LFC OF TWO INTERCONNECTED POWER SYSTEM USING INTELLIGENT CONTROLLER METHOD

Transcription

1 U.P.B. Sc. Bull., Seres C, Vol. 77, Iss. 1, ISSN LFC OF TWO INTERCONNECTED POWER SYSTEM USING INTELLIGENT CONTROLLER METHOD S.M. KAMALI NALLADURAI 1, R.S.D WAHIDA BANU 2 In ths paper the bran emotonal learnng based ntellgent controller (BELBIC) has been proposed for solvng the frequency varaton problems due to the uncertanty n loads n the two nterconnected power systems. It s a recent advanced controller based on emotonal processng method n the bran. The devatons n the frequency can be mnmzed by tunng the proposed controller. The steady state error n the nterconnected system can be elmnated by the PI controller, but t develops overshoot problems n the system. The llustrated results show that BELBIC method s more robustness wth acceptable response n the system parameters. Keywords: Load frequency control, Bran Emotonal Learnng Based Intellgent Controller, Ant Wndup Control 1. Introducton The prmary concern n the desgnng of nterconnecton of two power systems s the control of load frequency, whch becomes more sgnfcant when ncreasng the sze and the structure. It s necessary to mantan the frequency and the power flow n the te lne, where two power systems nterconnected together, wthout devaton even under load dsturbances [1]. The Load Frequency Control (LFC) attans the prmary functon are provdng better electrcty and safe system operaton. Hence, t necesstates desgnng a load frequency control for mantanng the better relablty n the electrc power system at a requred level. The commonly used technque s the PI controller for controllng such load frequency. But the selecton of ts gan becomes a major drawback because of the non-lnearty of the power system. The conventonal PI controller can able to mantan the constant load frequency under nconsstent load condtons n the two nterconnected power system. But t generates sudden overshoot problems, n order to provde an effcent soluton to the LFC problem. Many controllers such as Optmal Control (OC) [17]-[18] and Sldng Mode Controller (SMC) [19] are used. But n the SMC and OC nvolves more complex, and stablty of the power 1 Department of EEE, Mahendra Insttute of Technology, Namakkal , Inda, E-mal: kamalnalladura@gmal.com 2 Prncpal, Government College Of Engneerng Salem , Inda

2 74 S.M. Kamal Nalladura,.R.S.D Wahda Banu system s not assured. Hence, the desgnng of a controller wth less processng tme s necessary. As a result, the effcent, ntellgent controller s proposed to mnmze the problems such as overshoot and the hgh settlng tme. 2. Modellng Of The Proposed System The nter-area te lne power wth real oscllatons and the controllng of load frequency are the man aspects of ths paper. Area control error s a vtal role and acts as the nput sgnal for the BELBIC controller [6]. Also the BELBIC controller controls and mantans the load frequency of the two nterconnected power areas smultaneously through a te lne as shown n the Fg. 1. In the case of two or more nterconnected power system, the load frequency wll start to oscllate due to the dfferent load frequency of dfferent power system that leads to provde stablty problems. In order to avod these stablty problems, the Automatc Generaton Control (AGC) technque must be added to reduce the frequency devatons. Fg1. Block dagram of proposed system 2.1 Drawbacks n the Conventonal PI Controller The PI controller s the smplest controller wth real speed of response, whch can be referred to the followng equaton (1).

3 LFC of two nterconnected power system usng ntellgent controller method 75 K U ( s) = K P E( s) + E( s) (1) s But t produces low steady state error and long settlng tme. 2.2 Ant Wndup PI Controller The Ant-wndup PI controller s desgned by ntroducng the ntegrator wndup phenomenon n the PI controller wth feedback. In ths controller, to reduce the overshoot and the settlng tme a maxmum ntegrated output value s kept wthn the lmts. Hence, t can recover quckly from the nonlneartes. 2.3 Fuzzy gan schedulng PI Controller Intellgent fuzzy logc controller (FLC) s desgned to tune the gans of the PI controller n AGC. The fxed values of the gans n PI controller rrespectve of error may ncrease the settlng tme. It s referred as fuzzy gan schedulng (FGSPIC). In FGSPIC based PI receves ACE contnuously. It decdes the proportonal gan and ntegral gan of the PI controller based on present ACE and change n ACE. Proportonal Controller has the effect of reducng rse tme, but never elmnates the steady state Error. An ntegral control has the effect of elmnatng steady state error, but the hgh value of an ntegral controller may make the transent response worse. However, the overshoot controlled by the FGSPI controller s not up to the requred level. Thus to decrease the overshoot to a very low-value BELBIC s ntroduced n the AGC of the nterconnected power system. 3. BELBIC Controller Structure (Proposed controller) In 2004, Lucas ntroduced BELBIC, s a non-lnear and neuromorphc controller based on the computatonal learnng method to produce control actons. Snce BELBIC s a bo-nspred control method, the structure s based on the lmbc system of mammalan s bran. It has four man sectons such as the amygdala (A), orbtofrontal cortex (O), thalamus (TH) and sensory cortex (CX) as shown n Fg. 2 and Fg. 3. Amygdala plays an mportant role because the emotonal learnng process s takng place. Amygdala receves nputs from the Thalamus and cortcal areas whle the Orbtofrontal receves nputs from Amygdala and cortcal areas. The system also receves a Prmary Reward sgnal (REW) n addton to Sensory Cortex nputs. There s also one O node for each of the stmul n the Orbtofrontal except for the thalamc node. The output node, E, smply sums the outputs from the A node and

4 76 S.M. Kamal Nalladura,.R.S.D Wahda Banu then subtracts the nhbtory outputs from the O nodes. The result s the output from the model. The E node sums the outputs from A except A th and then subtracts from nhbtory outputs from the O nodes. E = A O (2) E ' = A O (3) The thalamc connecton s calculated as the maxmum of stmul nputs (S): A th =max(s ) (4) Unlke other nputs to the Amygdala, the thalamc nput s not planned as the Orbtofrontal part and cannot be nhabted. Fg 2. Smplfed model of BELBIC The output of the node s obtaned by multplyng the nput by a plastc connecton weght V for each A node. A =S V (6) The connecton weghts V are drectly proportonal to the dfference n the actvaton of the A nodes and prmary reward(pr). The learnng rule of Amygdala s gven as follow: ΔV = α S max(0, PR A ) (7) where α s a learnng rate parameter whch s used to adjust the learnng speed. Its value s set between 0 (no learnng) and 1 (nstant adaptaton). The Orbtofrontal learnng rule and the Amygdala rule are more smlar, but the weghts of the Orbtofrontal connecton should be ncreased and decreased to track the requred nhbton.

5 LFC of two nterconnected power system usng ntellgent controller method 77 Fg 3. A graphcal depcton of the computatonal model of emotonal learnng The learnng rule n Orbtofrontal cortex s calculated as follows: { S ( E' PR) } ΔW = β (8) where, ΔW = Change n the weght of Orbtofrontal connecton Β = Orbtofrontal learnng rate The Orbtofrontal connecton node values are calculated as follows: O =S W (9) 3.1 Desgn of Two Interconnected Power System The two nterconnecton power systems consst of more complex and mult-varable structures wth several nterconnected control areas. In each area, there s an own generator or group of generators, whch s responsble for ts load and scheduled nterchangng wth neghborng areas connected through te-lnes. The te-lnes provde support for the contractual exchange of power between areas and the case of abnormal condtons. Durng the varable load condtons, the power flow and frequency of the te lne wll be affected at the operatng pont. Hence for the stable operaton of power system wth varable load and abnormal system parameters, the devatons n the system frequency and te-lne power should be mnmzed as quckly as possble. Thus to ensure the qualty of power supply, a load frequency controller [2] s needed to mantan the system frequency at the desred nomnal value. The lnear combnatons of system frequency and the te lne power flow are known as area control error (ACE). The frequency and nterchanged power are kept at ther desred values by means of feedback of area

6 78 S.M. Kamal Nalladura,.R.S.D Wahda Banu control error contanng devaton n frequency and error n te lne power, and controllng the prme movers of generators. The area control error (ACE) for ts area s defned as: β = 1 D R + (10) Where R s the regulaton constant, and D s the dampng rato of ts system. Each nterconnected power system area has three major components namely turbne, governor, and generator. The transfer functon block dagram of uncontrolled twoarea system s llustrated n Fg. 6. The frequency devatons n area1 and area2 are f1 and f2 respectvely n Hz. Lkewse, PL1 and PL2 represent the load demand changes n areas 1 and two respectvely n per unt (p.u.) [2]. 3.2 Desgnng BELBIC Model for Load Frequency Controller The man am of the load frequency controller s to get the dampng frequency n the te-lne area for a better performance. For ths, BELBIC controller s consdered as controller n each control area. Accordng to Fg. 2, n each control area the ACE acts as the nput sgnal of the BELBIC controller that s used by the LFC system. Fg. 7 shows the block dagram of the new control system ncorporatng the emotonal controller (BELBIC). Fg. 4. Transfer functon block dagram of two nterconnected power systems

7 LFC of two nterconnected power system usng ntellgent controller method 79 Fg 5. Desgnng BELBIC model for Load Frequency Controller As t s llustrated n Fg. 2, sensory nput (SI) and reward sgnal (PR) can be the arbtrary functon of the reference, and the plant nput and output. It s all up to the desgner to fnd a proper functon for control. In ths work, the functons used n emotonal cues and sensory nput blocks are consdered as follow: d( ACE) RW = k1 ( ACE) + k2 + k3 ( ACE). dt (11) dt SI = k ACE + k ACE. Δf 4 5 (12) Where ACE represents the area control error, f represent the frequency devaton for the w th area and k 1,k 2, k 3, k 4, k 5 are gans, whch must be tuned for desgnng a satsfactory controller. The gan k 1 and k 2 are responsble for tunng the overshoot. The gan k 3 s responsble for tunng the settlng tme. The gan k 4 responsble for tunng the steady-state error and the gan k 5 s responsble for smoothng the begnnng of the response. 4. Results and Dscussons The smulaton results of the two nterconnected power system were analyzed usng MATLAB R2011b. Prmarly the system on a conventonal PI controller was focused. Then the same system s analyzed usng an Ant Wndup PI controller and BELBIC. The decsve factors consdered n ths paper are settlng tme and overshoot. To examne the performance of the plant, n ths paper two nterconnected power systems are dsturbed by 1% step load. Table 1 shows the parameters of the system consdered.

8 80 S.M. Kamal Nalladura,.R.S.D Wahda Banu Parameters of Power Plant R1, R2 2Hz/p.u.M.W T T 0.3 Sec T H 0.08 Sec K p 100 T p 20 Sec A 12-1 T B p.u.m.w / Hz Table 1 Fg. 6. shows the frequency devaton n system1 of the two areas nterconnected system wth varous controllers. Experments performed by varous controllers wth dfferent load perturbatons also done BELBIC (Proposed) AWPI Fuzzy-PI Conv-PI 0 Del-F1 (Hz) Tme (sec) Fg. 6. Frequency devaton n system 1 wth all controllers Frequency devaton of all controllers n system 1 and system 2 are performed and compared. Table II shows performance comparsons of varous controllers. Table 2 Comparson of controller performance Controllers Load (p.u) t s e ss %M p TT TT TT Conventonal PI FGSPIC AWPIC BELBIC

9 LFC of two nterconnected power system usng ntellgent controller method Concluson In ths paper, a new control system ncorporatng the emotonal controller (BELBIC) s used for control of frequency and dampng the te-lne power varaton n a mult-machne power system. The performance of desgned controller s expermented on a two-area power system wth consderng governor lmters and the results obtaned are compared wth the conventonal PI controller, FGSPI controller, AWPI controller. The robustness of the proposed method s tested aganst change of parameters. The smulaton studes show that the desgned controllers by BELBIC have a very desrable dynamc performance even when the system parameters change. R E F E R E N C E S [1]. H. Shayegh, A robust decentralzed power system load frequency control, Journal of Electrcal Engneerng, vol. 59, 2008, pp [2]. A. Khodabakhshan and M. Edrs, A new robust PID load frequency controller, Control Engneerng Practce, vol. 16,2008, pp [3]. M. R. Jamal, A. Aram, B. Hossen, B. Moshr, and C. Lucas, "Real tme emotonal control for ant-swng and postonng control of SIMO overhead travelng crane", Internatonal Journal of Innovatve Computng, Informaton and Control, vol. 4 No. 9, 2008, pp [4]. H. Shayegh and H. A. Shyanfar, Applcaton of ANN technque based on μ-synthess to load frequency control of nterconnected power system, Electrcal Power and Energy Systems, vol , pp [5]. A. Khodabakhshan and N. Golbon, Robust load frequency controller desgn for hydro power systems, n Proc. IEEE Conf. Control Applcatons, pp [6]. C. Lucas, D. Shahmrzad, and N. Shekholeslam, Introducng BELBIC: Bran Emotonal Learnng Based Intellgent Controller, Internatonal Journal of Intellgent Automaton and SoftComputng, vol. 10, 2004, pp [7]. R. Sngh, I. Sen, Tunng of PID controller based AGC system usng genetc algorthms, n Proc. TENCON'04, 2004, pp [8]. C. F. Juang, C. F. Lu, Power system load frequency control wth fuzzy gan schedulng desgned by new genetc algorthms, n Proc. IEEE Int. Conf. on Fuzzy Systems. 2002, pp [9]. P. L, H. Zhu, Y. L, Genetc algorthm optmzaton for AGC of mult-area power systems, n Proc. IEEE TENCON'02., 2002 pp [10]. J. Moren and C. Balkenus, A Computatonal Model of Emotonal Learnng n The Amygdala: From anmals to anmals. n Proc. Internatonal conference on the smulaton of adaptve behavor, Cambrdge, Mass., The MIT Press, 2000 pp [11]. D. K. Chaturved, P.S. Satsang and P. K. Kalra, Load frequency control: a generalzed neural network approach, Internatonal Journal of Electrcal Power & Energy Systems, vol. 21, 1999, pp [12]. G. Ray, A. N. Prasad, and G. D. Prasad, A new approach to the desgn of robust load frequency controller for large-scale power systems, Elect. Power Syst. Res., vol. 51, 1999, pp ,

10 82 S.M. Kamal Nalladura,.R.S.D Wahda Banu [13]. M. Azzam, Robust automatc generaton control, Energy Convers. Manage., vol. 40, 1999, pp [14]. J. Talaq and F. Al-Basr, Adaptve fuzzy gan schedulng for load frequency control, IEEE Trans. Power Syst., vol. 14, Feb 1999, pp [15]. H. Jang, H. Ca, J. F. Dorsey F, Z. Qu, Toward a globally robust decentralzed control for large-scale power systems, IEEE Trans. Control Syst Technol., vol. 5, 1997 pp [16]. N. N. Bengamn and W. C. Chan, Varable structure control of electrc power generaton, IEEE Trans. Power App. Syst., vol.pas-101, 1982, pp [17]. M. Calovc, Lnear regulator desgn for a load and frequency control theory, IEEE Trans. Power App. Syst., vol. PAS-91, 1972 pp [18]. C. E. Fosha and O. I. Elgerd, The megawatt-frequency control problem: A new approach va optmal control theory, IEEE Trans. Power App. Syst., vol. PAS-89, 1970, pp [19]. Mohammad Shamsodn, Rouholla Mane, Al Bekter, Babak Ranjbar & Samra Soltan, Desgn a New Fuzzy Optmze Robust Sldng Surface Gan n Nonlnear Controller, Internatonal Journal of Intellgent Systems and Applcatons(IJISA), vol.5, no. 12, Nov 2013, pp