Voltage-Based MaximumPower Point Tracking Controlof PV System

MUMMADI VEERACHARY, Student Member, IEEE

TOMONOBU SENJYU, Member, IEEE

KATSUMI UEZATOUniversity of the RyukyusJapan

Photovoltaic (PV) generators exhibit nonlinear v-i

characteristics and maximum power (MP) points that vary

with solar insolation. An intermediate converter can therefore

increase efficiency by matching the PV system to the load and

by operating the solar cell arrays (SCAs) at their maximum

power point. An MP point tracking algorithm is developed using

only SCA voltage information thus leading to current sensorless

tracking control. The inadequacy of a boost converter for array

voltage based MP point control is experimentally verified and an

improved converter system is proposed. The proposed converter

system results in low ripple content, which improves the array

performance and hence a lower value of capacitance is sufficient

on the solar array side. Simplified mathematical expressions for

a PV source are derived. A signal flow graph is employed for

modeling the converter system. Current sensorless peak power

tracking effectiveness is demonstrated through simulation results.

Experimental results are presented to validate the proposed

method.

Manuscript received March 14, 2000; revised July 3, 2001; releasedfor publication July 26, 2001.

IEEE Log No. T-AES/38/1/02589.

Refereeing of this contribution was handled by I. Batarseh.

Authors’ address: Dept. of Electrical and Electronics Engineering,Faculty of Engineering, University of the Ryukyus, 1 Senbaru,Nishihara-cho, Nakagami, Okinawa 903-0213, Japan.

0018-9251/02/$17.00 c° 2002 IEEE

I. NOMENCLATURE

C, Ca Filter, array capacitanced1, d2 Duty ratios of switch S1,S2DP1, DP2 Diodes of individual boost cellsIA SCA currentIAb SCA current with boost converterIAd SCA current with IDB converterIm SCA current at maximum power

operationIab Average load current with boost converterIaid Average load current with IDB converterIph Insolation dependent photo currentI0 Cell reverse saturation currentL1, L2 Inductances of individual boost cellsNs, Np Number of SCA cells in series, parallelPgb, Pgd SCA power output with boost, IDB

converterPm Maximum power of the SCAR1, R2 Inductor series resistancesRs Cell series resistanceR Load resistanceS1, S2 Switches of individual boost cellsVAb, VAd SCA voltage with boost, IDB converterVm SCA voltage at maximum power

operationVab, Vaid Average load voltage with boost, IDB

converter´b, ´id Efficiency of the boost, IDB converterª Solar insolation.

II. INTRODUCTION

The rapid trend of industrialization of nations andincreased interest in environmental issues recentlyled us to explore the use of renewable forms suchas solar energy. Photovoltaic (PV) generation isgaining increased importance as a renewable source[1—2] due to its advantages like absence of fuel cost,little maintenance, and no noise and wear due to theabsence of moving parts, etc. In particular, energyconversion from solar cell arrays (SCAs) receivedconsiderable attention in the last two decades. ThePV generator exhibits a nonlinear v-i characteristic,and its maximum power (MP) point varies with thesolar insolation and temperature. At a particular solarinsolation, there is a unique operating point of thePV generator at which its power output is maximum.Therefore, for maximum utilization efficiency, it isnecessary to match the PV generator to the load suchthat the equilibrium operating point coincides with theMP point of the PV source. However, since the MPpoint varies with insolation and seasons, it is difficultto maintain MP operation at all solar insolations.To overcome this problem, use of an intermediatedc-dc converter is proposed [3—5], which continuously

262 IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS VOL. 38, NO. 1 JANUARY 2002

adjusts the voltage, current levels and matches the PVsource to the load.The MP point tracking is applied to PV systems

to extract maximum available power from theSCAs at all solar insolations. Different methods ofpeak power tracking schemes have been proposedby using different control strategies [6—9]. Boostconverter based MP point tracking using fuzzylogic is reported [10]. These studies show that thefuzzy control algorithm is capable of improving thetracking performance as compared with conventionalmethods. However, in fuzzy implementation severalparameters are selected on a trial and error basis,which mainly depends on designer experience andintuition. To overcome some of the disadvantagesmentioned above, a fuzzy neural network based MPpoint tracking is proposed [11]. All these methodsdepend on the SCA power output and/or load powerdetection using the instantaneous voltage and currentinformation, requiring voltage and current sensors.Neural network based real time MP tracking controllerfor PV grid connected systems has been reported [12].The studies emphasize that the SCA operating pointis shifted to its MP point by using a voltage controltype inverter, which utilizes the array voltage togetherwith pilot cell voltage information. Array voltagebased MP point tracking using dc-dc converters iscurrently under investigation by various researchers.This method of MP tracking has advantages likestraightforward array voltage measurement (which isinexpensive as compared with the measurements ofsolar insolation, and other environmental factors), noneed of current sensors, (which introduces losses andcomplexity in the system) etc. The authors have triedto implement boost converter array voltage based MPpoint control. The experimental investigations showthat boost converters are not suitable for array voltagebased MP point tracking control as this leads to anundesirable operation. An alternate converter scheme(interleaved dual boost (IDB) converter) is proposedby the authors for array voltage based MP extraction,by simple addition of one more boost cell in parallelto the existing boost cell and controlling these twoboost cells in an interleaved fashion.The advantages of the present converter system

are 1) ripple cancellation both in the input and outputwaveforms to the maximum extent possible, 2) lowervalue of ripple amplitude, and high ripple frequencyin the resulting input and output waveforms, and 3)reduced electromagnetic interference (EMI) becauseof low ripple amplitude of SCA current. Althoughthe interleaving technique increases the number ofcomponents, the actual increase of cost may not besignificant. This is because more boost cells canshare the current flow in the inductors and switchingdevices, so lower current rating devices may beemployed. Further, parallel connection of convertershas many desirable properties such as reduced device

Fig. 1a. Experimental setup of PV system.

stresses, fault tolerance for the system, flexibility inthe system design, etc.This paper presents MP point tracking of SCA

employing an intermediate interleaved dual boost(IDB) converter using only the array voltageinformation, eliminating the array current detectionand hence achieving current sensorless peak powertracking. This paper is organized as follows.In Section III, we present the development ofmathematical models for the PV generator and powerconverters. Maximum power point tracking controlprocess is discussed in Section IV. Experimentalsystem description is given in Section V. Section VIpresents experimental results, and conclusions areprovided in Section VII.

III. MATHEMATICAL MODEL OF SYSTEM

Fig. 1(a) shows the overview of the combinedsystem, which mainly consists of SCA, IDB converter,and data acquisition system. The analysis of thesystem is carried out under the following assumptions.

1) Switching elements (MOSFET and diode) ofthe converter are assumed to be ideal, i.e., forwardvoltage drops and ON-state resistances of the switchesare neglected.2) The equivalent series resistance of the

capacitance and stray capacitances are neglected.3) Passive components (R,L,C) are assumed to be

linear, time invariant, and frequency independent.4) The two parallel boost cells are identical and

operate in the continuous inductor current mode.5) The switches (S1,S2) operate in interleaved

fashion.

Mathematical models for individual componentsare developed in the following sections.

A. PV Generator Model

The PV generator is formed by the combinationof many PV cells connected in series and parallel toprovide the desired output voltage and current. This

VEERACHARY: VOLTAGE-BASED MAXIMUM POWER POINT TRACKING CONTROL OF PV SYSTEM 263

PV generator exhibits a nonlinear insolation dependentv-i characteristic, mathematically expressed for theSCA consisting of Ns cells in series and Np cells inparallel [1] as

VA =¡IARsÃNsNp

!+µNs¤

¶ln

(1+

NpIph¡ IANpI0

)(1)

where ¤= (q=AKT), q–electric charge;A–completion factor; K–Boltzmann’s constant;T–absolute temperature; Rs–cell series resistance;Iph–photo current; I0–cell reverse saturation current;IA, VA are the SCA current and voltage, respectively.For given values of SCA parameters, the VA¡ IAcharacteristic depends on the solar insolation and theMP point varies with the solar insolation. Rewriting(1) as

VA =¡IARsÃNsNp

!+µNs¤

¶

£"lnµIphaI0

¶+ ln

Ã1¡ IA

NpIpha

!#(2)

where Ipha = Iph+ I0. Expanding the termln(1¡ (IA=NpIpha)) into Taylor series and neglectinghigher order terms [5] results in the followingequation

VA =¡IAÃRsg +

2Ns¤NpIpha

!+µNs¤

¶·lnµIphaI0

¶¸:

(3)

Simplifying the above equation for the SCA currentresults in the following equation.

IA =1Ã

Rsg +2Ns

¤NpIpha

! ·Ns¤lnµIphaI0

¶¡VA

¸(4)

where Rsg =NsRs=Np. The equations (3) and (4) areused in the simulation studies.

B. Boost Converter Model

The intermediate boost converter produces achopped output voltage and controls the average dcvoltage applied to the load. Further, the convertercontinuously matches the output characteristic of thePV generator to the input characteristic of the load.The steady-state voltage and current relations of theboost converter operating in continuous current modeare

Vab =VAb

(1¡ d) (5)

Iab = ´b(1¡ d)IAb (6)

where ´b is the efficiency of the boost converter,VAb,IAb are the array voltage and current, respectively.Transforming the load to the SCA side (Fig. 1(b)),

Fig. 1b. Equivalent circuit of system.

then the reflected equivalent load on the SCA side isgiven by the following equation

Req = ´b(1¡ d)2R (7)

i.e.,VAbIAb

= ´b(1¡d)2R: (8)

Power extracted by the boost converter from the SCAis

Pgb =V2Ab

´b(1¡d)2R: (9)

From the above expression the array power Pgbdepends on the load and converter duty ratio. For agiven load, the array power continuously increaseswith duty ratio, theoretically resulting in minimumpower at d = 0 and infinite power at d = 1, at whichthe SCA voltage collapses, leading to an undesirableoperation. Furthermore, the power Pgb is continuouslyincreasing with duty ratio, and hence with this powercomparison method it may not be possible to reachthe MP point. To overcome this disadvantage anidentical boost branch is connected in parallel tothe existing one and controlls these two branches incomplementary fashion. The analytical discussion ofthis converter is given in the following section.

C. IDB Converter Model

The intermediate IDB converter produces achopped output dc voltage and controls the averagedc voltage applied to the load. Further, the convertercontinuously matches the output characteristic of thePV generator to the input characteristic of the load sothat MP is extracted from the SCA. The steady-statevoltage and current relations of the IDB converteroperating in continuous current mode are derivedusing signal flow graph (SFG) technique [13]. Thesteady-state signal flow graph of the IDB converter isobtained as shown in Fig. 2. Voltage gain is derivedby employing the well-known Mason’s gain formula.To start with various possible forward paths and loopsare identified from the steady-state SFG (Fig. 2). Theforward paths transmittances formed by the nodes

264 IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS VOL. 38, NO. 1 JANUARY 2002

Fig. 2. Steady-state SFG of IDB converter.

(Vg ¡V1¡ IL1¡ I0¡V0), (Vg ¡V2¡ IL2¡ I0¡V0) are

P1 =d2R

R1(10)

P2 =d1R

R2: (11)

In this steady-state SFG two loops formed by thenodes (V2¡ IL2¡ I0¡V2), (V1¡ IL1¡ I0¡V0¡V1) existand their loop transmittances are

L1 =¡Rd21R2

(12)

L2 =¡Rd22R1

: (13)

Applying Mason’s gain formula the steady-statevoltage gain obtained as

VaidVAd

=R(R1d1 +R2d2)

R1R2 +R(R1d21 +R2d

22)

(14)

where d1,d2 are the duty ratios of the switchesS1,S2, respectively. The IDB converter switchingdevices (S1,S2) are activated in complementary modewith duty ratios d1 and d2, satisfying the relationd1 + d2 = 1. If the two parallel boost cells are identical(R1 = R2 = r; RÀ r) then the above equation becomes

Vaid =VAd

(d21 + d22): (15)

Substituting d1 = d, d2 = (1¡ d), the above equationcan be written as

Vaid =VAd

(2d2¡ 2d+1) : (16)

Using power balance, the current expression isobtained as

Iaid = ´id(d21 + d

22)IAd (17)

where ´id is the efficiency of the IDB converter,VAd,IAd are the array voltage and current, respectively.Transforming the load to the SCA side (Fig. 1b), thenthe reflected equivalent load on the SCA side is givenby the following equation

Req = ´id(2d2¡ 2d+1)2R (18)

i.e.,VAdIAd

= ´id(2d2¡ 2d+1)2R: (19)

Substituting the IAd = IA expression (from (4)) in theabove expression then

VAd =[´id(2d

2¡ 2d+1)2R]ÃRsg +

2Ns¤NpIpha

! ·Ns¤lnµIphaI0

¶¡VAd

¸:

(20)

On simplification the array voltage equation becomes

VAd =[´id(2d

2¡ 2d+1)RNs]¤[K + ´id(2d2¡2d+1)R]

·lnµIphaI0

¶¸(21)

where K = (Rsg +(2Ns=¤NpIpha)). Power extracted bythe IDB converter from the SCA is

Pgd =V2AdReq

: (22)

Substituting (18) and (21) in (22) then the resultingSCA power expression is

Pgd =(´idRN

2s )

¤2[K + ´id(2d2¡ 2d+1)R]2·lnµIphaI0

¶¸2:

(23)

From (23) it can be noticed that, for given valuesof the array parameters and load, the SCA power(Pgd) depends on the duty ratio of the IDB converter.Suitable adjustment of converter duty ratio results inVAd = Vm, which in turn results in extraction of MPfrom the SCA.

IV. MAXIMUM POWER POINT TRACKINGCONTROL PROCESS

The control flow chart is given in Fig. 3, whichcontrols the tracking process of the PV suppliedconverter system. The tracking process can be startedby outputting the command signal either 0 or 5 V tothe pulsewidth modulation (PWM) generator, whichcorresponds to duty ratio of zero or one, respectively.Whatever may be the duty ratio (0-1) the array power(Pg) is computed from the (22) using the already

VEERACHARY: VOLTAGE-BASED MAXIMUM POWER POINT TRACKING CONTROL OF PV SYSTEM 265

Fig. 3. Flow chart for MP point tracking.

sensed array voltage. Change (increase or decrease)the duty ratio and then measure the instantaneousarray power. This power is compared with theprevious power and a decision on whether to increaseor decrease the duty ratio is taken depending on thelocation of the operating point and direction of itsmovement as indicated in Fig. 4. As a consequence,there are four possibilities (two if the operatingpoint is left of the MP point, two if the operatingpoint is right of the MP point) for the operatingpoint movement. The duty ratio control signal iscontinuously adjusted to maximize the array powerby following the equation d = d§¢d. The sign ofthe incremental duty ratio (¢d) is determined bythe incremental power (¢P) and operating pointmovement as indicated in Fig. 4. If ¢P is positiveand the operating point is left of MP point, thend = (d+¢d), otherwise d = (d¡¢d). Along similarlines, if the ¢P is negative and the operating point isleft of the MP point, then d = (d+¢d), otherwise d =(d¡¢d). This tracking control process repeatsitself until the peak power point is reached andthen oscillates within an allowable range aboutthis point. In the simulated MP point trackingprocess the instantaneous array voltage and powerare computed employing the models developed inpreceding sections, whereas in real time computerimplementation the instantaneous array voltage,power information is obtained by means ofdata acquisition system. The MP point trackingprocess both in simulation and real time computerimplementation are same except the above mentioneddifference.

Fig. 4. Power tracking process with duty ratio.

V. EXPERIMENTAL SYSTEM DESCRIPTION

The basic configuration of the proposed PV systemis shown in Fig. 1(a). The data acquisition systemis set up by using PC, interface AZI-3503 card,which mainly consists of 8-channel 12-bit A/D, D/Aconverters. For power measurements a digital powermeter (YOKOGAWA-WT130) is used, through whicha GPIB interface is connected to the PC to record theSCA power data. The PWM modulator is a voltagecomparator made of LF311 operational amplifier.The reference signal to this comparator is the signalobtained from the D/A converter, generated by meansof the MPPT algorithm. A synthesized YOKOGAWAfunction generator (FG120) was used to obtain phasedisplaced triangular carrier signals to the PWMgenerator. The experimental prototype circuit was builtwith an International Rectifier IRF530N MOSFETwith suitable driver circuit, and the diode FML-32S.The artificial sun is realized in the laboratory bymeans of incandescent lamp set. Further, the solarinsolation level illuminated on the solar panel isadjusted by controlling the power to this incandescentlamp set.

VI. EXPERIMENTAL RESULTS AND DISCUSSIONS

The simple boost converter is not able to track MPpoint by sensing only the array voltage information.This is because the equivalent load impedance (in(7)) seen by the SCA is continuously decreasing withincreasing the duty ratio. Further, from the (9) for agiven load, the array power continuously increaseswith the duty ratio, resulting in minimum power atd = 0 and infinite power at d = 1, which is physicallyan unrealizable condition. This phenomena is verifiedexperimentally and the corresponding characteristicsare shown in Fig. 5. To overcome this disadvantageand to extract MP from the SCA using only thearray voltage, an identical boost cell is connectedin parallel to the existing boost cell as shown inthe experimental setup (Fig. 1(a)). These two boostbranches are controlled in an interleaved fashion usingphase shift between the gate signals. This converter

266 IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS VOL. 38, NO. 1 JANUARY 2002

Fig. 5. Array characteristics with boost converter.

Fig. 6. Experimental VA¡ IA characteristics of SCA module fordifferent solar insolations.

is capable of reducing the ripple in the source currentEMI and avoids the discontinuous input current modeeven though the individual boost branches enter intodiscontinuous current mode.Prototype PV-supplied converter (SCA and

converter parameters are given in Table I) systemwas bread boarded to study the array voltagebased maximum power point tracking. The VA¡ IAcharacteristics of the experimental PV generator forthree different solar insolations (ª2 = 30%, ª3 =60%, and ª5 = 100%) are shown in Fig. 6. The 100%solar insolation represents a standard intensity of1000 W/m2. The data acquisition system measuresthe instantaneous array voltage information. For agiven load, the MP point control algorithm computesthe SCA power (Pg) from the known instantaneousarray voltage information Vg. The algorithm tracksthe maximum power point continuously by adjustingthe converter duty ratio such that the array power ismaximum. At solar insolation ª5 the experimentalpower tracking characteristic is shown in Fig. 7. Forverification, the output power of the SCA is measuredby sensing the SCA voltage and current. The powercharacteristic so obtained is superimposed in the samefigure. These two characteristics are closely matchingeach other. Slight discrepancies may be due to errorsin the measuring system, drops in parasitics, etc.The simulation program was developed in

the MATLAB environment for the PV-supplied

Fig. 7. Comparison of experimental SCA power trackingcharacteristics.

TABLE ISCA and Converter Parameters

Np 1

Ns 27Rs 0.04 −

¤ 13.68 V¡1Io 0.00045 AR1 0.044 −

R2 0.063 −

L1 0.250 mHL2 0.250 mHC 220 ¹FCa 2200 ¹Ffs 25 kHzR 50 −

converter system (shown in Fig. 1(a)). Comprehensivesimulation studies were made to investigate theinfluence of IDB converter as an intermediateMP point tracker for the PV supplied system. Asimulation software is developed for MP pointtracking employing the equations derived in thepreceding sections and the control flow chart given inFig. 3. In these studies the PV array is simulated using(3) and (4). The simulated dynamic MP point trackingcharacteristics at 100% solar insolation are plotted inFigs. 8 and 9. At this solar insolation the experimentaldynamic MP tracking characteristics are also obtainedand they are superimposed in Figs. 8 and 9. Thesimulation and experimental results are in closeagreement. Discrepancies between simulation andexperimental results may be due to 1) the difficultiesin realizing the identical solar insolation conditionsin the experimental setup, and 2) the fact that theanalysis was made on the assumption that the twoboost branches are identical.The experimental array power tracking

characteristics for three different solar insolations (ª2,ª3 and ª5) are also obtained as shown in Fig. 10. Forverification of the MP points of the SCA, experimentswere conducted on the SCA by connecting a variableload resistance. The experimental MP points obtainedat different solar insolations are tabulated in Table

VEERACHARY: VOLTAGE-BASED MAXIMUM POWER POINT TRACKING CONTROL OF PV SYSTEM 267

Fig. 8. Comparison of experimental and simulated SCA powertracking characteristics.

Fig. 9. Comparison of experimental and simulated duty ratiotracking characteristics.

Fig. 10. SCA power tracking characteristics for different solarinsolations.

II. Comparing the tracking characteristics (Fig. 10)with MP points, it can be noticed that the duty ratio ofthe converter is so adjusted such that MP is extractedfrom the SCA. Experimental studies are also made toobserve the effectiveness of the developed trackingalgorithm for changing solar insolations. Experimentalobservations (Fig. 11) show that the developedalgorithm is capable of tracking MP point even forvariable solar insolations. The tracking capability ofthe IDB converter system is verified under partialshading conditions also. For illustration, array powertracking characteristics when few cells (4) are shadedby 50% are shown in Fig. 12. Under this condition the

TABLE IIExperimental Maximum Power Points of SCA

% Solar Insolation Maximum Power (W)

ª1 2.65ª2 8.17ª3 17.63ª4 21.44ª5 23.03

Fig. 11. SCA power tracking characteristic for variable solarinsolations.

Fig. 12. Experimental SCA power tracking characteristic forpartial shading conditions.

SCA power output decreases and settles to a new MPpoint as evidenced by Fig. 12.

VII. CONCLUSIONS

Current sensorless SCA voltage based on a MPpoint tracking algorithm is developed for an IDBconverter supplied PV system. Analytical expressionsfor the SCA, and power output expressions withconverters are derived. The SFG approach is usedin modeling the IDB converter. Simulation andexperimental results for MP tracking are presentedfor changing solar insolations and partial shadingconditions. The inadequacy of the boost converterfor array voltage based MP point tracking schemeis verified. The experimental results demonstratethat in the array voltage based peak power tracking

268 IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS VOL. 38, NO. 1 JANUARY 2002

scheme the IDB converter is suitable for extractingMP from the SCA as compared with boost convertersupplied PV system. Furthermore, the use of an IDBconverter avoids the discontinuous input current modeof operation and reduces the ripple in the array inputcurrent. As a consequence the reduced ripple in thearray current results in improved SCA performance.

ACKNOWLEDGMENTS

The first author wishes to acknowledge theGovernment of Japan for granting MONBUSHOscholarship and JNT University authorities forpermission to attend these research studies.

REFERENCES

[1] Appelbaum, J. (1986)Starting and steady-state characteristics of dc motorspowered by solar cell generators.IEEE Transactions on Energy Conversion, 1 (1986),17—23.

[2] Fam, W. Z., and Balachander, M. K. (1988)Dynamic performance of a dc shunt motor connected tophotovoltaic array.IEEE Transactions on Energy Conversion, 3 (1988),613—617.

[3] Salameh, Z., and Taylor, D. (1990)Step-up maximum power point tracker for photovoltaicarrays.Solar Energy, 44 (1990), 57—61.

[4] Alghuwainem, S. M. (1992)Steady-state performance of dc motors supplied fromphotovoltaic generators with step-up converter.IEEE Transactions on Energy Conversion, 7 (1992),267—272.

[5] Alghuwainem, S. M. (1997)A close form solution for the maximum power operatingpoint of a solar cell array.Solar Energy Materials and Solar Cells, 46 (1997),249—257.

[6] Kislovski, A. S. (1993)Power tracking methods in photovoltaic applications.Proceedings of Power Conversion, (1993), 513—528.

[7] Hua, C., Lin, J., and Shen, C. (1998)Implementation of a DSP controlled photovoltaic systemwith peak power tracking.IEEE Transactions on Industrial Electronics, 45 (1998),99—107.

[8] Matsui, M., Kitano, T., Xu, D.-H., and Yang, Z.-Q. (2000)New MPPT control scheme utilizing power balance at DClink instead of array power detection.In Proceedings of International Power ElectronicsConference (IPEC), 2000, 158—163.

[9] Sharif, M. F., Alonso, C., and Martinez, A. (2000)A simple and robust maximum power point control forground photovoltaic generators.Proceedings of International Power Electronics Conference(IPEC), 2000, 164—169.

[10] Won, C.-Y., Kim, D.-H., and Kim, S.-C. (1994)A new maximum power point tracker of photovoltaicarrays using fuzzy controller.In Proceedings of Power Electronic Specialist Conference,1994, 396—403.

[11] Senjyu, T., Arashiro, Y., Uezato, K., and Hee, H. K. (1998)Maximum power point tracking control of photovoltaicarray using fuzzy neural network.Proc. of International Conference on Power Electronics(ICPE), 1998, 987—992.

[12] Hiyama, T., Kouzuma, S., Imakubo, T., and Ortmeyer, T. H.(1995)Evaluation of neural network based real time maximumpower tracking controller for PV system.IEEE Transactions on Energy Conversion, 10 (1995),543—548.

[13] Smedley, K., and Cuk, S. (1994)Switching flow-graph nonlinear modeling technique.IEEE Transactions on Power Electronics, 42 (1994),405—413.

VEERACHARY: VOLTAGE-BASED MAXIMUM POWER POINT TRACKING CONTROL OF PV SYSTEM 269

Mummadi Veerachary was born in Survail, AP, India in 1968. He obtainedhis Bachelors degree from College of Engineering Anantapur, JNT University,Hyderabad, India, in 1992 and Master of Technology from Regional EngineeringCollege, Warangal, India in 1994.In 1994 he joined as an Assistant Professor in the Dept. of Electrical

Engineering, JNTU College of Engineering, Anantapur, India. Presently, heis at the Dept. of Electrical and Electronics Engineering, University of theRyukyus, Okinawa, Japan for his research studies. His fields of interest are powerelectronics, modeling and simulation of power electronics and application tophotovoltaic solar energy utilization.Mr. Veerachary was the recipient of the IEEE Industrial Electronics Society

student travel grant award for the year 2001.

Tomonobu Senjyu was born in Saga prefecture, Japan, in 1963. He receivedthe B.S. and M.S. degrees in electrical engineering from University of theRyukyus, Okinawa, Japan, in 1986 and 1988, respectively, and the Ph.D. degreein electrical engineering from Nagoya University, Nagoya, Japan, in 1994.Since 1988, he has been with the Department of Electrical and Electronics

Engineering, Faculty of Engineering, University of the Ryukyus, where he iscurrently a Professor. His research interests are in the areas of stability of acmachines, advanced control of electrical machines, and power electronics.Dr. Senjyu is a member of the Institute of Electrical Engineers of Japan.

Katsumi Uezato was born in Okinawa prefecture, Japan, in 1940. He receivedthe B.S. degree in electrical engineering from the University of the Ryukyus,Okinawa, Japan, in 1963, the M.S. degree in electrical engineering fromKagoshima University, Kagoshima, Japan, in 1972, and the Ph.D. degree inelectrical engineering from Nagoya University, Nagoya, Japan, in 1983.Since 1972, he has been with the Department of Electrical and Electronics

Engineering, Faculty of Engineering, University of the Ryukyus, where heis currently a Professor. He is engaged in research on stability, control ofsynchronous machines and power electronics.Dr. Uezato is a member of the Institute of Electrical Engineers of Japan.

270 IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS VOL. 38, NO. 1 JANUARY 2002