Feeder Reconfiguration for Loss Reduction in Unbalanced Distribution System Using Genetic Algorithm

Transcription

1 nterntionl Journl of Electricl nd Electronics Engineering 3: Feeder Reconfigurtion for Loss Reduction in Unblnced Distribution Syste Using Genetic Algorith Gnesh. ulsl, Sivngru. Sirigiri, Rn. Thiruveedul Abstrct This pper presents n efficient pproch to feeder reconfigurtion for power loss reduction nd voltge profile iprveent in unblnced rdil distribution systes (URDS). n this pper Genetic Algorith (GA) is used to obtin solution for reconfigurtion of rdil distribution systes to inii the losses. A forwrd nd bcwrd lgorith is used to clculte lod flows in unblnced distribution systes. By siulting the survivl of the fittest ong the strings, the optiu string is serched by rndoid infortion exchnge between strings by perforing crossover nd uttion. Results hve shown tht proposed lgorith hs dvntges over previous lgoriths The proposed ethod is effectively tested on 19 node nd 25 node unblnced rdil distribution systes. Keywords Distribution syste, Lod flows, Reconfigurtion, Genetic Algorith.. NTRODUCTON OWER distribution systes typiclly hve tie nd P sectionlizing switches whose sttes deterine the topologicl configurtion of the networ. The syste configurtion ffects the efficiency with which the power supplied by the substtion is trnsferred to the lod. Power copnies re interested in finding the ost efficient configurtion, the one which iniis the rel power loss of their three-phse distribution systes. t ws estited tht distribution systes cuse loss of bout 5 13% of the totl power generted [1] in developing countries. n [2]-[3] brnch exchnge ethod ws used in which pproxite forule provide the chnge in loss due to feeder reconfigurtion. Shirohdi nd Hong [4] proposed techniue in which the switches were opened one by one beginning fro fully eshed syste, bsed on n optil flow pttern. Goswi nd Bsu [5] proposed n lgorith bsed on optil flow pttern, of single loop, fored by closing norlly open switch, nd the switch with iniu current ws opened. Tylor nd Luben [6] developed n Gnesh. ulsl is with the Jwhrll Nehru Technologicl university, Anntpur, Andhr Prdesh, ndi (corresponding uthor to provide phone: ; fx: ; e-il: gni_vg@ yhoo.co). Sivngru. Sirigiri, ws with Jwhrll Nehru Technologicl University,Kind, Andhrprdesh, ndi. (e-il: sirigiri70@yhoo.co.in). Rn. Thiruveedul is with the RSYRN College of engineering nd Technology, Chirl, Andhr Prdesh, ndi. (e-il: trdy@yhoo.co.in). expert syste using heuristic rules to shrin the serch spce for reducing the coputtion tie. Gloocnin [7] used udrtic progring techniue to solve the reconfigurtion proble. Wgner et l. [8] proposed liner progring ethod nd heuristic serch ethod. Borozn et l. [9] presented ethod siilr to tht of [2]-[3], for solving reconfigurtion proble. Copenstion bsed power flow ethod ws used to obtin power flow solution for eshed syste. A survey on reconfigurtion ws presented in [10]. Srfi et l. [11] developed ethod bsed on prtitioning the distribution syste into group of lod buses, such tht the line section losses between groups of nodes were iniid. Royteln et l. [12] presented heuristic-bsed two stge solution pproch, in which weights were ssigned to ultiobective functions. n [13]-[14], lgoriths for distribution syste switch reconfigurtion nd cpcitor control hve been proposed. McDerott et l. [15] proposed constructive heuristic ethod tht strted with ll switches open, nd t ech step, the switch tht resulted in the lest increse in the obective function ws closed. Chen nd Cho [16] presented n pproch to derive optil switching pln to chieve energy loss iniiztion, for short nd long-ter opertion of distribution systes. Zhou et l. [17] proposed heuristic pproch for reconfigurtion, which reduced operting cost over specified tie period. Schidt et l. [18] forulted the proble s ixed integer, nonliner optiiztion proble. Newton ethod is used to copute brnch currents within the integer serch. n [19], ethod ws proposed to deterine the configurtion with iniu energy loss for given period. n [20] [23], solution strtegies hve been proposed for feeder reconfigurtion using siulted nneling. Morton et l. [24] developed grph theoretic techniues involving sei-sprse trnsfortions of current sensitivity trix. n [25] [31], different pproches were presented to obtin iniu loss configurtion of the distribution syste using genetic lgorith. Hsio, Y.T [32] developed lgoriths bsed on ulti obective evolution progring ethod for feeder reconfigurtion. A heuristic ethod for three-phse unblnced systes ws proposed in [33] bsed on pproches in [3, 4]. Wng et l. [34] proposed n lgorith to inii loss nd lod blncing for lrge-scle unblnced syste using networ reconfigurtion. n this ethod, explicit loss forule were developed for deterintion of switching opertions. 754

2 nterntionl Journl of Electricl nd Electronics Engineering 3: t cn be seen fro the literture heuristic bsed ethod hs been proposed for the proble under considertion. Hence, there ppers to be need for developing siple reconfigurtion ethod using GA for unblnced systes. n this pper, GA is presented for reconfigurtion of unblnced distribution systes for loss iniiztion. A siple ethod for deterining the open/closed sttes of sectionlizing nd tie switches to chieve loss reduction is explined. The perfornce of proposed ethod is tested with two different unblnced rdil distribution systes consisting of 19 nd 25 node unblnced rdil distribution systes. n this wor distribution syste considered s three-phse three wire line sections nd ll of the re phse-syetricl, but bus lods y not be eul for ll phses. Therefore, the distribution networ should be treted s unblnced. The gol of this wor describes siple reconfigurtion ethod for unblnced rdil distribution systes for loss iniiztion hs been proposed, which is resonbly siple. The effectiveness of the proposed ethod hs been copred with the ethods of Shirohdi nd Hong [4] nd Goswi nd Bsu [5] in respect of switching nd loss reduction.. PROBLEM FORMULATON LOSS MNMZATON n the rdil distribution syste, ech rdil feeder is divided into lod sections with sectionlizing switches nd is connected to other feeders vi severl tie switches. The distribution systes loss iniu reconfigurtion proble is to decide the position of sectionlizing switches to obtin the best iniu distribution losses with stisfying the following constrints. This proble cn be forulted s = nb Minii f P (1) = 1 loss where, nb is the nuber of brnches P is the loss in brnch loss Subect to the following constrints i. Rdil networ constrint Distribution networ should be coposed of rdil structure considering opertionl point of view. ii. Power source liit constrint The totl lods of certin prtil networ cnnot exceed the cpcity liit of the corresponding power source. x P s P (2) x Q s Q (3) iii. oltge constrint oltge gnitude t ech bus ust lie with their perissible rnges to intin power ulity. in x (4) iv. Current constrint Current gnitude of ech brnch ust lie within their perissible rnge. x (5) n the networ reconfigurtion proble the lod flow study is reuired to clculte the overll rel power loss for given syste configurtion in order to rn it ginst other configurtions. A. Three phse Lod flow n three phse unblnced lod flow of distribution syste the following coponents re odeled by their euivlent circuits in ters of inductnce, cpcitnce, resistnce nd inected current. Conductors ndividul phse representtion for both priry nd secondry with cpcitive line chrging on priry conductor only. Trnsforers A generl pproch is recoended where by ll trnsforer connections, including the coon core trnsforer, re represented s individul trnsforers. Cpcitors Cpcitors re represented by their euivlent inected currents. Lods The unblnced lods re bsiclly considered becuse of single phse, two phse nd uneul three phse lods which exist in different types viz. constnt power, constnt pednce nd constnt current. Shunt dittnce nd series ipednce re represented by the ctul phse untities. B. Forwrd- Bcwrd Sweep Lod Flow Method There re severl power flow ethods bsed on forwrd/bcwrd sweep techniue. They y be clssified s power sution ethods nd current sution ethods. current sution ethod is used in this pper s it is ore convenient nd fst thn power sution ethod becuse it uses only nd insted of P nd Q. C. Bcwrd Sweep The purpose of the bcwrd sweep is to updte brnch currents in ech section, by considering the previous itertion voltges t ech node. During bcwrd propgtion voltge vlues re held constnt t the vlues obtined in the forwrd pth nd updted brnch currents re trnsitted bcwrd long the feeder using bcwrd pth. Bcwrd sweep strts fro extree end brnch nd proceeds long the forwrd pth. Fig. 1 shows phse of three-phse syste where lines between nodes p nd feed the node nd ll the other lines connecting node drw current fro line between node p nd. 755

3 nterntionl Journl of Electricl nd Electronics Engineering 3: p Fig.1. Single phse line section with lod connected t node between to phse nd neutrl n. The prent brnch current feeds the lod t the th node nd the sub-lterls connected to the prent brnch. This current cn be clculted using En. (6). Where Y sh itertion. = L + M + M Y sh 1 (6) is the hlf line shunt dittnce of the brnch in th is the brnch current vector in line section in th itertion. is the current vector in brnch before updting in th itertion. - 1 is the voltge vector of the brnch in (-1) th itertion. M represents the set of line sections connected to th brnch f cpcitor bn is plced t the receiving end of the brnch then cpcitor current should lso be included. D. Forwrd Sweep The purpose of the forwrd sweep is to clculte the voltges t ech node strting fro the source node. The source node voltge is set s 1.0 per unit nd other node voltges re clculted s = p + Ze Y sh p (7) Where, p re the voltge vectors of phses for p th nd th nodes respectively in th itertion L S C Ze = b c b bb cb c bc cc is the current vector in th brnch in th itertion These clcultions will be crried out till the voltge t ech bus is within the specified liits. At this point the voltges t ech node, nd the currents flowing in ll line segents re nown, which re used to find the power losses in ech line segent. Therefore the rel nd rective power losses in the line between nodes p nd y be written s: ( )( ) S = (8) p Where S is vector of power loss with three, two or nd p re voltge vector of three single phse phses t nodes p nd is the brnch current vector of three phses for the section connected in between p th nd th nodes. SOLUTON METHODOLOGY Feeder reconfigurtion is perfored by opening/closing two types of switches; tie nd sectionlizing switches. A whole feeder or prt of feeder y be served fro nother feeder by closing tie switch lining the two while n pproprite sectionlizing switch ust be opened to intin rdil structure. n the context of loss reduction, the proble to be ddressed in this rticle is to identify tie nd sectionlizing switches tht should be closed nd opened, respectively, to chieve inii the syste rel losses. Genetic Algorith (GA) is used to obtin the solution for reconfigurtion of rdil distribution systes to inii the losses. The chnge in the losses cn esily be coputed fro the results of two lod flow studies siulting the syste configurtions before nd fter the feeder reconfigurtion.. MPLEMENTATON OF GENETC ALGORTHM A. Chroosoe Coding Strtegy Encoding strtegies re ethods, which effect genetic opertions especilly during crossover. Most of vilble GA pplictions in power systes re siply eploying GA nd no specil ttention hs been put on the chrcteristic of the solved proble, odeling nd solving techniue. n fct, the ost criticl proble in pplying GA is to find suitble encoding ethod for chroosoes in the proble doin. A good choice of chroosoe coding will e the serch esy by liiting the serch spce. A poor choice will result in 756

4 nterntionl Journl of Electricl nd Electronics Engineering 3: lrge serch spce nd divergence. There re two bsic principles for choosing GA coding: The first principle is to select encoding such tht the building blocs of the underlying designer sll nd reltively unrelted to building blocs t other positions. The second principle sttes tht the user should select the sllest lphbet tht perits n expnsion of the design so tht the nuber of exploitble schees is xiid. n the following section, by ing use of rdil feture of distribution systes, GA bsed lgorith is proposed. B. Shorten the Length of the Chroosoe To intin continuous power supply to ll the lods, the following set of rules to be dopted for selection of switches. Rule 1: All switches tht do not belong to ny loop re to be closed. Rule 2: All switches connected to the sources re to be closed. Rule3: Sectionlizing switches, those lie on priry side of the tie switches, re ten s opening options of the initil configurtion. Rule4: Selection of sectionlizing switches to be opened is done by ting ny one switch ong the vilble set of switches for prticulr tie switch closed. C. Genetic Opertions n this lgorith, Roulette wheel selection ethod is eployed. This ethod hs ore dvntges copred to rndo selection. Since here selection is ipleented s liner serch through Roulette wheel with slots weighted in proportion to string fitness vlues. Here chroosoes re rned in scending order ccording to their fitness vlues. The individul with best fitness re dopted the individul with best fitness re obtined for next genertion. n this proposed lgorith crossover opertions re defined s to exchnge between tepltes while uttion opertions re restricted within the teplte. D. Reproduction For selecting the prents to perfor crossover, Roulette wheel selection hs been dopted s it gives best results thn rndo selection. n this process fitness of ech string is clculted, sorted in scending order nd then ccording to their fitness, prents re selected to crossover. E. Crossover The unifor crossover process rndoly selects two prents to exchnge the bits of chroosoe with crossover rte P c. i.e. rndo nuber in the rnge [0 1] is generted. f tht vlue is less thn P c, s is generted with the length eul to string length. f bit of s is 1 prticulr bit is exchnged between two prents. Before crossover After crossover Ms F. Muttion Muttions re restricted in one string. Muttion is the occsionl rndo ltertion of vlue of string position i.e., chnging 1 to 0 nd vice vers. n the proposed ethod, rndo nuber between 0 nd 1 is generted if tht nuber is less thn P nd bits of prticulr string re ltered. Of course, loops ight shre brnches in coplicted power distribution syste nd tepltes y overlp, these overlps re eliinted in this lgorith. n this sitution, fesible solutions produced during GA opertion nd ute individul will be discrded. G. Fitness The fitness should be cpble of reflecting the obective nd directing the serch towrds optil solution. Lod flow clcultion is perfored to obtin totl ctive power loss. Reciprocl vlue of totl ctive power loss is proposed s fitness function to obtin optil ctive power loss. So xiiztion of fitness function of the GA cn be written s 1 Mxii F = (9) 1 + f Where = nb f P = 1 loss. LLUSTRATE EXAMPLES The perfornce of the proposed ethod is evluted on two test systes of unblnced rdil distribution syste to inii the power loss by networ reconfigurtion. The proposed ethod is illustrted with two test systes consisting of 19 node nd 25 node unblnced rdil distribution systes. A. Exple - 1 The 11, 19 node unblnced rdil distribution syste with bse reconfigurtion is shown in Fig. 2. Bse MA for 19 node unblnced rdil distribution syste is 1 MA. The line, lod nd tie switch dt re given in ppendix A1. The norlly opened switches re s19 nd s20. The GA control preters selected re length of chroosoe (12), popultion si (20), cross over probbility (0.8), nd uttion probbility (0.04). Fig. 2 Single line digr of 19 node URDS before networ reconfigurtion oltge profile nd sury of test results of 19 node URDS before nd fter networ reconfigurtion re given tbles 1 nd 2 respectively. Fro tbles 1 nd 2, it hs been observed tht the iniu voltge in phses A, B, C re iproved fro , nd to ,

5 nterntionl Journl of Electricl nd Electronics Engineering 3: nd p.u nd the ctive power loss in phses of A, B nd C re reduced fro 4.45, 4.45 nd 4.56 W to 2.83, 2.53 nd 2.86 W respectively. Hence, there is n iproveent in the iniu voltge nd reduction in power loss when copred with the before networ reconfigurtion nd fter networ reconfigurtion. TABLE 1 OLTAGE PROFLE OF 19 NODE URDS Before Reconfigurtion After Reconfigurtion Node No. b c b c Power loss Best Powerloss TABLE SUMMARY OF TEST RESULTS OF 19 NODE URDS Before Reconfigurtion After Reconfigurtion Description Phse Phse b Phse c Phse Phse b Phse c Open Switches s19, s20 s10, s11 Miniu oltge Mx. oltge regultion (%) proveent of Mx. oltge regultion (%) Totl Active Power Loss (W) Totl Active Power Loss reduction (%) Totl Rective Power Loss (Ar) Totl Rective Power Loss reduction (%) Totl Active Power Dend (W) Totl Relesed Dend (W) Totl Rective Power Dend (Ar) Totl Relesed Rective Power Dend (Ar) Totl Feeder Cpcity (A) Totl Relesed Feeder Cpcity (A) itertion no. Fig. 3 Totl Active power loss s Genertion nuber of 19 node URDS The norlly opened switches re s19 nd s20 before reconfigurtion nd fter reconfigurtion the open switches re s10 nd s11 for 19 node unblnced rdil distribution syste. Fro tble 2 it is observed tht the percentge reduction in ctive power loss fter reconfigurtion is 38.90%. The totl ctive power loss s genertion nuber for 19 node URDS is shown in Fig. 3. Reconfigured networ of 19 node unblnced rdil distribution syste is shown in the Fig. 4. B. Exple - 2 The 4.16, 25 node unblnced rdil distribution syste consists of three tie lines bse reconfigurtion is shown in Fig. 5. Bse MA for 25 node unblnced rdil distribution syste is 30 MA. The line, lod nd tie switch dt re given in ppendix A2. The norlly opened switches re s25, s26 nd s27. The GA control preters selected re length of chroosoe (12), popultion si (20), cross over probbility (0.8), nd uttion probbility (0.04). Fig. 4 Single line digr of 19 node URDS fter networ reconfigurtion 758

6 nterntionl Journl of Electricl nd Electronics Engineering 3: observed tht the percentge reduction in ctive power loss fter reconfigurtion is 11.08%. After the reconfigurtion of 25 node unblnced rdil distribution syste hs been shown in the Fig. 6. The totl ctive power loss s genertion nuber for 25 node URDS is shown in Fig. 7. Fig. 5 Single line digr of 25 node URDS before networ reconfigurtion Node No. TABLE OLTAGE PROFLE OF 25 NODE URDS Before Reconfigurtion After Reconfigurtion b c b c oltge profile nd sury of test results of 25 node URDS before nd fter networ reconfigurtion re given tbles 3 nd 4 respectively. Fro tbles 3 nd 4, it is observed tht the iniu voltge in phses A, B, C re iproved fro , nd to , nd nd the ctive power loss in phses of A, B nd C re reduced fro 52.82, nd W to 47.53, nd W respectively. The norlly opened switches s25, s26 nd s27 re for before configurtion nd open switches s22, s17 nd s15 re for fter reconfigurtion in the 25 node unblnced rdil distribution syste. Fro tble 4 it is Fig. 6 Single line digr of 25 node URDS fter networ reconfigurtion TABLE SUMMARY OF TEST RESULTS OF 25 NODE URDS Before Reconfigurtion After Reconfigurtion Description Phse Phse b Phse c Phse Phse b Phse c Open Switches s25, s26, s27 s15, s17, s22 Miniu oltge Mx. oltge regultion (%) proveent of Mx. oltge regultion (%) Totl Active Power Loss (W) Totl Active Power Loss reduction (%) Totl Rective Power Loss (Ar) Totl Rective Power Loss reduction (%) Totl Active Power Dend (W) Totl Relesed Dend (W) Totl Rective Power Dend (Ar) Totl Relesed Rective Power Dend (Ar) Totl Feeder Cpcity (A) Totl Relesed Feeder Cpcity (A)

7 nterntionl Journl of Electricl nd Electronics Engineering 3: Fig. 7 Totl Active Power loss s Genertion nuber of 25 node URDS TABLE COMPARSON OF TEST RESULTS FOR 19 AND 25 NODE URDS WTH THE EXSTNG METHODS 19 node URDS 25 node URDS Methods No of switching opertion Loss reduction (W) No of switching opertion Loss reduction (W) Shirohdi nd Hong ethod [4] Goswi nd Bsu ethod [5] Proposed Method Coprison of test results of the proposed ethod for 19 node nd 25 node URDS with the excising ethods re given in tble 5. The proposed ethod hs been copred with exciting ethods of Shirohdi nd Hong s ethod [4] nd Goswi nd Bsu s ethod [5] nd results re presented in Tble 5. Fro tble 3.10 the effectiveness of the proposed ethod copred with nuber of switching opertion nd loss reduction with two existing ethods. The totl ctive power loss reduction hs been iproved fter networ reconfigurtion. The nuber of switching opertions ws less for the proposed ethod when copred with the existing ethods. Hence the proposed ethod is superior when copred with the existing ethods.. CONCLUSON The results deonstrte the effectiveness of the proposed lgorith. n ddition to power loss reduction, voltge profile is lso iproved by the proposed ethod. The obective function of this pper hs been shown tht genetic lgorith cn be used to find the configurtion of the unblnced distribution networs which iniis overll power losses of the distribution syste. t hs been found tht the proposed ethod switching options re lesser thn the existing ethods. The effectiveness of proposed ethod hs been tested on two unblnced rdil distribution systes. brnch Sending End APPENDX TABLE A1 LOAD DATA AND LNE CONNECTTY OF 19 NODE UNBALANCED RADAL DSTRBUTON SYSTEM Receiving end lod in A Receiving End Conductor type Length, Phse A Phse B Phse C Type 1 b pednce in ohs/ b c c

8 nterntionl Journl of Electricl nd Electronics Engineering 3: brnch TABLE A2 LOAD DATA AND LNE CONNECTTY OF 25 NODE UNBALANCED RADAL DSTRBUTON SYSTEM Receiving end lod in A Sending End Receiving End Conductor type Length, ft Phse A Phse B Phse C ` Type 1 b 2 b 3 b pednce in ohs/ile b c c c c REFERENCES [1] Asin Developent Bn, Power systes efficiency through loss reduction nd lod ngeent, Proceedings of the Regionl Conference on Power Systes Efficiency through Loss Reduction nd Lod Mngeent, [2] S. Civnlr, J.J. Gringer, H. Yin nd S.h. Lee Distribution feeder reconfigurtion for loss reduction, EEE Trns. Power Deliv. 3, (3), pp ,1988. [3] M.E. Brn, nd F.F. Wu Networ reconfigurtion in distribution for loss reduction nd lod blncing, EEE Trns. Power Syst., 4, (3), pp , [4] D. Shirohdi nd H.W. Hong Reconfigurtion of electric distribution networs for resistive line loss reduction, EEE Trns. Power Deliv., 4, (2), pp ,1989. [5] S.K. Goswi nd S.K. Bsu new lgorith for the reconfigurtion of distribution feeders for loss iniiztion, EEE Trns. Power Deliv., 7, (3), pp , [6] T. Tylor nd D. Luben pleenttion of heuristic strtegies for distribution feeder reconfigurtion, EEE Trns. Power Deliv., 5, (1), pp , [7]. Gloocnin Optil loss reduction of distribution networs, EEE Trns. Power Syst., 5, (3), pp , [8] T.P. Wgner, A.Y. Chihni nd R.Hc Feeder reconfigurtion for loss reduction, EEE Trns. Power Deliv., 6, (4), pp , [9]. Borozn, D. Ricic nd R. Acovsi proved ethod for loss iniiztion in distribution networs, EEE Trns. Power Syst., 10, (3), pp , [10] R.J. Srfi, M.M.A. Sl nd A.Y. Chihni A survey of the stte of the rt in distribution syste reconfigurtion for syste loss reduction, Electr. Power Syst. Res., 31, pp , [11] R.J. Srfi, M.M.A. Sl nd A.Y. Chihni Distribution syste reconfigurtion for loss reduction: n lgorith bsed on networ prtitioning theory, EEE Trns. Power Syst., 11, (1), pp , [12]. Royteln,. Melni, S.S. H.Lee nd R.L. Lugtu Multi-obective feeder reconfigurtion by distribution syste ngeent syste, EEE Trns. Power Syst., 11, (2), pp , [13] D. Jing nd R. Bldic Optil electric distribution syste reconfigurtion nd cpcitor control, EEE Trns. Power Syst., 11, (2), pp , [14] G.J. Peponis, M.P. Ppdopoulos nd N.D. Htzirgyriou Optil opertion of distribution networs, EEE Trns. Power Syst., 11, (1), pp , [15] T.E. McDerott,. Drezg nd R.P. Brod Wter A heuristic nonliner constructive ethod for distribution syste reconfigurtion, EEE Trns. Power Syst., 14, (2), pp ,

9 nterntionl Journl of Electricl nd Electronics Engineering 3: [16] C.S. Chen nd M.Y. Cho Energy loss reduction by criticl switches, EEE Trns. Power Deliv., 8, (3), pp ,1993. [17] Q. Zhou, D. Shirohdi nd W.H.E Liu Distribution feeder reconfigurtion for opertion cost reduction, EEE Trns. Power Syst., 12, (2), pp , [18] H.P. Schidt, N. d, N. Kgn nd J.C. Gurldo Fst reconfigurtion of distribution systes considering loss iniiztion, EEE Trns. Power Syst., 20, (3), pp [19] R. Tlesi, nd D. Ricic Distribution networ reconfigurtion for energy loss reduction, EEE Trns. Power Syst., 12, (1), pp , [20] H.D. Ching nd R.M. Jen-Jeu Optil networ reconfigurtion in distribution systes, Prt 1: new forultion nd solution ethodology, EEE Trns. Power Deliv., 5, (4), pp , [21] H.D. Ching nd R.M. Jen-Jeu Optil networ reconfigurtion in distribution systes, Prt 2: solution lgoriths nd nuericl results, EEE Trns. Power Deliv., 5, (3), pp , [22] H.C. Cheng nd C.C. Kuo Networ reconfigurtion in distribution systes using siulted nneling, Electr. Power Syst. Res., 29, pp , [23] Y.J. Jeon, J.C. Ki, J.O. Ki, J.R. Shin nd K.Y. Lee An efficient siulted nneling lgorith for networ reconfigurtion in lrge-scle distribution systes, EEE Trns. Power Deliv., 17, (4), pp , [24] A.B. Morton nd.m. Mreels An efficient brute-force solution to the networ reconfigurtion proble, EEE Trns. Power Syst., 15, (3), pp , [25] L. Liu nd X.Y. Cheng, Reconfigurtion of Distribution networs bsed on fuzzy Genetic lgoriths, Proceedings of the CSEE,,, vol. 20, no. 2, pp , Februry [26] K. Nr, A. Shiose, M. Kitgw nd T. shihr pleenttion of genetic lgorith for distribution syste loss iniu reconfigurtion, EEE Trns. Power Syst., 7, (3), pp , [27] J.Z. Zhu Optil reconfigurtion of electricl distribution networ using the refined genetic lgorith, Electr. Power Syst. Res., 62, pp , [28] E. Lopez, H. Opzo, L. Grci nd P. Bstrd Online reconfigurtion considering vribility dend: pplictions to rel networs, EEE Trns. Power Syst., 19, (1), pp , [29] Y.Y. Hong nd S.Y. Ho Deterintion of networ configurtion considering ulti-obective in distribution systes using genetic lgoriths, EEE Trns. Power Syst., 20, (2), pp , [30] J.P.Chiou, C.F. Chung nd C.T.Su rible scling hybrid differentil evolution for solving networ reconfigurtion of distribution systes, EEE Trns. Power Syst., 20, (2), pp , [31] J. Mendoz, R.Lopez, D. Morles, E. Lopes, P. Dessnte nd R. Morg Minil loss reconfigurtion using genetic lgoriths with restricted popultion nd ddressed opertors: rel ppliction, EEE Trns. Power Syst., 21, (2), pp , [32] Y.T. Hsio Multi-obective evolution progring ethod for feeder reconfigurtion, EEE Trns. Power Syst., 19, (1), pp , [33]. Borozn, D. Ricic nd R. Acovsi Miniu loss reconfigurtion of unblnced distribution networs, EEE Trns. Power Deliv., 12, (1), pp , [34] J.C. Wng, H.D. Ching nd G.R. Drling An efficient lgorith for rel tie networ reconfigurtion in lrge scle unblnced distribution systes, EEE Trns. Power Syst., 11, (1), pp , [35] D.E. Goldberg, Genetic Algoriths in Serch, Optiiztion nd Mchine Lerning, Boston: Addison Wesley, [36] W.H. Kersting, Distribution Syste Modeling nd Anlysis, CRC press, Gnesh ulsl received the M. Tech. in Power Syste Opertion nd Control fro S University, Tirupti, Andhr Prdesh, ndi in Currently, he is Reserch Scholr in Electricl Engineering Deprtent of JNT University, Anntpur, ndi. He is n Assistnt Professor in the Deprtent of Electricl nd Electronics Engineering, JNTU College of Engineering, Anntpur, Andhr Prdesh, ndi. His res of reserch include Power Syste, Distribution Systes, Optiiztion Techniues, GA ppliction to Power Distribution Syste. Sivngru Sirigiri received the M. Tech. in Power Syste fro T, Krghpur, West Bengl, ndi in He received the Ph.D. in Electricl nd Electronics Engineering fro JNT University, Hyderbd in He is n Associte Professor in Electricl nd Electronics Engineering, JNTU College of Engineering, Kind. His res of reserch include Power Syste, FACTS, Distribution Systes, Optiiztion Techniues, GA pplictions to Power Systes nd FACTS. Rn Thiruveedul is pursuing the M. Tech. in Power nd ndustril Drives fro JNT University, Kind, Andhr Prdesh, ndi. He is n Assistnt Professor in the Deprtent of Electricl nd Electronics Engineering, RSYRN College of Engineering nd Technology, Chirl, Andhr Prdesh, ndi. His res of reserch include Power Syste, FACTS, Distribution Systes, Optiiztion Techniues, GA pplictions to Power Systes nd FACTS. 762