Published by : International Journal of Engineering Research & Technology (IJERT)

http://www.ijert.org ISSN: 2278-0181

Vol. 6 Issue 05, May - 2017

Design and Comparison of a Multi-Purpose PID

Controller using Classical and Zeigler-Nichols
Technique
Seemab Gul1, Muhammad Naeem Arbab2,
1 2
Deptt. of Electrical Engineering, Deptt. of Electrical Engineering,
University of Engineering & Technology University of Engineering & Technology
Peshawar, Pakistan Peshawar, Pakistan

Uzma Nawaz3
3
Deptt. of Electrical Engineering,
University of Engineering & Technology
Peshawar, Pakistan

Abstract-PID controller tuning is designed using II. PID CONTROLLER

Matlab (Simulink). The design also gives us optimization of The letters P, I and D stands for P – Proportional,
PID controller without too much mathematic calculations. I – Integral and D- Derivative. Transfer function for PID
Ziegler-Nichols closed loop method is used for the design of controller is written as
tuning PID controller. Some disadvantages have been found 𝐾
in this technique, it is time consuming method because trial 𝐺(𝑠) = 𝐾𝑝 + 𝑖 + 𝐾𝑑 𝑠 (1)
𝑠
and error procedure is involved. The traditional PID 𝐾𝑑 𝑠 2 +𝐾𝑝 𝑠+𝐾𝑖
controller is replaced by Ziegler Nichols tuning PID 𝐺(𝑠) = (2)
𝑠
controller so that is applied to wide range of processes and to Where 𝐾𝑝 = Proportional gain, 𝐾𝑖 = Integral gain and 𝐾𝑑 =
obtain the minimum steady state error, also to improve Derivative gain. All of these 𝐾𝑝 , 𝐾𝑖 and 𝐾𝑑 are tuning
the other dynamic behavior. The problem has been solved by
parameters.
using Matlab (Simulink) which has the ability to characterize
both compensated and uncompensated relationship and we
can learn this relationship from the data being modeled.
These results can then be marked authorized by using Matlab
(Simulink) and manual calculations. 𝑲𝒑 , 𝝉𝒊 , and 𝝉𝒅 used in
Ziegler-Nichols formula can be calculated manually.

Keywords: PID controller, Ziegler-Nichols, Matlab

(Simulink), (𝑲𝒑 ) Proportional gain, (𝝉𝒊 ) integral time, (𝝉𝒅 )
derivative time.
Figure 1: Block diagram of PID controller in cascade with the plant

I. INTRODUCTION We assume that controller in the Figure 1 is a closed-loop

PID controllers are commonly used in the process unity feedback system. The variable 𝑒(𝑡) represents the
industries for the reason of simplicity and outstanding error which is sent to PID controller. The signal 𝑢(𝑡) is
performance. More than 95% of closed loop process use equal to the proportional gain𝐾𝑝 time’s error signal plus the
PID controllers. Control systems are designed to achieve
integral gain 𝐾𝑖 time’s integral of the error signal plus the
specific objectives. For control system design some
derivative gain 𝐾𝑑 times derivative of the error signal [3].
characteristics are required. A good quality control system
has a lesser amount of error, excellent response, high 𝑡 𝑑𝑒(𝑡)
accuracy, damping that has no unnecessary overshoot and 𝑢(𝑡) = 𝐾𝑝 𝑒(𝑡) + 𝐾𝑖 ∫0 𝑒(𝑡) 𝑑𝑡 + 𝐾𝑑 (3)
𝑑𝑡
fine stability [1]. Several tuning methods have been
proposed up to now for getting more accurate and stable III. ZIEGLER-NICHOLS CLOSED LOOP METHOD
control response. Based on our requirement we want to Thismethod is a trial and error tuning based method on
characterize process dynamics by few features. Some continuous oscillations was first proposed by John G.
tuning methods considered only one feature as a condition Ziegler and Nathaniel B. Nichols in 1942. This method is
for their tuning algorithm. Some of these tuning methods widely used for tuning PID controllers and is also
considered more than one feature as a condition for their recognized as continuous cycling method or ultimate gain
tuning algorithms [2]. In this study we will give an idea tuning method. Ziegler and Nichols use ¼ decay ratio as a
about new tuning rules in spirit of Ziegler and Nichols. design criterion for this method.

IJERTV6IS050361 www.ijert.org 684

(This work is licensed under a Creative Commons Attribution 4.0 International License.)
Published by : International Journal of Engineering Research & Technology (IJERT)
http://www.ijert.org ISSN: 2278-0181
Vol. 6 Issue 05, May - 2017

In the Ziegler-Nichols closed loop method First of (i). Lag-Lead compensator

all the system is stabilized into steady state now put on the 𝐺1 (𝑠)
PID controller into P controller by setting 𝜏𝑖 = ∞ and 𝜏𝑑 =
616𝑠 2 + 1908𝑠 + 184.4
0 as shown in Figure 2. Increase 𝐾𝑝 until the system =
oscillates continuously. Use Table 1 to get the approximate 𝑠 5 + 27.21𝑠 4 + 207.672𝑠 3 + 432.07𝑠 2 + 764.29𝑠 + 7.6
values for the controller gains 𝐾𝑝 , 𝐾𝑖 and 𝐾𝑑 [4].

Figure 2: Closed loop system with Proportional gain (𝐾𝑝 )

Table 1: Controller parameter for closed loop Ziegler-Nichols method

Controller 𝐾𝑝 𝜏𝑖 𝜏𝑑
P 0.5𝐾𝑝𝑢 ∞ 0
PI 0.45𝐾𝑝𝑢 𝑃𝑢 0
1.2
PID 0.6𝐾𝑝𝑢 𝑃𝑢 𝑃𝑢
Figure 5: Simulink model of Lag- lead compensator
2 8
The gain that gives us these continuous oscillations is
ultimate gain 𝐾𝑝𝑢 , 𝑃𝑢 is the period of oscillation at 𝐾𝑝𝑢 as
shown in the Figure 3 [5].

Figure 6: Lag- lead compensator response

Figure 3: Sustained oscillation with period (𝑃𝑢 )
(ii). Ziegler-Nichols closed loop method
IV. GENERALIZED MODEL FOR PID CONTROLLER
By setting 𝜏𝑖 = ∞ and 𝜏𝑑 = 0 we obtain the
Observe that PID controller tuned by Ziegler-
closed loop transfer function as follows:
Nichols closed loop method gives:
Now from Eq. (1) 𝐶(𝑠) 𝐾𝑝
𝐾 =
𝐺(𝑠) = 𝐾𝑝 + 𝑖 + 𝐾𝑑 𝑠 (4) 𝑅(𝑠) (𝑠 + 10)(𝑠 2 + 2𝑠 + 5) + 𝐾𝑝
𝑠
𝐾𝑝 𝐾𝑑
Where 𝜏𝑖 = and 𝜏𝑑 =
𝐾𝑖 𝐾𝑝
Characteristic equation for the closed loop system is;
1
𝐺(𝑠) = 𝐾𝑝 (1 + + 𝜏𝑑 𝑠) (5)
𝜏𝑖 𝑠
𝑠 3 + 12𝑠 2 + 25𝑠 + (50 + 𝐾𝑝 ) = 0
Where 𝐾𝑝 is proportional gain, 𝜏𝑖 is integral time and 𝜏𝑑 is
referred as derivative time [6]. Using Routh’s stability criterion we find that continuous
For PID-controller oscillation will occur at the ultimate gain 𝐾𝑝𝑢 .
𝐾𝑝 .𝜏𝑖 .𝜏𝑑 𝑠 2 +𝐾𝑝 .𝜏𝑖 𝑠+𝐾𝑝
𝐺(𝑠) = (6)
𝜏𝑖 𝑠
V. SIMULATIONS AND RESULTS 𝐾𝑝𝑢 = 250
2𝜋 2𝜋
Consider the control system as shown in Figure 4 𝑃𝑢 = = = 1.25
𝜔 5
in which PID controller is used to control the system the
PID controller has the transfer function.
1
𝐺(𝑠) =
(𝑠 + 10)(𝑠 2 + 2𝑠 + 5)

Figure 4: PID controlled system Figure 7: Simulink diagram for Sustained Oscillations

IJERTV6IS050361 www.ijert.org 685

(This work is licensed under a Creative Commons Attribution 4.0 International License.)
Published by : International Journal of Engineering Research & Technology (IJERT)
http://www.ijert.org ISSN: 2278-0181
Vol. 6 Issue 05, May - 2017

Figure 8: Sustained oscillation for 𝐾𝑝𝑢 = 250

From Table 1 we determine 𝐾𝑝 , 𝜏𝑖 and 𝜏𝑑 as follows:

Figure 12: Lag-lead compensator and Ziegler-Nichols closed loop method
responses
𝐾𝑝 = 0.6𝐾𝑝𝑢 = 0.6* 250 = 150
𝜏𝑖 = 0.5𝑃𝑢 = 0.5*1.25= 0.625 VI. CONCLUSIONS
𝜏𝑑 = 0.125 𝑃𝑢 = 0.125*1.25= 0.15625 In this research, comparison between classical lag-
lead compensator and Ziegler-Nichols technique is applied
Using eq. (5) we get to an under damped system. The analysis is done based on
𝐾𝑝 = 150 mathematical calculations and then system response is
obtained from the SIMULINK. The result of simulation
𝐾𝑖 = 240 shows that both methods improve system performance to a
desired value. System response is mainly depends on both
𝐾𝑑 = 23.43 transient response and steady state response. In classical
lag-lead compensator, both transients and error is improved
to a significant value. While in Ziegler-Nichols technique,
steady state error is effectively improved while there is no
change in transient response. Ziegler-Nichols tuning is
applicable to specific applications while it is not considered
as optimal. This tuning gives maximum reduction in
disturbance parameter in PID loop. But the gain and
overshoot is high in this technique which is acceptable in
some applications. So it is concluded that both methods are
Figure 9: Simulink model of Ziegler-Nichols closed loop method applicable to tune system response as per desired values.

REFERENCES
[1] U.A. Bakhi, V.U. Bakshi, “Compensation of control
systems,” Control system Engineering, 1st ed. Mumbai,
India: Technical publication Pune, 2010, ch. 14, pp.14-1 to
14-2.
[2] M. Shahrokhi and A. Zomorrodi, "Comparison of PID
controller tuning methods," Department of chemical &
Petroleum Engineering sharif University of Technology, pp.
1-12, Apr 2013.
[3] H. Ahmad and A. Rajoriya, “Performance Assessment of
Figure 10: Ziegler-Nichols closed loop method response
Tuning Methods for PID Controller Parameter used for
Position Control of DC Motor,” International Journal of u-
(iii). Lag lead compensator and Ziegler Nichols closed and e-Service, Science and Technology, pp. 139-150, 2014.
loop method [4] C.A. Smith, “Feedback controllers,” Automated contineous
process control, New York, Wiley publishers, 2002, ch. 3,
pp. 38-54.
[5] http://www.cpdee.ufmg.br/~palhares/PID_Ogata.pdf
[6] R. Bansal, A. Patra, and V. Bhuria, "Design of PID
Controller for Plant Control and Comparison with ZN PID
Controller." International Journal of Emerging Technology
and Advanced Engineering, pp. 312-314, 2012.

Figure 11: Simulink model of Lag-lead compensator and Ziegler-Nichols

closed loop method

IJERTV6IS050361 www.ijert.org 686

(This work is licensed under a Creative Commons Attribution 4.0 International License.)