supression of low freqquency oscillations in the power system by neuro fuzzy logic controller

8/10/2019 supression of low freqquency oscillations in the power system by neuro fuzzy logic controller

http://slidepdf.com/reader/full/supression-of-low-freqquency-oscillations-in-the-power-system-by-neuro-fuzzy 1/6

[Sai babu, 5(1): January- March, 2015] ISSN: 2277-5528Impact Factor: 2.745 (SIJF)

Int. J. of Engg. Sci. & Mgmt. (IJESM), Vol. 5, Issue 1:Jan.-Mar.: 2015, 1-6

INTERNATIONAL JOURNAL OF ENGINEERING SCIENCES &MANAGEMENT

SUPRESSION OF LOW FREQUENCY OSCILLATIONS(LFO) IN THEPOWER SYSTEM BY NUERO FUZZY LOGIC CONTROLLER

K.V.V.Sai Babu*, U.Sravya

*II M.Tech, Department of EEE, DVR & Dr.HS MIC College of Technology, Kanchikacherla,Krishna DT

Assistant Professor, Department of EEE, DVR & Dr.HS MIC College of Technology,Kanchikacherla, Krishna DT

ABSTRACTLow-frequency oscillations (LPF) in electric power systems due to the uncontrollability of the disturbance in the

power system. These oscillations need to be controlled to maintain system stability. Several control devices, such as power system stabilizers, lead lag controllers are used to enhance power system stability. The UPFC can be used toeffectively control these low-frequency power system oscillations. It is done by designing a supplementary signal

based on either the real power ow along the transmission line to the series converter side or to the shunt converter

side through the modulation of the voltage magnitude reference signal. In this work the linearized model ofsynchronous machine connected to infinite bus (Single Machine-Infinite Bus: SMIB) with UPFC, adaptive neuro-fuzzy controller for UPFC is designed and simulated to suppress the low frequency oscillations. Simulation is

performed on the characteristics of the fuzzy controlled bus system. It is evident that the fuzzy logic controller haseffective operation than the conventional lead-lag controller through the obtained results.

Keywords : Neuro-Fuzzy Controller, Low Frequency Oscillations (LFO), Unified Power Flow Controller (UPFC),Single Machine-Infinite Bus (SMIB). .

INTRODUCTIONThere is no single acceptable denition of the termelectric power quality. The term generally applies to

the goodness of the electric power supply, voltageregulation, frequency, voltage wave shape, currentwave shape, level of impulses and noise, and theabsence of momentary outages [4]. The regulation ofthe network frequency constitutes one of the essentialelements for the “quality” of operation; excessiv efrequency variations would not, in fact, be tolerated

by many end-users, or by auxiliary equipment of thegenerating power stations themselves [9].The growthof the demand for electrical energy leads to loadingthe transmission system near their limits. Thus, theun-controlling of the system is leads to develop thefrequency oscillation. The damping torque is used to

provide the balancing in the system.FACTS controllers allow steady-state, quasi-steady-state, dynamic, transient control actions and they

provide important equipment and system protectionfunctions. The primarily use of UPFC is to controlthe power flow in power systems. The UPFC consistsof two Voltage source converters (VSC) each of them

has two control parameters namely me, δe, mb andδb. For systems which are without power systemstabilizer (PSS), excellent damping can be achieved

via proper controller design for UPFC parameters.It is usual that Heffron-Philips model [13] is used in

power system to study small signal stability. Thismodel has been used for many years providingreliable results. In recent years, the study of UPFCcontrol methods has attracted attentions so thatdifferent control approaches are presented for UPFCcontrol such as Fuzzy control, Conventional lead-lagcontrol, Genetic algorithm approach, and robustcontrol methods.In this work, the class of adaptive networks that ofthe same as fuzzy inference system in terms of

performance is used. The controller utilized with theabove structure is called Adaptive Neuro FuzzyInference System or briefly ANFIS. Applying neuralnetworks has many advantages such as the ability ofadapting to changes, fault tolerance capability,recovery capability. To show performance of thedesigned adaptive neuro-fuzzy controller, aconventional lead-lag controller that is designed in is





used and the simulation results for the power systemincluding these two controllers are compared witheach other.This paper is organized as follows: in Section II, themodel of the power system including UPFC is

presented. The proposed Adaptive Neuro-Fuzzycontroller and lead-lag compensator are explained inSection III. The results of the simulation are shown inSection IV. Finally conclusions are presented insection V

POWER SYSTEM MODELINCLUDING UPFCUPFC is one of the famous FACTS devices that areused to improve power system stability. Fig.1 showsa single-machine-infinite-bus (SMIB) system withUPFC. It is assumed that the UPFC performance is

based on pulse width modulation (PWM) converters.In figure 1 me, mb and δe, δb are the amplitudemodulation ratio and phase angle of the referencevoltage of each voltage source converter respectively[12].A linearized model of the system is used in dynamicstudies of power system [2]. In order to consider theeffect of UPFC in damping of LFO, the dynamicmodel of the UPFC is employed. In this model theresistance and transient of the transformers of theUPFC can be ignored [7]. The representation of the

power system equipped with the UPFC can berepresented as shown in equation (1) and torquegeneration of the Heffron Philips model is given byequation (2). The control variables of the system arethe speed variation, angle variation in the rotorcalculated from the characteristics of the singlemachine system. These variables are given as thecontrolling signals to the fuzzy controller [11].Generally, the study of the synchronous machine isdriven from a constant voltage-frequency ratio, themotor develops the same maximum torque at allspeeds. For the salient pole rotor this will occurs atthe load angle which is influenced by the relativevalues of L q and L d.. The steady state representationof the torque developed in the salient pole rotor isgiven as equation (2) .

F ig.1 A single machine connected to inf in ite buswith UPFC

∆∆∆∆∆

=0 0 0 0− − − 0 −−4 0 − −−

45 0−

6− −

7 0 8 0 9

∆∆∆∆∆

+

000

∆ 00 0 0 0 0

∆∆∆∆

Where ΔmE, ΔmB, ΔδE and Δδb are thedeviation of input control signals of the UPFC.

=⌈ + +( − ) −

+ ⌉ (2)

CONTROLLER DESIGNThe switching of the convertors must be reliable,accurate and effective in the operation in order tocontrol the frequency ranges in the loads, controlling

pulses require controlling the variable parameters.Power full control strategies to the convertorsrequired to obtain stable controlled switching with

low sensitivity and burden against the controlling parameters. A) Lead-Lag Controller Design:Lead lag compensation is used for the good stabilitymargin, having the poor steady state accuracy. Thecharacteristic of lead lag compensator is given by theequation (3).

= ++ (3)





The compensator selection and design basedupon the parameters as given below:

1. Compensator Selection2. Unity Gain Frequency3. Desired Phase Margin4. Compensator Zero-Pole Placement

5.

Compensator Gain Calculation6. Stability Margin Variation7. Results.

B). Adaptive Neuro-Fuzzy Controller DesignFuzzy logic controller is the new hybrid technologytotally depends upon the human thinking. Thecontrollers easily designed based on the controllingvariable are taken as the inputs by taking the samplevalues. Generally some samples of the values of thecontrolling parameter variation as the input to thecontroller and specific function is taken as theexpected function of the controller. The main steps inthe implementation of a fuzzy logic controller asfollows:

1. Fuzzification .The variation in the speed and angle of the rotor,

∆δ and ∆w as taken as the inputs to the fuzzycontroller. In Figure 2, a Sugeno type of fuzzy systemhas the rule base with rules such as follows [5]:

1. If Δδ is A1 and Δω is B1 then f1=p1Δδ+q1 Δω+r1.

2. If Δδ is A2 and Δω is B2 then f2=p2Δδ+q2 Δω+r2.

2. I nterference Engine:The result of the fuzzy control algorithm can be

obtained by using the control rules, membershipfunctions, and an interference engine.

a) Type: 'Sugeno' b) And Method: 'Prod'c) Or Method: 'Probor'd) De-Fuzzification Method: 'Wtaver'e) Implication Method: 'Prod'f) Aggregation Method: 'Sum'g) Input: [1x2

Struct]h) Output: [1x1

Struct]i) Rule: [1x20

Struct]

μA and μ B are the membership functions offuzzy sets Ai and Bi for i=1… 20. In evaluating therules, we choose product for T-norm (logical and).Then controller could be designed in following steps

F ig.2 ANF I S architecture for 2 in put Sugeno fu zzymodel wi th 20 r ules

1) Evaluating the rule premises :

= ∆ ∆, =1,2,3,…20. (3)

2) Evaluating the implication and the ruleconsequences :

∆ ,∆=∆ ,∆∆ ,∆+⋯….+∆ ,∆∆ ,∆∆ ,∆+⋯.+∆ ,∆ (4)

Or leaving the arguments out

=+ +⋯.+ +⋯ (5)

This can be separated to phases by first defining−= + +⋯, =1,2,3,…20 (6)

These are called normalized firing strengths.Then f can be written as

=− − ⋯.− (7)The above relation is linear with respect to P i, Qi

r i and i=1, 20. So parameters can be categorized into2 set of linear parameters and set of nonlinear

parameters. Now Hybrid learning algorithm can beapplied to obtain values of parameters. Hybridlearning algorithm is combination of linear andnonlinear parameters learning algorithm. Descriptionfor learning procedure can be found in [20]. Thisnetwork is called adaptive by Jang and it isfunctionally equivalent to Sugeno type of a fuzzysystem. It is not unique presentation. With regard tothe explanations presented and with the help ofMATLAB software, adaptive neuro-fuzzy controllercan be designed. The rules surface for designedcontroller shown in figure 3. The membershipfunctions for input variable Δω are p resented infigure 4and 5.One of the advantages of using neuro-fuzzycontroller is that we can utilize one of the designedcontrollers for instance ΔmE controller in place of theother controllers. While if we use conventional lead-lag controller, for each controls parameters, acontroller must be designed.





F ig.3 Th e rules sur face

F ig.4 The membership f uncti ons for in put variable Δω

F ig.5 The membership fu nctions for in put vari able Δδ

SIMULATION & RESULTS

F ig 6: M odeil li ng and Simu lati on of the power system in cluding UPF C





In this work, LFO is linearily controlled bycontrolling the angular velocity (speed) , rotor anglevariation . These values are given as a input to thelead-lag and fuzzy controllers. Its clearly shows formthe Fig 7 to that the lead-lag controller response isnot as good as neuro-fuzzy controller response and

also neuro-fuzzy controller decreases settling time. Inaddition maximum overshoot has decreased incomparison with lead-lag controller response.

A)Wave Form

F ig 7. Angular velocity vari ation in th e rotor bylead lag contr oller

F ig 8. Angle vari ation in th e rotor by lead lagcontroller

F ig 9. Angular velocity vari ation in th e rotor byfu zzy logic contr oller

F ig 10. Angle variation i n th e rotor by fuzzy logiccontroller

B)Tabular Form[6]

Time(sec)

Lead LagController

Fuzzy LogicController

AngleDeviation

SpeedDeviation

AngleDeviation

SpeedDeviation

0 1.52-0.000141

-1.579 0

0.1 -2.2 -0.01739 -1.887 -0.01025

0.2 -2.44 0.003215 -2.057 0.00213

8

0.3 -2.135 0.0103 -1.849 0.006653

0.4 -1.931-0.000740

-1.727-0.000142

0.5 -2.087 -0.00539 -1.1819 -

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1-0.025

-0.02

-0.015

-0.01

-0.005

0

0.005

0.01

0.015

time(sec)

roto

rspeed deviation

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1-2.5

-2.4

-2.3

-2.2

-2.1

-2

-1.9

-1.8

-1.7

-1.6

-1.5

time(sec)

rotorangle deviation(deg)

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1-12

-10

-8

-6

-4

-2

0

2

4

6

8x 10

-3

time(sec)

rotorspeed dev

iation(deg)

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1-2.1

-2

-1.9

-1.8

-1.7

-1.6

-1.5

-1.4

time(sec)

rotorangle deviation(deg)





0.003037

0.6 -2.198 4.89e-006 -1.881 0.00033

0.7 -2.137 0.002949 -1.832 0.00192

1

0.8 -2.061 0.0004826 -1.789 4.8e-005

0.9 -2.086-0.001322

-1.806 0.0009026

1 -2.125-0.000111

-1.825 0

CONCLUSIONBy the comparison of the results, fuzzy logiccontroller has the effective operation rather than theconvenctional lead-lag controllers. By using the

adaptive techniques, by the damping torque, lowfrequency oscillation are reduced and nullified to thesteady state operation of the system. Fuzzy logiccontrollers are desinged to nearest of the humanityand doesn’t require any system paprameters oroperating conditions. Fuzzy logic controllers canb beapply to various tpes of convertors and controllers tocontrol the capability, quality of the system. Also wecan utilize advantages of neural networks such as theability of adapting to changes, fault tolerancecapability, recovery capability, High-speed

processing because of parallel processing and abilityto build a DSP chip with VLSI Technology

REFERENCES1. Li Wang, Senio Dynamic Stability

Improvement of an IntegratedGrid-Connected Offshore Wind Farm andMarine- Current Farm using STATCOM,”,IEEE TRANSACTIONS ON POWERSYSTEMS, VOL. 26, NO. 2, MAY 2011.

2. Arrillaga and C. P. Arnold. “Computeranalysis of power systems”, ISBN 0 47192760 0.

3. Mats Larsson,” Coordinated VoltageControl in Electric Power Systems”,Department of Industrial ElectricalEngineering and Automation Lund Instituteof Technology.Lund Un iversity.

4. N. G. Hingorani and L. Gyugyi,Understanding FACTS: Concepts andTechnology of Flexible AC TransmissionSystem, IEEE Press, 2000.

5. Muhammad H. Rashid,” Power ElectronicsHandbook”, University of

Florida,University of West Florida JointProgram and Computer Engineering.

6. Dr.Shailendra Jain,”Modelling & Simulationusing Matlab- Simulink” Wiley IndiaPvt.Ltd.

7. L.L. Grigsby, Andrew Hanson,”Power

System Analysis and Simulation”, © 2001CRC Press LLC8. Om Nayak, Surya Santoso, Paul Buchanan,’

Power Electronics Spark New SimulationChallenges”.

9. Roger C Ducan, Mark F Mc Granaghan,Surya Santoso,” Electrical Power SystemsQuality, Second Edition”.

10. Simone Buso, Paolo Mattavelli,” DigitalControl in Power Electronics”, Morgan &Claypool Publishers.

11. L. Gyugyi, C.D. Schauder, S.L. Williams,T.R.Rietman, D.R. Torgerson, A. Edris,"The UnifiedPower Flow Controller: A

New Approach to PowerTransmissionControl", IEEE Trans., 1995, pp. 1085-1097.

12. H F Wang. ‘Damping Function of UnifiedPower Flow Controller.’ IEE Proceedings -C,vol 146, no 1, January 1999, p 81.

supression of low freqquency oscillations in the power system by neuro fuzzy logic controller

Documents