Load frequency control of interconnected hydro-thermal power system using conventional pi and fuzzy logic controller

Transcription

1 International Journal of Energy and Power Engineering 23; 2(5): 9-96 Publihed online October 3, 23 ( doi:.648/j.ijepe Load frequency control of interconnected hydro-thermal power ytem uing conventional pi and fuzzy logic controller Muhammad Ahan Zamee, Dipankar Mitra 2, Sadaf Yuuf Tahhan 2 Department of Electrical & Electronic Engineering, World Univerity of Bangladeh, Dhaka, Bangladeh 2 Department of Electrical & Electronic Engineering, Chittagong Univerity of Engineering and Technology, Chittagong, Bangladeh addre: zamee_mit@yahoo.com (M. A. Zamee), dipankar.mitra.dipu@gmail.com (D. Mitra), tahhanadaf@gmail.com (S. Y. Tahhan). To cite thi article: Muhammad Ahan Zamee, Dipankar Mitra, Sadaf Yuuf Tahhan. Load Frequency Control of Interconnected Hydro-Thermal Power Sytem Uing Conventional PI and Fuzzy Logic Controller. International Journal of Energy and Power Engineering. Vol. 2, No. 5, 23, pp doi:.648/j.ijepe Abtract: In indutry or any area increaing load i a vat problem for power generation plant due to increae in demand for power. So making balance between generation and demand i the operating principle of load frequency control (LFC). The reliable operation of a large interconnected power ytem necearily require an Automatic Generation Control (AGC). The objective of AGC i to regulate the power output of Generator within a pecified area in repone to change in the ytem frequency, tie line power or relation of the two to each other, o a to maintain the cheduled ytem frequency and power interchange in the other are within the precribed limit. Thi paper preent the ue of conventional PI controller and artificial intelligence to tudy the load frequency control of interconnected power ytem. In the propoed cheme, a control methodology i developed uing conventional PI controller and Fuzzy Logic controller (FLC) for interconnected hydro-thermal power ytem. The control trategie guarantee that the teady tate error of frequencie and inadvertent interchange of tie-line power are maintained in a given tolerance limitation. The performance of the controller are imulated uing MATLAB/SIMULINK package. A comparion of Fuzzy controller and PI controller baed approache how the uperiority of propoed Fuzzy logic controller for tep change in loading condition. The imulation reult alo tabulated a a comparative performance in view of ettling time and peak over hoot. Keyword: Load Frequency Control, Fuzzy Logic Controller, PI controller, MATLAB/SIMULINK. Introduction Due to increae in ytem load; turbine peed drop before the governor can adjut the input. A the change in the value of peed decreae, the error ignal become maller and the poition of governor valve get cloe to the required poition, to maintain the contant peed. However the contant peed will not be the et point and there will be an offet, to overcome thi problem an integrator i added, which will automatically adjut the generation to retore the frequency to it nominal value. Thi cheme i called automatic generation control (AGC). The role of AGC i to divide the load among the ytem, tation and generator to achieve maximum economy and accurate control of the cheduled interchange of tie-line power while maintaining a reaonability uniform frequency. Automatic generation control (AGC) play a very important role in power ytem a it main role i to maintain the ytem frequency and tie line flow at their cheduled value during normal operating period. Automatic generation control with primary peed control action, a change in ytem load will reult in a teady tate frequency deviation, depending upon governor droop characteritic and frequency enitivity of the load. Retoration of the ytem frequency to nominal value require upplementary control action which adjut the load reference et point. Therefore the primary objective of the automatic generation control are to regulate frequency to the nominal value and to maintain the interchange power between control area at the cheduled value by adjuting the output of elected generator. Thi function i commonly referred to a load frequency control. A econdary objective i to ditribute the required change in generation among the unit to minimize the operating cot. A control ignal made up of tie line flow deviation added to frequency deviation weighted by a bia factor

2 92 Muhammad Ahan Zamee et al.: Load Frequency Control of Interconnected Hydro-Thermal Power Sytem Uing Conventional PI and Fuzzy Logic Controller would accomplih the deired objective. Thi control ignal i known a area control error (ACE).ACE erve to indicate when total generation mut be raied or lowered in a control area. In an interconnection, there are many control area, each of which perform it AGC with the objective of maintaining the magnitude of ACE (area Control Error) ufficiently cloe to uing variou criteria. In order to maintain the frequency ufficiently cloe to it ynchronou value over the entire interconnection, the coordination of the control area action i required. Any wide deviation from the nominal value of frequency or voltage will lead the ytem to total collape. Hence AGC ha gained importance with the growth of interconnected ytem and with rie in ize of interconnected ytem automation of the control ytem have aroued. A number of control trategie exit to achieve better performance. [7] The mot applied controller i Conventional Proportional Integral (PI) [3, 6]. It i eaier but uually give large ettling time. Mot reearch going on now i baed on artificial intelligent ytem (fuzzy and neural network). The inherent gain of thee technique i that they do not require the ytem model and identification but depend on human expertie knowledge of the behavior. In thi paper, a fuzzy logic controller along with PI controller i propoed and performance comparion i carried out for conventional PI. 2. Load Frequency Control Theory 2.. The Invetigated Power Sytem The detailed block diagram modeling of two area thermal-hydro power ytem for load frequency control invetigated i hown in figure. An extended power ytem can be divided into a number of load frequency control area interconnected by mean of tie line. Without lo of generality one can conider a two- area cae connected by ingle tie line (Surya-Prakah et al.29). The control objective are a follow: Each control area a for a poible hould upply it own load demand and power tranfer through tie line hould be on mutual agreement. Both control area hould controllable to the frequency control. In an iolated control area cae the incremental power (G D) wa accounted for by the rate of increae of tored kinetic energy and increae in area load caued by increae in frequency. Since a tie line tranport power in or out of an area, thi fact mut be accounted for in the incremental power balance equation of each area Modeling of the Tie-Line The power tranfer equation through tie line i given by, V V2 P ( ) Sin δ δ () 2 x P Power tranferred from area to 2 through tie line. Conidering area ha urplu power and tranfer to area 2. V V2 P ( ) Sin δ δ2 (2) X Therefore, Power tranferred from Area to Area 2 i given by the following equation T Torque produced 2.3. Tie line Control 2πT ) ( 2 ( f ( ) f ( )) In normal operation the power on the tie-line follow from the equation i.e. (3) If 2H [ ( ) ( ) P ( )] f( ) T E f + B f( ) 2H f ( ) B + f B 2HB f K P (4) (5) Equation (4) can be written a f ) G ( )[ ( ) ( ) ( )] (6) ( P T E Fig. : Block diagram model of hydro-thermal reheat power ytem.

3 International Journal of Energy and Power Engineering 23; 2(5): G P KP ( ) + T Where E i real load change Due to the action of turbine controller, the generator increae it output by the amount T. The net urplu power T E will be aborbed by the ytem. Tie-line bia control i ued to eliminate teady tate error in frequency in tie-line power flow. Thi tate that the each control area mut contribute their hare to frequency control in addition for taking care of their own net interchange. Let ACE area control error of area ACE2 Area control error of area 2 In thee control area, ACE and ACE 2 are made linear combination of frequency and tie line power error. 2 P (7) ACE P + b f (8) ACE2 P2 + b2 f2 (9) Where, the contant b & b 2 are called area frequency bia of area and area 2 repectively. Now R and R 2 are mode integral of ACE and ACE 2 repectively. t i ( ) R K + b f dt () t 2 i 2 ( 2 2 2) R K + b f dt () 2πT f f 2 (input of integrating block ) (6) tie, and tie, 2 Therefore T T 2 a tie, tie,2 Contant (7) Hence tie, tie, 2 R R 2 And f f 2 Thu, under teady condition change in the tie- line power and frequency of each area i zero. Thi ha been achieved by integration of ACE in the feedback loop of each area (Surya-Prakah et al. 29). Control methodology ued (FLC & PI) i mentioned in next preceding ection. [, 2, 8] 3. Controller Ued a) Conventional PI Controller b) Fuzzy Logic Controller. 3.. Conventional PI Controller When an integral controller i added to each area of the uncontrolled plant in forward path the teady tate error in the frequency become zero. The tak of load frequency controller i to generate a control ignal u that maintain ytem frequency and tie-line interchange power at predetermined value [2]. The block diagram of PI controller i hown in figure2. Taking Laplace tranform of the above equation, we get Ki PR( ) [ ( ) + b f( )] () Ki2 PR2( ) [ 2( ) + b2 f2( )] (3) The tep change D and D2 are applied imultaneouly in control area and 2 repectively. When teady tate condition are reached, the output ignal of all integrating block will be contant and their input ignal mut become zero. K i.e. + b f (input of integrating block i )(4) K 2 + b 2 f 2 (input of integrating block i2 )(5) Fig. 2: Conventional PI controller Conventional Proportional plu Integral controller (PI) provide zero teady tate frequency deviation, but it exhibit poor dynamic performance (uch a number of ocillation and more ettling time), epecially in the preence of parameter variation and nonlinearity [].In PI Controller Proportionality contant provide implicity, reliability, directne etc. The diadvantage of offet in it i eliminated by integration but thi ytem will have ome ocillatory offet. The control ignal can be written a: U K p. ACE K i ʃ ACE dt U2 K p. ACE2 K i ʃ ACE2 dt Where Kp and Ki are proportional and integral gain,

4 94 Muhammad Ahan Zamee et al.: Load Frequency Control of Interconnected Hydro-Thermal Power Sytem Uing Conventional PI and Fuzzy Logic Controller repectively. For conventional PI controller, the gain K p and K i ha been optimized uing integral quare error (ISE) criterion. For ISE technique, the objective function ued i, t J ( F + F + + ) dt 2 tie Where F Change in frequency Ptie Change in tie line power PI and Fuzzy Logic Controller. For fuzzy logic controller, fuzzy logic toolbox ha been ued to et up the memberhip function and interference rule Fuzzy Logic Controller Fuzzy logic i a problem-olving control methodology incorporated in control ytem engineering, to control ytem when input are either imprecie or the mathematical model are not preent at all. There are three principal element to a fuzzy logic controller. Fuzzification module (Fuzzifier) 2. Inference rule engine 3. Defuzzification module (Defuzzifier) For Load Frequency Control the proce operator i aumed to repond to repond to variable error (e) and change of error (ce) (that i frequency deviation and change in frequency deviation). five number of triangular memberhip function (MF) which provide better dynamic repone with the range on input (error in frequency deviation and change in frequency deviation) i.e. univere of dicoure i -.25 to.25. The number of rule are 25. The dynamic repone are obtained and compared to thoe obtained with conventional integral controller. Fig. 4: Setting up fuzzy rule in MATLAB/SIMULINK Fig. 3: Memberhip function for control input variable Fig. 5: Frequency repone without uing any controller. Table : Fuzzy Interference rule for fuzzy logic controller Input e(k) NB NM ZE PM PB ce(k) NB NB NB NM NM ZE NM NB NB NM ZE ZE ZE NM NM ZE PM PM PM ZE PM PM PB PB PB ZE ZE PM PB PB 4. Reult and Dicuion MATLAB/SIMULINK oftware ha been ued for evaluating the performance of propoed controller for both Fig. 6: Frequency Repone uing PI Controller

5 International Journal of Energy and Power Engineering 23; 2(5): New controller can be deigned with other Artificial Intelligence algorithm for better control performance in term of frequency and tie line power deviation. 3. More than one controller can be ued uch a conventional PI, PID and FLC or Artificial Neural Network in erial or parallel for reducing tranient repone and peak overhoot. Fig. 7: Frequency repone uing fuzzy logic controller 4.. Comparion of Sytem Performance Frequency deviation Table 2: Performance evaluation without uing any controller Settling time Frequency deviation Settling time Frequency deviation Thermal Hydro Combined ~ Stable at deviated frequency Stable at deviated frequency Stable at deviated frequency Table3: Performance evaluation uing PI controller Thermal Hydro Combined -3.5~ ~ ~ Sec 45 ec 5 Sec Table 4: Performance evaluation uing Fuzzy Logic controller Thermal Hydro Combined -2.5~ ~ ~.2 Settling time Concluion From the above reearch it can be een concluded that the tranient repone, ettling time and peak overhoot in cae of fuzzy logic controller i leer compared to the conventional PI controller. Thu imulation reult of FLC have better control performance over conventional PI when ome diturbance in load i given to the ytem. In hort we can ay that FLC i adequate for better quality and reliable electric power upply. 6. Future Scope. More than two area uch a thermal, hydro, ga etc can be interconnected and controlled for automatic generation of controlled power. Appendix Parameter are a follow: f 5 Hz, R R2 2.4 Hz/ per unit MW, Tg.8 ec, Tp2 ec P tie, max 2 MW Tr ec kr.5, H H2 5 ec Pr Pr2 2MW Tt.3 ec KpKp2 Hz.p.u/MW Kd 4. ki 5. Tw. ec D D * -3 p.u MW/Hz. Nomenclature F : Nominal ytem frequency Pri : Area rated power, Hi: Inertia contant P Di : Incremental load change Pg i : Incremental generation change T : Synchronizing coefficient, Tg:Steam governor time contant Kr : Reheat contant, Tr : Reheat time contant Tt : Steam turbine time contant Ri : Governor peed regulation parameter Bi: Frequency bia contant Reference [] Anand B., Ebenezer A. Jeyakumar. 29. Load frequency control with fuzzy logic controller conidering non-linearitie and boiler dynamic, ICGST-ACSE Journal, ISSN , Volume 8, iue, pp 5-2. [2] Aravindan P., Sanavullah M.Y. 29. Fuzzy Logic Baed Automatic Load Frequency Control of Two Area Power Sytem With GRC, International Journal of Computational Intelligence Reearch, Volume 5, Number. pp [3] Jawat, T. and Fadel, A, B. Adaptive Fuzzy Gain Scheduling for Load frequency control, IEEE Tran. on PAS, vol. 4, No, February 999. [4] Ibraheem, Kumar P., Kothari D.P., 25. Recent Philoophie of Automatic Generation Control trategie in Power ytem, IEEE Tranaction on Power Sytem Vol.2, No., pp [5] Surya-Prakah, Sinha S.K., Brijeh-Singh, Pandey A.S. 29. Impact of lider gain on Load Frequency Control uing Fuzzy Logic Controller, ARPN Journal of Engineering and Applied Science, Vol. 4, No 7.pp [6] Edion, B, and Ilie, M, Advanced Generation control: Technical Enhancement, cot, and Repone of market Driven Demand Proc. of the 57th Annual American Power

6 96 Muhammad Ahan Zamee et al.: Load Frequency Control of Interconnected Hydro-Thermal Power Sytem Uing Conventional PI and Fuzzy Logic Controller conf; vol. 57, No. 2, 995, pp [7] Amit Kumar, Aziz Ahmad, Ahwani Grover, Umeh Gupta, Load Frequency Control Uing Fuzzy Logic, International Journal of Scientific and Reearch Publication, Volume 2, Iue 7, July 2 [8] Surya Prakah, S. K. Sinha, Application of artificial intelligence in load frequency control of interconnected power ytem, International Journal of Engineering, Science and Technology. [9] A. Mangla and J. Nanda, Automatic Generation Control of an Interconnected Hydro-Thermal Sytem Uing Conventional Integral and Fuzzy Logic controller, International conference on electrical utility, deregulation, detructuring, and power technologie, pp , April 24. [] Q. P. Ha A Fuzzy liding mode controller, International Conference of Knowledge baed Intelligent Electronic Sytem. Adelaide, Autralia. 23 2nd April.