Received July 8, 2019, accepted July 29, 2019, date of publication August 2, 2019, date of current version August

15, 2019.
Digital Object Identifier 10.1109/ACCESS.2019.2932694

Design and Hardware Implementation of New

Adaptive Fuzzy Logic-Based MPPT Control
Method for Photovoltaic Applications
HEGAZY REZK 1,2 , MOKHTAR ALY 3,4 , (Member, IEEE), MUJAHED AL-DHAIFALLAH 5,

AND MASAHITO SHOYAMA 6 , (Senior Member, IEEE)

1 College of Engineering at Wadi Addawasir, Prince Sattam Bin Abdulaziz University, Wadi Aldawaser 11991, Saudi Arabia
2 ElectricalEngineering Department, Faculty of Engineering, Minia University, Minia 61111, Egypt
3 ElectricalEngineering Department, Faculty of Engineering, Aswan University, Aswan 81542, Egypt
4 Solar Energy Research Center (SERC-Chile), Universidad Técnica Federico Santa María, Valparaíso 2390123, Chile
5 Systems Engineering Department, King Fahd University of Petroleum and Minerals, Dhahran 31261, Saudi Arabia
6 Department of Electrical and Electronic Engineering, Kyushu University, Fukuoka 819-0395, Japan

Corresponding authors: Hegazy Rezk ([email protected]) and Mujahed Al-Dhaifallah ([email protected])

This work was supported in part by the Solar Energy Research Center (SERC-Chile) under Grant CONICYT/FONDAP/15110019, and in
part by the Advanced Center for Electrical and Electronic Engineering (AC3E) under Grant CONICYT/BASAL/FB0008.

ABSTRACT An adaptive fuzzy logic (FL)-based new maximum power point (MPP) tracking (MPPT)
methodology for controlling photovoltaic (PV) systems is proposed, designed, and implemented in this
paper. The existing methods for implementing FL-based MPPTs lack for adaptivity with the operating point,
which varies in wide range in practical PV systems with operating irradiance and ambient temperature.
The new proposed adaptive FL-based MPPT (AFL-MPPT) algorithm is simple, accurate, and provides
faster convergence to optimal operating point. The effectiveness and feasibility verifications of the proposed
AFL-MPPT methodology are validated with considering various operating conditions at slow and fast
change of solar radiation. In addition, the simplified implementation of the proposed algorithm is carried out
using C-block in PSIM software environment, wherein the proposed algorithm and system are simulated.
Additionally, experimental results are performed using a floating-point digital signal processing (DSP)
controller (TMS320F28335) for verifying the feasibility of the proposed AFL-MPPT methodology. The
results of simulations and experimental prototypes show great consistency and prove the capability of the
new AFL-MPPT methodology to extract MPPT rapidly and precisely. The new proposed AFL-MPPT method
achieves accurate output power of the PV system with smooth and low ripple. In addition, the new proposed
AFL-MPPT method benefits fast dynamics and it reaches steady state within 0.01 s.

INDEX TERMS DSP controller, energy efficiency, fuzzy logic (FL), MPPT, photovoltaic systems.

NOMENCLATURE
PV GENERATION SIDE EG The bang-gap energy
Vpv Output voltage of the PV cell VOC Open-circuit voltage of the PV
Ipv Output current of the PV cell ISC Short-circuit current of the PV
Iph Photocurrent of the PV cell NS Number of series cells in solar PV
Io Saturation current of PV cell NP Number of parallel cells in solar PV
Vt Thermal-voltage of the PV cell
a Ideality factor of the diode
Rpm and Rsm The parallel and series resistances of the PV BOOST CONVERTER SIDE
cell model, respectively Vin Input voltage to the boost converter
Irs Reverse saturated current of the PV cell Vout Output voltage of the boost converter
D Duty cycle of boost converter
The associate editor coordinating the review of this manuscript and DMPPT Duty cycle of boost converter at MPPT
approving it for publication was Wei Wang. VMPPT PV panel voltage to the boost at MPPT
2169-3536 2019 IEEE. Translations and content mining are permitted for academic research only.
VOLUME 7, 2019 Personal use is also permitted, but republication/redistribution requires IEEE permission. 106427
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
 H. Rezk et al.: Design and Hardware Implementation of New AFL-MPPT Control Method for PV Applications

PMPPT PV panel power to the boost at MPPT and variable step size techniques [15]. Among AI-based
Rload Load resistance of the boost converter MPPT methods, FL has proven itself as an effective solution
for such nonlinear PV systems without the need for accu-
rate system data [18]–[21]. In accordance, enhancing and
AFL-MPPT CONTROLLER SIDE achieving better performance of FL based MPPT methods has
P, 1P The PV power and the change of power to MPPT become an important research topic.
controller, respectively Several combinations and design methodologies for
V , 1V The PV voltage and the change of voltage to FL-based MPPT schemes have been introduced in literature.
MPPT controller, respectively In [22], the FL was introduced for achieving MPPT of PV
E, 1E The error and the change of error inputs to the systems. The FL control has achieved a good operation of
AFL-MPPT controller, respectively MPPT compared to conventional methods. A modified design
of FL control has been proposed in [21] by using convergent
I. INTRODUCTION distribution type of the membership functions (MFs) instead
Solar energy and other renewables like geothermal, biomass, of the symmetrical distribution type of MFs. Higher precision
photovoltaic (PV), and wind energy can minimize the emis- and lower voltage fluctuations have been achieved. Although
sions of the CO2 and other harmful gases resulting from these design and implementation methods lack for adaptivity
fossil fuel. Recently, the yearly growth rates of PV generation with various operating points of PV systems.
systems have been largely increased attaining total gener- From another side, hybrid MPPT methods with FL have
ation of 402 GW in 2017 compared with total generation been introduced in the literature. These hybrid MPPT meth-
of 303 GW in 2016. It is expected that the increase PV ods possess the best features of the combined methods. The
generation penetration levels continues and achieves between FL control was combined with conventional P&O in [23].
1760 to 2500 GW by the year 2030 [1], [2]. Being avail- A lower overshoot with a stable operation has been achieved.
able everywhere, continual decrease in their cost per watt, In [18], a combined approach of FL and hill climbing (HC)
and environmental friendly have made wide spreading of methods was presented. The FL control has enabled an
PV system in comparison with the other renewable energy accurate and better performance rather than conventional
sources [3]–[5]. Concerns of energy efficiency and better HC method. Additionally, dual MPPT method using fuzzy
energy harvesting have been arisen in order to maximize their and P&O method was introduced in [19]. This method has
power densities, and to reduce the cost of their generated achieved a good performance regarding adaptivity with the
kilowatt-hour. The most conventional and effective technique operating point. However, higher complexity and slower per-
for better energy efficiency of PV systems is the maximum formance are obtained, especially in fast changing irradi-
power point tracking (MPPT) control [6]. This is due to the ance conditions. Additional techniques for FL-based MPPT
fluctuating nature of PV generation. The output power of PV combined with particle swarm optimization (PSO) have been
systems is highly dependent on their operating irradiance and presented in [24], [25]. The modified sine-cosine optimized
ambient temperature. MPPT method has been presented in [26] with using adaptive
Several MPPT extraction schemes have been developed FL method integrated with the sliding mode control. How-
in literature, including hill climbing, perturb and observe ever, their implementations lack for dynamic adjustment of
(P&O), incremental conductance and incremental resis- FL membership functions boundaries and type.
tance [7]–[12]. Surveys and comparison between MPPT The fractional order (FO) was combined with FL MPPT
methods have been introduced in [13]–[16]. The P&O repre- in [20], wherein the robustness of FL is added to the accu-
sents the most widely used method because of its simplicity racy of fractional order so as to enhance the performance of
and generality. However, fast tracking has to be compromised their tracking efficiency. The alpha factor of fractional order
with the steady state fluctuations in defining the proper step control is set to large or small value based on the dynamic
size for the P&O method. Generally, fixed step size MPPT range of the fuzzy controller operation. Therefore, wide or
methods have to consider the trade-off dilemma into con- narrow range of the input MFs of the FL controller is obtained
sideration [17]. The situation becomes worse in case of fast through adapting the alpha factor of fractional order control.
and continuous changes in the operating environmental con- Nevertheless, the complexity of implementing the fractional
ditions, including solar irradiance and ambient temperature. order controller limits this type of MPPT controllers. Addi-
Therefore, several variable step size MPPT methods have tionally, there are several complexities in the calculations of
been developed for mitigating such dilemma [11]. Although, the required alpha factor for the PV operating point.
these methods suffer from dependence of step size on the Motivated by the aforementioned weaknesses of existing
data of PV panel, which is highly related to environmental MPPT methods in the literature, a new adaptive FL-based
operating point. MPPT (AFL-MPPT) methodology is proposed in this paper.
From another side, smart artificial intelligence (AI)-based The proposed design and implementation features the high
schemes, such as the neural network, fuzzy logic (FL), neuro- accuracy of extracting MPPT, low fluctuation in the PV
fuzzy, and genetic algorithms were developed for MPPT voltages, and adaptivity with wide range of operating point
extraction and for avoiding the associated problems to fixed of PV systems. Moreover, a simplified implementation of

106428 VOLUME 7, 2019

H. Rezk et al.: Design and Hardware Implementation of New AFL-MPPT Control Method for PV Applications

FIGURE 1. Schematic diagram of PV system and control method.

the proposed controller is achieved using PSIM program in considered in this paper [28], [29]. The configuration of the
order to provide an adaptive scheme of the input MFs of the two stages PV system and controller are illustrated in Fig. 1.
proposed AFL-MPPT controller. The boost dc-dc power converter represents the first stage,
The main contribution of this paper can be summarized as wherein the maximum available energy is harvested from
following: the PV module. The boost power converter possesses the
• A new adaptive MPPT control methodology is proposed, benefits of voltage boosting, less component, and continuous
designed, and implemented based on fuzzy logic control current of the solar PV panel side. The control of this stage
method. is composed of the MPPT and the PWM modulator that
• A simplified implementation of adaptive algorithm for transfers the MPPT signal to gating pulses for the boost dc-dc
fuzzy logic based MPPT method is presented using the converter.
C-block programming in PSIM simulation program. The full bridge dc-ac converter is usually utilized in the
• A generalized implementation of fuzzy logic-based second power stage. This stage is responsible for modulat-
MPPT controllers is presented. The membership func- ing the dc output of the first stage into ac output to the
tions can be shaped in static or dynamic manner, whether utility grid. By this process, the extracted maximum power
they are symmetrical or asymmetrical types. is harvested and injected to the utility grid. The reference
The rest of paper is prepared as following: Section II current for the current controller is calculated according to
explains various PV system components and its control the MPPT power in addition to the demanded reactive power
method. The design procedures for the implementation of the by the grid system. Additionally, a phase locked loop (PLL)
new proposed AFL-MPPT technique are presented in details is employed for synchronizing the PV system with the utility
in Section III. The simulation and experimental obtained grid. There are several approaches that were developed in the
results for validating the AFL-MPPT technique under differ- literature to achieve the current control of the dc-ac power
ent conditions are introduced in Section IV. The outlines of inverter stage [30], which is not covered in this paper as it is
the research conclusion are highlighted in Section V. out of scope of this paper.
The 75-watt PV module is selected in this paper as a case
II. DESCRIPTION OF PV SYSTEM study and the PV module characteristics are tabulated and
The PV power system (PVPS) is mainly composed of PV listed in Table 1. The simulation of the PV panel is performed
panels, which are often connected in series and/or parallel in using PSIM program using 110 series connected solar PV
order to increase their generated voltage and output power. cells. The mathematical models and detailed analysis of PV
The existing PV systems in the literature can be mainly modeling can be found in [13], [31]. The component and
sectioned into two categories; 1) Single stage-based PV sys- the controller for the first power stage using the boost dc-dc
tems, wherein only one dc-ac converter stage is employed; power converter are presented in the following:
2) Two-stages-based PV systems, in which series cascaded
dc-dc converter stage and dc-ac converter stage are utilized to
connect the PV system with the utility grid [27]. According A. PV ARRAY MODEL
to the various published papers in literature to distinguish and The single diode-based PV cell model has been widely
compare the two PV systems, considering efficiency, voltage used in the literature for modeling PV cells, as shown in
boost ratio, and system cost, the two stages PV systems are Fig. 2 [13]. The PV output current IPV can be represented

VOLUME 7, 2019 106429

 H. Rezk et al.: Design and Hardware Implementation of New AFL-MPPT Control Method for PV Applications

TABLE 1. PV system parameters employed in the selected case study.

FIGURE 2. Single diode model for PV equivalent circuit.

as fellows [32], [33]:

   
VPV + Rs I PV VPV +Rsm IPV
I PV = Iph − Io exp −1 −
Vt a Rpm
(1)
where Vpv and Ipv denote to the output voltage and current of
the cell; Iph and Io represent the photocurrents and saturation;
FIGURE 3. Circuit and operation of boost dc-dc converter.
Vt represents the thermal-voltage of the PV cell; a is the
diode’s ideality factor. Whereas, Rpm and Rsm are the parallel
and series resistances of the PV cell.
important key element for selecting the boost converter com-
The photocurrent is mostly dependent on the solar PV
ponents and for controlling of the dc-dc power converter to
radiation intensity and the PV cell temperature. It can be
maximize the output harvested power of the solar PV system.
formulated as fellow:


The output and input voltages are related as following:
I ph = Isc + ki Tc − Tref R (2) 1
Vout = × Vin (4)
Moreover, the following relation can be used to estimate 1−D
the diode saturation current [34]: where Vin denotes to the input voltage to the converter from

Tc 3
 
qEG

1 1
 the PV system, and Vout is the output voltage of the boost
I o = Irs exp − (3) dc-dc power converter [37]. The duty cycle corresponding
Tref kA Tref Tc
to MPPT DMPPT of the boost converter can be estimated as
where Irs denotes to the reverse saturated current, and EG following [36]:
denotes to the bang-gap energy.
VMPPT
DMPPT = 1 − √ (5)
B. BOOST DC-DC POWER CONVERTER PMPPT × Rload
Fig. 3(a) illustrates the ideal circuit of the boost dc-dc where, VMPPT denotes to the PV panel voltage that corre-
converter. It is powered by a dc voltage source from the sponds to MPPT operating point, PMPPT is the PV panel
photovoltaic solar panels. The circuit output is controlled power of at MPPT, Rload is the equivalent resistance for the
through the duty cycle of the power semiconductor MOSFET boost power converter output load and DMPPT represents
device, which is controlled through the pulse-width modula- operating duty-cycle for the boost power converter at the
tion (PWM) [35]. The converter duty cycle D is the most MPPT operating point.

106430 VOLUME 7, 2019

H. Rezk et al.: Design and Hardware Implementation of New AFL-MPPT Control Method for PV Applications

FIGURE 4. Circuit Schematic diagram for the new proposed AFL-MPPT method for PV systems.

III. THE PROPOSED AFL-MPPT METHOD signal and 1E are estimated, respectively, as follows:
Employing FL control methods in several applications has
been increased because it is simple, does not require data P (i) − p (i − 1)
E (n) = (6)
regarding mathematical modeling, and can address the non- V (i) − V (i − 1)
linearity of systems. The nonlinear nature of solar PV panels 1E (i) = E (i) − E (i − 1) (7)
and the climate conditions result in quite complicated track-
ing behavior. Therefore, FL-based MPPT methods can be where P(i), and P(i − 1) represent the current and previ-
employed for tracking the MPP in the solar PV system with ous samples of the PV measured output power, respectively.
simplified implementation and less required data [21]. There Whereas, V (i), and V (i − 1) denote to the measured current
are several FL algorithms have been developed in literature and previous samples of the PV output voltage, respectively.
for extracting the MPPT in PV systems [18]–[21]. E(i), and E(i − 1) represent to the current and previous
The operation and design of the FL controller can be samples of the error variable, respectively. Then, the E and
summarized by three main stages. These stages include 1E values are estimated using the measured output power
fuzzification, rule evaluation, and defuzzification steps. Dur- and voltage of the solar panel based on (6) and (7).
ing the fuzzification stage, the measured changes in the In the proposed AFL-MPPT method, there are two input
PV output voltage and current are employed for determin- MFs belonging to the estimated E and 1E. The variable
ing the input membership functions (MFs) of the FL-based inputs and output MFs are divided to seven different fuzzy
MPPT controller. The MFs are assigned for each input of subsets: Pos3 (denotes to Positive and Big subset), Pos2
the FL controller in order to convert the measured change (denotes to Positive and Medium subset), Pos1 (denotes to
into suitable inputs for the FL MPPT controller. The num- Positive and Small subset), Zer0 (denotes to Zero Error sub-
ber of input MFs, which are assigned to the FL MPPT set), Neg1 (denotes to Negative and Small subset), Neg2
controller, defines the accuracy of the FL control system. (denotes to Negative and Medium subset), and Neg3 (denotes
In the rule evaluation stage, the control action is deter- to Negative and Big subset). Therefore, there are total
mined based on FLC linguistic rules, which link the logic 49 fuzzy control rules in the proposed AFL-MPPT method
functions between the input and output MFs. Result of the in implementing the algorithm. Once the proposed algorithm
rule evaluation step is a fuzzy output MF for every type estimates the values of E and 1E, the FL interface converts
of the consequent action of input MFs. In the defuzzifi- these values into linguistic variables. Lastly, the proposed
cation stage of FL controller, the predictable value of an AFC-MPPT method outputs the duty cycle demand as the
output MF is obtained and outputted to the rest of the output MF of the system. The rules that links the input and
system. output MFs for the proposed AFL-MPPT method are shown
The inputs to the FL MPPT controller include the error (E), in Table 2. The schematic diagram for the new proposed
and the change of error (1E) quantities. The error signal can AFL-MPPT method is shown in Fig. 4. Fig. 5 shows the three-
be estimated as the change in the PV extracted output power dimensional (3D) surface (surface function) representation of
divided by the change in the PV output voltage. The error the input-output MFs for the proposed AFL-MPPT method.

VOLUME 7, 2019 106431

 H. Rezk et al.: Design and Hardware Implementation of New AFL-MPPT Control Method for PV Applications

TABLE 2. The FL rules between the input and output MFs in the proposed
AFL-MPPT method.

FIGURE 6. The main types of membership functions for the Fuzzy MPPT
controller.

describe the relationship relating both the input and output

MFs. In case of partial shading condition of series connected
PV modules, the new proposed AFL-MPPT method can be
integrated with existing global MPPT searching algorithms in
literature [24], [37]. The integration of the moth-flame, arti-
ficial bee colony (ABC), particle swarm optimization (PSO),
ant colony optimization algorithm (ACO), and other optimiz-
FIGURE 5. Three-dimensional 3D surface function for the proposed ers [38], [39] with the new proposed AFL-MPPT method
AFL-MPPT method. would provide enhanced performance at steady state for
global MPPT methods in series-connected PV modules.
Fig. 6 shows the symmetrical and asymmetrical Fuzzy
In the proposed AFL-MPPT algorithm, triangular mem- membership function for the inputs and outputs of the Fuzzy
bership functions are used for both inputs and output MFs controller. The boundaries and shapes of the membership
due to their implementation simplicity using low cost digital functions in fuzzy logic-based MPPT methods are usually
controller systems. Linguistic variables, which are assigned defined by the power level and parameters of the power
for the 1D at various combinations of the E and 1E input conversion system. The new proposed implementation using
MFs, are relying on the rating and type of the employed power the C-code enables the easy adaptivity of the type of the mem-
converter, in addition to the user knowledge. The last stage bership function. In particular, the points X1, X2, ..,X7 can
of the proposed AFL-MPPT method is the defuzzification, be adjusted in a static or dynamic manner according to the
wherein the output of the proposed controller is transformed required response. This in turn enables the adaptive imple-
to numerical output variable using the controller linguistic mentation for the MPPT controller.
output variable. The output of this stage is analog signal Fig. 7 shows the flowchart of the new proposed
that denotes to the duty cycle command 1D for the boost AFL-MPPT implementation. In stage 1, the parameters for
converter. the new AFL-MPPT control method are calculated using the
The output of the fuzzy logic controller is compared with measured current and voltage of the PV module. In stage 2,
sawtooth triangular signal, and the comparator generates the membership functions types and boundaries (X for inputs
the operating duty cycle for the power converter in accor- and Y for outputs) are defined according to the system
dance. Based on Table 2, there are 49 IF-THEN fuzzy rules parameters and the required system response. In stage 3,
have been employed in the proposed AFL-MPPT method to the estimated parameters and the defined membership

106432 VOLUME 7, 2019

H. Rezk et al.: Design and Hardware Implementation of New AFL-MPPT Control Method for PV Applications

FIGURE 8. Simulation results for the new proposed AFL-MPPT method at

transient starting.

FIGURE 7. The flowchart of the new proposed AFL-MPPT controller

implementation.
Firstly, the proposed AFL-MPPT method is tested under
the transient response condition at starting the PV systems.
functions are fed into the proposed AFL-MPPT controller, Fig. 8 shows the simulated waveforms of the PV output
which outputs the duty cycle for the boost converter. current IPV , voltage VPV , and power PPV at starting from
zero irradiance level to 1000 W/m2 irradiance level. The
IV. RESULTS AND DISCUSSIONS new proposed AFL-MPPT controller reaches the steady state
A. SIMULATIONS RESULTS operating point in less than 0.01 seconds. In addition, it has
Simulation programs and experimental prototypes are become clear that the proposed AFL-MPPT tracks accurately
employed for verifying the superior performance of the new the reference maximum output power of the solar PV system.
proposed AFL-MPPT controller for PV systems. The PSIM The extracted power in the solar PV system is also smooth
software is employed for designing and investigating the sim- without fluctuations in the steady state operating point. This
ulation performance for the proposed AFL-MPPT method. is superior performance for the new proposed AFL-MPPT
The PVPS, shown in Fig. 1, is simulated using one solar PV controller in comparison with the widely used conventional
panel with 75 W maximum output power. The dc/dc boost MPPT methods in the literature.
power converter is designed so as to operate in the continuous The proposed AFL-MPPT method has been tested at sud-
conduction current mode, and the switching frequency is den change in the radiance in order to verify its tracking abil-
selected to be 30 kHz. The parameters of the boost power ity of the operating point of the PV system. Fig. 9 shows the
converter are as follows: the input inductance LB is 1 mH, simulation results at sudden decrease of the solar PV irradi-
the boost output capacitor Cdc is 47 µF, and output load ance from 1000 W/m2 to 600 W/m2 at the time of 0.1 seconds.
is 50  resistive loads. For achieving MPPT of the solar The results show that the new proposed AFL-MPPT achieves
PV panel, the panel voltage and the current are sensed and fast transient response in addition to accurate tracking for
employed as inputs for the MPPT controller block, as shown the maximum output power operating point. This in turn
in Fig. 4. The output of the MPPT controller block represents verifies the tracking efficiency in the proposed AFL-MPPT
the gating pulse, which is employed for driving the IGBT for PV systems with fast changing irradiance. The proposed
switch of the boost dc-dc converter. AFL-MPPT achieved low fluctuations at the sudden change

VOLUME 7, 2019 106433

 H. Rezk et al.: Design and Hardware Implementation of New AFL-MPPT Control Method for PV Applications

FIGURE 9. Simulation results for the new proposed AFL-MPPT method at

step change in the radiance.

FIGURE 10. Simulations results of the new proposed AFL-MPPT method

in addition to elimination the power fluctuations at steady at slow tracking response of the radiance.
state operating point, regardless of the irradiance level.
In several applications, there are gradual increase/decrease
of the solar irradiance, which requires proper design of the B. EXPERIMENTAL RESULTS
MPPT algorithm. Therefore, the proposed AFL-MPPT con- Experimental prototype was developed in laboratory to val-
troller has been tested at gradual change of the irradiance, idate the tracking performance for the new AFL-MPPT
by linear decrease of the solar PV irradiance level from method, as illustrated in Fig. 11. The AFL-MPPT controller
1000 W/m2 to 600 W/m2 in 0.1 seconds. The results in algorithm has been implemented by using floating point digi-
Fig. 10 show that the proposed AFL-MPPT method can track tal signal processor (TMS320F28335) DSP unit. The Agilent
the slow changing irradiance at high efficiency, which makes PV simulator is used experimentally for emulating the solar
the PV system capable of producing the maximum available PV system and the changes in the irradiance levels. The
power in the whole operating period of the solar PV system. output voltage and current quantities of the PV system are
The proposed controller can achieve MPPT with elimi- measured using LEM sensors with LEM LA 50-P, and LEM
nating power fluctuations, which disadvantages most of the LV 25-P transducers, respectively.
existing MPPT methods in the literature. In addition, the pro- The new AFL-MPPT controller has been tested exper-
posed AFL-MPPT method has been tested at sudden increase imentally for transient starting operating point with start-
of the solar irradiance at time 0.25 seconds from 600 W/m2 to ing solar radiation from 0 W/m2 level to 1000 W/m2
1000 W/m2 . It is clear that the proposed controller achieved level. Fig. 12 displays the obtained experimental results for
fast transient tracking of the irradiance increase with high the new AFL-MPPT controller at transient starting point.
amount in short time. In addition, the proposed controller It has become clear that the proposed AFL-MPPT controller
can mitigate the fluctuated power of the conventional MPPT achieves fast and accurate tracking of the operating point
method. Therefore, the simulations results confirm that the of the solar PV system. The steady state point has been
new AFL-MPPT controller and design method achieve fast reached after 0.03 seconds with small power fluctuations and
MPPT with low fluctuations during the steady state. In addi- stable operation of the system. The results of the transient
tion, to the ability to track the various fluctuated operating starting coincide with the simulated test case study of the
pointing of practical PV system. PV system.

106434 VOLUME 7, 2019

H. Rezk et al.: Design and Hardware Implementation of New AFL-MPPT Control Method for PV Applications

FIGURE 11. Experimental setup of PV system with new AFL-MPPT and

controller.

FIGURE 13. Experimental results for the new proposed AFL-MPPT

method at steady state irradiance.

TABLE 3. Circuit and PV parameters for the selected case study.

FIGURE 12. Experimental results of the proposed AFL-MPPT method at

transient starting.

In addition, the smooth operation with small power fluc-

tuations of the new proposed AFL-MPPT is experimentally
investigated. Zoomed-in results of the output power, voltage,
and current waveforms of the PV system are illustrated in
Fig. 13(a) at steady state 1000 W/m2 irradiance level. The
results confirm that steady state operating performance of
the new AFL-MPPT algorithm achieves the maximum the-
oretical output power for the solar PV system. Moreover,
mitigated power fluctuations are obtainable using the new The new proposed AFL-MPPT method has been tested
proposed AFL-MPPT method compared to traditional design experimentally at changing irradiance levels from 1000 W/m2
methods in the literature. Moreover, the stable and prober to 600 W/m2 using the Agilent PV emulator. Fig. 14 shows
design of the boost power stage are investigated experimen- the experimental results of the output voltage, current, and
tally. The input and output waveforms of the deigned boost extracted output power of the PV system at different irra-
dc-dc converter are illustrated in Fig. 13(b). It can be seen diance levels. The new proposed AFL-MPPT controller
the stepping up ability for the boost converter with low ripple achieves fast and accurate tracking of optimal operating point.
in the output voltage. The proposed PV system and power In addition, the proposed controller possesses smooth steady
converter can also achieve stable operation of the PV system. state output performance with eliminating the fluctuated

VOLUME 7, 2019 106435

 H. Rezk et al.: Design and Hardware Implementation of New AFL-MPPT Control Method for PV Applications

TABLE 4. Performance comparison of MPPT methods.

clear that the new proposed AFL-MPPT method possesses

superior performance over the existing fuzzy logic methods
in the literature.
The main advantages of the new proposed AFL-MPPT
controller over the traditional methods in the literature can
be summarized as following:
• The new proposed controller is simple, accurate and
provides faster convergence to optimal operating point.
• In addition, the simplified implementation of the pro-
posed algorithm is carried out using the C-block,
which makes the new proposed algorithm general, and
adaptive.
• The shapes of input and output membership functions
for the AFL-MPPT method can be adjusted in static and
dynamic manner in the proposed algorithm for achieving
specific system response.
FIGURE 14. Experimental results for the new proposed AFLC-MPPT
method at step change in solar irradiance.
V. CONCLUSION
This paper presented a new modified controller and design
method for FL-based MPPT tracking for PV systems. The
power output, which exists in the conventional tracking meth- new proposed method represents an adaptive FL-based MPPT
ods. The obtained experimental results match with the simu- (AFL-MPPT method). The main advantages in the pro-
lated results of the new AFL-MPPT controller for the various posed AFL-MPPT method are accurate and adaptive tracking
test cases and operating conditions. performance of the operating maximum power extraction
point of the solar PV system, and the mitigation of power
C. PERFORMANCE COMPARISON fluctuations in transient and steady state operating points.
Table 4 presents performance comparison of the new pro- Moreover, the proposed AFL-MPPT method achieves faster
posed AFL-MPPT with the MPPT methods in the literature MPPT convergence with simple implementation. The pro-
based on the fuzzy logic principle. The comparison includes posed AFL-MPPT controller can effectively overcome the
the principle of operation, implementation complexity, appli- demerits of the existing MPPT methods in the literature. The
cation to symmetrical (Sym.) and/or asymmetrical (Asym.) proposed AFL-MPPT method has been implemented using
membership functions, training necessities, and adaptivity. C-block in PSIM environment and verified by experimental
In the traditional fuzzy based MPPT methods in the litera- prototyping of 75-watt PV module. The obtained experi-
ture, a compromise between the implementation complexity mental results coincide with the obtained simulation results,
and generalization is required. However, the new proposed which verify the superior performance for the new proposed
AFL-MPPT method benefits simple, accurate and provides control method and design procedures over the conventional
faster convergence to optimal operating point. It has become MPPT extraction schemes in the literature. The new proposed

106436 VOLUME 7, 2019

H. Rezk et al.: Design and Hardware Implementation of New AFL-MPPT Control Method for PV Applications

AFL-MPPT method benefits high tracking efficiency and fast [19] A. Al Nabulsi and R. Dhaouadi, ‘‘Efficiency optimization of a DSP-based
dynamics by reaching steady state point within 0.01 seconds. standalone PV system using fuzzy logic and dual-MPPT control,’’ IEEE
Trans. Ind. Informat., vol. 8, no. 3, pp. 573–584, Aug. 2012.
Additionally, the new proposed implementation method can [20] S. Tang, Y. Sun, Y. Chen, Y. Zhao, Y. Yang, and W. Szeto, ‘‘An enhanced
be easily integrated with the existing global MPPT searching MPPT method combining fractional-order and fuzzy logic control,’’ IEEE
algorithms. J. Photovolt., vol. 7, no. 2, pp. 640–650, Mar. 2017.
[21] A. El Khateb, N. A. Rahim, J. Selvaraj, and M. N. Uddin, ‘‘Fuzzy-logic-
controller-based SEPIC converter for maximum power point tracking,’’
REFERENCES IEEE Trans. Ind. Appl., vol. 50, no. 4, pp. 2349–2358, Jul./Aug. 2014.
[22] T. L. Kottas, Y. S. Boutalis, and A. D. Karlis, ‘‘New maximum power
[1] (2015). REN21—Renewables 2016 Global Status Report. [Online]. Avail-
point tracker for PV arrays using fuzzy controller in close cooperation with
able: http://www.ren21.net/status-of-renewables/global-status-report/
fuzzy cognitive networks,’’ IEEE Trans. Energy Convers., vol. 21, no. 3,
[2] A. Elmelegi, M. Aly, E. M. Ahmed, and A. G. Alharbi, ‘‘A simpli-
pp. 793–803, Sep. 2006.
fied phase-shift PWM-based feedforward distributed MPPT method for [23] M. A. A. M. Zainuri, M. A. M. Radzi, A. C. Soh, and N. A. Rahim,
grid-connected cascaded PV inverters,’’ Sol. Energy, vol. 187, pp. 1–12, ‘‘Development of adaptive perturb and observe-fuzzy control maximum
Jul. 2019. power point tracking for photovoltaic boost dc-dc converter,’’ IET Renew.
[3] M. B. Shadmand, M. Mosa, R. S. Balog, and H. A. Rub, ‘‘Maximum Power Gener., vol. 8, no. 2, pp. 183–194, Mar. 2014.
power point tracking of grid connected photovoltaic system employing [24] N. Priyadarshi, S. Padmanaban, P. K. Maroti, and A. Sharma, ‘‘An exten-
model predictive control,’’ in Proc. IEEE Appl. Power Electron. Conf. sive practical investigation of FPSO-based MPPT for grid integrated PV
Expo. (APEC), Mar. 2015, pp. 3067–3074. system under variable operating conditions with anti-islanding protection,’’
[4] M. Aly, E. M. Ahmed, and M. Shoyama, ‘‘Modulation method for improv- IEEE Syst. J., vol. 13, no. 2, pp. 1861–1871, Jun. 2018.
ing reliability of multilevel T-type inverter in PV systems,’’ IEEE J. Emerg. [25] N. Priyadarshi, S. Padmanaban, M. S. Bhaskar, F. Blaabjerg, and
Sel. Topics Power Electron., to be published. A. Sharma, ‘‘Fuzzy SVPWM-based inverter control realisation of
[5] K. S. Tey and S. Mekhilef, ‘‘Modified incremental conductance algo- grid integrated photovoltaic-wind system with fuzzy particle swarm
rithm for photovoltaic system under partial shading conditions and load optimisation maximum power point tracking algorithm for a grid-
variation,’’ IEEE Trans. Ind. Electron., vol. 61, no. 10, pp. 5384–5392, connected PV/wind power generation system: Hardware implemen-
Oct. 2014. tation,’’ IET Electr. Power Appl., vol. 12, no. 7, pp. 962–971,
[6] R. B. A. Koad, A. F. Zobaa, and A. El-Shahat, ‘‘A novel MPPT algorithm Apr. 2018.
based on particle swarm optimization for photovoltaic systems,’’ IEEE [26] S. Padmanaban, N. Priyadarshi, J. B. Holm-Nielsen, M. S. Bhaskar,
Trans. Sustain. Energy, vol. 8, no. 2, pp. 468–476, Apr. 2017. F. Azam, A. K. Sharma, and E. Hossain, ‘‘A novel modified sine-cosine
[7] T. K. Soon and S. Mekhilef, ‘‘A fast-converging MPPT technique for pho- optimized MPPT algorithm for grid integrated PV system under real
tovoltaic system under fast-varying solar irradiation and load resistance,’’ operating conditions,’’ IEEE Access, vol. 7, pp. 10467–10477, 2019.
IEEE Trans Ind. Informat., vol. 11, no. 1, pp. 176–186, Feb. 2015. [27] R. Hasan, S. Mekhilef, M. Seyedmahmoudian, and B. Horan, ‘‘Grid-
[8] A. Elmelegi, M. Aly, and E. M. Ahmed, ‘‘Developing phase-shift PWM- connected isolated PV microinverters: A review,’’ Renew. Sustain. Energy
based distributed MPPT technique for photovoltaic systems,’’ in Proc. Rev., vol. 67, pp. 1065–1080, Jan. 2017.
Int. Conf. Innov. Trends Comput. Eng. (ITCE), Aswan, Egypt, Feb. 2019, [28] Y. Zhu, J. Yao, and D. Wu, ‘‘Comparative study of two stages and single
pp. 492–497. stage topologies for grid-tie photovoltaic generation by PSCAD/EMTDC,’’
[9] T. Radjai, L. Rahmani, S. Mekhilef, and J. P. Gaubert, ‘‘Implementation of in Proc. Int. Conf. Adv. Power Syst. Automat. Protection, Beijing, China,
a modified incremental conductance MPPT algorithm with direct control Oct. 2011, pp. 1304–1309.
based on a fuzzy duty cycle change estimator using dSPACE,’’ Sol. Energy, [29] P. K. Sahu, P. Shaw, and S. Maity, ‘‘Modeling and control of grid-
vol. 110, pp. 325–337, Dec. 2014. connected DC/AC converters for single-phase micro-inverter application,’’
[10] G. M. Dousoky and M. Shoyama, ‘‘New parameter for current-sensorless in Proc. Annu. IEEE India Conf. (INDICON), New Delhi, India, Dec. 2015,
MPPT in grid-connected photovoltaic VSIs,’’ Sol. Energy, vol. 143, pp. 1–6.
pp. 113–119, Feb. 2017. [30] P. Sanchis, A. Ursaea, E. Gubia, and L. Marroyo, ‘‘Boost DC-AC inverter:
[11] E. M. Ahmed and M. Shoyama, ‘‘Scaling factor design based variable A new control strategy,’’ IEEE Trans. Power Electron., vol. 20, no. 2,
step size incremental resistance maximum power point tracking for PV pp. 343–353, Mar. 2005.
systems,’’ J. Power Electron., vol. 12, no. 1, pp. 164–171, Jan. 2012. [31] A. Fathy and H. Rezk, ‘‘A novel methodology for simulating maximum
[12] K. S. Tey, S. Mekhilef, M. Seyedmahmoudian, B. Horan, A. T. Oo, and power point trackers using mine blast optimization and teaching learning
A. Stojcevski, ‘‘Improved differential evolution-based MPPT algorithm based optimization algorithms for partially shaded photovoltaic system,’’
using SEPIC for PV systems under partial shading conditions and load J. Renew. Sustain. Energy, vol. 8, no. 2, Mar. 2016, Art. no. 023503.
variation,’’ IEEE Trans. Ind. Informat., vol. 14, no. 10, pp. 4322–4333, [32] M. Nabil, S. M. Allam, and E. M. Rashad, ‘‘Modeling and design consid-
Oct. 2018. erations of a photovoltaic energy source feeding a synchronous reluctance
[13] H. Rezk and A. M. Eltamaly, ‘‘A comprehensive comparison of differ- motor suitable for pumping systems,’’ Ain Shams Eng. J., vol. 3, no. 4,
ent MPPT techniques for photovoltaic systems,’’ Sol. Energy, vol. 112, pp. 375–382, Dec. 2012.
[33] M. Nabil, S. M. Allam, and E. M. Rashad, ‘‘Performance improvement
pp. 1–11, Feb. 2015.
of a photovoltaic pumping system using a synchronous reluctance motor,’’
[14] N. Karami, N. Moubayed, and R. Outbib, ‘‘General review and classifica-
Electr. Power Compon. Syst., vol. 41, no. 4, pp. 447–464, Feb. 2013.
tion of different MPPT Techniques,’’ Renew. Sustain. Energy Rev., vol. 68,
[34] M. G. Villalva, J. R. Gazoli, and E. R. Filho, ‘‘Comprehensive approach
pp. 1–18, Feb. 2017.
to modeling and simulation of photovoltaic arrays,’’ IEEE Trans. Power
[15] L. M. Elobaid, A. K. Abdelsalam, and E. E. Zakzouk, ‘‘Artificial neural Electron., vol. 24, no. 5, pp. 1198–1208, May 2009.
network-based photovoltaic maximum power point tracking techniques: [35] B. Hauke, ‘‘Basic calculation of a boost converter’s power stage,’’ Texas
A survey,’’ IET Renew. Power Generat., vol. 9, no. 8, pp. 1043–1063, Instrum., Dallas, TX, USA, Appl. Rep. SLVA372C, Aug. 2015, pp. 1–7.
Nov. 2015. [36] Y. S. Kumar and R. Gupta, ‘‘Maximum power point tracking of multiple
[16] M. Seyedmahmoudian, B. Horan, T. K. Soon, R. Rahmani, A. M. T. Oo, photovoltaic arrays,’’ in Proc. Students Conf. Eng. Syst., Allahabad, India,
S. Mekhilef, and A. Stojcevski, ‘‘State of the art artificial intelligence- Mar. 2012, pp. 1–6.
based MPPT techniques for mitigating partial shading effects on PV [37] S. Padmanaban, N. Priyadarshi, M. S. Bhaskar, J. B. Holm-Nielsen,
systems–A review,’’ Renew. Sustain. Energy Rev., vol. 64, pp. 435–455, E. Hossain, and F. Azam, ‘‘A hybrid photovoltaic-fuel cell for grid inte-
Oct. 2016. gration with jaya-based maximum power point tracking: Experimental
[17] J. Kivimäki, S. Kolesnik, M. Sitbon, T. Suntio, and A. Kuperman, ‘‘Revis- performance evaluation,’’ IEEE Access, vol. 7, pp. 82978–82990, 2019.
ited perturbation frequency design guideline for direct fixed-step max- [38] H. Rezk, A. Fathy, and A. Y. Abdelaziz, ‘‘A comparison of different global
imum power point tracking algorithms,’’ IEEE Trans. Ind. Electron., MPPT techniques based on meta-heuristic algorithms for photovoltaic
vol. 64, no. 6, pp. 4601–4609, Jun. 2017. system subjected to partial shading conditions,’’ Renew. Sustain. Energy
[18] B. N. Alajmi, K. H. Ahmed, S. J. Finney, and B. W. Williams, ‘‘Fuzzy- Rev., vol. 74, pp. 377–386, Jul. 2017.
logic-control approach of a modified hill-climbing method for maximum [39] M. A. Mohamed, A. A. Z. Diab, and H. Rezk, ‘‘Partial shading mitigation
power point in microgrid standalone photovoltaic system,’’ IEEE Trans. of PV systems via different meta-heuristic techniques,’’ Renew. Energy,
Power Electron., vol. 26, no. 4, pp. 1022–1030, Apr. 2011. vol. 130, pp. 1159–1175, Jan. 2019.

VOLUME 7, 2019 106437

 H. Rezk et al.: Design and Hardware Implementation of New AFL-MPPT Control Method for PV Applications

HEGAZY REZK received the B. Eng. and M. Eng. MUJAHED AL-DHAIFALLAH received the B.Sc.
degrees in electrical engineering from Minia Uni- and M.Sc. degrees in systems engineering from
versity, Egypt, in 2001 and 2006, respectively, and King Fahd University of Petroleum and Minerals,
the Ph.D. degree from Moscow Power Engineer- Dhahran, Saudi Arabia, and the Ph.D. degree in
ing Institute, Moscow. He is an Associate Pro- electrical and computer engineering from the Uni-
fessor (on leave) with the Electrical Engineering versity of Calgary, Calgary, AB, Canada. He has
Department, Minia University. He was a Post- been an Assistant Professor of systems engineer-
doctoral Research Fellow with the Moscow State ing with King Fahd University of Petroleum and
University of Mechanical Engineering, Russia, for Minerals, since 2009. His current research inter-
six months. He was a Visiting Researcher with ests include nonlinear systems identification, con-
Kyushu University, Japan, for one year. He is currently an Associate Profes- trol systems, optimization, artificial intelligence, and renewable energy.
sor with the Electrical Engineering Department, College of Engineering at
Wadi Addwaser, Prince Sattam University, Saudi Arabia. He has authored
more than 50 technical papers. His research interests include renewable
energy, smart grid, hybrid systems, power electronics, optimization, and
artificial intelligence.

MOKHTAR ALY (S’14–M’18) received the B.Sc.

and M.Sc. degrees in electrical engineering from
Aswan University, Aswan, Egypt, in 2007 and MASAHITO SHOYAMA (M’93–SM’06) received
2012, respectively, and the Ph.D. degree from the B.S. degree in electrical engineering and the
the Department of Electrical Engineering, Faculty Dr.Eng. degree from Kyushu University, Fukuoka,
of Information Science and Electrical Engineer- Japan, in 1981 and 1986, respectively, where
ing, Kyushu University, Japan, in 2017. In 2008, he joined the Department of Electronics, as a
he joined the Department of Electrical Engineer- Research Associate, in 1986. He has been an Asso-
ing, Aswan University, as an Assistant Lecturer, ciate Professor, since 1990, and he has been a Pro-
where he has been an Assistant Professor with the fessor, since 2010. Since 2009, he has been with
Faculty of Engineering, since 2017. He is currently a Postdoctoral Researcher the Department of Electrical Engineering, Faculty
with the Solar Energy Research Center (SERC-Chile), Universidad Técnica of Information Science and Electrical Engineer-
Federico Santa María, Chile. His current research interests include reliability ing, Kyushu University. He has been active in the field of power elec-
of power electronics systems especially in renewable energy applications, tronics, especially in the areas of bi-directional converters for dc/ac power
multi-level inverters, fault tolerant control, electric vehicles, and light emit- systems, high-frequency switching converters for renewable energy sources,
ting diode (LED) lamp drivers. He is a member of the IEEE Power Electron- power factor correction (PFC) converters, and electromagnetic compatibility
ics Society (PELS), the IEEE Industrial Electronics Society (IES), and the (EMC). He is a member of IEICE, IEEJ, and SICE.
IEEE Power and Energy Society (PES).

106438 VOLUME 7, 2019