You are on page 1of 19

STUDY OF PROCESS CONTROL - I

Summary
The main objectives of this experiment were to learn a First Order Plus Time Delay Model
(FOPTD) for a process, to observe the output response curve in comparison with the input curve
using a step change data and then to design and tune Proportional Integral (PI) and Proportional
Integral Derivative (PID) controllers applying the Internal Model Control (IMC) method in
MatLab Simulink. Firstly three process parameters were assumed as process gain, Kp = 5, time

delay, θ = 10 s and time constant, 𝜏 = 134 s to define a process. Then a step function with step

time = 100s, initial value =0 and final value = 1 was introduced with the process to generate an
open loop response curve (process without a controller). Then in accordance with the IMC
method, the controller parameters, controller gain, Kc, integral time, 𝜏I and derivative time, 𝜏D

were calculated (for PI Kc and 𝜏I only) using the process parameters. The obtained controller
parameters were used to generate the closed loop response curve (process with a controller).
These values were the initial guess for the tuning process and tuning was done by changing these
parameters to obtain a output response curve with a much faster response without any
overshooting and oscillation. The Integral of Absolute Errors (IAE)for both PI and PID were
noted and also minimized in the tuning process.The best tuning parameters for PI were:

Controller gain, Kc = 0.492, integral time, 𝜏I = 130 s with a minimum IAE of 0.054. The best
tuning parameters for PID were: Controller gain, Kc = 0.6, integral time, 𝜏I = 135 s and

derivative time, 𝜏D = 5 s with a minimum IAE of 0.0462.

1
STUDY OF PROCESS CONTROL - I

Simulink Block Diagram

Display

5
173s+1
Step TransferFcn Transport
Delay Scope

simoutOLBD

ToWorkspace

Figure 01: Simulink block diagram for open loop

Figure 02: Simulink block diagram for closed loop

2
STUDY OF PROCESS CONTROL - I

Observed Data
Given,

Step time =100 s

Initial value of input = 0

Final value of input = 1

Initial output value = 0

Final Output value = 5


𝐶ℎ𝑎𝑛𝑔𝑒 𝑜𝑓 𝑂𝑢𝑡𝑝𝑢𝑡 5−0
Process gain, Kp = = =5
𝐶ℎ𝑎𝑛𝑔𝑒 𝑜𝑓 𝑖𝑛𝑝𝑢𝑡 1−0

Time delay, θ = (110 – 100) s = 10 s

Time constant, 𝜏 = 134

𝐾𝑝 𝑒 −𝜃𝑠
in the Transfer Function, G(s) =
τs+1

5𝑒 −10𝑠
=
134s + 1

The open loop response curve is presented next page with detailed information.

3
STUDY OF PROCESS CONTROL - I

Figure 03: Open loop response curve and different process parameters

4
STUDY OF PROCESS CONTROL - I

Tuning Data
Table 01: Tuning parameters for PI and PID controllers

Integral time, Derivative Integral


Type of No. of Controller
Absolute
controller observations gain, Kc 𝜏I (s) time, 𝜏D(s) Error (IAE)

01 (IMC) 0.49 134 0 0.05519

02 (Tuning 1) 0.495 132 0 0.05512


PI
03 (Tuning 2) 0.492 130 0 0.05411

04 (Tuning 3) 0.4905 134 0 0.05414

01 (IMC) 0.56 139 4.82 0.05019

02 (Tuning 1) 0.600 135 4.80 0.0462


PID
03 (Tuning 2) 0.565 135 4.80 0.04894

04 (Tuning 3) 0.56 138.5 4.84 0.05004

5
STUDY OF PROCESS CONTROL - I

Sample Calculation
Calculation of controller parameters by IMC method

1. For PI controller

1 𝜏
Controller gain, Kc =
𝐾𝑝 𝜃 + 𝜏𝑐

1 134
= = 0.05519
5 10 + 134
3

Integral time or reset time, 𝜏I =𝜏= 134 s

2. For PID controller

𝜃
1 𝜏+2
Controller gain, Kc =
𝐾𝑝 𝜏𝑐 + 𝜃
2

10
1 134+ 2
= 134 10= 0.0509
5 + 2
2

𝜃
Integral time or reset time, 𝜏I =𝜏+ = 134 + 5 = 139 s
2

𝜏𝜃
Derivative time, 𝜏D =
2𝜏 + 𝜃
134 ×10
= = 4.82 s
2 ×134 + 10

6
STUDY OF PROCESS CONTROL - I

Graphical Representation

Figure 04: Closed loop response curve for PI controller IMC

7
STUDY OF PROCESS CONTROL - I

Figure 05: Closed loop response curve for PI controller Tuning 1

8
STUDY OF PROCESS CONTROL - I

Figure 06: Closed loop response curve for PI controller Tuning 2

9
STUDY OF PROCESS CONTROL - I

Figure 07: Closed loop response curve for PI controller Tuning 3

10
STUDY OF PROCESS CONTROL - I

Figure 08: Closed loop response curve for PID controller IMC

11
STUDY OF PROCESS CONTROL - I

Figure 09: Closed loop response curve for PID controller Tuning 1

12
STUDY OF PROCESS CONTROL - I

Figure 10: Closed loop response curve for PID controller Tuning 2

13
STUDY OF PROCESS CONTROL - I

Figure 11: Closed loop response curve for PID controller Tuning 3

14
STUDY OF PROCESS CONTROL - I

Results and Discussion


Results

The controller parameters for PI and PID controller are calculated in the sample calculation
section in accordance with IMC (Internal Model Control) method. Then PI and PID controllers
were tuned to find the best tuning parameters for which IAE (Integral of Absolute Errors) were
minimum without overshooting and oscillation.

PI Controller:

The best tuning parameters were: Controller gain, Kc = 0.492, integral time, 𝜏I = 130 s with a
minimum IAE of 0.054

PID Controller:

The best tuning parameters were: Controller gain, Kc = 0.6, integral time, 𝜏I = 135 s and
derivative time, 𝜏D = 5 s with a minimum IAE of 0.0462

Discussion

The open loop response curve was generated by simulation in Simulink to observe how the
output response curve would look like in comparison with the input response curve,for a process
defined by its three parameters: process gain, time delay and time constant. The input curve was
generated by using a step function. The open loop response curve shows the pattern of output
response without any controller. The process defined here is a First Order Plus Time Delay
(FOPTD) model, which is of first order with respect to ‘s’ domain and a delay time is attributed
here too. If the time delay is not attributed then the model is called First Order (FO) model.

The standard form of a first order plus time delay model is


Kpe s
G( s) 
s  1
Where, KP = Process gain
∆ 𝑂𝑢𝑡𝑝𝑢𝑡
=
∆ 𝐼𝑛𝑝𝑢𝑡

Time delay, θ is the amount of time, by which the output response curve is lagging the input
response curve.

15
STUDY OF PROCESS CONTROL - I

Time constant, 𝜏is the difference between the time required for 63.2% of output change and the
time required for response with delay, which is illustrated in figure 03. 63.2% of output change is
chosen because the output response curve is steepest upto 63.2%.

Thumb rule is, time required for complete output response curve is almost five times the time

constant, 𝜏.Theoretically the process output never reaches the new steady-state value except as
t→∞. It does only approximate the final steady-state value when t ≈ 5𝜏.

The FOPTD model of the process was represented using a transfer function and a transport delay
block in Simulink. The transfer function accounted for process gain in the numerator coefficient
and time constant in the denominator coefficient and the transport delay accounted for the time
delay term.

The process defined for open loop was then simulated to be controlled by PI and PID controllers.
The control parameters for both the controllers were calculated in accordance with IMC method
to obtain the first guess. Calculations using IMC method are presented in the sample calculation
section.

The PID controller has three principal control effects. The proportional (P) action gives a change
in the input (manipulated variable) directly proportional to the control error. But this action
cannot eliminate the control error entirely, there is always a difference between the input and
output response curve which is called ‘offset’. The integral (I) action gives a change in the input
proportional to the integrated error, and its main purpose is to eliminate offset. The less
commonly used derivative (D) action is used in some cases to speed up the response or to
stabilize the system, and it gives a change in the input proportional to the derivative of the
controlled variable. The overall controller output is the sum of the contributions from these three
terms. The corresponding three tunable PID parameters are:

 Controller gain K C (increase of K C value gives more proportional action and faster control)

 Integral time  I (decrease of  I value gives more integral action and faster control)

 Derivative time  D (increase of  D value gives more derivative action and faster control)

16
STUDY OF PROCESS CONTROL - I

Tuning action is executed in accordance with these relationships. For PI controller to be


executed, PID controller block was selected in simulation and then derivative time,𝜏D was set to
zero.

The standard form of a PI controller is


1
GC  K C (1  )
I s
The standard form of a PID controller is
1
GC  K C (1    D s)
I s
Where,

GC  controller transfer function


K C  controller gain
 I  integral time or reset time
 D  derivative time

17
STUDY OF PROCESS CONTROL - I

Conclusion
In all the modern plants and industries, nowadays, automatedcontrol system is a vital part for
unhindered, smooth and safer unit and process operations. Considering this fact, having a fine-
tuned knowledge in process control is must for a chemical engineer. This experiment was set to
introduce process control for a First Order Plus Time Delay (FOPTD) model using simulation in
Simulink. This experiment might not provide with the knowledge of advance and complex
control system used in modern plants but it paved the way to the basics of process control system
in simulation level.

18
STUDY OF PROCESS CONTROL - I

Marking Scheme

Report on
‘STUDY OF PROCESS CONTROL – I’

Name: Atique Ahmed Khan


Student ID: 1302004

Section and % marks allocated Marks

Summary (10%)

Experimental set up (20%)

Observed Data (10%)

Calculated Data (10%)

Sample Calculation (10%)

Results and discussion (20%)

Quality of Tables and figures (10%)

Overall presentation (10%)

Total (100%)

19

You might also like