Hindawi Publishing CorporationInternational Journal of PhotoenergyVolume 2013, Article ID 549273, 11 pageshttp://dx.doi.org/10.1155/2013/549273
Research ArticleDSPACE Real-Time Implementation ofMPPT-Based FLC Method
Abdullah M. Noman,1 Khaled E. Addoweesh,1 and Hussein M. Mashaly2
1 Electrical Engineering Department, King Saud University, Saudi Arabia2 Sustainable Energy Technologies Center, King Saud University, Saudi Arabia
Correspondence should be addressed to Abdullah M. Noman; [email protected]
Received 22 December 2012; Revised 23 February 2013; Accepted 27 February 2013
Academic Editor: Sih-Li Chen
Copyright Β© 2013 Abdullah M. Noman et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.
Maximum power point trackers are so important in photovoltaic systems to improve their overall efficiency. This paper presents aphotovoltaic system with maximum power point tracking facility. An intelligent fuzzy logic controller method is proposed in thispaper to achieve the maximum power point tracking of PVmodules.The system consists of a photovoltaic solar module connectedto a DC-DC buck-boost converter. The system is modeled using MATLAB/SIMULINK. The system has been experienced underdisturbance in the photovoltaic temperature and irradiation levels.The simulation results show that the proposedmaximum powertracker tracks the maximum power accurately and successfully in all conditions tested. The MPPT system is then experimentallyimplemented. DSPACE is used in the implementation of the MPPT hardware setup for real-time control. Data acquisition andcontrol system is implemented using dSPACE 1104 software and digital signal processor card. The simulation and practical resultsshow that the proposed system tracked the maximum power accurately and successfully under all atmospheric conditions.
1. Introduction
Energy is important for the human life and economy. Con-sequently, due to the increase in the industrial revolution,the world energy demand has also increased. In the lateryears, irritation about the energy crisis has been increased.Fossil fuels have started to be gradually depleted. On theother hand, people are more concerned about the fossilfuel exhaustion and other environment problems which area result of conventional power generation. It is a globalchallenge to generate a secure, available, and reliable energyand at the same time reduce the greenhouse gas emission [1].Energy saving was suggested by the researchers to meet theworldwide energy demand. But thismethod is a cost-effectivesolution. One of the most effective and suitable solutions isthe renewable energy supplies. Renewable energy sources areconsidered as a technological option for generating clean,green, environment-friendly, and sustainable energy [1, 2].
Photovoltaic (PV) system has taken a great attentionsince it appears to be one of the most promising renewableenergy sources. The photovoltaic (PV) solar generation is
preferred over the other renewable energy sources due toadvantages such as the absence of fuel cost, cleanness, beingpollution-free, little maintenance, and causing no noise dueto absence of moving parts. However, two important factorslimit the implementation of photovoltaic systems. These arehigh installation cost and low efficiency of energy conversion[1]. In order to reduce photovoltaic power system costsand to increase the utilization efficiency of solar energy,the maximum power point tracking system of photovoltaicmodules is one of the effectivemethods [3]. Maximum powerpoint tracking, frequently referred to as MPPT, is a systemused to extract the maximum power of the PV moduleto deliver it to the load [4]. Thus, the overall efficiency isincreased [4].
Since the power generated from the photovoltaic moduledepends on the temperature and the solar radiation, thesefactors must be taken into account while designing themaximum power point tracker. The main goal of the MPPTis to move the module operating voltage close to the voltageat which the PV produces the maximum power under all
2 International Journal of Photoenergy
atmospheric conditions. MPPT is very important in PV sys-tems. Different techniques have been developed to maximizethe output power of the photovoltaic module. They haveadvantages and limitations over the others. These techniquesvary in complexity, in the number of sensors required, in theirconvergence speed, and in their cost. In the literature someof MPPT methods are introduced such as feedback voltagemethod, and incremental conductancemethod, perturbationand observation method [1, 4β8]. The open-circuit voltagemethod is based on (1) which states that the voltage of thePVmodule at maximum power point is linearly proportionalto the open circuit voltage [9β12]. The proportional constantπΎ depends on the meteorological conditions, fabrication ofthe PV cell and on the fill factor of the PV cell [8]:
πΎ =πMPPπocβ constant < 1. (1)
The proportional constantπΎ has been reported to be between0.71 and 0.78 [13]. The common value of πΎ is about 0.76(within Β±2%) [12].
In order to implement the constant voltage algorithm,PV modules must be interrupted with a certain frequencyto measure the open-circuit voltage of the PV module. Themeasured voltage is then multiplied by the factorπΎ to obtainthe voltage at maximum power point. Then the operatingvoltage of the PVmodule is adjusted to the calculated voltagein order to obtain the maximum power.This process must berepeated periodically [8]. Although this method is simple toimplement, it has a drawback which is high power losses dueto periodically interrupting the system operation. Anotherdrawback is that it is difficult to choose an optimal value ofthe constant K [8, 10].
The other method is the constant voltage (current)method.The constant voltage (current)method compares themeasured voltage (current) of the PVmodulewith a referencevoltage (current) to continuously adjust the duty cycle ofthe DC-DC converter and hence operate the PV module atthe predetermined point close to the MPP [8]. Although theconstant voltage (current) tracking method is very simple,this method is not able to track the maximum power pointwith changing environment conditions specially when thetemperature changes. That means it cannot be applied in ageneralized fashion in systems which do not consider theeffect of variations of the irradiation and temperature of thePV panels [8, 14].
Perturbation and observation (P&O) method is an alter-native method to obtain the maximum power point of thePV module. It measures the voltage, current, and power ofthe PV module and then perturbs the voltage to encounterthe change direction [9]. Figure 1 shows the π-π curve of thePV module. As shown in the left-hand-side MPP the powerof the PV is increased with increasing the voltage of the PVmodule until the MPP is reached. In the right-hand-side ofthe MPP with increasing the voltage the power is decreased.Thatmeans if there is an increase in the power, the subsequentperturbation should be kept in the same direction until MPPis reached. If there is a decrease in the power, the perturbationshould be reversed [4, 5, 8, 15, 16].
Table 1: The operation of P&O algorithm.
ΞπPV ΞπPV Perturbation>0 >0 Increase π>0 <0 Decrease π<0 >0 Decrease π<0 <0 Increase π
0
20
40
60
80
100
120
0 5 10 15 20 25πPV (V)
πPV
(W)
ππ/ππ > 0
ππ/ππ < 0
ππ/ππ = 0
Figure 1: Variation of ππ/ππ in the π-π characteristic of the PVmodule.
The maximum power point is reached when ππ/ππ =0. The flow chart of P&O method is shown in Figure 2. Inorder to implement the P&O MPPT method, the PV voltageand current must be initially measured. The change in thevoltage (Ξπ) and the change in the power (Ξπ) must thenbe calculated. The PV voltage must then be perturbed bya constant value. If the perturbation in the voltage causesthe power to increase, the next perturbation must be keptin the same direction; otherwise, the next perturbation mustbe reversed. Table 1 summaries the operation of the P&Oalgorithm [15].
In this paper, a new method based on fuzzy logiccontroller (FLC) is proposed to achieve maximum powerpoint tracking.The proposed method depends on measuringthe change in the PV voltage and power of the PV module.The performance of the FLC method is evaluated by MAT-LAB/SIMULINK. The proposed system is then experimen-tally implemented. DSPACE real-time control is used in theimplementation of the MPPT hardware setup. Data acquisi-tion and the control system is implemented by using dSPACE1104 software and digital signal processor card on PC.
2. Characteristics of Solar Module
In order to model the PV module, a PV cell model must beinitially established. An equivalent electrical circuit makes itpossible to model the characteristic of a PV cell. In a practicalPV cell, there are two resistances: series resistance andparallelresistance. Series resistance accounts for the losses in thecurrent path due to the metal grid, contacts, and current-collecting bus. Parallel resistance due to the loss is associated
International Journal of Photoenergy 3
Start
Return
No
No
NoNo
Yes
Yes
YesYes
Measure π(π), πΌ(π)
Calculate power π(π)
π(π) β π(π) = 0
π(π) β π(π β 1) > 0
π(π) β π(π β 1) > 0π(π) β π(π β 1) < 0
π· = π· + Ξπ· π· = π·β Ξπ· π· = π·+ Ξπ· π· = π·β Ξπ·
Figure 2: The flow chart of the P&O algorithm.
π
+
β
β
βπΌ
πΌPVπ πΏ
π π
Figure 3: Equivalent circuit of PV cell simulation.
with a small leakage of current through a resistive path inparallel with the intrinsic device. Parallel resistance is largeand its effect is negligible. The equivalent circuit of the PVcell is shown in Figure 3.
The output current delivered to the load can be expressedas follows [4, 17, 18]:
πΌ = πΌPV β πΌπ (π(π(π+πΌπ
π )/(πππ
π))
β 1) , (2)
where πΌ is the output current of the solar module (A) andπ isthe output voltage of the solar cell (V), which can be obtainedby dividing the output voltage of the PV module by thenumber of cells in series, πΌPV is the current source of the solarmodule by solar irradiance (A), πΌ
πis the reverse saturation
current of a diode (A), ππis the series connection number
of the solar module, n is the ideality factor of the diode (π =1βΌ2), q is the electric charge of an electron (1.6 Γ πβ19c), k is
Boltzmannβs constant (1.38 Γ 10β23 j/K), and T is the absolutetemperature of the solar cell (βK).
To model the PV module using MATLAB, the currentgenerated by the incident light which is also called short-circuit current (πΌsc) at a given temperature (π
π) must be
calculated as follows [17β19]:
πΌPV = πΌscn (1 + π (ππ β ππ))πΊ
πΊπ, (3)
where πΌscn is the short-circuit current at normal conditions(25βC, 1000W/m2), π
πis the given temperature (βK), πΌPV is
the short-circuit current at a given cell temperature (ππ), π is
the temperature coefficient of πΌsc, andπΊπ is the nominal valueof irradiance, which is normally 1000W/m2.
On the other hand, the reverse saturation current of diode(πΌπ) at the reference temperature (π
π) is given as follows [17,
18]:
πΌππ=
πΌscnπ(ππocn/(ππππ)) β 1
, (4)
where πocn is the open-circuit voltage at normal conditions.The reverse saturation current at a given cell temperature (π
π)
can be expressed as follows [18]:
πΌπ= πΌππ(ππ
ππ
)
(3/π)
π((βππΈ
π/ππΎ)(1/π
πβ1/ππ))
. (5)
4 International Journal of Photoenergy
Table 2: PV module parameters.
Maximum power (πmax) 115WVoltage at πmax (πmp) 17.1 VCurrent at πmax (πΌmp) 6.7 AOpen-circuit voltage (πoc) 21.8 VShort-circuit current (πΌsc) 7.5 ATemperature coefficient of πΌsc 0.065 Β± 0.015%/βC
0 5 10 15 20 250
20
40
60
80
100
120
π(W
)
1000 W/m2
800 W/m2
600 W/m 2
400 W/m 2
200 W/m2
π (V)
Figure 4: π-π curves under changing solar radiation.
TheBP3115 PVmodule is used in this paper.The PVmod-ule parameters under the reference conditions (1000W/m2,25βC) are listed in Table 2.The PVmodule is simulated usingMATLAB. Figure 4 shows the simulated P-V curves of thePV module under changing solar radiation from 200W/m2to 1000W/m2 while keeping the temperature constant at25βC. On the other hand, Figure 5 shows the simulationresults of the π-π curves of the PV module under changingtemperature from 10βC to 50βC while keeping the solarradiation constant at 1000W/m2.
3. DC-DC Buck-Boost Converter
DC conversion has gained the great importance in manyapplications, starting from low-power applications to high-power applications. In this paper, buck-boost converter ischosen to be used in theMPPT system. Buck-boost converteris used to step down and step up the DC voltage by changingthe duty ratio of theMOSFET. If the duty ratio is less than 0.5,the output voltage is less than the input voltage; while if theduty ratio is greater than 0.5, the output voltage is greater thanthe input voltage. Duty ratio is the time at which theMOSFETis on to the total switching time. The buck-boost converter isshown in Figure 6.
The relation between the input and the output voltages ofthe buck-boost converter is given as follows:
πout =βπ·
1 β π·πin, (6)
Table 3
Buck-boost converter parametersπΏ 1mHπΆ1
1000πFπΆ2
330 πFππ
40KHZResistive load RL 5Ξ©
Controller type: dSPACE 1104 DSPMOSFET type: IRF3710Diode type: BYV32-200
Components used in the measurement circuitCurrent transducer LTS 25-NP
Voltage divider
Two 120KΞ© and 39KΞ©resistors are connected inseries. The voltage is takenacross 39KΞ© resistor.
0 5 10 15 20 250
20
40
60
80
100
120
140
π(W
)
π (V)
10βC25βC
40βC
55βC
Figure 5: π-π curves under changing temperature.
where π· is the duty cycle of the converter which is given asfollows:
π· =πonππ
, (7)
where πon is the on-state time of the MOSFET while ππis the
switching time.The buck-boost converter is designed and simulated
using MATLAB/SIMULINK. The converter componentsused in the simulation and in the hardware setup are shownin Table 3.
4. MPPT-Based FLC Method
PV systems have relatively high initial cost. Approximately57% is spent on the PV modules, 30% on the batteries,7% on the MPPT controllers and inverters, and 6% onthe installation [20]. Therefore, introducing a high-efficientMPPT controller can help in decreasing the total cost of thePV systems. FLC can be used as a controller to obtain themaximum power that the PV modules capable of producing
International Journal of Photoenergy 5
MOSFET
Gate drive circuit
Diode
+
++
+
+
β
5 ohmπoutπin
[π]
πΏ
π·
ππ
πΆ2π πΏ
Figure 6: The buck-boost converter circuit.
Fuzzification
Inference
Defuzzification
Rules
Crisp inputs
Crisp outputs
Figure 7: The stages of the FLC.
under changing weather conditions. The use of fuzzy logiccontrollers has been increased over the last decade becausethey are simplicity, deal with imprecise inputs, do not need anaccurate mathematical model, and can handle nonlinearity[21]. The nonlinear nature of the PV modules and the envi-ronment conditions make the tracking behavior so difficult.Thus, the FLC is an interesting tool to achieve the maximumpower and eliminate the complexity in the computation sinceit is simple, does not need the mathematical model, and doesnot need any reference MPP parameters [20].
The process of FLC can be classified into three stages,fuzzification, rule evaluation, and defuzzification.These com-ponents and the general architecture of an FLC are shown inFigure 7. The fuzzification step involves taking a crisp input,such as the change in the voltage reading, and combining itwith stored membership function to produce fuzzy inputs.To transform the crisp inputs into fuzzy inputs, membershipfunction must be first assigned for each input. The numberof membership functions used depends on the accuracy ofthe controller, but it usually varies between 5 and 7 [10]. Thesecond step of fuzzy logic processing is the rule evaluation inwhich the fuzzy processor uses linguistic rules to determinewhat control action should occur in response to a given setof input values. The result of rule evaluation is a fuzzy outputfor each type of consequent action.
The last step in fuzzy logic processing is defuzzificationin which the expected value of an output variable is derivedby isolating a crisp value in the universe of discourse of theoutput fuzzy sets. In this process, all of the fuzzy outputvalues effectively modify their respective output membershipfunction. One of the most commonly used defuzzificationtechniques is called center of gravity (COG) or centroidmethod.
Fuzzy logic was applied in designing different MPPTcontrollers [14, 20β27]. They apply a set of linguistic rules toobtain the required duty cycle.The input variables of the FLCdiffer from one paper to another. In [22β25] the inputs to theFLC are the error (E) and the change in error (ΞπΈ). The error(E) is calculated as the change in the PV power to the changein the PVvoltage (Ξπ/Ξπ).The change in the duty cycle is theoutput fromFLC. In other cases, the change in current insteadof the change in the PV voltage to calculate the error (E) isused as in [26]. Some other papers use other inputs to the FLCsuch as the change in the voltage (ΞV) and the change in thepower (Ξπ) while the output from FLC is either the changein the duty ratio of the power converter (Ξπ·) or the changein the reference voltage (Ξπ). Li andWang use Ξπ and Ξπ asthe input variables while the output variable is the change inthe reference voltage [14]. An adaptive fuzzy logic controllerforMPPTwas presented in [27] to adjust the duty cycle of thedefuzzification to enhance the controller performance underchanging atmospheric conditions.
In this paper, a new method-based FLC is proposed toachieve tracking the maximum power of the PV moduleunder changing weather conditions. The proposed inputs ofthe FLC are the change in the voltage of the PVmodule (Ξπ)and the change in the power of the PV module (Ξπ). Theproposed output from FLC is (Ξπ) which corresponds to themodulation signal which is applied to the PWM modulatorin order to produce the switching pulses.
5. The Proposed MPPT Fuzzy LogicBase Method
The input variables are defined as in (8). During fuzzification,the numerical input variables which are converted into
6 International Journal of Photoenergy
0.04
0.00
80 0.2
0
1
β0.2
β0.08
β0.04
β0.001
0.001
Figure 8: The membership function of the input variable (Ξπ).
0
0
1
β0.04
β0.0135
β0.0067
0.001
β0.001
0.00
67
0.01
35
0.04
Figure 9: The membership function of the input variable (Ξπ).
linguistic variables are based on the membership func-tions. Figures 8, 9, and 10 show the membership of Ξπ,Ξπ, and Ξπ, respectively. Five fuzzy levels are used for all theinput and output variables: NB (negative big), NS (negativesmall), ZE (zero), PS (positive small), and PB (positive big):
Ξπ = π (πΎ) β π (πΎ β 1) ,
Ξπ = π (πΎ) β π (πΎ β 1) .
(8)
The theoretical design of the rules is based on the factthat if the change in the voltage causes the power to increase,the moving of the next change is kept in the same direction;otherwise the next change is reversed. After the theoreticaldesign, all the MFs and the rules were adjusted by the trialand error to obtain the desired performance.
The proposed rules are shown in Table 4. The fuzzyrules are designed to track the maximum power point ofthe photovoltaic system under changing weather conditions.Rapidly changing solar radiation is taken into account whiledesigning these rules.
6. Simulation Results
In order to verify that the proposed MPP tracker tracksthe maximum power point successfully, the controller istested under changing weather conditions. It is importantto test the proposed MPPT system under different ambientconditions in order to validate the designed system. Thedesign and the simulation performance are done usingMAT-LAB/SIMULINK. The model used for simulation is shownin Figure 11. In this system, the PV module is connected
Table 4: Rule base used in the fuzzy logic controller.
ΞπΞπ
NB NS Z PS PBNB PB PS NB NS NSNS PS PS NB NS NSZ NS NS NS PB PBPS NS PB PS NB PBPB NB NB PB PS PB
0.48
0.61
0.22
0.09
0.36
3
0.33
7 1
0
1
0.350
β0.3
Figure 10: The membership function of the output variable (Ξπ).
to the DC-DC buck-boost converter. The output of theconverter is the 5Ξ© load resistor. In order to start trackingthe maximum power, the output voltage and the outputcurrent of the PV module must be measured to be usedas an input to the MPPT control block. The output of theMPPT control block is the gating signal which is used to drivethe MOSFET. The proposed FLC MPPT method is testedunder changing weather conditions as shown in Figure 12.As shown in this figure the solar radiation is changed as aconstant value 300W/m2 until 0.03 sec. The solar radiationis then assumed to be changed as a ramp function withpositive slope to account for changing the solar radiationin the sunrise periods. Then the irradiance is changed as aunit step function to account for changing the solar radiationrapidly. In practical point of view, the solar radiation isdecreased as a ramp function with a negative slope duringthe sunset periods. On the other hand, the temperature iskept constant at 25βC and then raised up rapidly to 50βCat 0.06 sec. It is clear from Figure 12 that the proposedsystem will be tested under all expected ambient conditions.These are constant solar radiation, rapidly changing solarradiation, and changing solar radiation as a ramp function.The FLC-based MPPT method is tested under these ambientconditions. The proposed method tracked the maximumpower effectively and accurately as shown in Figure 13. Thecontroller tracked the maximum power under all ambientconditions listed above. The proposed system tracked themaximumpower under changing solar radiation as a positiveand negative ramp function. On the other hand, it followsthe maximum power under rapidly changing solar radiationaccurately. The tracking efficiency using FLC is 98.13%. Thetracking efficiency can be calculated as the energy generatedfrom the PV module divided by the theoretical maximumenergy. Comparing the tracking behavior of the proposedFLC MPPT method shown in Figure 13 with the tracking
International Journal of Photoenergy 7
[πPV ]
[πPV ]
[π]
π
π
[πΊ]
πΊ
πΊ
[π]
[πΊ]
πΏ+β+
β
+
+
β
+β
π
In1
PV module
In2
[πmax ]
Out1 [π]
[π]
ππ·
π
π
Temperature
Irradiation
MeasurementsFL controller
Goto5
7
4
MOSFETπΌPV
[πΌPV ]
πPV
[πPV ]
πΌ
ππPV1
β
πΆ1
πΆ2
πΌ
πΌPV
πPV
[πΌPV ]
[ππ΅]
[πΌπΏ]
Figure 11: MPPT system used for simulation.
0
500
1000
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1Time (s)
πΊ(W
/m2)
(a)
0204060
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1Time (s)
π(β
C)
(b)
Figure 12: Changing ambient condition.
behavior of the P&O MPPT method shown in Figure 14,the tracking efficiency using MPPT fuzzy logic controller is98.13% which is higher than the tracking efficiency obtainedusing P&O MPPT method which is 97.15%. On the otherhand the oscillation around the MPP when the P&O is usedis much higher than that when the FLC is used for MPPtracking. The system performance shows that the proposedsystem is well functioning to obtain the maximum powerthat the PV module is capable of producing under differentambient conditions.
7. Experimental Setup
The implementation of the MPPT hardware setup is done byusing dSPACE real-time control. Figure 15 shows the blockdiagram of the hardware step while Figure 16 shows thehardware setup of the MPPT system. In the hardware setup,one BP 3115J PV module is connected to the DC-DC buck-boost converter. Data acquisition and the control system
0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.100
50100150
π(W
)
Time (s)
(a)
05
101520
0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10
π(V
)
Time (s)
(b)
0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1026
10
πΌ(A
)
Time (s)
(c)
Figure 13: MPP tracking with FLC.
are implemented by using dSPACE 1104 software and digitalsignal processor card on PC. The PV voltage and the PVcurrentmust be initially measured. In this system, the voltageis measured by using the voltage divider while the PV currentis measured by using the LTS 25-NP current sensor. Theanalog measured quantities of the PV voltage and PV currentwhich are fed to the A/D converter of the dSPACE in order tobe used in the SIMULINK MPPT control block. The MPPTcontrol which is constructed on MATLAB/SIMULINK isshown in Figure 17.
The signal applied to a dSPACE A/D channel must be inthe range from β10V to +10V. A signal of +10V gives aninternal value of 1.00 within SIMULINK.
8 International Journal of Photoenergy
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1Time (s)
050
100150
π(W
)
(a)
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1Time (s)
01020
π(V
)
(b)
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1Time (s)
26
10
πΌ(A
)
(c)
Figure 14: MPP tracking with P&O method.
IRF3710
1000
BYV32-200
Current sensor
dSPACE 1104 system
Gate drivecircuit
25+ 273
PV moduleBP 3115 J
39 K
120 K 100 V, 57 A16 A
5 ohm120 V
πΏ
+
+
+ ++
+
ππ
π·
Solar radiation (W/m2) πΆ1πΆ2
π πΏ
Temperature (βK)
Figure 15: Block diagram of the hardware setup.
Figure 16: The hardware setup of the system.
Every signal came fromA/D convertermust bemultipliedby 10. The filters are used for removing any high-frequencynoise or any switching noise that appears in the signals. Asshown in Figure 17 the instantaneous measured voltage andcurrent are then multiplied by each other to obtain the PVinstantaneous power. The PV voltage and the PV power arethen applied to the MPPT algorithm to generate the requiredduty cycle. The output signal of the MPPT algorithm is then
applied to the DS1104SL DSP PWM block which is used togenerate the required switching signal to drive the MOSFET.The generated PWMsignals should not be connected directlyto the MOSFET since the maximum current drown fromdSPACE board must not exceed 13mA. For this reason andfor the isolation purposes a 6N137 optocoupler is used. ThePWM generated signal from the dSPACE is connected to the6N137 optocoupler and the output of the optocoupler is thenconnected to the MOSFET gate on the buck-boost converterand manage the on-off time of the switch.
To verify the function and the performance of theproposed FLC MPPT method, the method is experimentallyimplemented by using dSPACE 1104 data acquisition system.In order to start real-time tracking of the MPP of thePV module, the SIMULINK MPPT control block, must bedownloaded to the dSPACE board to generate C code ofthe MPPT control block. To successfully track the MPP,some modifications were taken into consideration whenthe proposed method is experimentally implemented. Themembership function of the input variable Ξπ is modified asshown in Figure 18. On the other hand, some of rules are also
International Journal of Photoenergy 9
Voltage sensorcalibration
3
1
π·
MPPTalgorithm
Product
10Gain3
10Gain2
In1
Current sensorcalibration
0
1
0
Butter
Analogfilter design5
Butter
Analogfilter design1
9
3
In1ADCchannel
ADCchannel
Out1
Out2
2
[π]
[π]
+βπ
π
DS1104ADC C6
DS1104ADC C5
DS1104SL DSP PWM
πΌPVa
πPVa
[πPV ]
[πPV ]
πPV
Γ
Figure 17: MPPT SIMULINK model implemented in dSPACE 1104.
0.2
0.40 0.1 1
0
1
β0.1
β0.2
β0.4
β1
Figure 18: Modified membership function of the input variable Ξπ.
0400800
πΊ(W
/m2)
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Time (s)
08:38AM
02:38PM
12:00PM
2.2Γ104
(a)
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 20
204060
Time (s)
08:38AM
02:38PM
12:00PM
πPV(β
C)
2.2Γ104
(b)
Figure 19: Changing the ambient conditions on October 2, 2012.
tuned in order to obtain a better performance. Table 5 showsthe modified FLC rules.
After doing the above-mentioned modifications, FLCMPPT method is tested under some different ambientconditions. Figure 19 shows the changing in the ambient
Table 5: Modified rule base used in the fuzzy logic controller.
ΞπΞπ
NB NS Z PS PBNB PB NB NB NS NSNS PB NB NB NS NSZ NS NS NS PB PBPS NS PB PS NB PBPB NB NB PB PS PB
conditions on 02-10-2012 starting from 08:38 AM to 02:38PM. Figure 20(a) shows the changing in the solar radia-tion while the lower plot shows the changing in the PVtemperature. FLC MPPT method tracked the MPP of thePV module successfully as shown in Figure 20. The upperplot in this figure shows the maximum power tracked.Figures 20(b) and 20(c) show the PV voltage and the PVcurrent at the maximum power. Figure 20(d) the duty ratiowhich is generated by the proposed FLC method. Duty ratiois measured at the output of the MPPT block which is thendirected to the PWM in order to generate the switchingpulses of the MOSFET. It is noted that the proposed FLCMPPTmethod tracked themaximumpower successfully andaccurately with fast response.
Having a deep investigation on the proposed MPPTsystem performance under rapidly changing solar radiation,the PV module is covered by an opaque cloth to prevent theincidence of the solar radiation on the PV module. Variationof the power, the voltage, and the current of the system isshown in Figure 21. As shown in this figure, the proposedMPPT method has tracked the maximum power effectivelyand accurately under rapidly changing solar radiation.
10 International Journal of Photoenergy
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 20
50100
Time (s)
02:38PMPM
12:00AM
08:38
πPV
max
(W)
2.2Γ104
(a)
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 20
1018
Time (s)
02:38PMPM
12:00AM
08:38
πPV
max
(V)
2.2Γ104
(b)
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2058
Time (s)
02:38PMPM
12:00AM
08:38
πΌ PV
max
(A)
2.2Γ104
(c)
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Time (s)
00.5
1
Dut
y ra
tio
02:38PMPM
12:00AM
08:38
2.2Γ104
(d)
Figure 20: Experimental tracking behavior of the FLC MPPT. 02-10-2012.
Time (s)0 50 100 150 200 250 300 350 400
050
100
πPV
max
(W)
(a)
Time (s)0 50 100 150 200 250 300 350 400
01020
πPV
max
(V)
(b)
Time (s)0 50 100 150 200 250 300 350 400
05
πΌ PV
max
(A)
(c)
Time (s)0 50 100 150 200 250 300 350 400
00.5
1
Dut
y ra
tio
(d)
Figure 21: Performance of the FLC MPPT method under rapidly changing solar radiation.
8. Conclusion
Photovoltaic model using MATLAB/SIMULINK and thedesign of appropriate DC-DC buck-boost converter with amaximum power point tracking facility are presented in thispaper. MPPT is achieved using fuzzy logic controller whichenhanced the performance of the MPPT and eliminated thecomplexity in the computation needed.The proposed systemis simulated using MATLAB/SIMULINK and tested underdifferent ambient conditions to show the tracking behavior.The tracking behavior shows that the proposed systemsuccessfully and accurately tracked the maximum powerpoint with better performance than that of conventionalmethod. Experimental implementation of the MPPT systemis presented in this paper where data acquisition and thecontrol of the proposed FLC MPPT method are achieved bydSPACE 1104. The practical results show that the proposedmethod tracked the MPP effectively and accurately with fastresponse. Furthermore, tests verified that the proposed FLCmethod is well functioning with a good performance onrapidly changing atmospheric conditions.The results indicatethat the designed MPP tracker is capable of tracking the PVmodule maximum power and hence improves the efficiencyof the PV system.
Conflict of Interests
The authors of this paper assure that they do not have anyinterest in dSPACE 1104 board and its software. They useddSPACE 1104 for research purposes onlywithout any relationswith the manufacturer or dealer.
Acknowledgment
This work was financially supported by NPST program, KingSaud University, Project no. 09 ENE 741-02.
References
[1] J. L. Santos, F. Antunes, A. Chehab, and C. Cruz, βA maximumpower point tracker for PV systems using a high performanceboost converter,β Solar Energy, vol. 80, no. 7, pp. 772β778, 2006.
[2] B. Khan, Non-Conventional Energy Resources, Tata McGraw-Hill Education, Noida, India, 2006.
[3] V. A. Chaudhari, Automatic peak power tracker for solar pvmodules using dspacer software [thesis], Maulana Azad NationalInstitute of Technology (Deemed University), 2005.
[4] T. C. Yu and T. S. Chien, βAnalysis and simulation of charac-teristics and maximum power point tracking for photovoltaic
International Journal of Photoenergy 11
systems,β in Proceedings of the International Conference onPower Electronics and Drive Systems (PEDS β09), pp. 1339β1344,January 2009.
[5] A. M. Bazzi and S. H. Karaki, βSimulation of a new maxi-mum power point tracking technique for multiple photovoltaicarrays,β in Proceedings of the IEEE International Conferenceon Electro/Information Technology (EIT β08), pp. 175β178, May2008.
[6] C. Liu, B. Wu, and R. Cheung, βAdvanced algorithm for MPPTcontrol of photovoltaic systems,β in Canadian Solar BuildingsConference Montreal, 2004.
[7] A. Yafaoui, B. Wu, and R. Cheung, βImplementation of max-imum power point tracking algorithm for residential pho-tovoltaic systems,β in Proceedings of the 2nd Canadian SolarBuilding Conference, 2007.
[8] V. Salas, E. OlΔ±as, A. Barrado, and A. Lazaro, βReview of themaximum power point tracking algorithms for stand-alonephotovoltaic systems,β Solar Energy Materials and Solar Cells,vol. 90, no. 11, pp. 1555β1578, 2006.
[9] D. Hohm and M. Ropp, βComparative study of maximumpower point tracking algorithms using an experimental, pro-grammable, maximum power point tracking test bed,β inProceedings of the 28th IEEE Photovoltaic Specialists Conference,pp. 1699β1702, 2000.
[10] T. Esram andP. L. Chapman, βComparison of photovoltaic arraymaximum power point tracking techniques,β IEEE Transactionson Energy Conversion, vol. 22, no. 2, pp. 439β449, 2007.
[11] J. H. R. Enslin,M. S.Wolf, D. B. Snyman, andW. Swiegers, βInte-grated photovoltaic maximum power point tracking converter,βIEEE Transactions on Industrial Electronics, vol. 44, no. 6, pp.769β773, 1997.
[12] D. Y. Lee, H. J. Noh, D. S. Hyun, and I. Choy, βAn improvedMPPT converter using current compensation method for smallscaled PV-applications,β in Proceedings of the 18th Annual IEEEApplied Power Electronics Conference and Exposition, vol. 1, pp.540β545, February 2003.
[13] Y. E.Wu, C. L. Shen, and C. Y.Wu, βResearch and improvementof maximum power point tracking for photovoltaic systems,β inProceedings of the International Conference on Power Electronicsand Drive Systems (PEDS β09), pp. 1308β1312, January 2009.
[14] J. Li and H. Wang, βMaximum power point tracking of pho-tovoltaic generation based on the fuzzy control method,β inProceedings of the 1st International Conference on SustainablePower Generation and Supply (SUPERGEN β09), pp. 1β14, April2009.
[15] I. Houssamo, F. Locment, and M. Sechilariu, βMaximumpower tracking for photovoltaic power system: developmentand experimental comparison of two algorithms,β RenewableEnergy, vol. 35, no. 10, pp. 2381β2387, 2010.
[16] J. Zhang, T. Wang, and H. Ran, βA maximum power pointtracking algorithm based on gradient descent method,β inProceedings of the IEEE Power and Energy Society GeneralMeeting (PES β09), pp. 1β5, July 2009.
[17] M.G. Villalva, J. R. Gazoli, and E. Ruppert Filho, βModeling andcircuit-based simulation of photovoltaic arrays,β in Proceedingsof the Brazilian Power Electronics Conference (COBEP β09), pp.1244β1254, October 2009.
[18] G.Walker, βEvaluatingMPPT converter topologies using amat-lab PVmodel,β Journal of Electrical and Electronics Engineering,Australia, vol. 21, no. 1, pp. 49β55, 2001.
[19] X. Liu, An improved perturbation and observation maximumpower point tracking algorithm for PV panels [M.S. thesis],Concordia University, 2004.
[20] A. M. Z. Alabedin, E. F. El-Saadany, and M. M. M. SalamaA, βMaximum power point tracking for Photovoltaic systemsusing fuzzy logic and artificial neural networks,β in Power andEnergy Society General Meeting, pp. 1β9, 2011.
[21] C. Larbes, S. M. AΔ±t Cheikh, T. Obeidi, and A. Zerguerras,βGenetic algorithms optimized fuzzy logic control for the max-imum power point tracking in photovoltaic system,β RenewableEnergy, vol. 34, no. 10, pp. 2093β2100, 2009.
[22] F. Chekired, C. Larbes, D. Rekioua, and F. Haddad, βImple-mentation of a MPPT fuzzy controller for photovoltaic systemson FPGA circuit,β Proceedings of the Impact of IntegratedClean Energy on the Future of the Mediterranean Environment(MEDGREEN β11), pp. 541β549, April 2011.
[23] M. Adly, H. El-Sherif, and M. Ibrahim, βMaximum powerpoint tracker for a PV cell using a fuzzy agent adapted by thefractional open circuit voltage technique,β in IEEE InternationalConference on Fuzzy Systems, pp. 1918β1922, 2011.
[24] M. A. Islam, A. B. Talukdar, N. Mohammad, and P. K. S.Khan, βMaximum power point tracking of photovoltaic arraysin matlab using fuzzy logic controller,β in Proceedings of theAnnual IEEE India Conference: Green Energy, Computing andCommunication (INDICON β10), pp. 1β4, December 2010.
[25] M. S. Ngan and C. W. Tan, βA study of maximum power pointtracking algorithms for stand-alone photovoltaic systems,β inProceedings of the IEEE Applied Power Electronics Colloquium(IAPEC β11), pp. 22β27, April 2011.
[26] I. Purnama, Y.-K. K. Lo, and H.-J. J. Chiu, βA fuzzy controlmaximum power point tracking photovoltaic system,β in IEEEInternational Conference on Fuzzy Systems, pp. 2432β2439, 2011.
[27] Y. H. Chang and W. F. Hsu, βA maximum power point trackingof PV system by adaptive fuzzy logic control,β in InternationalMultiConference of Engineers and Computer Scientists (IMECSβ11), pp. 975β980, March 2011.
Submit your manuscripts athttp://www.hindawi.com
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Inorganic ChemistryInternational Journal of
Hindawi Publishing Corporation http://www.hindawi.com Volume 2014
International Journal ofPhotoenergy
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Carbohydrate Chemistry
International Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Journal of
Chemistry
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Advances in
Physical Chemistry
Hindawi Publishing Corporationhttp://www.hindawi.com
Analytical Methods in Chemistry
Journal of
Volume 2014
Bioinorganic Chemistry and ApplicationsHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
SpectroscopyInternational Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014
Medicinal ChemistryInternational Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Chromatography Research International
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Applied ChemistryJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Theoretical ChemistryJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Journal of
Spectroscopy
Analytical ChemistryInternational Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Quantum Chemistry
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Organic Chemistry International
ElectrochemistryInternational Journal of
Hindawi Publishing Corporation http://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
CatalystsJournal of