Optimal Energy Management Algorithm for Plug in Hybrid Electric Vehicles

Transcription

1 Optmal Energy Management Algorthm for Plug n Hybrd Electrc Vehcles Dr.K.Lakshm 1, M.Kaleeswar 2 Professor, Dept. of EEE, K.S.Rangasamy College of Technology, Truchengode, Tamlnadu, Inda 1 PG Student [Power Systems Engneerng], Dept. of EEE, K.S.Rangasamy College of Technology, Truchengode, Tamlnadu, Inda 2 ABSTRACT: Plug n Hybrd Electrc Vehcles (PHEVs) chargng and dschargng, renewable energy resource generaton and utlzaton s most mportant n future power system control. Proper ntegraton of these energy sources gves soluton to the challenges. In ths paper Mxed Integer Lnear Programmng (MILP) was proposed for plug n Hybrd Electrc Vehcles chargng and dschargng n a chargng staton. Chargng staton conssts of Photo Voltac (PV) system wth practcal constrants, power balance constrants, battery chargng and dschargng constrants are consdered. The results proposed that the proposed algorthm mnmze the chargng cost of PHEVs and optmal power flow for the grd connected PHEVs systems. Lkewse, PHEVs owners could yeld more proft by dschargng ther vehcles to the grd n addton to havng preferred charge n the departure tme. KEYWORDS: Mxed Integer Lnear Programmng, Plug n Hybrd Electrc Vehcles, Chargng Park, Voltage Stablty. I.INTRODUCTION The mportance of energy savngs s ncreasng and governments are encouragng the use of renewable energy. A clear correlaton can be observed between vehcle densty (cars per 1000 nhabtants) and a country s GDP (Gross Domestc Product); ths suggests that as densely populated countres such as Chna, Inda, and Brazl acheve hgher economc status, t can be expected that the demand for personal transportaton wll ncrease accordngly. Today, ths demand can be drectly translated nto ncreased demand for petroleum, a fact that s hardly compatble wth current data on ol producton. Plug-In Electrc Vehcles are recevng a great deal of nterest n the Unted States due to ther energy effcency, convenent and low-cost rechargng capabltes and reduced use of petroleum. PHEVs are an mportant of an electrc power system n the upcomng days. PHEVs gve the soluton to the fossl fuel shortage and ar polluton problems [1].The emsson reducton s acheved by usng the PHEVs wth combnaton of renewable energy resources. Beyond these advantages, the power system may face sgnfcant challenges due to the huge electrcty demand of these loads [2-4]. PHEVs use battery as an energy storage system by usng ths battery PHEVs supply the power to the electrc drve motor. Whenever PHEVs connected to chargng staton t wll be operated n two modes such as Vehcle to Grd (V2G) mode and Grd to Vehcle (G2V) mode [5-6]. In V2G mode the state of charge of battery can go up or down, depends on the power demand. PHEVs owner get proft by usng V2G capabltes. The desgn of PHEVs energy storage s manly for transportaton sector. So t provdes suffcent energy to drve the vehcle. In order to maxmze the customer satsfacton and mnmze the grd dsturbance, PHEVs chargng staton gves the soluton for the energy management challenges [7-8]. An estmaton of dstrbuton algorthm to schedule the large number of PHEVs chargng n a chargng staton has been proposed. The method optmzes the energy allocaton to the PHEVs n the real tme whle consderng varous constrants. Ths paper has only proposed the chargng method of PHEVs and the V2G opton does not consder [9].The authors n [10] proposed smulated annealng approach and heurstc techncal valdaton of the obtaned solutons to solve the energy resources schedulng. In ths paper, chargng staton wth PV system on the roof, bdrectonal utlty grd for chargng and dschargng of PHEVs are presented. The grd connecton s to satsfy the demand. Excess PV output s sent to the grd durng peak hours. An energy management system wth PV based Copyrght to IJAREEIE /jareee

2 chargng staton s proposed here n whch the PV generaton uncertanty and V2G capablty of PHEVs are consdered [11]. Moreover, the proposed model consders system constrants and customer's preferences. The contrbutons of the proposed method are hghlghted as follows: The rest of the paper s organzed as follows: n Secton 2, the proposed system components are ntroduced. Secton3 presents the problem formulaton; ncludng the resources and PHEVs constrants. Analyss of the results s shown n Secton4. Fnally, concluson s presented n Secton5. II.COMPONENTS OF PROPOSED SYSTEM Ths secton deals wth the archtecture of the proposed system whch ncludes the multple PV panel on the roof of the chargng staton and PHEVs are shown n fg 1. In addton there s a pont of connecton to the utlty grd to enable the electrcty tradng wth utlty grd. In ths paper the chargng staton plays the mportant role for the Energy Management System n (EMS) PHEVs. PHEVs are parked n the chargng staton can delver the power to the grd or absorb the power from t accordng to the State of Charge (SOC) presents on ther battery. The PHEV s owner not only uses the space for parkng the vehcle and also beneft from V2G capabltes. A. Chargng Staton The chargng staton s compared to conventonal one present s new faclty to PHEV s owners and the utlty grd. Ths energy management system automatcally receves and sends data to vehcles and makes a smart decson regardng the schedulng of chargng and dschargng of the PHEVs. The PHEVs owner can submt ther desred parameters of chargng as prevous day by usng smart phone applcaton. The chargng staton receves parameters from each PHEVs owner, such as arrval tme, duraton of presence n the chargng staton approxmately and the mnmum requred state of charge at the departure tme. These parameters are consdered as nput data. Chargng staton frst receves the day-ahead electrcty prces, PHEVs owners preferences and the forecastng data of solar radatons as the nput data. Then PV power and PHEVs charge/dscharge program s determned by chargng staton control tself. Fnally the result of optmum charge/dscharge schedulng s sent to each PHEVs owner. Fg. 1 Archtecture of Chargng staton. B. B-Drectonal Converter The purpose of the B-Drectonal converter n the PHEVs charger system s to nterface the battery wth the system enablng t to charge and dscharge when needed. In a buck mode, ths converter wll lower the output voltage gong to the nput of the battery to a safe level enablng the battery to safely charge to a full level. The B-Drectonal converter s shown n fg 2.If the system s not chargng; the converter can be swtch nto ts next mode. In ths mode, the battery wll dscharge and ncrease the voltage to drve the load. Fg. 2 B-Drectonal Converter Copyrght to IJAREEIE /jareee

3 The b-drectonal functon of ths crcut s gven by two swtchng transstors, T1 and T2. These transstors wll receve control sgnals from the controller drectng the flow of current. When the current through the nductor s postve means t wll act as buck converter. When the current through the nductor s negatve means t wll act as boost converter. C. Photovoltac Panels Solar power vares n the day-tme as a result of the changng poston of the sun and the moton of clouds. Such varablty and uncertanty should be carefully consdered n the proposed energy management system desgn. In recent years, a large number of technques have been proposed for trackng the Maxmum Power Pont (MPP). Maxmum Power Pont Trackng (MPPT) s used n PV systems to maxmze the photovoltac array output power. PV array output power s used to drectly control the DC to DC converter, thus reducng the dffculty of the system. Incremental conductance of MPPT flow chart s shown n fg 3. The method s based on use of an Incremental conductance of the PV to determne an optmum operatng current for the maxmum output power. In ncremental conductance method the array termnal voltage s always adjusted accordng to the MPP voltage t s based on the ncremental and nstantaneous conductance of the PV module. The MPPT regulates the PWM control sgnal of the Dc to DC boost converter untl the condton: ( I/ V) + (I/V) = 0 s satsfed. D. L-Ion Battery Fg. 3 Incremental Conductance of MPPT Flow Chart One of the prmary goals s to optmally charge a hgh power battery. A sngle Lthum cell break-down of the chargng stages s shown n fg 4. The selected battery conssts of 400 Cell for chargng purpose (288V). Lthum-Ion battery wth a total rated capacty of 13.9 [Ah] s used for chargng the PHEVs. Each lthum-ion cells has a voltage of 4.2V at maxmum capacty. Mantanng maxmum state of charge on a L-Ion battery puts a large amount of stress on the cells, and shortens the overall lfespan of the battery. Copyrght to IJAREEIE /jareee

4 Fg. 4 Chargng Stages of a Lthum-Ion Battery III. IMPLEMENTATION OF MIXED INTEGER LINEAR PROGRAMMING METHOD TO SOLVE THE EMS PROBLEM The objectve functon and varous constrants for solvng mxed nteger lnear programmng are formulated as follows: OBJ = x x=1 prob x T t T t t t=1 ( P C,PHEV Π Ch ) + t=1( P D,PHEV ( C OM C Dch )) P x,t t N UG C OM =1( λbnce )) t prob x represents the probablty of each scenaro, P x,t UG represents the transferred power between the grd and PHEVs n perod t under scenaro x. Its postve values determne sold power to the utlty whle negatve values determne purchased power from the utlty. P C,PHEV and P D,PHEV are the charge or dscharge powers of the th PHEVs n perod t under scenaro x respectvely. C t t t Ch, C OM and C Dch are the open market electrcty prce and the PHEVs specfed chargng and dschargng prce n perod t respectvely,bnce s the departure stored energy devaton from the customer preferences of the th PHEVs n perod t under scenaro x, λ s the penalty cost of uncharged batteres, N ndcate the number of PHEVs. T s the schedulng tme horzon. A. Constrants The man objectve functon s chargng cost mnmzaton wth respect to varous constrants [12].That constrans are explaned as follows: a) Power Balance Constrant: x,t P PV P x,t x UG + P,t PV + N =1 P D,PHEV N + BNC =1 N = P C,PHEV s the produced photo voltac power n perod t under scenaro x and BNC s the battery not charged n perod t under scenaro x respectvely. b) Remanng Battery Power for each PHEVs: The stored energy n the battery s consdered together wth the energy remanng from the prevous perod and the charge or dscharge n the perod t. E PHEV = E 1 PHEV + η G2V P C,PHEV t 1 P η D,PHEV t; ( 2) V2G Where E PHEV s the stored energy n the battery of PHEVs n perod t under scenaro x and η G2V and η V2G are PHEVs battery chargng and dschargng effcences. c) SoC Lmts: SoC mn SoC s,,t SoC max ; x,, t; (4) SoC max s maxmum SoC of th PHEVs. SoC mn s mnmum SoC of th PHEVs. d) Chargng/Dschargng Rate Lmts: SoC max SoC max ; x,, t; (5) Where SoC max s the change n maxmum SoC. e) Battery Chargng Constrant:,t (P C,PHEV + R dn,phev ) η G2V t E PHEV ; (6) Copyrght to IJAREEIE /jareee =1 (1)

5 f) Battery Dschargng Constrant:,t 1 (P Dh,PHEV + R up,phev ) t E η PHEV ; (7) V2G g) Departure SoC Constrant: E x,,t PHEV SoC Desred Cap BNC (8) SoC Desred s the desred SoC at departure tme of the th PHEVs. h) Transmtted Power Lmts: x,t P UG P max UG ; (9) max P UG s the maxmum transmtted power between the PHEVs and the grd. To solve the EMS problem the MILP method has two specfc advantages compared to other optmzaton methods. In the proposed model, the MILP optmzaton guarantees to fnd the globally optmum soluton [13]. Also, The MILP optmzaton fnds the optmum soluton n lower runtme [14, 15]. Moreover, the proposed model s a day-ahead energy and reserve schedulng. The computaton tme s also an mportant aspect of the applcablty of the proposed method. For a real sze chargng parks wth large number of EVs the MILP shows ts benefts n a lght executon tme. IV.RESULTS AND DISCUSSION The smulnk model for the proposed MILP method were developed usng MATLAB 7.10 software package and the system confguraton s Intel Core M Processor wth 2.90 GHz speed and 4 GB RAM. In proposed work two energy sources are consdered. Computatonal results of EMS problem attaned by the proposed MILP method for the two energy sources analyzed. A. Waveform of Utlty Grd Utlty grd connects wth the chargng staton through the DC bus. The output voltage waveform of utlty grd s shown n fg. 5. It produces 20 kw output voltages from generatng staton. Output s transferred to DC bus through the b drectonal converter. Utlty grd s synchronzed by usng VSC control B. Waveform of DC Bus Fg. 5 Voltage Waveform of Utlty Grd Fg. 6 Voltage Waveform of DC Bus Copyrght to IJAREEIE /jareee

6 Photo voltac system s maxmum power pont s contnuously tracked and ntegrated nto the dc-bus lnkng the PHEVs batteres to the man grd. The output voltage waveform of DC bus s shown n fg. 6. DC bus s connected wth utlty grd and chargng staton. DC bus voltage s mantaned at 480v. C. V2G and G2V Power output of the varous V2G and G2V condtons s shown n Table I. the man objectve of the proposed method s to mnmze the chargng cost of plug n hybrd electrc vehcles. The varous algorthm technques and ts chargng cost of PHEVs are shown n Table II. Table I. Power Flow Condtons n PHEVs Plug-n Pv Soc (%) Power W(V2G) W (V2G) Kw (G2V) Table II. Chargng Cost of PHEVs Method Chargng Cost ($/day) Dynamc Programmng 0.46 to 0.40 Mxed Integer Lnear Programmng 0.38 to 0.35 Fuzzy 0.22 to 0.18 V. CONCLUSION The proposed method provdes the optmal power flow and chargng cost mnmzaton of plug n hybrd electrc vehcles. It mproves the performance of V2G and G2V functonaltes by usng renewable sources. The photovoltac and Fuel cell model was desgned usng MATLAB. The excess energy whch s produced from photovoltac and fuel cell power s transferred to the electrcal network. PHEVs system s desgned and chargng cost s mnmzed by usng MILP method. The proposed method satsfes the optmal power flow n the PHEVs system and also mproves the system effcency and mnmzes the losses. The excess energy from renewable energy resources can also be sent to the utlty grd and also satsfes the power demand. REFERENCES 1. H.A.Sher, K.E.Addoweesh, M.E.Webber, Power storage optons for hybrd electrc vehcles a survey, J. Renew. Sustan. Energy, vol. 4, no. 5, pp. 1-11, Q.Zhang, N.K.Ishhara, C.B.Mclellan, T.Tezuka, Scenaro analyss on electrcty supply and demand n Japan, Sustan.Energy, vol. 38, no. 1, pp , M. Pantos, Stochastc optmal chargng of electrc drve vehcles wth renewable energy, Sustan. Energy, vol. 36, no. 11, pp , B. Soares, M.C.Borba, A.Szklo, R.Schaeffer, Plug n hybrd electrc vehcles as a way to maxmze the ntegraton of varable renewable energy n power system: the case of wnd generaton n north eastern Brazl, Sustan Energy, vol. 37, no. 1, pp , Z. Ma, D. Callaway, I. Hskens, Decentralzed chargng control for large populatons of plug-n electrc vehcles: applcaton of the nash certanty equvalence prncple, IEEE Internatonal Conference on Control Applcatons (CCA), Yokohama, Japan, Copyrght to IJAREEIE /jareee

7 6. S.Rezaee,E.Farjah, B.Khorramdel, Probablstc analyss of plug-n electrc vehcles mpact on electrcal grd through homes and parkng lots, IEEETrans.Sustan.Energy, vol. 4, no. 4, pp , M. Honarmand, A.Zakarazadeh, S.Jadd, Optmal schedulng of electrc vehcles n an ntellgent parkng lot consderng vehcle-to-grd concept and battery condton, Renewable Energy, vol. 65, pp , S. Shahdnejad, S.Flzadeh, E.Bbeau, Profle of chargng load on the grd due to plug-n vehcles, IEEE Trans. Smart Grd, vol. 3, no. 1, pp , W. Su, M.Y.Chow, Performance evaluaton of an EDA-based large-scale plug-n hybrd electrc vehcle chargng algorthm, IEEE Trans. Smart Grd, vol. 3, no. 1, pp , T. Sousa, H.Moras, Z.Vale,P.Fara, J.Soares, Intellgent energy resource management consderng vehcle-to- Grd: a smulated annealng approach, IEEE Trans. Smart Grd, vol. 3, no. 1, pp , Ahmed Mohamed, Vahd Saleh, Osama A. Mohammed, Real-Tme Energy Management Algorthm for Plug-In Hybrd Electrc Vehcle Chargng Parks Involvng Sustanable Energy, IEEE Trans. Sustanable Energy, vol. 5, no. 2, pp Apr Masoud Honarmanda, Alreza Zakarazadeha, Shahram Jadd, Self-schedulng of electrc vehcles n an ntellgent parkng lot usng stochastc optmzaton, scencedrect, Journal of the Frankln Insttute,Jan T.Koch, T.Achterberg, E.Andersen, O.Bastert, T.Berthold, R.E.Bxby, etal., MIPLIB2010,Math.Program. Comput., no. 3, pp , March A. Zakara zadeh,s.jadd,p.sano, Economc-envronmental energy and reserve schedulng of smart dstrbuton systems: a mult objectve mathematcal programmng approach, Energy Convers. Mag, vol. 78, pp Aprl H.Falsaf, A.Zakara zadeh, S.Jadd, The role of demand response n sngle and mult-objectve wnd-thermal generaton schedulng: a stochastc programmng, Energy Convers.Mag, vol. 64, pp , June Copyrght to IJAREEIE /jareee