system, the capacity of the
system is very large in terms of air volume, and
the overall effect of the heater or cooler is
relatively slow.
The process under two-position control must
allow continued oscillation in the controlled
variable because this mode always produces such
oscillation.
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 13
Example 3.5
With the heater off, the temperature in a water tank drops at 2°K per
min. With the heater on, the temperature raises 4°K per min. The
setpoint is 323°K and the neutral zone is ±4% of the setpoint. There is
a 0.5 min. lag at both the on and off switch points. Find the period of
oscillation.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 14

Solution
The neutral zone in temperature is ∆T = ＋ 4% (323 K) = ＋
12.92 K. Therefore, the switching temperature are about 336 K
and 310 K.
When the system is cooling the temperature can be written,
Tc(t) = T(t0) - 2(t-t0)
where t0 is the stating time and T(t0) is the temperture at that
time.
Similarly, when the system is heating the temperature can be
written,
Th(t) = T(t1) + 4(t-t1)
Since there is a 0.5 min. lag at both the on and off switch
point, the actual starting points for heating and cooling are 309
K and 338 K respectively.
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 15
Solution
Assume at t = 0, the temperature has reached the real upper
trip point of 338 K and now is cooling.
The time when it reaches 309 K can be determined as follow:
309 = 338 -2t1
so, t1 = 14.5 min.
Therefore, at 14.5 minutes it starts heating. Similarly, the time
when it reaches 338 can be found as follow:
338 = 309 + 4(t2 - 14.5)
so, t2 = 21.75 min.
Therefore, the total time for one cycle, which is the period, is
21.75 minutes.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 16

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 17
Multi-Position Mode
100 e p > +e1

p =  50 − e1 < e p < +e1

 0 e p < −e1

 A logical extension of the previous two-position control mode is to

provide several intermediate, rather than only two.
 As long as the error is between -e1 and +e1 of the setpoint, the
controller stays at 50% of the output;
 If the error exceeds +e1, the output is increased to 100%;
 If the error is less then –e1, the output decreased to 0%;
 This requires more complicated final control element, because it
must have more than two settings.
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 18
 3-Position Mode
Figure in the next slide shows a graphs of
dynamic variable and final control element
setting versus time of 3-position control.
The response of 3-position mode control has
less oscillation as compared to the 2-
position mode control.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 19

 3-Position Mode

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 20

 Floating-Control Mode
 In the previous modes of controller action, the output
was uniquely determined by the magnitude of the error
input.
 If the error exceeded some preset limit, the output was
changed to a new setting as quickly as possible.
 In floating control, the specific output of the controller is
not uniquely determined by the error.
 If the error is zero (within the neutral zone), the output
does not change but remains (floats) at whatever setting
it was when the error went to zero.
 When the error moves off zero, the controller output
begins to change in certain rate.
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 21
 Single Speed Floating Control
 In the single-speed floating control mode, the output of
the control element changes at a fixed rate when the
error exceeds the neutral zone.
dp
= ±KF ; e p > ∆e p
dt
dp
where = rate of change of controller output with time
dt
K F = rate constant (%/s)

 If the equation above is integrated for the actual

controller output, we get
p = ± K F + p ( 0) e p > ∆e p
where p(0) = controller output at t = 0 s
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 22
 (a) Single speed floating
controller action as
the output rate of
change to input error

(b) An example of error

and controller
response

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 23

Example 3.6
Suppose a process error lies within the neutral zone with p = 25%. At t
= 0 s, the error falls below the neutral zone. If K = +2% per second, find
the time when the output saturates.

The relation between controller output and time is

p = KFt + p(0)
When p = 100
100 = (2%/s)(t) + 25%
so, t = 37.5 s

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 24

 Multiple Speed Floating Control
In this mode, there are several possible rates are
changed by controller output.
Usually, the rate increases as the deviation
exceeds certain limits.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 25

 Floating-Control Mode Application
 Primary applications of the floating-control mode are for
the single-speed controllers with a neutral zone.
 An example of single-speed floating control is a liquid flow
rate through a control valve as shown in the figure.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 26

 Continuous Controller Modes
Continuous controller modes include:
1 Proportional-control mode, P
2 Integral-control mode, I
3 Derivative-control mode, D
4 Composite control modes, PI, PD, PID

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 27

 Proportional-Control Mode
The two-position mode had the controller
output of either 100% or 0%, depending on
the error being greater or less than the
neutral zone.
In multiple-step modes, more divisions of
controller outputs versus error are
developed.
The natural extension of this concept is the
proportional mode.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 28

 Proportional-Control Mode
 A smooth, linear relationship exists between the
controller output and error.
 Each value of error has a unique value of controller
output in one-to-one correspondence.
 The range of error to cover the 0% to 100% is called the
proportional band (PB), because the one-to-one
correspondence exists only for errors in this range.
 This mode can be expressed by:
p = K p e p + p0 (equation P)
where
Kp = proportional gain
Po = controller output with no error
ep = error expressed as percent of span 29
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation
 Direct and Reverse Action
 The error in equation P is expressed using the difference
between setpoint and the measurement, r – b.
 This means that as the measured value increases above
the setpoint, the error will be –ve and the output will
decrease.
 That is, the term Kpep will subtract from po.
 Thus, equation P represents reverse action.
 Direct action would be provided by putting a –ve sign in
front of the correction term.
 A plot of the P mode output versus error is shown in the
figure in next slide.
100
 The proportional band is defined as PB =
KP 30
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation
 The Proportional Band of P Controller
 po is set to 50 %,
 proportional
band is
dependent on
the gain,
 The system
response is
faster with
higher gain but
the PB becomes
narrow and vice
versa.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 31

 Characteristics of the P Mode
1. If the error is 0, the output is a constant
equal to p0.
2. If there is error, for every 1% of error, a
correction of Kp percent is added to or
subtracted from p0, depending on the sign
of the error.
3. There is a band of error about zero of
magnitude PB within which the output is
not saturated at 0% or 100%.
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 32
 Offset
 An important characteristic of the P control mode is that it
produces a permanent residual error in the operating point
of the controlled variable when a change in load occurs.
 The error is referred as offset.
 It can be minimized by a larger constant, Kp, which also
reduces the P band.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 33

Example 3.7

Valve A is linear with flow scale factor of 10 m3/hr per percent of

controller output. This controller output is nominally 50% with KP
= 10% per percent. A load change occurs when flow through
valve B change to 600 m3/hr. Find the new controller output and
offset error.
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 34
Solution
Certainly, valve A must move to a new position of 600 m3/hr flow, if
not the tank will empty. This can be done by a 60% new controller
output because  3
 10m / h
QA =  (60%) = 600m 3 / h
 % 
Because this is a proportional controller, we have
p = Kpep+ po
with the nominal condition po = 50%. Thus
p − po 60 − 50
ep = = %
Kp 10
e p = 1%
so a 1 % offset error occured because the load change.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 35

 P-Controller Application
use in processes which need manual reset of
the operating point.
use in processes where large load changes are
unlikely.
if the gain is too large, PB becomes very small,
it then just acts like a ON/OFF controller.
with high gain, a P-controller causes
oscillations of the error.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 36

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 37
 Integral-Control Mode
Offset error in P-controller occurs because it can’t
adapt to the external loads.
P-controller can reduce the error but cannot
eliminate it.
Offset can be eliminated with integral action
controller by changing the zero-error output.
Integral action is provided by summing the error
over time, multiplying by a gain, and adding the
result to the controller output.
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 38
 Integral-Control Mode
I-controller does not give any effect to the
random excursion error because the net error
sum is zero.
The magnitude of the output of an integrator is
not proportional to the input, the rate of change
of the output depends on the error.
This can maintain the output even after the input
decreased to zero.
When the error is constant, the output is changed
linearly (ramp) as a function of time.
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 39
 The I-Controller equation
t dp
p (t ) = K I ∫ e p dt + p (0) = K I eP
0 dt
KI = Integration gain in %/s
p(0) = Controller output when
the integral action starts.
This can be express in the relation of the rate at which the controller
output changes:
- When an error occurs, the controller begins to increase or decrease
its output at a rate that depends upon the size of the error and the
gain.
- If the error is zero, the controller output is not changed.
- If there is a constanterror, the controller output begins to ramp up
(increases linearly).
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 40
 dp
= K I eP
dt
(a) The rate of change of
controller output depends
on the value of the gain
and the error.

(b) When a constant error

occurs, the controller
output depends on the
gain, it eventually saturates
at 100 %.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 41

 I-Controller Characteristics
When the error is zero, the output stays fixed at a
value equal to what it was when the error
becomes zero.
If the error is not zero, the output will begin to
increase or decrease at a rate of KI % per second
for every 1 % of the error.
I-controller is not used alone, it is normally used
together with the P-controller.
It can be used for systems with small process lags
and correspondingly small capacities.
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 42
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 43
 Integral mode
output and error,
showing the
effect of process
and control lag.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 44

Example 3.8
An Integral mode speed controller operates in a range of 10 to 15 rpm
with setpoint 12 rpm. Initially the controller output is 22%. KI = –
0.15% controller output per sec. per percentage error. If the speed
jumps to 13.5 rpm, Find the controller output after 2 sec. for a
constant ep.
r −b 12 − 13.5
Solution: ep = ×100% = ×100% = −30%
bmax − bmin 15 − 10

dp
= K I e p = (−0.15s −1 )(−30%) = 4.5% / s
dt
t
p = K I ∫ e p dt + p(0)
0

p = K I e p t + p (0) (since e p is constant)

After 2 s, we have,
p = (0.15)(30%)(2) + 22
p = 31%
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 45
 Derivative-Control Mode
 One solution to the overshoot problem is to include
derivative control.
 Derivative control “applies the brakes”, slowing the controlled
variable just before it reaches its destination.
 Suppose you were in charge of controlling some variable, and
at some time, t0, your helper yelled out, “ The error is zero.
What action do you want to take?”
 Well, it would seem perfectly rational to answer “None”
because the error was zero.
 …wait, suppose you have the error look like this:

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 46

 Derivative-Control Mode
 You can clearly see that even though the error at t0 is zero,
it is changing in time and will certainly not be zero in the
following time.
 Therefore, some action should be taken even though the
error is zero!
 This scenario describes the nature of derivative action.
 When a sudden need arises in a process which requires a
significant addition of energy input, the disturbance may
take a long time for a slow controller to eliminate it.
 To overcome this, an accelerated action is required to put
the system under control more quickly.
 A corrective action that is initially large but tapers off as
time goes on is desirable.
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 47
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 48
 Derivative-Control Mode
 This can be achieved by using a derivative (D) controller.
 The output of a D-controller is proportional to the rate of
change of its input.
de p
p(t ) = K D
dt
KD = gain for how much percent to change the
controller output for every % per second
rate of change of error.
 It is not used alone because it provides no output when
the error is constant.
 The gain is small, because a rapid rate of change of error
can cause very large, sudden change of controller output
which can lead to instability.
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 49
 D- Controller Characteristic
1. If the error is zero, the mode provides no
output.
2. If the error is constant in time, the mode
provides no output.
3. If the error is changing in time, the controller
gives an output of KD percent for every 1% per
second rate of change of error.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 50

 - For this example, we assume that the controller output without error is
50%.
- When a rapid positive error occurs, the controller produces large output.
- When no error occur, controller output back to 50 %.
- When the error is decreasing, the controller produces a negative output.
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 51
  By analyzing some typical inputs, it is possible to predict the
response of the output of a D-controller.
 If the input is a step function, a steep output will be produce.
 If the input is a linear function, the output will maintain a
constant value that is equal to the slop of the function.
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 52
 Composite Control Modes
This is commonly used in complex processes.
A combination of the control modes is used to
eliminate some limitations for the control mode
when used individually.
A stand alone I-controller has very poor transient
response.
When the error shoots up to a high value, the
output of the I-controller only begins to ramp.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 53

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 54
 Composite Control Modes
Even though the error signal will be reduced to
zero, the system reacts sluggishly to dynamic
system disturbances.
P-controller reacts quickly but cannot reduce the
error to zero.
I-controller reacts slowly but over a period of
time can eliminate the error.
P+ I can combine both merits of each controller.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 55

 Proportional-Integral (PI) Controller

t
p = K P e p + K P K I ∫ e p dt + pI (0)
0

PI (0) = Integral term value at t = 0 (initial value).

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 56

 PI Controller
t
p = K P e p + K P K I ∫ e p dt + pI (0)
0

 The advantage: one-to-one proportional mode is

available and the integral mode eliminates the inherent
offset.
 From the equation, KP can effect KI but KI can be adjusted
independently.
 With P mode alone, offset occurs when load changes.
This offset can only be eliminated by the I mode.
 For PI mode, the integral function provides zero error
output after the load changed.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 57

 Summary of the PI’s characteristic
1. When the error is zero, the controller output is
fixed at the value that the Integral term had when
the error went to zero.
2. If the error is not zero, the P-term contributes a
correction, and the I –term begins to increase or
decrease the accumulated value, depending on the
sign of the error.
t
p = K P e p + K P K I ∫ e p dt + pI (0)
0

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 58

 Reverse Action PI
Controller
 Error occurred at t1.
 The controller
produces correction
output which is sum of
proportional plus
integral action.
 This corrective output
leads the error to zero
eventually.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 59

 Disadvantages:
Because of the integration time, the process must
have relatively slow changes in load to prevent
oscillations induced by the integral overshoot.
During start-up of a batch process, the integral action
causes a considerable overshoot of the error and
output before settling to the operation point.
Pl mode can shift the whole Proportional Band. (PB is
defined as that +ve and –ve error for which the
output will be driven to 0% and 100%). Therefore, it
changes the amount of error that will bring such
saturation by the proportional term.
The output saturates whenever the error exceeds the
PB limits.
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 60
 Overshoot and cycling
of PI mode during
start-up (dashed lines
show the PB).

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 61

Example 3.9
From the given error as shown below, plot the PI controller
output. KP = 5, KI = 1.0 s-1 and p(0) = 20%.
t
p = K P e p + K P K I ∫ e p dt + pI (0)
0

For 0 ≤ t ≤ 1, ep = t
t
p1 = 5t + 5∫ t dt + 20
0
t
t  2
p1 = 5t + 5  + 20
 2 0
p1 = 5t + 2.5t 2 + 20
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 62
 t
Solution p = K P e p + K P K I ∫ e p dt + pI (0)
0

For 1 ≤ t ≤ 3, ep = 1 pI (1) = 2.5t 2 + 20 = 22.5%

t
p2 = 5 + 5∫ 1 dt + 22.5
1

p2 = 5 + 5[t ] 1t + 22.5
p2 = 5 + 5[t − 1] + 22.5

For t ≥ 3, e p = 0 pI (3) = 5[t − 1] + 22.5 = 32.5%

t
p3 = 5[0] + 5∫ 0 dt + 32.5
3

p3 = 32.5

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 63

Solution... e = t ep = 1
p
ep = 0

p2 = 5 + 5(t − 1) + 22.5

p3 = 32.5%

p2 (1) = 27.5%

p1 = 5t + 2.5t 2 + 20
p(0) = 20%

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 64

 Proportional-Derivative (PD) Control
 A PD mode has many industrial applications.
 It involves the serial used of the P and D modes.
de p
p = K Pep + K P K D + p0
dt
 Derivative action moves the controller output in relation
to the error rate of change.
 This system cannot eliminate the offset of proportional
controllers.
 It can handle fast process load changes as long as load
change offset error is acceptable.
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 65
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 66
Example 3.10
A PD controller with KP = 5, KD = 0.5 s, and p0 = 20%. Draw
the controller output.

de p
p = K Pep + K P K D + p0
dt
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 67
Solution de p
p = K Pep + K P K D + p0
For 0 ≤ t ≤ 1, ep = t dt
p1 = 5t + 2.5 + 20

For 1 ≤ t ≤ 3, ep = 1
p2 = 5 + 20 = 25

For 3 ≤ t ≤ 5, e p = −0.5t + 2.5

p3 = −2.5 + 12.5 − 12.5 + 20
p3 = −2.5t + 31.25
For t ≥ 5, e p = 0
p4 = 20

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 68

 p1 = 5t + 2.5 + 20
p2 = 5 + 20 = 25

p3 = −2.5t + 31.25

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 69

 PID controller
It is one of the most powerful but complex
controller mode. It can be expressed as:
t de p
p = K P e p + K P K I ∫ e p dt + K P K D + p I ( 0)
0 dt

This mode eliminates the offset of the

proportional mode and still provides fast
response.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 70

 PID 3-Mode Controller

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 71

 PID Controller
 PID action:
– P-controller provides fast response to system
disturbances,
– D-controller ensures that sudden disturbances will be
met with a quick correction of the error,
– I-controller provides a means of eventually
elimination of the error altogether.

 The tuning of PID controller depends on:

– The configuration of the controller.
– The characteristic of the process being controlled.
– The desired controller performance.
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 72
 PID Controller
 Different tuning is needed for different process.
 Two precautions:
– If the D control action is too large, this can cause
the system to oscillate.
– If the I control action is too large, this can cause
the system to overshoot.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 73

 PID Control
Action

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 74

Example 3.11
Draw the controller output for a 3-mode PID controller
with KP = 5, KI = 0.7 s-1, KD = 0.5 s, and pI(0) = 20%.
t de p
p = K P e p + K P K I ∫ e p dt + K P K D + pI (0)
0 dt

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 75

Solution t de p
p = K P e p + K P K I ∫ e p dt + K P K D + p I ( 0)
0 dt
t de p
p = 5e p + 3.5∫ e p dt + 2.5 + 20
0 dt

For 0 ≤ t ≤ 1, ep = t
t
p1 = 5t + 3.5∫ t dt + 2.5 + 20
0

p1 = 5t + 1.75t 2 + 22.5

2
For 1 ≤ t ≤ 3, ep = 1 pI (1) = 1.75t + 20 = 21.75%
t
p2 = 5 + 3.5∫ (1) dt + 0 + 21.75
1

p2 = 3.5(t − 1) + 26.75
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 76
 t de p
Solution p = K P e p + K P K I ∫ e p dt + K P K D + PI (0)
0 dt

pI (3) = 3.5(t − 1) + 21.75 = 28.75%

For 3 ≤ t ≤ 5, e p = −0.5t + 2.5
t
p3 = 5(−0.5t + 2.5) + 3.5∫ (−0.5t + 2.5) dt − 1.25 + 28.75
3

p3 = −0.875t 2 + 6.25t + 21.625

pI (5) = 3.5(−0.25t 2 + 2.5t − 5.25) + 28.75 = 32.25%

For t ≥ 5, e p = 0
p4 = 32.25%
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 77
 p3 = −0.875t 2 + 6.25t + 21.625

pI = 32.25%

p2 = 3.5(t − 1) + 26.75

p1 = 5t + 1.75t 2 + 22.5

78
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 78
 p3 = −0.875t 2 + 6.25t + 21.625

pI =
p2 = 3.5(t − 1) + 26.75 32.25
%

p1 = 5t + 1.75t 2 + 22.5

PID Controller

PD Controller 79
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 79
 PID tuning using Matlab

The Matlab optimization tool can help to

obtain PID’s gain value.

You may type “pidtune_demo” in the Matlab

command window.

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 80

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 81
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 82
Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 83
 Q&A

Prepared by Wong Chee Woon UEEA2413 Process Control and Instrumentation 84