Development of an Intelligent Controller for Vehicle to Grid (V2G) System

Transcription

1 Dvlopmnt of an Intllignt Controllr for Vhicl to Grid (V2G) Systm [1] Hriday Ranjan, [2] Dr.Sindhu M R [1] Dpartmnt of Elctronics and Communication Einri,Amrita School of Einri, Coimbator, Amrita Vishwa Vidyaptham, Amrita Univrsity, India [2] Dpartmnt of Elctrical and Elctronics Einri, Amrita School of Einri, Coimbator, Amrita Vishwa Vidyaptham, Amrita Univrsity, India Abstract: Elctric Vhicl (EV) is gaini popularity du to growi concrn on pollution. But a larg numbr of EV can hav an advrs ffct on powr systm as it acts as a hug load on th grid. Th solution lis in havi a co-ordinat systm of chargi and dischargi which can b ffctivly usd to support grid. Hr a controllr has bn proposd which analyzs currnt grid status, EV battry stat of charg (SOC), lctricity tariff offrd by powr utility company alo with EV ownr prfrnc. An analysis illustrats how th V2G fasibility is dtrmind by th chargi station controllr and thn basd on grid status and EV battry SOC, chargi and dischargi rat of currnt is prdictd by V2G controllr. Th proposd control algorithm has bn simulatd usi MATLAB SIMULINK. Indx Trms: Elctric Vhicl, V2G,Fuzzy Logic Controllr I. INTRODUCTION Global warmi has attractd th attntion of whol world du to its advrs impact on Earth s climat. Tmpratur ris du to global warmi can b attributd to burni of fossil ful which producs carbon dioxid and othr grnhous gass. Growi concrn on climat cha has ld to vry strint mission norms across automotiv industris. Hnc at prsnt mphasis is on th nd to hav a clan sourc of nrgy with minimum pollution. It has contributd to incras in popularity of Elctric Vhicl (EV).EV uss rchargabl battry, lctric motor and controllr for propulsion systm. EV battry can b chargd or dischargd usi off -board or on-board chargr. Thr ar diffrnt typs of battris usd in th EV, but Lithium ion battris ar prfrrd ovr othr battris du to high rfficincy, gratr nrgy, bttr cycl lif, highr cll voltag. Highr cll voltag of Lithium ion battris rducs th numbr of clls, which in turn rducs th siz of battry. Du to rapid advancmnt in battry tchnology now it is bi manufacturd typically in th ra of 15 to 85 kwh. Whn EV is connctd to grid it acts as a hug load on grid. But on th othr sid EV battry du to its larg nrgy storag capacity can also act as an altrnativ sourc of nrgy which can b usd duri pak dmand to support grid. Thus thr is a nd to dvlop an intllignt controllr for vhicl to Grid (V2G) systm which taks car of currnt status of EV and grid. In this papr Sction-II dscribs V2G systm, Sction-III provids th dtail xplanation of proposd controllr, followd by Simulation Rsult & Prformanc Analysis in Sction IV and Conclusion and Futur Scop in sction-v. II. VEHICLE TO GRID SYSTEM Intrconnction of EV and grid is calld as V2G systm. It supports bi-dirctional powr transfr i.. from vhicl-to-grid and grid-to-vhicl [1]. Fig.1 shows th typical diagram of V2G systm which has bn xplaind by W. Kmpton and J. Tomic [2]. Fig.1. Vhicl to Grid Systm (V2G) 224

2 In consists of grid havi rnwabl (thrmal powr) and non-rnwabl (wind powr) sourcs of nrgy. EV whn connctd to th grid acts as a hug load to th grid which can hav an advrs impact on th powr quality. Govrnmnt of India has announcd to incras th shar of rnwabl nrgy sourcs by incrasi th currntly installd capacity of Solar and Wind nrgy. But it will also caus uncrtaintis du to intrmittncy in th gnration of th rnwabl nrgy. With incras in EV pntration in India V2G can act as a support to grid for mitigation th powr quality issus [3].It can b don by usi co-ordinat chargi Typ Lvl- I Lvl- II Lvl- III mthodology. In this mthodology chargi and dischargi rat of currnt varis basd on th currnt status of EV battry and grid. For xampl, lctric vhicl Nissan Laf has capacity of 30 kwh which is ngligibl whn compard to grid fluctuation (in MW) [14]. But flt of EV can b usd to rgulat th grid both duri pak hour and off hour. Duri pak hour whn dmand is mor and gnration is lss, it can actas an altrnativ sourc of nrgy to mt th dmand by dischargi th stord nrgy in EV battry to grid. Similarly duri off hour whn gnration is mor and load is lss on grid EV can act as load by chargi th battry by taki nrgy from grid [4]-[9]. As pr SAE J1772 standard EV chargr has bn classifid into thr catgoris, Lvl 1, Lvl 2, and Lvl 3 for both ac and dc supply[10], [11], [13]. Summary of diffrnt powr lvl ar givn in Tabl I. Tabl I EV Chargr Classification Voltag Lvl 120 Vac(US) 230 Vac(EU) 240 Vac(US) 400 Vac(EU) Vac or Vdc For optimizd chargi and dischargi currnt rat svral control algorithms hav bn dscribd in th litratur. Grid status has bn monitord ithr by masuri voltag or frquncy and EV battry currnt status is masurd usi various SOC stimation Chargr Location On board 1- Phas On board 1 or 3 Phas Off board 3- Phas Expctd Powr Lvl mthods [15]-[17]. Basd on th abov rsarch findis, it can b concludd that vhicl to grid chargi and dischargi & its impact on grid status hav bn studid. But thr is nd to do th fasibility analysis bfor conncti th EV to grid. Hr an intllignt controllr is prsntd which first analyzs th fasibility of EV connction to grid basd on currnt grid status, EV battry SOC, currnt lctricity tariff offrd by powr utility company and EV ownr drivi nds & prfrrd pric for chargi or dischargi. If chargi station controllr output is fasibl thn only fuzzy basd V2G controllr prdicts th optimizd chargi or dischargi currnt rat basd on currnt grid status and EV battry SOC. III. CHARGING STATION AND V2G CONTROLLER This sction provids th worki principl of th proposd controllr and its simulation in MATLAB Simulink. It consists of a chargi station controllr and a fuzzy basd V2G controllr. A. Chargi Station Controllr Chargi Tim 1.4kW(12A) 4-11 hour 4kW(17A) 8kW(32A) 19.2kW(80A) 50 kw 100 kw 1-4 hour 2-6 hour 2-3 hour 0.4-1hour hour Vhicl Typ PHEV(5-15 kwh) EV(16-50 kwh) PHEV(5-15 kwh) EV(16-30 kwh) EV(3-50 kwh) EV(20-50 kwh) Chargi station controllr is acti as an intrfac btwn EV ownr, Aggrgator and Powr Utility Company as shown in Fig.2. Hr Aggrgator can b thought as a plac which provids chargi and dischargi facility to many EV simultanously. Powr Utility Company facilitats th Aggrgator prsnt 225

3 chargi and dischargi tariff as pr th agrmnt signd amo thm. Chargi and dischargi pric offrd by Powr Utility Company dynamically chas with grid status. For Indian scnario tariff with grid status has bn rsarchd by BhanuBhusan in th papr ABC of ABT: A PRIMER ON AVAILABILITY TARIFF in grat dtail [6]. Chargi and dischargi pric pr unit offrd by Powr utility company with chai grid status has bn mulatd in our projct basd on this papr. Whn grid is in powr surplus condition i.. chargi point voltag is qual to or mor than 230 V and EV ownr prici critria is matchi with prsnt chargi pric offrd by Powr Utility company thn, powr flow will b in G2V mod (grid to vhicl mod). 3) Dischargi (V2G mod) Whn grid is in powr dficint condition i.. chargi point voltag is lss than 230 V and EV ownr prici critria is matchi with prsnt dischargi pric offrd by Powr Utility Company thn, powr flow will b in V2G mod (grid to vhicl mod). Hr Vhicl will stop supplyi powr to Grid onc it rachs SOC limit as prfrrd by th EV ownr. Fig.2. MATLAB Simulink modl of proposd unifid controllr Whn EV ownr parks his vhicl in th Aggrgator for V2G support, EV ownr can spcify th prfrnc. Paramtrs spcifid by EV ownr ar maximum chargi pric(pc) EV ownr is willi to pay for chargi his vhicl, minimum dischargi pric (Pd) EV ownr would lik to hav in cas of V2G mod. Battry SOC limit (lsoc) EV ownr would lik to maintain taki car of futur drivi nds i.. blow which battry cannot b dischargd. Control logic of Chargi Station Controllr has bn xplaind in dtail with th hlp of flow chart sown in Fig.3. Basd on th input from EV ownr and chargi point voltag (grid status) Chargi Station Controllr valuats th V2G support fasibility by compari it with prsnt offrd pric by th Powr utility company for chargi or dischargi. Thr can b thr possibl outputs from Chargi Station Controllr. 1) V2G support is not fasibl If EV ownr prfrnc is not matchi with prsnt grid status thn Chargi Station Controllr will notify th EV ownr to wait as th V2G critria ar not matchi. Mismatch in th critria may happn du to maximum dischargi pric(pd) st by EV ownr is highr than th prsnt offrd pric (Pud) by Powr utility company for dischargi. 2) Chargi (G2V mod) Fig.3. Flow-chart of chargi station controllr B. V2G Controllr V2G controllr works on th principl of fuzzy logic which is usd to calculat th optimizd chargi or dischargi rat basd on EV battry SOC and grid status as shown in Fig-2. Output of chargi station controllr dcids V2G or G2V support is fasibl or not. If it is fasibl thn only fuzzy logic basd V2G controllr prdicts th optimizd chargi or dischargi rat of currnt basd on grid status (chargi point voltag) and currnt SOC of battry. Hr positiv sign of V2G controllr output indicats dischargi i.. V2G mod & ngativ sign indicats chargi i.. G2V mod. Actual powr flow btwn vhicl and grid is dcidd by th magnitud of th output of fuzzy basd V2G controllr i.. on th basis of optimizd chargi or dischargi rat. So hr chargi/dischargi rat of th battry will b dynamically chai basd on battry SOC and grid status i.. chargi point voltag. 226

4 1) Introduction to fuzzy logic control Concpt of fuzzy logic was givn by Lotfi A Zadh in In binary logic w hav two valus i.. ithr TRUE/FALSE or 0/1 but in fuzzy logic w hav multi valu. Fuzzy logic variabls ar xprssd on th scal of 0 to 1 dpndi on th irtruthnss. It works on th basis of simpl IF-Thn rul bas and thus abolishi th nd of complx mathmatical modli to implmnt th givn systm bhavior. It is widly usd for thos systms whos mathmatical modl is vry complicatd and thus tough to implmnt. Mamdani typ fuzzy logic controllr has bn dsignd in this papr. A typical architctur of fuzzy logic control systm is shown in Fig.4.Thr ar four major moduls in th fuzzy control systmfuzzification, infrnc logic, rul st and dfuzzification modul. Fuzzification and dfuzzification moduls ar complimntary in natur. First on transforms th crisp valu into fuzzy variabls and scond on transforms th fuzzy variabls back to crisp valus. It is obtaind by mans of a mmbrship function. Rul is a st of IF-THEN statmnts which implmnts th control algorithm. Th output from ach rul in th rul bas is dducd by th infrnc logic to arriv at a valu for ach output mmbrship function. Th cntr of gravity (COG) mthod is usd for dfuzzification which givs a crisp output. Fig.5. SOC mmbrship function Chargi point voltag whr EV is connctd to th aggrgator has bn first convrtd into pr unit systm. Pr unit systm is dfind as th ratio of actual valu of any quantity dividd by th bas valu of th masurd quantity. Q pu = Q Q bas (1) Whr Q pu is th pr unit quantity (dimnsionlss) Qis quantity in normal units Q bas is th bas valu of th quantity in normal unit In this papr 230 V has bn takn as th bas voltag and hnc in pr unit systm it is rprsntd as 1 pu. Similarly pr unit valu of 250 V is pu. Hr voltag ra has bn takn from 0.7 to 1.2 in pr unit systm.mmbrship function of chargi point voltag which shows th grid condition has bn fuzzifid into 5 stags namly Grid Vry Low (GVL), Grid Low (GL), Grid Mdium (GM), Grid High (GH) and Grid Vry High (GVH). Fig- 6 shows th mmbrship function of chargi point voltag alo with its fiv stags. Fig.4. Fuzzy logic V2G control systm 2) Fuzzy mmbrship function Dfini th mmbrship function is critical for fuzzification. Triaular mmbrship function is mostly usd du simplicity in calculation for acadmic purpos. In this papr mmbrship function of battry SOC has bn fuzzifid into 5 stags namly Battry Vry Low (BVL), Battry Low (BL), Battry Mdium (BM), Battry High (BH) and Battry Vry High (BVH). Fig-5 shows th mmbrship function of SOC alo with its fiv stags. Fig.6. Chargi point voltag mmbrship function Output mmbrship function CRATE can hav positiv or ngativ valu. Positiv valu mans powr flow is from Vhicl to Grid (V2G) i.. battry is dischargi. And ngativ valu mans EV battry is chargi i.. powr flow is from Grid to Voltag (G2V). It has bn fuzzifid into 9 stags basd on CRATE namly Ngativ Highcurrnt (cnh), Ngativ Mdiumcurrnt (cnm), Ngativ Low 227

5 currnt(cnl), Vry Ngativ Lowcurrnt (cvnl),vry Positiv Lowcurrnt(cVPL), Positiv Lowcurrnt(cPL), Positiv Mdiumcurrnt(cPM), Positiv Highcurrnt(cPH) and Vry Positiv Highcurrnt(cVPH). Fig.7 shows th mmbrship function of CRATE alo with its nin stags. Fig.7. CRATE mmbrship function 3) Fuzzy control rul st A fuzzy control rul st is formd on th basis of human xprtis in th actual application. It is a st of form of IF-THEN ruls which dtrmins th systm output for a givn st of input. Hr ruls can b xprssd in liuistic way. In this papr for our application rul is basd on followi infrnc: Tabl II Ruls of V2G controllr SOC ( ) BVL BL BM BH BVH Voltag( ) GVL cvpl cpl cpm cph cvp H GL cvpl cpl cpm cph cph GM cvn cvp cpl cp cpm L L M GH cnh cnh cn cnl cvnl M GVH cnh cnh cn M cnl cvnl a) If grid is in high nd of powr (GVL) and EV battry SOC is maximum (BVH) thn, dischargi rat (+v) should b maximum (cvph). b) If grid is in high surplus powr (GVH) and EV battry SOC is minimum (BVL) thn, chargi rat (-v) should b maximum (cnh). Fig.8. CRATE output aftr dfuzzification 4) Dfuzzification Most popular dfuzzification tchniqu usd is cntr of gravity (COG) typ. By applyi COG mthod w can convrts fuzzy systm output into crisp output i.. dfuzzifid output.in this papr all mmbrship function has triaular shap and now if trial is partitiond into two parts by passi a straight lin horizontally dpndi on mmbrship valu thn bas portion of trial will hav trapzoidal shap. Each rul follows th abov procss and corrspondi to ths trapzoids shaps ar formd. Suprimposi of all ths trapzoids form a gomtrical shaps whos cntroid is calculatd. Co-ordinat of th cntroid is th dfuzzifid valu.blow is th sampl output for chargi point voltag as 0.8 p.u. & SOC = 90. Fig.8 shows th calculatd valu of CRATE in rul viwr intrfac of MatlabSimulik. IV. SIMULATION RESULT & PERFORMANCE ANALYSIS Proposd Controllr has bn simulatd in Matlab Simulink. Tabl III shows th various input and output paramtrs of chargi station controllr and fuzzy basd V2Gcontrollr which mulat th ral tim grid scnario for pak hour, off hour and balancd condition. 228

6 Tabl III Controllr input and output paramtrs Ch argi Poi nt Vol Cu rr nt S O C Ba tt ry S O C Ch argi pric ow Ch argi pric utili Disc hargi pric own r Disc hargi pric utilit y Cr at (A m p) tag lim it nr (Pc) ty (Puc ) (Pd) (Pud) Fig.9. Input paramtrs of Chargi Station Controllr and V2G Controllr Chargi station controllr output has two possibilitis. It can b ithr fasibl for which on valu has bn shown or not fasibl for which zro valu has bn shown in Fig.10. Basd on chargi station controllr output V2G fuzzy basd controllr can hav thr diffrnt output valus. 1) Zro Whn Chargi Station Controllr output is not fasibl i.. zro thn fuzzy basd V2G controllr output is also zro. 2) Positiv numric valu Positiv numric valu of V2G controllr output mans EV is oprati in V2G mod i.. dischargi mod. 3) Ngativ numric valu Ngativ numric valu of V2G controllr output mans EV is oprati in G2V mod i.. chargi mod. Considri all possibl scnario of Chargi station Controllr and V2G controllr, hr four diffrnt cass hav bn simulatd. Input and output paramtrs of chargi station and V2G controllr ar shown in Fig.9 & Fig.10. Blow is th summary of four diffrnt cass: Cas-I: Whn Chargi point voltag is 235 V i.. grid is in powr surplus condition (G2V mod). In this scnario chargi station controllr output is fasibl i.. at t=0 to 0.2 scond and optimizd chargi currnt rat prdictd by V2G controllr is -50 amp. Fig.10. Chargi Station and V2G Controllr Output Cas-II: Whn Chargi point voltag is 220 V i.. grid is in powr dficint condition (V2G mod). In this scnario chargi station controllr output is not fasibl i.. at t=0.1 to 0.2 scond bcaus hr minimum dischargi pric st by EV ownr is Rs 6 and currnt dischargi pric offrd by powr utility company is Rs 5 only which violats th prdfind condition of chargi station controllr. Hr V2G controllr output will also b zro. Cas-III: Whn Chargi point voltag is 180 V i.. grid is in powr dficint condition (V2G mod). In this scnario chargi station controllr output is fasibl i.. at t=0.2 to 0.3 and optimizd dischargi currnt rat prdictd by V2G controllr is 70 amp. Cas-IV: Whn Chargi point voltag is 200 V i.. grid is in powr dficint condition (V2G mod). In this scnario chargi station controllr output is not fasibl i.. at t=0.3 to 0.4 scond bcaus hr minimum lvl of SOC of EV battry prfrrd by EV ownr is 51and currnt SOC of battry is 40 which violats th prdfind condition of chargi 229

7 station controllr. Hr V2G controllr output will also b zro. V. CONCLUSION AND FUTURE WORK Proposd controllr was simulatd usi MATLAB Simulink undr diffrnt condition. Chargi station controllr has bn implmntd to analyz th fasibility of V2G or G2V support for a givn EV basd on currnt grid condition. If fasibl thn only fuzzy basd V2G controllr is usd to prdict th optimizd chargi or dischargi rat of currnt basd on grid status and EV battry SOC. It is usful in balanci th uncrtaintis introducd by th intrmittncy of th rnwabl nrgy which in turn incrass th rliability and stability of powr systm. It also givs an conomic bnfit to th EV ownr as thy can discharg th stord nrgy in EV battry back to grid (V2G) duri pak dmand as lctricity tariff will b high & can charg th EV battry duri off hour (G2V)whn lctricity tariff is low. In futur proposd controllr in this papr can b intgratd with a bidirctional EV chargr to study th ral tim bhavior of an individual EV and its intraction with grid to assss th impact on powr systm. Optimum chargi/dischargi currnt prdictd by th controllr can b takn as a rfrnc to th EV chargr in ordr to minimiz th advrs impact on grid and thus improvi th powr quality. REFERENCES [1] C. Guill and G. Gross, A concptual framwork for th vhicl-to-grid (V2G) implmntation, Enrgy Policy, vol. 37, no. 11, pp , [2] W. Kmpton and J. Tomić, Vhicl-to-grid powr implmntation: From stabilizi th grid to supporti larg-scal rnwabl nrgy, Journal of Powr Sourcs, vol. 144, no. 1, pp , [3] A. Kumar, M.Anamol and V. Akhil, A stratgy to Enhanc Elctric Vhicl Pntration Lvl in India, Procdia Tchnology, vol. 21, no. 1, pp , Aug [4] E. Sortomm, M.Hindi, S. MacPhrson, and S. Vnkata, Coordinatd chargi of plug-in hybrid lctric vhicls to minimiz distribution systm losss, IEEE Trans. Smart Grid, vol. 2, no. 1, pp , Mar [5] K. Clmnt-Nyns, E. Hasn, and J. Drisn, Th impact of chargi plug-in hybrid lctric vhicls on a rsidntial distribution grid, Trans. Powr Syst., vol. 25, no. 1, pp , Fb [6] B. Bhusan, A Primr on availability tariff, ABC of ABT, vol., No., pp. 1-36,2005 [7] D. Xu, G. Joos, M. Lvsqu t al., Intgratd V2G, G2V, and rnwabl nrgy sourcs coordination ovr a convrgd fibr-wirlss broadband accss ntwork, IEEE Transactions on Smart Grid, vol. 4, No.3, pp ,2013 [8] K.C.Nyns, E.Hasn,andJ.Drisn, Th impact of chargi plug-in hybrid lctric vhicls on a rsidntial distribution grid, IEEE Trans. Powr Syst., vol. 25, no. 1, pp , [9] Z.Zha,K.Chau, Puls-Width- ModulationBasdElctromagn tic IntrfrncMitigation of Bidirctional Gridconnctd Convrtrs for Elctric Vhicls, IEEE Transactions on Smart Grid, vol. 3053, no. c, pp ,2016 [10] M. Yilmaz and Philip T. Krin, Rviw of chargi powr lvls and infrastructur for plug-in lctric and hybrid vhicls, IEE Transactions on Powr Elctronics, vol. 28, no. 5, pp , [11] N. Wo and M. Kazrani, A rviw of bidirctional on-board chargr topologis for plugin vhicls, IEEE Canadian Confrnc on Elctrical and Computr Einri: Vision for a Grnr Futur, CCECE, vol., no., pp.0-5, [12] Z. Wa, S. Wa, "Grid powr pak shavi and vally filli usi vhicl-to-grid systms, IEEE Transactions on Powr Dlivry, vol. 28, No.3, pp ,2013 [13] SAE Intrnational, Chargi configurations and ratis trminology [Onlin].Availabl: g/smartgrid/chargispds.pdf [14] NissanLaf[Onlin].Availablhttps:// usa.com/lctric-cars/laf/vrsions-spcs/ [15] M. Datta, Fuzzy logic basd frquncy control by V2G aggrgators, IEEE 5th Intrnational Symposium on Powr Elctronics for Distributd Gnration Systms (PEDG), vol., No. 1, pp. 1-8,2014 [16] D. Karun, T. Sindhu, Fuzzy logic basd load frquncy control of grid connctd distributd gnrators, IEEE Intrnational Confrnc on Tchnological Advancmnts in Powr & Enrgy, vol., No., pp ,2015 [17] M. Sih, P. Kumar, I. Kar, Implmntation of Vhicl to Grid Infrastructur Usi Fuzzy Logic Controllr, IEEE Transactions on Smart Grid, vol. 3, No. 1, pp ,