An Investigation of Scaled-FLC Using PSO for Multi-area Power System Load Frequency Control

doi:10.4236/epe.2013.54B088

Paper Menu >>

Journal Menu >>

Energy and Power Engineering, 2013, 5, 458-462

doi:10.4236/epe.2013.54B088 Published Online July 2013 (http://www.scirp.org/journal/epe)

An Investigation of Scaled-FLC Using PSO for Multi-area

Power System Load Frequency Control

Aqeel S. Jaber1, A.Z. Ahmad1,2, Ahmed N Abdalla1

1Faculty of Electrical and Electronics Engineering, University Malaysia Pahang, Pekan, Malaysia

2Sustainable En.& Power Elec. Res. (SuPER) Group, University Malaysia Pahang, Pekan, Malaysia

Email: [email protected], zahari[email protected], [email protected]

Received April, 2013

ABSTRACT

Load Frequency Control (LFC) is one of power systems important requirements which maintain the zero steady-state

errors in the frequency changing and restoring the natural frequency to its normal position. Many problems are subject

to LFC such as suddenly large load or suddenly disconnecting generating unit by the protection device. In this paper,

multi-area Frequency Control by using the combination of PSO and fuzzy logic control (FLC) technique. PSO optimi-

zation method is used to tuning the fuzzy controller input and output gains. Four of an interconnected electrical power

system used as a testing the effectiveness of the proposed method compared to a conventional PI controller and

scaled-fuzzy controller. The simulation result has been shown that the controller can generate the best dynamic response

in multi-load conditions.

Keywords: Fuzzy Control; PSO; Load Frequency Control

1. Introduction

One of the major requirements in parallel operation of

interconnected power systems is the Load Frequency

Control (LFC) which is responsible for scheduled power

transfers between the interconnected areas at any distur-

bance in the case of the connecting or disconnecting ge-

nerating unit or suddenly large load. Various controllers

have been used in different areas could not efficiently

control the frequency and rather slow for the output re-

sponse due to fact of non-linearity in system components

[1-2], time invariant and governed by strong

cross-couplings of the input variables. Therefore, the

controllers have to be designed with taking into account

the nonlinearities and disturbances.

Many of control methodologies have been suggested

to solve LFC problem. Static Output Feedback gains and

Linear Matrix Inequality are the most efficient and effec-

tive tool which stabilizes the system which used to cal-

culate the gains of PID controller [3]. The Robust adap-

tive control also has been used to deal with the change in

a system parametric [4]. Optimization techniques have

been done to solve LFC, but they require information

about the entire system rather than local information [5].

Other control approaches such as PID-ANN, PI-fuzzy

and optimal control applied to LFC has been reported in

[6]. Using genetic algorithm to scale of PI fuzzy control-

ler in LFC has been reported in [7].

This paper presents the FLC using PI-fuzzy controllers.

The proposed controller is tuned using PSO to obtain the

controller gains in order to get an efficient fuzzy control

on four of an interconnected electrical power system.

This is a new approach to optimize the fuzzy controller

that differentiates to other's methods. The simulation

results are carried out in term frequency response for its

damping under different load conditions and compared it

to the effectiveness of proposed controllers with other

controllers. Simulation results show that the undershot

and settling times with the proposed controller are better

and guarantees robust performance under a wide range of

operating conditions.

2. Theoretical Background

Power systems have multi-variable and complex struc-

tures and consist of different control blocks and deal with

nonlinear and/or non-minimum phase systems [8]. Power

systems are divided into control areas connected by tie

lines and all generators are supposed to constitute a co-

herent group in each control area.

2.1. Load Frequency Control

The aim of LFC is to maintain real power balance in the

system through control of system frequency. Small

changes in real power are mainly dependent on changes

in rotor angle δ and, thus, the frequency f. whenever the

A. S. JABER ET AL. 459

real power demands changes, a frequency change occurs.

However, the change in angle δ is caused by momentary

change in generator speed. This frequency error is ampli-

fied, mixed and changed to a command signal which is

sent to turbine governor. The governor operates to restore

the balance between the input and output by changing the

turbine output. This method is also referred as Megawatt

frequency or Power-frequency (P-f) control [9].

2.2. Fuzzy Logic

According to many researchers, there are some reasons

which present popularity of fuzzy logic control such as

easily applied for most applications in industry. Besides,

it can deal with intrinsic uncertainties by changing the

controller parameters. On the other hand, their robustness

and reliability make fuzzy controllers useful in solving a

wide range of control problems [10]. The fuzzy control-

ler for the single input, single output type of systems is

shown in Figure 1 [7]. Fuzzy logic shows experience

and preference through its membership functions. These

functions have different shapes depending on system

experts’ experience [11].

2.3. PSO Algorithm

PSO was introduced by Eberhart and Kennedy as a new

heuristic method [12,13]. PSO was inspired by the food-

searching behaviours of fish and their activities or a flock

of birds. In D-dimensional search space. The best indi-

vidual position of particle i and the best position of the

entire swarm are represented by

(1)

(2)

where:-

Pi=(pi1, pi2,…, piD) and G=(g1, g2,…, gD), respec-

tively, ω is inertia weight parameter and c1, c2 is accel-

eration coefficients. In each iteration the particles will

using eq. 1&2 to update their position (xi) and velocity

(vi) [12].

3. Four Area LFC Model

The net power (ΔP) due to disturbance (ΔPD) is when the

changes in power generation. Where the ΔPG is described

as(3).

(3)

This change will absorbed by changing in kinetic en-

ergy (Wkin,) load consumption and export of power (ΔPtie)

so ΔP for ith area is as follows;

Figure 1. Fuzzy controller block diagram.

(4)

where, Di is power regulation and equal to ΔP/Δf. By

taking Laplace transformation

(5)

where, , (H) is inertia constant and (f) is the

frequency. If the line losses are neglected, the individual

ΔPtie ij can be written as

(6)

(7)

where, and δ is load an-

gle. Upon Laplace transforming (5), one gets

(8)

The transfer of generator turbine (Gtf) is written by

(9)

where, TT are turbine time constant and TG speed gover-

nor time constant. The parameters can be represented

such in the Figure 2.

From Figure 2, the bias factor (Bi ) is suitable value

can be computed as follows

(10)

ACEi, Ri are area control error, speed droop charac-

teristic of area (i) respectively.

Figures 3 & 4, show ith area block diagram of and

illustrate the tie line block diagram of interconnected

power system [10].

Figure 4 shows the method of interconnection be-

tween four areas that have been used in this paper.

Figure 2. LFC Model for one area of system.

A. S. JABER ET AL.

460

Figure 3. Model of tie-line power control are

Figure 4. Four-area interconnection power system.

4. Proposed Method

bership functions that are ad-

dium positive, SP: small positive, SN: small

The boundary of the mem

justed based on expert in classical Fuzzy methods, per-

son’s experiences may be does not guarantee the sys-

tems’ performance. The boundaries of the membership

functions are tuned by PSO to select the best boundaries

by finding suitable gains (scaled fuzzy parameters) for

the inputs and output fuzzy controller. These gains obtain

by three parameters Gin1, Gin2 and Gout that shown in

Figure 5 are defined the uncertain range by PSO algo-

rithms. The fuzzy rule has been designed as in Table 1

hat based on the number of membership function from

the inputs and the output (as in Figure 7).The flow chart

of PSO algorithm to optimize the scaled fuzzy parame-

ters is shown in Figure 6.

where:

MP: me

gative, Z: zero and MN: medium negative.

Table 1. Fuzzy controller rules.

MP SP Z SN MN

MP MP SP SP Z Z

Sp SP SP Z Z NS

Z SP Z Z NS NS

SN Z Z NS NS MN

MN Z NS NS MN MN

Figure 6. Optimizing fuzzy parameter using PSO.

Figure 7. Member ship function for input & output of fu

5. Result and Discussion

ess was tested in order to

zzy

controller.

The proposed method effectiven

investigate the system performance by using the MAT-

LAB 7.1. Tables 2 &3, list all system parameter and tie

line parameter.

Figure 5. Scaled Fuzzy PI controller diagram.

A. S. JABER ET AL. 461

The scaled fuzzy type controller was designed and

re 10 and

mum KP and KI for PI control-

Table 2. Four area model parameters.

Area R TG T

T T

P K

P B

mpared with the classical fuzzy and PSO-PID for LFC

under system uncertainties (controller robustness) in

multi load conditions. The frequency response results are

shown in Figure 8 and Figure 9 respectively.

The tie power response is shown if Figu

gure 11 respectively.

Table 4, show the opti

rs parameters using PSO, and the optimum values of

the scaled fuzzy parameters that are computed by using

PSO algorithms is shown in Table 5.

1 2.4 0.08 0.030 20.08 1200.401

2 2.1 0.091 0.025 17.24 1110.3

3 2.9 0.072 0.044 22.97 1350.48

1 4 .9950.044 0.044 53.19 1060.391

Table 3. Tie Line Parameters.

T12 T13 34 T14 T23 T24

0.425 0.5 0.4 0.455 0.523 0.6

Figure 8. Frequency deviations of 30% load change.

Figure 10. Tie power transfer of 3% load change.

Figure 11. Tie power transfer of 5% load change.

Tabl4. PI controller values.

Area Kp Ki

1 0.51 0.631

2 0.432 0.551

3 0.552 0.681

4 0.601 0.61

Table 5. Scaled Fuzzy Parameters.

Area Gin1 Gin2 Gout

1 0.138 0.0. 074 0114

2 0.129 0.057 0.0960

3 0.139 0.0234 0.1181

4 0.039 0.128 0.1172

Table 6 shows for the frequency deviation of peak

ot, St (s): Settling times (s),

der shoot & and settling time for scaled fuzzy-PI con-

troller and conventional PI controller for each intercon-

nected power system area.

where: PUS: Peak undersho

Figure 9. Frequency deviations of 50% load change. L.Ch: %load change

A. S. JABER ET AL.

462

Response Comparison For Scaled Table 6. Frequency

Fuzzy-PI Controller And Conventional PI Controller.

PI controller PSO fuzzy controller

L.Ch

Ps)

US St (PUS S.t (s)

1 0. 0 0818 21.2 .004514.23

2 0.0233 23.1 0.0081 14.97

3 0.0361 23.7 0.0132 15.12

4 0.0472 24.5 0.0178 15.43

5 0.0605 24.81 0.0227 15.91

able 7. Power transfer response Comparison of scaled

oller

fuzzy-PI controller and conventional PI controller.

PI controller PSO fuzzy contr

L.Ch

PUS*1 US

0^-3 St (s) P St (s)

1 3 5.342 22.4 .32117.14

2 10.938 23.2 6.712 18.52

3 16.311 23.7 9.131 18.79

4 22.331 23.9 12.342 19.31

5 27.211 24.8 16.201 20.54

Table 7 shows for the total power transfer deviation of

d method performance is

ak under shoot & and settling time for scaled fuzzy-PI

controller and conventional PI controller for each inter-

connected power system area.

The robustness of the propose

monstrated based on ITAE that is under step change in

the different demands as

Finally, from tables (6,7) and figures (8 to 11) of

6. Conclusions

ces PSO-FLC to improve the

step

ange, the scaled Fuzzy controller has better perform-

ance than the optimized PI controller at all operating

conditions. Therefore, the performance comparison be-

tween both controllers indicates that the frequency re-

sponse of the proposed method has smaller undershoot

and shorter in settling time with respect to PI controller.

This paper introduper-

formance of four-area power system and the linearization

in errors is considered as parametric uncertainties. Each

area consists of the turbine, governor and power system

which modelled by first-order transfer functions. In addi-

tion, PSO was used to adjust the input and the output of

FLC memberships. Simulation results proved that the

proposed scaled FLC has obtained fast response and less

undershoots compared to conventional PI controller.

REFERENCES

[1] Y. Wang, R. Zhou and C. Wen, “Robust Load-frequency

Controller Design for Power Systems,” Generation,

Transmission and, Vol. 140, No. 1, 1993.

[2] E. Çam and İ. Kocaarslan, “Load Frequency Control in

Two Area Power Systems using Fuzzy Logic Controller,”

Energy Conversion and Management, Vol. 46, No. 2,

2005, pp. 233-243. doi:10.1016/j.enconman.2004.02.022

[3] A. A. Ghany, “Design of Static Output Feedback PID

controller via ILMI method for a Power System Stabi-

lizer,” Power System Conference, MEPCON, 2008.

[4] A. Stankovic, “On Robust Control Analysis and Design

for Load Frequency Regulation,” Power Systems, IEEE,

Vol. 13, No. 2, 1998, pp. 449-455. doi:10.1109/59.667367

[5] K. Parmar, S. Majhi and D. Kothari, “Multi-area Load

Frequency Control in a Power System using Optimal Out-

put Feedback Method,”(PEDES) & 2010 Power, 2010.

[6] C. Chang and W. Fu, “Area Load Frequency Control

using Fuzzy Gain Scheduling of PI Controllers,” Electric

Power Systems Research, 1997.

doi:10.1016/S0378-7796(96)01199-6

[7] S. Broujeni, S. Hemmati and H. Fayazi, “Load Frequency

Control in Multi Area Electric Power System using Ge-

netic Scaled Fuzzy Logic,” Int. J. Phys. Sci, 2011.

[8] J. Hertz, A. Krogh and R. Palmer, “Introduction to the

Theory of Neural Computation,” 1991.

[9] H. Saadat, Power System Analysis, Mcgraw-Hill, 2005.

[10] T. W.Yang, R. Zhang and C. Zhang, “Research of Real-

izing PID Controller Based On FPGA,” Microcomputer

Information, 2005, pp. 2-11.

[11] K Tomsovic “Fuzzy,” in International Conference On

Intelligent System Application, 1999.

[12] J. Kennedy and R. Eberhart, “Particle Swarm Optimiza-

tion,” Neural Networks, 1995. Proceedings, pp. 2-6.

[13] A. Jaber, A. Ahmad and A. Abdalla, “Advance Two-Area

Load Frequency Control Using Particle Swarm Optimiza-

tion Scaled Fuzzy Logic,” Advanced Materials Research,

Vol. 622-623, 2013, pp. 80-85.