Fuzzy Logic Controller for Superheated Steam

Temperature Control
Mohammad Shahid and Prof. Ibraheem and Prof. Zaheeruddin
Department of Electrical Engineering
Faculty of Engineering and Technology
Jamia Millia Islamia
New Delhi-110025 India
mshahid.shazeb@gmail.com
ibraheem_2k@yahoo.com
zaheer_2k@hotmail.com

Abstract— The precision control of superheated steam fed to The efficiency of boiler mainly depends on superheated
turbines for the generation of electrical power has been a steam temperature. With the development of advanced control
challenging task for control engineers for a long time. There are systems in the conventional controllers are getting replaced for
several limitations that are associated with conventional control superheated steam temperature control. In power boilers the
philosophies used for this purpose. The modern control temperatures of superheated steam depends on the amount of
techniques are being preferred due to their inherent merits over fuel input, angle of the flame burner ,load etc. spray water
the conventional control techniques in thermal power plants in which is a manipulating variable is used to control the
general and for superheated steam temperature control in temperature of the superheated steam. The overall cycle
particular. These days, fuzzy logic control concept is one of the
efficiency is very sensitive to this parameter and a tight control
intelligent control techniques which has been the revolutionary
applications in power system operation and control. In this
of this parameter would also ensure a better life of boiler
paper, an attempt has been made to design fuzzy logic based PID components. For example, a one percent fall in thermal
controllers for superheater temperature control of a boiler. The efficiency of a 500MW unit would amount to burning an extra
PID controllers based on are also designed. Considering a small coal of about 10,000 tons per annum. Therefore, it becomes
change in inlet temperature as the disturbance, the PID very important to increase the efficiency of thermal turbines by
controllers designed in the study are implemented and the implementing properly designed regulators for the turbine
dynamic response plots are obtained. Standard SIMULINK outputs. The temperature control of superheated steam should
Software is used on MATLAB platform to get the results. The be considered as an important parameter as a first step in this
investigations of simulation studies demonstrate the superiority regard.
of fuzzy logic based controllers over the conventional PID
controllers designed using ZEIGLER-NICHOLS tuning In the past few years there has been a tremendous increase
in the popularity of PID controllers. The test of the evolution of
technique.
the PID is that, actually most of the classical industrial
controllers have procedure to automate its parameters. Then, if
Keywords- Superheated steam, Intelligent techniques, PID
controller, Ziegler-Nichols tuning, Fuzzy logic controller
we can get a good model of the process, given by analytic
linear equation, direct technique of control are the simplest and
less cost alternatives. The classical PID controller provides an
I. INTRODUCTION (HEADING 1) accurate and efficient solution to linear control problems. But
The electrical energy consumption has been considered as a the involved process are in general complex, time variant, with
prime index to determine the trend of development of a delays and non-linearity’s and very often, with a poorly-
country. Therefore, for overall development of a country, defined dynamics. When the processes are too complex to be
availability of electric energy is one of the important described by analytic models, they are hardly controlled by
requirements. As far as our country is concerned, there is drastic approaches that simplifies them but do not get the
always a gap exists between requirement of energy and its required efficiency. To circumvent, some of these problems,
availability. Major part of electrical power generation in India modern control techniques have emerged for their applications
comes from fossil fuel fired power plants. These fossil reserves in power systems. However, optimal control technique based
are depleting rapidly. In order to conserve fossil reserves, the on full state vector feedback has its limitations for practical
power plant should run more efficiently. In the chain of thermal applications due to its requirement of availability of all system
cycle, since boiler is the front end requirement, its role plays a states for their measurement.
major importance in the overall reliability and efficiency of a The evolution of intelligent control concepts like, artificial
thermal power plant. In turn the efficiency of the boiler Neural Network, Fuzzy logic, Genetic Algorithm and Neuro-
depends on several operating parameters, and superheated Fuzzy control system has changed the whole scenario of
steam temperature is one amongst them, which is very control systems as far the design and implantation of
important. controllers for practical systems are concerned. Fuzzy logic
control concept (FLC) is one of the fittest tools to counter the III. DESIGN of PID CONTROLLER
vagueness associated in defining/describing the physical
systems. The parts of physical power system are having various The design of controller consists of the following two steps:
parameters which can not be modeled accurately. The (1) Selection of the mode of the controller, say proportional
mathematical modeling of superheater used for reheat the
or integral or derivative control or their combination (P,
steam available is done with some assumptions to simplify the
PI, or PID).
mathematical manipulations. Therefore, the analysis/
investigations carried out with these approximated models do (2) Specification of the value of the parameters associated
not serve the purpose of providing accurate and realistic with the selected control mode (Kp, Ki, and Kd).
results. With the application of fuzzy logic control concept in The above two steps are to be executed by properly
the design of controllers for superheated steam temperature, designing Kp, Ki, and Kd of in order to meet the specification
more realistic studies can be carried out. In this work, FLC is of the closed-loop system such as overshoot/ undershoot, decay
considered as one of the intelligent control concepts for ratio, settling time, rise time, etc, in time domain(or) gain
application to superheated steam temperature control. margin, phase margin, bandwidth, etc, in frequency domain.

II. DESIGN PARAMETERS OF SUPERHEATING SYSTEM A. Tuning of conventional controllers

For the present study, a superheated steam turbine is The conventional PID controllers may be tuned using well
considered for the investigation. The block diagram of the known Zeigler-Nichols Technique as;
system is shown in Fig.1. A brief description of various parts
and their design parameters are explained in the ensuing Zeigler-Nichols Technique
paragraphs.  Increase the proportional gain until the system oscillates;
that gain is the ultimate gain Ku.
 Read the time between peaks Tcr at this setting.
 Table- 1 gives approximate values for the controller gains.

TABLE I. ZEIGLER-NICHOLAS RULES

Controller Gain K Integration time Ti Derivation time Td

P 0.5Kcr

Figure 1. Control System with Single Controller. PI 0.5Kcr 0.85Tcr

PID 0.6Kcr 0.5Tcr 0.125Tcr

Desuperheater : The response of outlet temperature to the
changes in spray water is instantaneous. Hence, there is no The Simulink model of PID controller implementation on
dynamics involved the coefficient connecting the spray change superheater is shown in Fig.2.
to change in outlet temperature at 100% load has been derived.

Super heater : In this system the inlet temperature of the super

heater is increased by a factor of 1.3882 at 100% load for
a500MW boiler because of heat input from the furnace. This
model has been made available for simulation purpose for this
project which is given below:
(1.3882/19^5)/(s+0.0526) ^5
Transducer : It measures the outlet temperature of the
Figure 2. Super heater with PID controller.
superheated steam whose time constant is of the order of 10
sec. this time delay is added on to the delay of the
superheating system for simulation purpose. IV. DESIGN OF FUZZY LOGIC CONTROLLER
FLC is mainly used for system whose mathematical models
Control valve : The time constant of control valve is different are difficult to derive from, first principle. The heart of the FLC
for opening and closing. This value is of the order of 10 is to form knowledge base which can be obtained from the
seconds for closing and 5 seconds for openings per data experts in that field. In power boilers, the temperature of
provided by the manufacturers. The overall steady state superheated steam depends upon the amount of fuel input,
coefficient for control valve and desuperheater is found to be angle of the flame burner, load etc. spray water which is
0.556 relating the change at the superheated outlet temperature manipulating variable, is used to control the temperature with
to the change in the controller design. spray water. This time constant varies inversely with load.
 Block diagram of this control system with a single FLC
controller is given in Fig.3.

Figure 3. Simulink model of superheater with Fuzzy logic controller.

The FLC contains two input variables (error and error rate)
and one output variable. At every time instant FLC computes Figure 7. Output.
the change of output and adds to the previous output to get the
current value. The membership functions and their ranges for
all the input and output variables are as detailed below:

A. Mamdani Model

Figure 4. Mamdani Model.

B. Membership Function Figure 8. Control Surface.

No of membership functions: 3
Labels are small, medium and large. C. Expert Rules
The following 9 rules are applicable for this controller:
(1) If (ERROR is SMALL) and (ERROR RATE is SMALL)
then (CONTROL is SMALL).
(2) If (ERROR is SMALL) and (ERROR RATE is MEDIUM)
then (CONTROL is SMALL).
(3) If (ERROR is SMALL) and (ERROR RATE is LARGE)
then (CONTROL is MEDIUM).
(4) If (ERROR is MEDIUM) and (ERROR RATE is SMALL)
then (CONTROL is SMALL).
Figure 5. Error. (5) If (ERROR is MEDIUM) and (ERROR RATE is
MEDIUM) then (CONTROL is MEDIUM).
(6) If (ERROR is MEDIUM) and (ERROR RATE is LARGE)
then (CONTROL is LARGE).
(7) If (ERROR is LARGE) and (ERROR RATE is SMALL)
then (CONTROL is MEDIUM).
(8) If (ERROR is LARGE) and (ERROR RATE is MEDIUM)
then (CONTROL is LARGE).
(9) If (ERROR is LARGE) and (ERROR RATE is LARGE)
then (CONTROL is LARGE).

Figure 6. Error rate.

 V. SIMULATION RESULTS VII. REFERENCES
This paper presents design of PID controller using Ziegler- [1] H.Ying, Fuzzy Control & Modeling:Analytical foundations and
Nichols (ZN) technique for higher order system. A Fuzzy logic applications. New York, IEEE Press, 2000.
controller using simple approach & smaller rule set is [2] Fuzzy logic toolbox for use with MATLAB,user’s guide,rev.2(1995)
proposed. Simulation results are demonstrated. Performance [3] J. Zhang, N. Wang and S. Wang, “A developed method of tuning PID
controllers with fuzzy rules for integrating process,” Proceedings of the
analysis shows the effectiveness of the proposed Fuzzy logic American Control Conference,Boston, 2004, pp. 1109-1114.
controller as compared to the ZN tuned PID controller. The
[4] D.M WIberg, fuzzy logic and control.schaum’s outline series.Mc-Graw-
result shows the fuzzy logic controller has less overshoot and Hill.New York,NY2007.
settling time as compared to conventional PID controller. It [5] K.H. Ang, G. Chong and Y. Li, “PID control system analysis, design
shows the overall improvement in the performance of the and technology,” IEEE transaction on Control System Technology,
system. Fig.9.shows the natural response of the superheater, Vol.13, No.4, 2005, pp. 559-576.
and the controlled output is shown in Fig.9. and Fig.10. [6] P. Cominos and N. Munro. “PID controllers: recent tuning methods and
design to specification, ” IEE proceedings,Control Theory and
Applications, 2002, pp. 46-53.
[7] L. A. Zadeh and C.A. fuzzy logic and state functions: A static space
approach. Mc Graw Hill. New York,1990
[8] S.Chopra, R.Mitra and V.Kumar, “Fuzzy controller:Choosing an
appropriate & smallest rule set, ”International Journal of Computational
Cognition, Vol.3, No.4, 2005, pp. 73-78.
[9] S.R.Vaishnav, Z.J.Khan, Design and Performance of PID and Fuzzy
Logic Controller with Smaller Rule Set for Higher Order System,”
Figure 9. Natural response of superheater with 50c step change (in input). WCECS 2007, October 24-26, 2007, San Francisco, USA”
[10] K. Astrom and T. Hagglund, PID controllers: Theory, design and tuning.
2nd ed., The Instrumentation, Systems and Automation Society
(ISA),1995.
[11] Y.Yongquan, H.Ying and Z.Bi, “The dynamic fuzzy method to tune the
weight factors of Neural fuzzy PID controller,” IEEE, 2004, pp.2397-
2402.
[12] Z.Y. Zhao, M. Tomizuka and S. Isaka, “ Fuzzy gain scheduling of PID
controllers, ” IEEE transactions on Systems, Man and
Cybernetics,Vol.23, No.5, 1993, pp. 1392 -1398.
[13] Zaheeruddin, “A Simple Technique for Generation and Minimization of
Figure 10. Controlled output superheated using PID controller. Fuzzy Rules”, Proceedings of The IEEE International Conference on
Fuzzy Systems (FUZZ-2005), pp. 489-494, held at Reno, Nevada, USA,
May 22-25, 2005

Figure 11. Controlled output of superheater using Fuzzy logic controller.

VI. CONCLUSION
The paper presented an overview of PID controller, design
of PID controller using Z-N technique and design of fuzzy
logic controller for higher order system. Simulation results
using
MATLAB / SIMULINK are discussed for Ziegler Nichols
tuned PID controller, fine tuned PID controller and the Fuzzy
logic controller, and this conventional controller’s performance
is compared with the new FLC. The FLC gives the overall
better performance than that of PID controllers.
For the output of total system the FLC gives negligibly less
overshoot and also gives overall less settling time for the
disturbances. These are the major advantages observed from
the fuzzy controller design. For this controller design the effect
of feed forward disturbances on the system are vary less and
neglected. For further research work, one may take the feed
forward disturbances also into account.