Received January 24, 2019, accepted February 12, 2019, date of publication February 15, 2019, date of current

version March 7, 2019.

Digital Object Identifier 10.1109/ACCESS.2019.2899610

Fuzzy Logic Inverter Controller in Photovoltaic

Applications: Issues and Recommendations
MAHAMMAD A. HANNAN 1 , (Senior Member, IEEE), ZAMRE ABD. GHANI2 ,
MD. MURSHADUL HOQUE 3 , (Student Member, IEEE), PIN JERN KER 1 ,
AINI HUSSAIN4 , (Member, IEEE), AND AZAH MOHAMED4 , (Senior Member, IEEE)
1 College of Engineering, Institute of Power Engineering, Universiti Tenaga Nasional, Kajang 43000, Malaysia
2 Department of Electronic and Computer Engineering, Universiti Teknikal Malaysia, Melaka 76100, Malaysia
3 Department of Electrical and Electronic Engineering, University of Chittagong, Chittagong 4331, Bangladesh
4 Centre for Integrated Systems Engineering and Advanced Technologies, Universiti Kebangsaan Malaysia, Bangi 43600, Malaysia

Corresponding author: Mahammad A. Hannan (hannan@uniten.edu.my)

This work was supported in part by the Ministry of Higher Education Grant 20190101LRGS under the Universiti Tenaga Nasional,
Malaysia, and in part by the Universiti Kebangsaan Malaysia under Grant DIP-2015-012.

ABSTRACT Fossil fuels and other conventional energy sources used to generate electricity are finite.
Therefore, alternative energy sources should be pursued to meet present and future energy demands. The
photovoltaic (PV) is a promising renewable energy source, especially for the remote areas. The PV is a DC
power source that needs to be converted into usable AC power using an inverter. However, its nonlinearity
and output fluctuation pose challenges in the design of PV based inverter. In this paper, a PV inverter
controller system with the fundamentals of a fuzzy logic controller (FLC) and its applications and execution
are reviewed. The different fuzzy controllers, inverter control algorithms, and switching techniques are
studied. The findings indicate that the fuzzy logic controls have been gaining attention in the area of power
control engineering, especially in inverter controller design for PV applications and generation. The FLC
has a flexible and intelligent design, expedient user interface, easy computation and learning system, and
combinations of different control algorithms. The FLC is also verifiable for completeness, redundancy, and
consistency. However, finding the boundaries of membership functions and other rules of FLC requires
manual tuning, long computation time, and considerable effort. This paper comprehensively reviews the
FLC-based inverter control system to minimize PV output fluctuations, which cause inverter issues related to
output harmonics, power factor, switching schemes, losses, and system implementation. The inverter system
and its control strategy for future PV applications and generation require further research and development.
Consequently, this review focuses on many factors and challenges and provides recommendations for
designing capable and efficient inverter control systems for converting PV power to usable AC power. All
the highlighted insights of this review will hopefully lead to increased efforts toward the development of the
advanced inverter control systems for PV applications for AC loads and the utility grid.

INDEX TERMS Power converter, fuzzy logic, fuzzy logic controller, inverter, photovoltaic system.

NOMENCLATURE DG Distributed Generation

ANFIS Adaptive Neuro-Fuzzy Inference System DSP Digital Signal Processor
ANN Artificial Neural Network
FLC Fuzzy Logic Controller
AVR Automatic Voltage Regulator
FNN Fuzzy Neural Network
BJT Bipolar Junction Transistor
FPGA Field-Programmable Logic Array
CCM Continuous Current Mode
GTO Gate Turn-Off Thyristor
COG Center-of-Gravity
IGBT Insulated-Gate Bipolar Transistor
CSI Current Source Inverter
IT2FLC Interval T2FLC
The associate editor coordinating the review of this manuscript and I-V Current vs. Voltage
approving it for publication was Heng Zhang. LC Inductor-Capacitor
2169-3536 2019 IEEE. Translations and content mining are permitted for academic research only.
24934 Personal use is also permitted, but republication/redistribution requires IEEE permission. VOLUME 7, 2019
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
M. A. Hannan et al.: Fuzzy Logic Inverter Controller in PV Applications

LOH Lowest-Order Harmonics Depending on their feasibility, advantages, and disadvan-

MF Membership Function tages, these energy sources can be harnessed for alternative
MIMO Multi-Input Multi-Output energy to meet the rising energy demand.
ML Multilevel PV is a promising renewable energy source that has been
MOSFET Metal-Oxide-Semiconductor Field-Effect applied in various sectors, such as construction, transporta-
Transistor tion, and residential, especially in remote or rural areas, over
MPPT Maximum Power Point Tracker the past few decades [1]. As a cost-effective and pollution-
NB Negative Big free energy source [4], PV can produce energy that can be
NM Negative Medium delivered to the utility grid in a grid-connected system with
NS Negative Small proper equipment and method. Unlike fossil fuels, PV is
P&O Perturbation and Observation abundantly available and can be maintenance-free. However,
PAS Power-Assisted Steering PV produces DC power that has to be converted into AC
PB Positive Big power using an inverter before it can be used, particularly in
PI Proportional-Integral AC-related applications [5]. The evaluation of PV technol-
PID Proportional-Integral-Derivative ogy depends on the development of materials and structures
PLL Phase-Locked Loop with the consideration of maximum power at minimum cost.
PM Positive Medium Silicon crystalline structure such as mono-crystalline, poly-
PS Positive Small crystalline, and emitter wrap through is used to build PV
PV Photovoltaic modules in first generation technology [6]. Mono-crystalline
PWM Pulse-Width Modulation PV occupies 80% of world market, however it has 24.7%
SHE Selective Harmonic Elimination efficiency. Poly-crystalline PV has been attracted in large
SOPWM Synchronous Optimal Pulse-Width Modula- scale production due to reduction of manufacturing cost and
tion defects in metal contamination and crystal structure, however
SPWM Sinusoidal Pulse-Width Modulation it has less efficiency (15%) as compared to mono crystalline
SVM Space-Vector Modulation based PV. Emitter wrap-though cells provide full surface area
SVPWM Space-Vector Pulse-Width Modulation for absorbing solar radiation that improves efficiency with
T1FLS Type-1 Fuzzy Logic System a rise of 15–20% without material changes [6]. However,
T2FLC Type-2 Fuzzy Logic Control this technology faces high resistance problem and poor fill
T2FLS Type-2 Fuzzy Logic System density [7]. In the next generation of PV technology, thin
THD Total Harmonic Distortion film technology is used with deposition of thin layer of certain
UPS Uninterruptible Power Supply materials in glass or SS substrates that makes PV modules
VSI Voltage Source Inverter flexible. Hence, it reduces the manufacturing cost. However,
ZERO Zero due to its thin structure the efficiency is comparatively low.
Efficiency can be increased by depositing different alloy
I. INTRODUCTION materials. Thin film PV modules have 15-20% market share
The growth of human population has resulted in significant for its versatility [6]. These are available with the technolo-
increase in energy demand. Fossil fuels are a conventional gies of amorphous Si multiple-junction structured cell, thin
energy source used to generate electricity. This resource poly-crystalline Si on a low cost substrate, copper indium dis-
poses harm to the environment directly or indirectly due to elenide/cadmium sulphide hetero-junction cell, and cadmium
carbon gases emissions [1]. Thus, alternative energy sources telluride/cadmium sulphide hetero-junction cell [6].
should be harnessed to address present and future energy In amorphous Si the bigger bandgap in a junction absorbs
demands. Following the current consumption trend, less than visible portion that infrared portion of solar radiation. This
50 years’ worth of petroleum reserves remain [2]. Fossil- technology uses several variations of glass or flexible SS,
based energy generation indirectly degrades air quality in tandem junction, double and triple junctions of different
the environment. Fossil energy sources include petroleum, characteristics to improve the efficiency and define degra-
natural gas, and coal [3]. Consumption of fossil-based energy dation hitches [8]. Cadmium telluride or cadmium sul-
by fuel-based machines and engines, such as automobiles phide/cadmium telluride is a promising PV material for thin
and generators, releases harmful carbon dioxide to the atmo- film technology with high absorption coefficient. However,
sphere, leading to air pollution and global warming. The the toxicity of cadmium has an environmental issue. Con-
preservation of the green, clean, and pollution-free envi- versely, this introduces first PV recycling and reduces the
ronment for future generations can be realized through the manufacturing cost [6], [9]. Copper indium diselenide or
utilization and implementation of pollution-free renewable copper indium gallium diselenide has high optical absorption
energy sources, such as solar photovoltaics (PV). Extensive coefficient and electrical characteristics for device tuning.
research on renewable energy technology has led to the This technology attracts researchers and developers in its
development and utilization of green energy sources, such deposition processes to make it more cost effective to com-
as PV, biomass, geothermal energy, hydropower, and wind. pete the world market with Si based PV modules. [10].

VOLUME 7, 2019 24935

 M. A. Hannan et al.: Fuzzy Logic Inverter Controller in PV Applications

Compound semiconductor is being used for PV technolo- controllers are required to generate power efficiently in
gies that have width band absorption of solar radiation. PV systems. Controllers, such as sinusoidal pulse–width
Moreover, it shows the robustness in cost reduction and effi- modulation (SPWM) and space–vector pulse–width modula-
ciency around 41% by introducing dual and triple junction tion (SVPWM), have been utilized in PV inverter applications
devices, and widely used in space satellites and large scale to obtain regulated inverter output. The SVPWM technique
solar plants [6]. However, this technology has some negative is more difficult to implement than the SPWM technique,
issues such as heat, UV light, and interaction of solvents which is applied in multilevel inverters [22]. APV inverter
within the encapsulation of the cell. Other solar cells such as system uses analog or digital control circuits, microcon-
organic solar cells and polymer solar cells are made from thin trollers or microcomputers, FPGAs, or digital signal proces-
films of organic semiconductors pentacene, polyphenylene sors (DSPs) [23]. Research and development continues to
vinylene, copper phthalocyanine and carbon fullerenes that is improve the control algorithm of PV inverters for sustainable
attracted for its mechanical flexibility and disposability. For renewable energy applications.
the use plastic materials, this technology is cheap. ‘‘The year PV inverters can be employed as stand-alone power
2007 has been a turning point for PV thin film technology at systems to provide AC power for residential appliances, such
least for US-based PV manufacturing with US thin film ship- as fluorescent lamps, fans, and refrigerators [24]. Inverters
ments reaching a market share of about 65%’’ [11]. However, with high-quality power output can be the main compo-
for superior efficiency and cheap production nanotechnology nents in automatic voltage regulator (AVR), uninterruptible
shows a great prospect in PV development. With a control on power supply (UPS), and power-assisted steering (PAS) sys-
the energy bandgap the nanotechnology lessens deficiencies tems [25]. These inverters can also be used individually
of other PV technologies [6]. As reported in [12], nanoma- or as part of the distributed generation (DG) system in the
terials have the most promising sector in solar energy conver- power system, which interfaces and feeds alternative energy
sion in the device formation of carbon nanotubes, quantum power to the grid system [26], [27]. Therefore, the power
dots (QDs), and ‘‘hot carrier’’ solar cells [13]–[15]. These electronic-based conversion system (inverter) should deliver
have higher efficiency of solar energy conversion, however, the proper amount of energy to provide an effective and
the research and development is conducting materials compo- efficient connection for the grid [28].
sition, recycling, cost effective production toward sustainable The main issues concerning PV inverters include the
energy generation. DC-output nature of PV devices, intermittent fluctuations,
The inverter, which is an electronic device that converts DC output voltage regulation, harmonic content, and conversion
power to AC power at a desired output voltage or current and efficiency [21]. The intermittent voltage, power, and DC
frequency, plays an important role in ensuring the operabil- nature of the PV output produce unstable inverter output
ity of the abovementioned products and applications in the unexpected by sophisticated loads. Therefore, these issues
field of renewable energy, such as PV. In these applications, need to be considered in designing PV inverters [29]. Many
the inverter should produce a clean output sine waveform methods have been introduced to lessen and control PV
at a precise voltage over controlled frequency ranges and deficiency with enhanced inverter control systems that reg-
within certain load variations. Certain features have to be ulate output voltage and power of PV inverters. On the
considered when designing PV inverters. The PV inverter contrary, self-power dissipation affects overall inverter effi-
system design should be simple (not complex or bulky), ciency, which is a significant issue in system design even
stable, reliable, efficient, and low-cost. The inverter design with the proper switching frequency [30]. Designing the
should allow for flexible control, ease of maintenance and filter to generate quality output waveform in the inverter
testing, low waveform total harmonic distortion (THD), and is easy with a high switching frequency, which produces a
unity power factor [16]. In addition, inverter systems should high level of harmonic components that cause power loss
be small and light [17]. The nonlinearity and output fluc- and efficiency reduction. Therefore, the switching frequency
tuation of PV pose challenges to PV inverter design [18]. should be appropriately selected to produce quality inverter
The inverter control system should minimize the impact of output and prevent power loss.
output fluctuation on the stability and power quality of the A suitable inverter controller has to be selected for the
inverter. For a PV with a low output voltage, a DC-to-DC inverter design to optimize the inverter parameters effec-
boost converter is incorporated into the power conversion tively. Both the quality of the inverters and the flexibil-
system [19]. ity of the inverter control system design are important in
The progress of PV system applications is associated providing the inverter designer a platform for further prod-
mainly with the power electronic interface, such as invert- uct research and design development, which commercially
ers and controllers. Inverter and controller technologies are available inverters lack. This feature enables the control
upgraded continuously to enhance the efficiency of PV systems to be assessed for upgrades to inverter parameters,
systems because inverter errors and failures are the pri- such as voltage, current harmonic content, and power factor.
mary sources of failure of PV systems [20]. Power con- Various types of inverter controller platforms have been pre-
version in a PV system depends on inverter controllers sented in the literature, including the microcomputer [23],
and control algorithms [21]. Accordingly, quality inverter DSP [31], and FPGA [32]. The utilization of controllers,

24936 VOLUME 7, 2019

M. A. Hannan et al.: Fuzzy Logic Inverter Controller in PV Applications

such as microcontrollers [33], microprocessors [34], and A. PHOTOVOLTAIC SYSTEM

analog and digital circuit controllers [35], has also been Different power sources, such as PV, wind generators, fuel
reported. Most studied controllers require users to have soft- cells, and batteries, can be utilized for the energy conversion
ware programming knowledge to develop inverter control system and integrated with the inverter using different meth-
algorithms, which is time consuming. The design platform ods. However, connecting the different power sources to the
should allow for a simple and short design and development inverter may require additional stages or devices. For exam-
process for inverter control system algorithms. ple, a wind turbine generator, which produces AC power,
A novel improved design method for PV system inverter requires an additional AC-to-DC converter (e.g., a rectifier) to
controllers needs to be developed to acquire results with low connect to an inverter [45]. Unlike a wind turbine, a PV panel
harmonic content, unity power factor, simple hardware imple- produces DC power and therefore can be connected directly
mentation, low-loss switching scheme, easy system imple- to the inverter, thereby reducing the overall cost of the PV
mentation, and short product design period. With high-quality system [46]. A PV cell is a semiconductor with a p-n junction
output, the inverter can be connected and feed PV power diode that converts light directly into electricity [47]. When
to the utility grid. This review can lead to more research the common junctions of the p-n diode are struck by sunlight,
opportunities on further PV system improvement. which contains photons, the energy is absorbed by the elec-
tron system of the material. Electron–hole pairs, as charge
II. OVERVIEW OF PV INVERTER CONTROLLER SYSTEMS carriers, are created with a potential barrier at the junction
Today, power electronic converters are widely used in various that generates a voltage that allows current to flow through
applications as single-phase and three-phase inverters [36]. the external circuit. This process, called the photovoltaic
They may operate as stand-alone units that connect and link effect [48], is illustrated in Fig. 2.
DC power to AC loads, such as AC motors [36]. They
can also be used as UPS [37], AVR [38], and many other
systems. Normally, three-phase inverters are preferred for
medium- to high-level applications [37]. In a utility grid-
connected system, inverters are used as interface devices to
feed DC power to the grid [39]. In this regard, components
such as relays or switches are required to enable the system
to connect to the grid [40]. An inverter’s design is simple
with fixed input voltage and loads. However, fluctuating PV
output and varying load disturbances affect the performance
of the inverter [41], [42]. Thus, inverter control systems
should be designed and developed to minimize the effect
of these disturbances on the performance of inverters and
their control systems for PV applications. Fig. 1 shows the
general block diagram of a PV conversion system, which FIGURE 2. Photovoltaic effect of a PV cell.
consists of a system controller, PV panel, filter, and power
converters [43], [44]. The integration of these elements
Cell arrangements, such as series, parallel, and series-
enables the proper utilization of solar energy in the energy
parallel, form a PV module that acquires the specified power
conversion system. The system is a feedback control system
capacity [49]. Similarly, an array or panel is constructed
that runs the output (voltage) of the process. The controller
by aeries-parallel combination of modules to obtain higher
and power converters ensure a stable voltage supply for the
power capacity [50]. The output voltage generated in the PV
load.
cell is depends on the photocurrent and level of the real-time
solar irradiation. [51]. Equation (1) presents the output cur-
rent of the solar cell Ic [52].
h q i
Ic = Iph − Io = Iph − Isat e AKTc (V +IRS ) − 1 (1)

where Iph is the photocurrent, Io is the diode current, Isat is

the p-n junction reverse saturation current, q is the electron
charge, A is the ideality factor, K is the Boltzman constant, Tc
is the cell temperature in Kelvin, and Rs is the stack internal
resistance.
The required PV voltage VPV , and current IPV are generated
by considering the series-parallel combination of modules Ns
FIGURE 1. Generalized block diagram of a PV energy conversion system. and Np . The voltage and current of PV arrays are expressed

VOLUME 7, 2019 24937

 M. A. Hannan et al.: Fuzzy Logic Inverter Controller in PV Applications

by Equations (2) and (3).

VPV = Ns Vref − β T − Tref − Rs T − Tref




(2)
    
G G
IPV = Np Iref + α


T −Tref +Isc −1
1000 1000
(3)
where α and β are the current and voltage temperature coeffi-
cients, T is the actual temperature, Tref is the reference stack
FIGURE 3. DC-to-DC boost converter with constant DC voltage source.
temperature,Vref and Iref are the PV reference voltages and
current, Isc is the short circuit current, and Gis the irradiance
on the PV surface. Equation (5) shows that the DC converter output depends on
Equation (4) represents the PV’s output power, which is the its input voltage and IGBT duty ratio. Increasing the duty
product of the output current and the output voltage of the PV ratio increases the output voltage. Nonetheless, this DC-to-
array [4,50]. DC boost converter only increases the output voltage higher
PPV = IPV × VPV (4) than that of the input. The ratio is modulated and forms
a PWM switching signal. This signal is generated by the
Some of the disadvantages of PV are its high initial instal-
boost converter controller and fed to the gate of the IGBT,
lation cost and low-efficiency energy conversion [53]–[55].
as shown in Fig. 3. Increasing the duty ratio increases the
In a PV system, which utilizes power converters, such
converter output voltage. As for example, the available DC
as inverters and DC-to-DC converters, the control system
input voltage is 450 V and the DC output voltage requirement
plays a vital role in enhancing system performance and
is 750 V, the boost factor would be 1.67 and the appropriate
efficiency [56].
constant IGBT duty ratio would be 0.4. Thus, the fixed duty
B. POWER CONVERTERS ratio can maintain a higher constant DC input voltage forthe
inverter. Parameter values, such as the inductor, capacitor,
A DC-to-DC converter converts one DC power level to
and switching frequency,should be selected properly when
another that has the advantage of regulating its output voltage
designing an inverter to allow the converter to operate in
even with variations in the DC input voltage [57]. This
continuous current mode (CCM). This mode ensures that the
converter is suitable for use in a PV system with unstable and
inductor current flows continuously (i.e., the minimum induc-
fluctuating PV output. In a PV system that utilizes both DC
tor current is higher than zero) and has low output current
and AC converters, DC output voltage stability is enhanced
ripple [61]. Therefore, with selected load resistance Rand
using the DC-link capacitor of the inverter, thus minimizing
switching frequency f , the minimum value of the required
the effect of PV fluctuation on the inverter AC output [57].
inductor Lmin is described by Equation (6) [60], [62].
Depending on the size or capacity of the PV output voltage,
these converters may be utilized along with DC voltage regu- D(1 − D)2 R
lation and stabilization capabilities to provide proper voltage Lmin = (6)
2f
level for the inverter input stage [54]. As the first stage in
The selection value of the capacitorC, which determines the
a PV system, a DC converter receives input power from the
peak-to-peak voltage ripple, is based on [60, eq. (7)].
PV panel, and then amplifies and feeds the said power to the
inverter. The boost converter steps up the DC input voltage 1 Vo D
= (7)
to the voltage level required by the inverter. Utilizing this Vo RCf
converter is cheaper and more preferable than investing in Hence, solving for C, the value of the required capacitance is
additional PV panels. Increasing the quantity of PV panels described by Equation (8):
tends to increase the overall PV system cost [39], [53].
D Vo
A boost converter is not necessary when the DC output volt- C= (8)
age from the PV panel meets the inverter DC input voltage Rf 1Vo
requirement [58]. Considering the fluctuating nature of the where 1Vo is the change of the output voltage or peak-to-peak
PV array, a boost converter is integrated with the inverter to voltage ripple of the capacitor voltage.
produce stable power for the loads [59]. Fig. 3 shows a basic Considering the variations in PV panel output power in
DC-to-DC boost converter with constant DC voltage source. different sun irradiances, the boost converter includes an
However, utilization of this converter increases system cost MPPT feature with a controller that ensures that the PV array
and slightly decreases system efficiency due to the losses that always operates at maximum power [63], [64]. The MPPT
occur in the switching devices. The relationship among the is essential especially in a stand-alone system with a battery
converter input voltage Vdc , output voltage Vout , and the duty back-up system and a grid-connected PV system that allows
ratio D of the IGBT are expressed by [60, eq. (5)]. for maximum power extracted to be stored in batteries and
1 exported to the grid system [65], [66]. The load determines
Vout = Vdc (5) the required amount of energy for extraction, provided that an
1−D
24938 VOLUME 7, 2019
M. A. Hannan et al.: Fuzzy Logic Inverter Controller in PV Applications

adequate supply of energy is available [65], [66]. However, located on the PV I-V curve, thereby enabling the PV to oper-
the MPPT is not recommended for a stand-alone PV system ate at its maximum power point at different solar irradiances.
without a battery back-up system because the former will An inverter performs a DC-to-AC power conversion [36]
cause a sustained energy imbalance [65]. A PV DC-to-DC and mainly produces a sinusoidal AC output with controllable
boost converter with MPPT is illustrated in Fig. 4. magnitude and frequency [60]. Inverters can be categorized
based on their DC input source. Inverters with DC current
input sources are called current source inverters (CSIs) and
are used for high-power AC motor drive applications. Invert-
ers with DC voltage input sources recalled voltage source
inverters (VSIs); a three-phase VSI is considered for PV
application. As shown in Fig. 6, the three-phase inverter con-
sists of a DC input voltage source Vic , DC-link capacitor C,
and power stage with three legs, namely, leg 1, leg 2, and leg 3.
The DC input voltage source is a constant voltage source
connected to the DC-link capacitor C. The required level of
the DC input voltage depends on the specifications of the
inverter. In a power system for a three-phase inverter, the DC
input voltage level must be higher than the peak voltage of
FIGURE 4. DC-to-DC boost converter with MPPT controller. the inverter’s line-to-line AC output voltage. In a fluctuating
DC power source, such as a PV panel, the specifications
Hill climbing, P&O, and incremental conductance meth- of the capacitor are essential in maintaining the stability of
ods are some of the MPPT methods that have been reported the DC-link voltage during power device switching [70].
in the literature [67]. The most common and popular method The DC-link capacitance should be large enough to reduce
is the P&O [68], where the PV panel parameter (voltage or the DC-link voltage ripple, resulting in good inverter output
current) is increased or decreased to find the highest power power quality [71]. Based on the energy flow of the system,
available in the PV panel. As illustrated in Fig. 4, the MPPT the suitable capacitance for the three-phase system can be
controller accepts PV voltage and current as inputs to gen- expressed as [72, eq. (9)].
erate the PWM switching signal for the gate of the IGBT.
In the MPPT algorithm, the sensed PV voltage and current PLoad,max
Cdc = , (9)
determine the available PV output power. Fig. 5 illustrates 2Vdc × 1Vdc × 2π × f
the position of the maximum power point on the I-V curve where PLoad,max is the maximum load rating, 1Vdc, is the
for a certain solar irradiance of 1000 watt/m2 at 25 ◦ C. The input voltage variation, and f is the output voltage frequency.
variations in solar irradiation and ambient temperature pro-
duce several maximum output power levels. Thus, the MPPT
controller should extract and determine the maximum power
available in the PV panel at different solar irradiances [69].
PV output current and voltage change with the IGBT duty
ratio, thereby varying the PV output power. By continuously
varying the duty ratio, the maximum power point can be

FIGURE 6. Schematic diagram of a three-phase PV inverter.

Switching devices play an important role in DC-to-AC

conversion. Each leg consists of two IGBTs that have anti-
parallel diodes and are never turned on at the same time [36].
The type of switching devices to be used depends on the
power capacity and switching speed of the system. Generally,
a MOSFET is utilized in systems with low power capacity
and high switching speed; meanwhile, IGBTs are designed
FIGURE 5. Solar panel of I-V curve showing the maximum power point for systems with high power capacity and moderate switching
location. speed [60]. The legs constitute the phase of the inverter

VOLUME 7, 2019 24939

 M. A. Hannan et al.: Fuzzy Logic Inverter Controller in PV Applications

output. Leg 1, leg 2, and leg 3 are regarded as output phase vref (t) and the inverter output voltage vinv (t), as expressed by
a, phase b, and phase c of the inverter, respectively. The gates [76, eq. (10)]. A control signal is then generated based on the
of the IGBTs receive the PWM switching pulses from the error signal level to determine the specific control signals fed
controller. The control algorithm developed in the control to the actuator, which, in this case, comprises the switching
system generates these PWM switching signals so that the devices, e.g., the MOSFET and IGBT. After the control signal
desired output voltage pattern and quality, i.e., sinusoidal is generated, it regulates the required voltage (DC input) to
waveform, is produced at the IGBTs collector terminals, a, make the latter suitable for the plant according to the control
b, and c [73]. The control strategies for the PV inverter strategy. This is accomplished through the control of the turn-
controller are explained in the following sections. on duration of the switching devices.
e(t) = vref (t) − vinv (t) (10)
III. INVERTER CONTROL STRATEGIES
An inverter control system aims to maintain the AC output Plant dynamics need to be modified to meet the system
voltage at a desired level and frequency and achieve minimum performance specifications [77]. Unfortunately, in the real-
THD [73]. This role is played by the controller using a suit- world environment, most plants are fixed and unalterable.
able control strategy that maintained the voltage at a specific System performance can be improved by inserting a con-
reference input or set point. With a proper control strategy troller or a compensator. The compensator manages to over-
design, the harmonic content of the output voltage may be come the drawback of a fixed plant in the control system.
reduced [74]. The duty cycle of the power devices is one of The commonly used controllers are proportional–integral (PI)
the control parameters in power converters. The duty cycle and proportional–integral–derivative (PID) controllers. The
is controlled to enable the inverter output voltage to track PI controller algorithm in the time domain is described by
the reference voltage signal [73]. Tracking accuracy usually [76, eq. (11)].
depends on the capability of the controller in the control Zt
system. Fig.7 depicts the general concept of the closed-loop
u(t) = Kp e(t) + Ki e(τ ) dτ , (11)
control system of an inverter.
0
where u(t) is the output of the PI controller and Kp and Ki are
the gains of the proportional and integral terms.
The values of the gains determine the behavior of the
system, which in turn, performs satisfactorily. The system
keeps the output voltage as close as possible to the reference
voltage and reduces steady-state errors, thereby regulating the
output voltage. Hence, the output voltage of the inverter may
be controlled and stabilized.
System performance is achievable through the execution of
the proper control strategy integrated with the PWM switch-
ing technique. To get a sinusoidal AC output waveform,
an SPWM switching technique is employed in the inverter
control algorithm. Moreover, controlling the inverter switch-
ing scheme can significantly reduce the harmonic content
of the output voltage, especially the lowest-order harmon-
FIGURE 7. General concept of an inverter closed-loop control system. ics (LOH) [36], [37]. PWM techniques to inverter control
produce the output power with low loss, high efficiency,
Based on the concept of the control system, the combina- low distortion, short computation time, and minimal harmon-
tion of the inverter and the filter is the main component to be ics [78]. PWM switching techniques are sinusoidal PWM
controlled. Thus, the output of the filter, which is the variable, (SPWM), hysteresis band PWM (HBPWM), and random
needs to be controlled so that it resembles and tracks the pulse width modulation (RPWM). These switching tech-
desired output. In controlling the output, the inverter output niques are applied to have various systems with satisfactory
waveform is fed to the control system input by a sensor. performance. However, in some cases, high computational
Depending on the control system input requirement, different time, harmonic distortion and switching loss cause problems
types of sensors (e.g., voltage sensors and current sensors) in PV inverter control. The space vector pulse width modula-
are utilized. The control algorithm is designed based on the tion (SVPWM) switching technique has robustness in mini-
property (e.g., voltage level) of the input signal to the control mizing switching loss and harmonic output signals from the
system so that the plant output meets the desired output. The inverter controller [79]. Besides, as an advanced computation
comparison of the input parameter with the desired output intensive PWM method, the SVPWM technique is applied for
parameter in the algorithm generates an error signal [75]. inverter control applications to minimize harmonic distortion.
The error is the difference between the reference signal Many researchers have used the SVPWM with the V/f control

24940 VOLUME 7, 2019

M. A. Hannan et al.: Fuzzy Logic Inverter Controller in PV Applications

TABLE 1. Comparison of AI control methods.

criteria. Whenever the reference signal voltage is higher than

that of the carrier signal, a ‘‘high’’ pulse signal is generated,
which turns on the upper IGBT of the corresponding leg.
Otherwise, a ‘‘low’’ pulse signal is produced, which turns the
IGBT off. The comparison of each reference signal produces
individual PWM switching signals for the IGBT of each leg.
The complementary of this signal should be fed to the IGBT
in the lower side of the corresponding leg. As a result, the DC
input voltage appears at the output phases of a, b, and c.
Fig.9 illustrates the general relationship among the average
FIGURE 8. PWM signal generation mechanism.
value of the generated inverter output voltage Van , reference
signal waveform vreference , carrier signal waveform vcarrier ,
strategy based on the ANN controller in the TIM [80]. The and DC input voltage Vdc [60]. A bipolar voltage switching
SVPWM is applied in IFOC control for a TIM and is executed technique may be applied on the inverter to simplify the
with an FLC to control the PV grid inverter system [81]. analysis.
V/f control is used the SVPWM technique for three-level The average DC output of the inverter can be regulated
VSIs is used in the V/f control for high-power variable speed by controlling the duty ratio of the IGBT. Only one switch-
drive [82]. The SVPWM is also employed with the FLC ing cycle is considered to simplify the analysis because the
to adjust the position loop for the vector control of the IM entire process requires constant switching. The reference
drive [83]. The DTC is used with the SVPWM technique signal resembles a constant signal with the amplitude of
to improve and minimize the torque ripple for an induction the carrier signal waveform in the case of the sinusoidal
machine [84]. The PMSM drives use the SVPWM to control reference signal waveform with high switching frequency.
torque and flux in the vector control [85]. The ANN with Thus, inverter output can be generalized in terms of the three
the SVPWM based on sensorless control estimates the rotor parameters, namely, reference signal waveform vreference , car-
speed in the vector control of the TIM drive [86]. However, rier signal waveform vcarrier , and DC input voltage Vdc . The
the SVPWM requires a complex online computation, which average value of the line-to-line voltage Vab is represented by
leads to difficulty in real-time implementation. Therefore, Equation (12).
the conventional SVPWM requires additional memory that vreference
Vab = Van − Vbn = Vdc (DIGBT 1 _DIGBT 3 ) = Vdc ,
limits the selection of switching frequency, thereby reducing vcarrier,peak
its accuracy [87]. Some researchers have solved this problem (12)
and have improved the power quality problem by utiliz-
Vab = ma Vdc , (13)
ing the GA-based SVPWM, ANN-based SVPWM for two-
level inverter, ANN-based SVPWM for three-level inverter, where DIGBT 1 is the duty ratio and ma is the modulation
and the ANFIS-based SVPWM used for two-level invert- index.
ers [87], [88]. Table 1 shows the comparison of AI control Equation (13) shows that the output voltage Vab depends on
methods of PV inverter system. modulation index ma and DC input voltage Vdc . This equation
For generation of simple PWM switching signal, as shown is also valid for the unipolar switching scheme [60]. The
in Fig.8 voltage comparison among reference signals A, B, modulation index or amplitude modulation ratio is the ratio of
and C and the carrier signal fc , is based on the following the peak amplitude of the reference waveform to the carrier

VOLUME 7, 2019 24941

 M. A. Hannan et al.: Fuzzy Logic Inverter Controller in PV Applications

FIGURE 9. Illustration of inverter output voltage generation.

waveform [101]. The value of the modulation index should The fuzzy logic controller (FLC) is the simplest among
be maintained below 1 to avoid over-modulation, which all intelligent controllers for inverter control application. The
results in an output waveform with a high output voltage FLC is much better than conventional controllers in terms
value and a considerable amount of harmonics. Similar to of response and settling time, sensitivity to parameter and
the periodic function, the line voltage waveform acquires a load changes, and robustness [108]. An inverter without a
time period T and a fundamental frequency f in a steady transformer can conveniently improve its size and weight
state. In addition to a dominant component at the fundamen- issues [109]. The complexities in implementing the PV sys-
tal frequency, the line voltage waveform contains unwanted tem are related to the inverter type and switching topology
harmonic components of the fundamental frequency [60]. used in the condition of the system environment. Different
The three-phase inverter line voltage is considered a non- types such as two or three level inverters can be imple-
sinusoidal periodic function waveform. mented easily in terms of their inverter topology and con-
Considerable research has been conducted on inverter trol algorithm. On the contrary, multilevel inverters require
control issues, internal power losses, system nonlinearity, a complex control algorithm in generating switching signal
power fluctuation, environmental variation, low PV effi- and system coordination, which provide extra computational
ciency, harmonic distortion, DC-bus voltage fluctuation, hassle to the processing unit [26]. The inverter system also
electromagnetic interface, and unity power factor mainte- becomes costly because of the increase in semiconductor
nance [102]. Therefore, developing an efficient inverter con- switches [110]. In [89] FLC based inverter control is pro-
troller to enhance the overall performance of a PV system posed for PV systems through simulation and experimental
is challenging [103]–[107]. A number of PID and PI con- performances having 2.5% voltage THD and 3.5% current
trollers has been implemented in PV inverter systems [100]. THD. An experimental analysis for grid connected VSI is
However, the conventional PID, PI, and other controllers shown with optimum LCL filter design using complex PWM
have demerits in terms of large-scale output filtering, grid control having 2% THD of inverter current [111]. Using
frequency, and intelligence application. The utilization of predictive power control algorithm a grid integration PV
adaptive control in PV inverter systems drive researchers and qZSI system is developed in [112]. The system is designed
developers to practice intelligence application in controllers, with a complex SVPWM switching having 3.2% current
such as neural network, fuzzy logic, and neuro-fuzzy [89]. THD. The experiment is going on the reduction of THD
A supervisory control employing an adaptive neuro-fuzzy of PV inverter, however, the control algorithm and design
interface system (ANFIS) can be developed for three-phase approach is becoming more complex and memory use as the
grid-connected PV inverter systems to address the demand optimization algorithm is being incorporated [99]. It might
for grid power. Nevertheless, the ANFIS has certain network be recommended for developing a robust inverter controller
training problems in a large number of data sets [91]. Mean- the compromise between the improved algorithm and system
while, intelligent controllers may be in demand in PV inverter complexities is necessary.
systems because they reduce the mathematical model analysis The performance of inverter systems is summarized
problem and the complexity of handling the nonlinearity of in Table 2, which includes the parameters of THD in the
the system. output voltage and current, power factor value, switching

24942 VOLUME 7, 2019

M. A. Hannan et al.: Fuzzy Logic Inverter Controller in PV Applications

TABLE 2. Summary of the inverter system performance comparison.

technique use, and designing platform. The design and imple- its membership degree, µA(u), as shown in [118, eq. (14)].
mentation of the inverter system face difficulties in terms The universe of discourse temperature level. These terms are
of its design, switching topologies, and design environment. described in terms of sets called fuzzy sets. A specific control
The multilevel inverters should be implemented simply for action is generated and does not have binary membership
efficient inverter design and control. The design of these characteristics [119]. Scenario is the basic concept of fuzzy
inverter types appears to be more feasible because it simpli- logic application in a control system. Through these fuzzy
fies the overall system algorithm, weight, and size. However, sets, which are associated with the corresponding MFs and
inverters with five levels and more require complex switching integrate with the fuzzy rules. This Fuzzy logic is determined
control and algorithm, which then produce computational by the thought refers to the range of possible values regarded
difficulties. The increased number of switches also causes the for fuzzy set members [120].
inverter system to become bulky and costly.
A = { (u, µA (u)) : u ∈ U } (14)
IV. CONCEPT OF FLC SYSTEM
An inverter is an essential device for interfacing a DC power Fuzzy set theory is ideated in such manner that an element
source, such as PV to an AC load. With the integration of belongs to a fuzzy set with a certain degree of member-
the essential components of the system such as the control ship (associated with an MF), which has a real number in
algorithm, which are embedded into the controller or sys- interval [0, 1]. MF provides the degree of membership of
tem processor, the inverter can perform an effective energy any element of the universe of discourse within the set.
conversion. With the appropriate and proper design of a The types and quantity of MF employed contribute to con-
control system algorithm, successful power conversion, and troller performance, memory space occupation, and execu-
effective filtering, the inverter manages to convert the DC tion speed [121]. They map the elements of the universe on
input voltage to AC sinusoidal waveform to power up the to numerical values in the interval [0, 1] [119]. In particular,
AC loads. The inverter controller plays an important role in the MF converts the degree of fuzziness into the normalized
ensuring a successful and effective energy conversion. Hence, interval [0,1] [120], [122]. Unlike an ordinary set, where
the selection of a suitable and effective controller is the key to each element either belongs or does not belong to the set, par-
the excellent performance of the power converter. Fuzzy logic tial membership in a fuzzy set is possible [119]. The example
is a theory of fuzzy sets that calibrates ambiguity. A fuzzy of the MF, µA(u), of set A is expressed in Equation (15).
set is defined without crisp boundaries Current research on The fuzzy set and its MF are illustrated in Fig. 10.
the FLC implementation in the application of a DC to AC
for u < 4

converter (inverter) to convert the PV power for AC loads has 
0,
 u−4
received considerable attention [115]. , for 4 ≤ u < 14



µA (u) = 10
u − 24 (15)
A. FUZZY SET AND MEMBERSHIP FUNCTIONS (MFS)

− , for 14 ≤ u ≤ 24
10



In the FLC system, the controller mimics how the human for u > 24

0,
brain thinks by interpreting the incoming information as
imprecise terms [116]. Several examples of the imprecise For example, number 10 in Fig. 10 is the element of fuzzy
terms are warm, hot, and cool, which describe the states of set A with a membership degree of µA (10) = 6/10, which
the degrees of truth [117]. Fuzzy set A in the universal set U is equal to 0.6. The closer this value is to 1, the more the u
(also called universe of discourse) is a set of ordered pairs of (or element) is considered to belong to set A [119]. The fuzzy
a generic element u (generally not numerical quantities) and set and its MF are determined and formed based on expert

VOLUME 7, 2019 24943

 M. A. Hannan et al.: Fuzzy Logic Inverter Controller in PV Applications

FIGURE 10. Illustration of a fuzzy set.

knowledge [117]. Among the commonly utilized MF forms

are trapezoidal, Gaussian, and bell shaped [120]. FIGURE 11. Basic type-1 FLC architecture.

B. LINGUISTIC VARIABLE AND RULES

A linguistic variable is a fuzzy variable that is not a number
The output of the fuzzy system can be obtained with the
and is represented in the form of a word or sentence. It is
defined fuzzy sets of the input and output variables, as well
applied in the formation of fuzzy rules of a fuzzy rule base,
as the fuzzy rules. The fuzzy expert system aggregates all
which can be found in the fuzzy controller [117], [120].
fuzzy output sets into a single output fuzzy set and then
Generally, the variables are named according to the sys-
produces a single crisp solution for the output variable [117].
tems input and output. The rule is a conditional statement
This condition is accomplished through the inference process,
in the form of IF–THEN, and make the algorithm and
which consists of fuzzification, rule evaluation, aggregations
fuzzy controller intelligent based on the human-like deci-
of the rule outputs, and defuzzifications [121].
sion [120], [123]. For example, an air-conditioning control
system has the rule ‘‘IF temperature is high and humidity is
A. FUZZY INTERFERENCE
low THEN speed is fast’’, where the words ‘‘temperature,’’
‘‘humidity,’’ and ‘‘speed’’ are the linguistic variables, and The inference process produces a fuzzy output set according
‘‘high,’’ ‘‘low,’’ and ‘‘fast’ are the linguistic values. These val- to the IF–THEN rule logic. In particular, it maps the input
ues are determined by fuzzy sets in the universe of discourses, to an output by utilizing the theory of fuzzy sets [120].
namely, ‘‘temperature,’’ ‘‘humidity,’’ and ‘‘speed’’ [118]. The fuzzy expert system combines all fuzzy output sets
In this rule, the ‘‘IF. . . ’’ part is called the antecedent (causes), into a fuzzy set and generates crisp results for the output
whereas the ‘‘THEN. . ..’’ part is called the consequent. variable [117]. The system defuzzifies the resulting fuzzy
Examples of the linguistic variables are defined as follows: set into a single number. The Mamdani method is a more
popular method applied for the inference process compared
temperature = {very high, high, medium, low, very low} with Sugeno style [120].
humidity = {very high, high, medium, low, very low}
B. FUZZIFICATION
speed = {very fast, fast, medium, slow, and very slow}
Before input data can be processed by the fuzzy system,
Given that each variable (e.g., temperature or humidity) the degree to which the physical values (crisps) of each
contains five linguistic values, 25 possible rules are consid- input belong in the appropriate fuzzy sets should be deter-
ered. mined [117]. Thus, they are transformed into a degree of
membership by the MF [124]. For example, the input is
V. FLC ARCHITECTURE fuzzified over all MFs, which yields a membership degree
Fuzzy logic is suitable for the control system because of (value) (e.g., µA(u) = 0.5) applied to the antecedents of the
the non-linear nature of a power converter, such as an fuzzy rules.
inverter [18]. The general architecture for the FLC used in
a control system is shown in [89, Fig. 11]. This FLC acts C. DEFUZZIFICATION
as the controller, which controls the plant (process) to attain Defuzzification, which is the final step in the inference
the plant requirements and specifications. Fig. 11 shows the process, extracts the crisp output value from a fuzzy output
basic type-1 FLC architecture with three main components: set [120]. This process takes the aggregate output fuzzy set
(i) fuzzifier unit, (ii) knowledge based rules and inference and produces a single-number output. The centroid technique
engine, and (iii) defuzzifier unit. The other elements associ- is one of the most popular methods of defuzzification. This
ated with the FLC control system are the reference (desired) method identifies a point that represents the center-of-gravity
signal, and the input and output signals from the plant or (COG) of the fuzzy set at a certain interval of k = 1
process to be controlled. to n [117].

24944 VOLUME 7, 2019

M. A. Hannan et al.: Fuzzy Logic Inverter Controller in PV Applications

VI. FUZZY LOGIC INVERTER CONTROLLER

FLC (fuzzy control) can be intelligent control, knowledge-
based control, or expert control [119]. A fuzzy logic may
model human behavior in terms of intelligence and sense
in smart machines [117]. By implementing user-defined
rules in performing the system processes, FLC is considered
as flexible, and rules can be added and modified easily to
improve or alter system performance [129]. Fuzzy logic is
the simplest to integrate among the different intelligent con-
trollers [130]. Unlike the conventional controller, the design
and application of the FLC algorithm in a highly complex
system is more favorable because of its simplicity [116].
A system with fuzzy logic as the controller also provides
FIGURE 12. Type-2 membership function with the shaded region as
footprint of uncertainty [126].
high quality and reliable output with low cost, better system
response, and excellent performance [131]. The nonlinear
behavior of the fuzzy controller may result in higher system
robustness. This system may comprise nonlinear elements
with a simple control structure [116].
Unlike conventional controllers, such as PID controllers,
which are based on the mathematical models of the process
under control, the FLC uses the concept of an ‘‘inference
engine.’’ This method is considered the heart of the system
operation. The FLC utilizes the concept of fuzzy sets theory
(fuzzy logic), set of IF–THEN fuzzy rules, linguistic vari-
ables, and fuzzy proposition in its system operation, such as
FIGURE 13. Type-2 fuzzy logic system overview [126]. in decision-making [116]. FLC is applicable to any nonlinear
system control (e.g., the PV inverter system), which does not
require any distinct process of design, such as pole placement,
D. TYPE-2 FUZZY LOGIC SYSTEM frequency response, and root-locus designs because of the
The conventional FLC as known as type-1 fuzzy logic sys- nonlinear behavior of rules.
tem (T1FLS) might have good controlling ability in vari- The design of a PV inverter control system should consider
ous applications, however it lags to define the uncertain- the fluctuation nature of the PV itself. This scenario ensures
ties of the systems’ noises, nonlinearities and environmental that the effect of the PV output variation toward the inverter
variations. Type-2 fuzzy logic system (T2FLS) defines the output is minimized. Given DC input power fluctuation, volt-
system’s uncertainty through a bounded region in member- age control that maintains a constant inverter output should be
ship function (MF) as upper and lower MFs as shown in obtained. This role is played by the effective inverter control
[125, Fig. 12]. The uncertainty appears due to the measure- system. Fig. 14 presents a block diagram of the PV inverter
ment noise in the system and can be captured by minimizing controller that utilizes FLC.
the negative effect on the system. Like T1FLS, it contains
a rule-base, a fuzzifier, an inference-engine and an output
processor. The output processor is comprised with defuzzifier
and type reducer that produces type-1 fuzzy output as shown
in [126, Fig. 13]. So, according to Hisdal [127], ‘‘increased
fuzziness in a description means increased ability to handle
inexact information in a logically correct manner.’’ According
to John [128], ‘‘Type-2 fuzzy sets allow for linguistic grades
of membership, thus assisting in knowledge representation
and they also offer improvement on inferencing with type-1 FIGURE 14. Fuzzy logic controller of a PV inverter system.
sets.’’ [126]
In T2FLS, the structure of rules are same as T1FLS except The selections of system variables, MFs, and fuzzy rules
that they involve in type-2, he inference process requires development are necessary to develop the FLC-based control
additionally to obtain unions and intersections of type-2 sets, algorithm for a system. The input and output variables used
and compositions of type-2 relations, the output process is in the FLC system should be initially determined. In an
involved a type-reduction process as similar process to type- electrical system, inputs to the FLC system are the param-
1 defuzzification to generate type-1 sets from type-2 sets and eters (electrical signals) originating from the ‘‘process to be
then defuzzified to crisp output [126]. controlled’’. However, the FLC output is the control process

VOLUME 7, 2019 24945

 M. A. Hannan et al.: Fuzzy Logic Inverter Controller in PV Applications

of the input parameters. Different types of system output TABLE 3. Fuzzy rule table for the rules base system.
parameters exist depending on the applications. The output
network of the inverter is considered to be the plant to
be controlled. The fuzzy logic inverter control algorithm is
designed based on the FLC architecture as shown in Fig. 11.
In the fuzzy logic-based inverter control system, an error
in the discrete time, e(k) [Equation (16)], and its rate of
change, de/dt, are selected as input variables for the FLC.
The controller output, u(k), is selected as the FLC output
variable as illustrated in Fig. 15. The inverter output voltage
is the output variable from the process. The process output in
an equilibrium or steady-state condition should be extremely
close to the desired value. (Table 3), to simplify the controller design.Table 3 shows a list
e(k) = r(k) − y(k), (16) of rules that governs the controller decision. Several examples
of the system rules are presented as follows [89]:
where y(k) is the process (inverter) output variable, and r(k) Rule 1: IF error is NB AND change of error is NB THEN
is the desired or reference signal. output, u is NB
Rule 2: IF error is NM AND change of error is NB THEN
output, u is NB
Rule 3: IF error is NS AND change of error is NB THEN
output, u is NM
...
...
Rule 49: IF error is PB AND change of error is PB THEN
output, u is PB
Equation (17) shows that the output of the fuzzy controller
changes for every sample of time, Ts , as in discrete-time,
FIGURE 15. FLC input and output variables.
u(k + 1)Ts , until it reaches steady-state [132].

u(k + 1)Ts = u(kTs ) + 1u(kTs ) (17)

A. RULE-BASED FUZZY CONTROLLER
The FLC structure is a double-input-single-output type. For where 1u(kTs ) is the new value of the controller output
the proper and accurate fuzzy system operation, the input control at t = kTs , and k is the value of (0, 1, 2, 3. . . ).
and output variables of the universe of discourse should The defuzzification process is implemented to convert or
be normalized to fit into the interval value of −1 and +1. extract crisp, u, (real signal or non-fuzzy output) output
Each input variable obtains five triangular-type MFs and two value from a fuzzy controller output [116]. The process then
trapezoidal-type MFs as shown in Fig. 16. The input and converts the numerical output to a control signal that drives
output variables associated with the fuzzy sets are NB, NM, the process to meet the desired outcomes. The method of
NS, ZERO, PS, PM, and PB. translating the rule firings into the crisp output values is
the COG. The method is implemented as a value of crisp
control action, which represents the fuzzy controller output as
shown in [117, eq. (18)]. This method obtains the equilibrium
position of the fuzzy region by determining the weighted
mean of the fuzzy region [133].
µk (uk ) × uk
P
COG = k P , (18)
k µk (uk )
where uk is the discrete element of output value (fuzzy set),
and µk (uk ) is its MF (value).

B. INVERTER CONTROL ALGORITHM

FIGURE 16. MFs for the error, change of error, and change of output. The purpose of the inverter control system is to generate a
desired and stabilized 50 Hz sinusoidal output waveform.
Given that seven MFs (fuzzy sets) of each input variable Eliminating the contents of the harmonic at the output and
exist, the fuzzy system obtains 49 possible rules. These control the constant DC input voltage are the main challenges
rules are presented in table form, called the fuzzy rule table for the inverter controller. These issues can be addressed

24946 VOLUME 7, 2019

M. A. Hannan et al.: Fuzzy Logic Inverter Controller in PV Applications

by generating PWM signals to gate the switching devices

of the inverter. Output voltage and current waveform must
be ensured to attain the phase condition called the unity
power factor. These objectives are accomplished with the
PWM technique. The duty cycles of the actuator or switching
devices are controlled to accomplish the control objectives.
The control approach is introduced to the control algo-
rithm, which is embedded in a system controller or DSP.
Fig. 17 presents the functional flowchart of the PWM control
algorithm of the inverter [134].
The flowchart shows that the final process requires gen-
eration of the IGBT duty cycle, which can be accomplished
through the PWM signal generation. The IGBT duty cycle
of the boost converter, PWMboost , is initialized to supply the
desired DC voltage to the inverter input [89]. The output
voltages of the three-phase inverter, va , vb , and vc , are then
sensed and transformed to the direct-quadrature (dq) refer-
ence frame to improve voltage regulation. An error signal is
generated by comparing voltages vd and vq to their respective
reference voltages, vdref and vqref , at dq reference frame. This
error signal guides the generation of the duty cycle of the
inverter, PWMinv , where a positive error contributes to higher
duty cycle, whereas a negative error indicates a lower duty
cycle [135]. This loop continues until a minimum error of
the voltage is attained to reach an equilibrium state.
Fig. 17 shows the control algorithm, which is com- FIGURE 17. Inverter control algorithm flowchart.

posed of different components and functional blocks, such

as PLL-type frequency synchronization, three-axes abc to
two-axes dq transformation, voltage regulation that utilize Three-phase quantity-like voltages in the synchronous
FLC controllers, and PWM signal generation and signal reference frame (abc) must be transformed into the static
conversion. The PLL block provides a stable frequency of reference frame (dq) to apply the fuzzy controller to the PV
reference for the system control to stabilize the output fre- inverter system. This mechanism simplifies the development
quency at its desired level. The standalone configuration of the control system. The three-phase coordinate system
provides the PV generated power directly to the local loads. equations are expressed in a two-phase coordinate system
The voltage-control scheme is used for inverter operation, that rotates at the fundamental angular frequency, ω and
where the output voltages with unity power factor and dis- presented in Fig. 19. In most cases, the three-phase variables
placed 120◦ to one another are sampled and transformed to have no zero sequence components. However, the abc to dq
dq components. This scenario ensures that these parameters transformation is reduced to d and q components only, which
can be employed by controllers for control purposes. The may be applied in the balanced three-phase load condition.
three AC voltage quantities in balanced three-phase sys- Therefore, the dq coordinate frame contains νd and νq terms
tems are transformed to two imaginary DC quantities. This only by disregarding the ν◦ term. By applying the transforma-
transformation can conduct a simple calculation to analyze tion matrix, the AC varying quantities become DC quantities,
the three-phase synchronous machines or control the three- which are easier for the controller to address [134]. The FLCs
phase inverters before recovering the three-phase AC results can contain knowledge on the reference voltage in transient
by inverse transformation. The Park transformation [136] is and steady-state conditions for error minimization. The volt-
utilized in this axes transformation with a 50 Hz PLL internal ages νd and νq are regulated at reference voltages νdref and
or local synchronizing signal. νqref, respectively, by comparing the course that generates
By obtaining the output voltage or current feedback sam- error voltages. These error voltages are fed to the FLCs to
ples and implementing the control algorithm, the inverter regulate the output voltage of the inverter. The dq quanti-
controller can transfer the maximum PV power to the lead and ties can be retransformed into the three-phase quantities by
maintain the output voltage, current, and frequency at their multiplying with the inverse transformation matrix [134] to
desired levels. The switching signal generation that employs ensure that the inverter produces three-phase output quanti-
the voltage control approach for a standalone three-phase ties, such as voltage and current. Therefore, the controller
inverter is illustrated in [89, Fig. 18]. The fuzzy logic control establishes and stabilizes the phase voltage, va , of 240 Vrms
strategy is implemented in the special DSP called dSPACE for instance or line voltage, vab , of 415 Vrms and frequency
DS1104 [134], [135]. of 50 Hz.

VOLUME 7, 2019 24947

 M. A. Hannan et al.: Fuzzy Logic Inverter Controller in PV Applications

FIGURE 18. Control strategy for a standalone three-phase inverter.

of a filter, an output waveform of a near sinusoidal shape

is obtained. This filter has low impedance characteristics
for lower frequencies and high impedance characteristics for
higher order harmonics. The cut-off frequency of the filter
may be assigned as 1 or 2 octaves above the fundamental
switching frequency of 200 Hz [137]. Thus,
ftriangular
fharmonic = kMf = k (19)
fcontrol
where k is an integer, ftriangular is the frequency of the tri-
angular waveform, fcontrol is the frequency of the sinusoidal
control signal, and Mf is the frequency modulation index.
FIGURE 19. Relationship between frames abc and dq.
The RMS value of the line-to-line inverter output voltage
is described as follows:
C. INVERTER SWITCHING SIGNAL GENERATION Vline, RMS = 0.612 Ma Vdc (20)
At the final stage of the control algorithm depicted in Fig. 18,
the dq components are converted into PWM signals. At this where ftriangular is the frequency of the triangular waveform,
point, the components are retransformed to the abc coordinate Vdc is the DC input voltage, and Ma is the amplitude modu-
frame and then compared to the 6050 Hz carrier triangular lation index.
wave. The switching frequency of the inverter switches is The amplitude modulation index is the control variable
established by this triangular wave [60], thereby generating responsible for the RMS value if the line-line inverter output
the switching signals called SPWM. SPWM is a technique (fundamental) voltage. The modulation index is the ampli-
that provides switching frequency to the inverter IGBTs to tude ratio of the control signal, Vcontrol , to the triangular
produce the inverter AC outputs as a pure sine wave [36]. signal, Vtriangular , as shown in Equation (21). By varying
The duty cycles of the IGBTs switching are modulated by the index from 0 to 1, the output voltage varies from 0
the sinusoidal control signal, which is the desired voltage and Vdc to 0.612 Vdc . It corresponds to the varying of the duty
frequency of the inverter. cycle or width of the switching time. An over-modulation
The voltage that results from the preceding action has har- phenomenon may occur due to the increase in the index value
monic contents that appear as sidebands around the switching of 1. This scenario may result in excessive harmonics in the
frequency, which are expressed in [49, eq. (19)]. With the aid inverter outputs, which should be avoided. Thus, the index

24948 VOLUME 7, 2019

M. A. Hannan et al.: Fuzzy Logic Inverter Controller in PV Applications

value of 0.9 should be maintained [137]. the application of PV systems and their inverters. The details
AVcontrol of the challenges and the corresponding recommendations are
Ma = (21) as presented as follows.
AVtri
Using the desired three-phase RMS line output volt-
age of 415 V and maximum modulation index of 0.9, A. GRID-CONNECTED PV INVERTER
Equation (21) is utilized to determine the minimum require- PV-based grid-tied inverters have been developed for com-
ment of the inverter DC input voltage [137]. The filter output mercial viability and practical implementation, which could
is connected to the three-phase load. The controller manages enable power electronic controls to evolve to address the tech-
the overall operations, including controlling the switching of nical demand of interfacing PV inverters with the grid [144].
the power devices and regulating the voltage and current. A PV inverter with grid integration could provide voltage
regulation by sourcing or sinking reactive power. The tradi-
D. FLC BASED ADVANCED CONTROLLER tional PV inverter requires modifications with this feature by
Many research and developments have been done to con- oversizing the power electronic properly to support reactive
trol nonlinear systems with uncertainties. A controller has requirements and avoid the problem of excessive ripples
been developed for buck and boost converters to study the during high VAR production or absorption [145]. However,
uncertainties of the system using T2FLC [130]. It has been the voltage regulation and VAR support from the PV inverter
implemented in three phase inverter to upgrade the perfor- system has yet to be defined. The addition of this feature
mance of the system compared to T1FLC [138]. A design to the PV inverter system would also incur considerable
has been proposed by Bay for DC-DC boost converter that costs. Hence, determining how the capabilities of future PV
enhances the PV output voltage to three phase inverter input inverters toward solar energy grid integration systems can be
using Interval T2FLC (IT2FLC) [139]. It has been shown that enhanced is important.
IT2FLC utilized converter system produces superior perfor- An increase in PV penetration levels increase require that
mance that T1FLC based converter control in handling the certain issues be addressed, such as voltage upswing, volt-
uncertainties of the system. An adaptive neuro-fuzzy logic age regulation for cloud-induction, transient problems due
control has been presented to a grid connected multilevel to mass tripping of PV at low voltage or frequency pro-
PV inverter system [140] as a robust controller, however, files, power quality issues for anti-islanding, undue harmonic
it lags to define the performance to environmental sensi- contamination, and coordination problem of defensive relays
tivities of PV modules. Fei and Zhu [141] proposed an and fuses [146]. Future solar energy grid integration sys-
adaptive fuzzy sliding mode controller for grid connected tem inverters should have critical capabilities to work under
PV inverter system. In this adaptive fuzzy controller, T2FLC enhanced services and controls to microgrid integration, as
has been incorporated to estimate upper bound of uncer- well as serve as an interface with the control communication
tainties of noise due to ambient conditions that results to bus system of power systems [147]. Ensuring the reliability
adjust the nonlinearities for reliable control of inverter [141]. and consistency of the inverter hardware concerns is highly
With fast nonlinear learning capability the fuzzy neural net- imperative. The mean time between inverter failure and PV
work (FNN) has been applied to nonlinear systems in many module hardware has a large gap, and thus, inverters should
fields of control engineering including industrial processes. be substituted once or twice over their life span [39]. Inverter
A fault detection method using FNNs for nonlinear systems failure also causes a missed-opportunity cost for energy not
was demonstrated by Xue et al. [92]. Lin et al. [93] and generated. Thus, enhancing the efficiency to ensure the long
Wen et al. [94] presented an adaptive FNN to control perma- service life of inverters would result in lower energy genera-
nent magnet motor drive system and multi-input multi-output tion costs.
(MIMO) nonlinear systems, respectively. A novel adaptive
hybrid control system was proposed by El-Sousy [95] for B. INTERMITTENCY OF PV INVERTERS
permanent-magnet servo drive. A wavelet Petri fuzzy neu- Policies for injecting PV power back to the grid are chal-
ral network (WPFNN) was used for squirrel-cage induction lenging because of the large penetration of intermittent PV
generator system in [96]. Li et al. [90] presented an adap- generation without fast acting inverters. Fluctuations in solar
tive fuzzy output-feedback dynamic surface controller for irradiance and available PV power would lead to the follow-
nonlinear systems. Many advanced FLC schemes have been ing issues;
proposed and demonstrated in [97], [98], [142], and [143].
• Continuous switching of active and reactive power
VII. CHALLENGES AND RECOMMENDATIONS resources;
A potential power inverter must be able to produce a sinu- • Decreased equipment lifetime and increased mainte-
soidal waveform of constant magnitude irrespective of load nance costs;
type. It should also provide fast recovery from transients • Transient changes in PV generated voltage on a sudden
caused by external instabilities without hampering the power trip, particularly if the system undergoes continuous
quality. However, several challenges could raise difficulty in voltage regulation; and

VOLUME 7, 2019 24949

 M. A. Hannan et al.: Fuzzy Logic Inverter Controller in PV Applications

• Harmonics generation in the outputs of PV inverters, Different PWM modulation strategies have been devel-
and inverter interconnection with the distribution system oped, including carrier-based sinusoidal PWM, selective har-
during a possible resonant interaction. monic elimination, and space vector modulation, to decrease
A sophisticated forecast model should be developed to sus- harmonics and offer flexibility to select optimal ML
tain satisfactory levels of consistency and viability in power inverter switching. Computational complexity occurs with
systems to address PV intermittence issues. A smart inverter the increase in inverter levels. Therefore, fast computation,
with bidirectional communication capabilities should be used which conversely performs with poor accuracy, is neces-
to curtail transient information of the forecasted model and sary [121]. A new approach that offers possibilities of low
avoid disturbances, thereby overcoming the forecasted drop switching frequency while maintaining acceptable THD to
in PV generation. For the load demand management corre- include advanced control strategies could be adopted [150]
sponding to the flexible power generation in a grid-connected to address the issue. Other programming approaches, such
PV system, incentive-based demand response management as gradient-based methods, genetic algorithms, and search
may provide valuable assistance. This management system optimizations, have also been used as ML inverter control
can mitigate the predicted variations or inconsistencies in strategies. However, these approaches undergo a long com-
PV generation. putational process and suffer from optimization convergence
and sensitivity at difficult levels.
C. INVERTER CONTROLLER AND ITS ALGORITHM The selection of a PWM switching method for an inverter is
Many inverter control techniques to solve the problems of the a trade-off between switching frequency and harmonic distor-
PV system integration with AC loads and utility grids have tion. Therefore, considering the optimization of modulation
been recommended. PI controllers are commonly employed techniques that allows a lower switching frequency with har-
in inverter controls [31]. However, these controllers require monic reduction is noteworthy. Several approaches have been
a trapezoidal sum approximation to transform the function considered to define these issues. Nevertheless, most of these
from the continuous-time to the discrete-time domain for approaches have demerits in terms of the wide modulation
suitable processing and performance [103]. Optimization index use or increment in switching losses. Accordingly,
techniques have been recommended to tune the gains of the SVPWM and synchronous optimal PWM, which consider
PI controller and improve the performance [148]. Artificial low switching frequencies, algorithm simplification, and
intelligence-based control techniques have been applied in neutral-point balancing, are utilized in ML inverter con-
inverters with high efficiency and superior dynamics, such trollers. These approaches synchronize the switching and
as the artificial neural network and ANFIS. Nonetheless, fundamental frequencies, in which the optimized frequency
these controllers require large training data before they are minimizes the harmonic distortion. Thus, optimal modula-
trained and implemented in the controller. Presently, FLCs are tion techniques are more suitable. However, the optimization
applied in inverter control design because they are simple and technique is invalid for transient conditions and high over
can be adapted to any complex systems without mathematical currents are produced subsequently along with pulse pattern
model. However, FLCs are implemented with several rules changes. Thus, the challenge in minimizing switching loss
and MFs. Generally, these rules and MFs are determined by and waveform distortion with an extremely low switching
a trial and error procedure, which is time consuming and frequency could be an optimum switching control method
inefficient [149]. Therefore, the optimization algorithm is that may be useful in reaching this goal.
recommended to optimize FLC rules and MFs that enhance
the efficiency of the FLC design and performance and maxi- E. COMPLEX COMPUTATION DUE TO FUZZY
mize power generation in a grid-connected PV inverter. MF AND RULES
As a convention, the selection of MF number and limits of
D. MULTI-LEVEL INVERTER SWITCHING each MF in the FLC structure should be in a suitable position
The switching losses of a multi-level (ML) inverter comprise set obtained by means of a trial-and-error procedure until
the maximum percentage of total device losses. A decrease the FLC provides a favorable result. However, this proce-
in switching frequency produces more harmonic distortion in dure can be difficult to accomplish because it incurs long
inverter output waveforms. ML inverters can provide output computational time and determining the MF boundaries is
voltage waveforms with less harmonic distortion. However, difficult [149]. Thus, an adaptive FLC design technique that
more switching devices translate to lower efficiency and can control the inverter system by adopting the optimization
consistency of the power converter. Thus, optimizing switch- algorithm should be utilized, thereby avoiding the rigorous
ing frequencies that lessen the harmonic distortion in output outdated trial-and-error approach and obtaining the mini-
waveforms is essential to ensure the high power quality and mized or best MFs [148], [151]. The generated adaptive MFs
fast dynamic responses of ML inverters. should be implemented in the input and output of the fuzzy
Low switching frequency methods are necessary to lower inverter controller to solve the problems of the scalar or vector
the switching losses for high power applications. How- control. FLC rules are also formed by the IF–THEN condition
ever, high switching frequency methods are appropriate for to develop the control algorithm. These rules are executed
high dynamic range applications with high power quality. to implement the control decisions of a controller. The more

24950 VOLUME 7, 2019

M. A. Hannan et al.: Fuzzy Logic Inverter Controller in PV Applications

linguistic variables with more linguistic values, the more pos- fast turn ‘‘on’’ and ‘‘off’’ feature of the MOSFET [60]. These
sible rules are generated [117], [123]. Therefore, a processor features are suitable for designing an inverter that employs
with high speed and space may be required to ensure efficient high switching frequency to produce suitable power quality
intelligent control decision for FLC and its operation. output with low conduction loss. A drive circuit is required
to switch the transistor state to either the ‘‘on’’ or ‘‘off’’
F. PASSIVE FILTER INTEGRATION state. The operation of the gate driver is essential because it
An efficient inverter system produces a stable and smooth affects the electrical performance of power converters, such
output voltage signal with low THD value in high-power as reliability and efficiency [156].
applications [150]. The three-phase inverter system faces a
high harmonics problem that harms the load system [152]. VIII. CONCLUSIONS
Therefore, the inverter system requires filters to lower the har- A power inverter system is utilized to serve as the interface
monics effect to the loads. The quality of the inverter output for the DC power source from the PV to the utility grid
also depends on the design and performance of the filter along or AC load. The optimal design of the inverter system and
with the DC input voltage quality [71]. Hence, the use of a its robust intelligent control has vigorous PV applications.
filter at the input side of the inverter system may enhance the The inverter with FLC controls the fluctuation of DC power
output quality. An LC passive filter of higher order may be from the PV and generates the quality and state of the out-
utilized in an inverter system application to lower higher har- put to loads. This study reviews and discusses the control
monics components and system design cost [153]. The plant system technology of a three-phase inverter and PV. Many
parameters should be controlled to integrate the filter in the controller algorithms and methods, which have been imple-
inverter system. This mechanism could also be implemented mented in PV power converters (e.g., boost converters and
by using voltage and current sensors, which are then fed to inverters), have been investigated to address issues such as
the input of the control system to track the desired output of system stability and output power quality. Different types
the filter through the control algorithm. of inverter system platforms have been presented, and have
been utilized and incorporated for inverter improvement. The
G. INVERTER POWER QUALITY ISSUES overall design concept of the fuzzy logic-based PV inverter
In power grid applications with PV system, maintaining the control algorithm was also discussed. A comparison of AI
power quality is an issue because of the fluctuations of the control methods for PV inverter system is presented with their
DC power source. A high value DC-link capacitor may be operations, advantages and disadvantages. Moreover, the per-
essential to stabilize the DC-link voltage and decrease the formance of inverter systems is reviewed in tabular format to
voltage ripple at the switching time of power devices, which show the control robustness and the application progresses.
can result in suitable inverter power quality [70], [71]. Main- This review includes the development of fuzzy rules and
taining the output signal of the inverter system with low system control strategy to accomplish the control objective
THD level is another challenge for the power quality issue, of converting DC input voltage to AC sinusoidal waveform to
which complies with the standard requirement (e.g., IEEE power up AC loads. The concept of PWM signal generation
Std 519-1992) [22], [154]. Given the power conversion has also been presented to accomplish the aforementioned
process, voltage and current qualities may be concerned with control objective. The current issues and challenges of a fuzzy
deviations from the ideal of a single-frequency sine wave logic inverter system in PV applications and generations have
of constant amplitude and frequency. Several power quality been highlighted to describe a grid-connected PV inverter,
issues include harmonics, voltage dips, voltage flickers, and intermittency of PV inverters, inverter controller and its algo-
noise [32]. Harmonics because of waveform distortion result rithm, multi-level inverter switching, complex computation
in equipment overheating, increased losses, decreased volt– because of fuzzy MF and rules, passive filter integration,
ampere capacity, and voltage and current waveform distor- and inverter power quality issues. This study presents several
tion [152]. The harmonic level of the output waveforms of a recommendations to enhance and improve the FLC-based
system is determined by the THD value [43]. The selection inverter system, as well as direct the future development of
of the controller is essential because it determines the amount intelligent inverter systems in PV applications and genera-
of plant output and desired value deviation [77]. In particular, tions. Conventional energy sources can be exhausted, and
a suitable controller tracks the reference value accurately, hence, finding alternative energy sources, such as PV, has
lowers the deviation to zero or the smallest value possible, become highly significant. Inverter controllers are considered
and sustains the fast transient response without overshooting a significant field in PV systems.
(above the desired output) and steady-state response with
REFERENCES
minimal error [155]. The control system must also have
[1] V. Salas and E. Olias, ‘‘Overview of the state of technique for PV inverters
a system stability criterion that allows the system natural used in low voltage grid-connected PV systems: Inverters below 10 kW,’’
response to decay to zero eventually as time approaches Renew. Sustain. Energy Rev., vol. 13, pp. 1541–1550, Aug./Sep. 2009.
infinity or undergo oscillation. The IGBT may be considered [2] J. Sheehan, V. Camobreco, J. Duffield, H. Shapouri, M. Graboski, and
K. S. Tyson, ‘‘An overview of biodiesel and petroleum diesel life cycles,’’
a controllable switch for the inverter, which has the feature in Agriculture and Energy. Golden, CO, USA: National Renewable
of lower conduction losses (in ‘‘on’’ state) of the BJT and the Energy Laboratory, U.S. Department of Energy, 1998, pp. 1–35.

VOLUME 7, 2019 24951

 M. A. Hannan et al.: Fuzzy Logic Inverter Controller in PV Applications

[3] A. Demirbas, ‘‘Global renewable energy projections,’’ Energy Sources, [27] M. A. Eltawil and Z. Zhao, ‘‘Grid-connected photovoltaic power systems:
B, Econ., Planning, Policy, vol. 4, no. 2, pp. 212–224, 2009. Technical and potential problems—A review,’’ Renew. Sustain. Energy
[4] G. E. Ahmad, H. M. S. Hussein, and H. H. El-Ghetany, ‘‘Theoretical Rev., vol. 14, no. 1, pp. 112–129, 2010.
analysis and experimental verification of PV modules,’’ Renew. Energy, [28] M. I. Marei, E. F. El-Saadany, and M. M. A. Salama, ‘‘A novel control
vol. 28, pp. 1159–1168, Jul. 2003. algorithm for the DG interface to mitigate power quality problems,’’ IEEE
[5] F. Blaabjerg, Z. Chen, and S. B. Kjaer, ‘‘Power electronics as efficient Trans. Power Del., vol. 19, no. 3, pp. 1384–1392, Jul. 2004.
interface in dispersed power generation systems,’’ IEEE Trans. Power [29] M. V. Rajkumar, P. S. Manoharan, A. Ravi, ‘‘Simulation and an experi-
Electron., vol. 19, no. 5, pp. 1184–1194, Sep. 2004. mental investigation of SVPWM technique on a multilevel voltage source
[6] L. El Chaar, L. A. Lamont, and N. El Zein, ‘‘Review of photovoltaic inverter for photovoltaic systems,’’ Int. J. Elect. Power Energy Syst.,
technologies,’’ Renew. Sustain. Energy Rev., vol. 15, pp. 2165–2175, vol. 52, pp. 116–131, Nov. 2013.
Jun. 2011. [30] K. Bandara, T. Sweet, and J. Ekanayake, ‘‘Photovoltaic applications for
[7] T. K. Manna and S. M. Mahajan, ‘‘Nanotechnology in the development off-grid electrification using novel multi-level inverter technology with
of photovoltaic cells,’’ in Proc. Int. Conf. Clean Elect. Power, May 2007, energy storage,’’ Renew. Energy, vol. 37, pp. 82–88, Jan. 2012.
pp. 379–386. [31] J. Selvaraj and N. A. Rahim, ‘‘Multilevel inverter for grid-connected
[8] D. L. Staebler and C. R. Wronski, ‘‘Reversible conductivity changes PV system employing digital PI controller,’’ IEEE Trans. Ind. Electron.,
in discharge-produced amorphous Si,’’ Appl. Phys. Lett., vol. 31, no. 4, vol. 56, no. 1, pp. 149–158, Jan. 2009.
pp. 292–293, 1977. [32] L. Hassaine, E. Olías, J. Quintero, and A. Barrado, ‘‘Power control for
[9] J. Britt and C. Ferekides, ‘‘Thin-film CdS/CdTe solar cell with 15.8 grid connected applications based on the phase shifting of the inverter
percent efficiency,’’ Appl. Phys. Lett., vol. 62, no. 22, pp. 2851–2852, output voltage with respect to the grid voltage,’’ Int. J. Elect. Power
1993. Energy Syst., vol. 57, pp. 250–260, May 2014.
[10] M. Powalla and D. Bonnet, ‘‘Thin-film solar cells based on the polycrys- [33] N. Yousefpoor, S. H. Fathi, N. Farokhnia, and H. A. Abyaneh, ‘‘THD
talline compound semiconductors CIS and CdTe,’’ Adv. OptoElectron., minimization applied directly on the line-to-line voltage of multilevel
vol. 2007, Jul. 2007, Art. no. 97545. inverters,’’ IEEE Trans. Ind. Electron., vol. 59, no. 1, pp. 373–380,
[11] PV NEWS. (Mar. 2008). Greentech Media and Promotheus Jan. 2012.
Institute. [Online]. Available: http://www.pvtech.org/news/a/ [34] K. O. Lin, H. M. Tun, Z. M. Naing, and W. K. Moe, ‘‘Microcontroller
concentratingsolarpowera200billionbusinessthrough2020/ based SPWM single-phase inverter for wind power application,’’ Int.
[12] A. Luque et al., ‘‘FULLSPECTRUM: A new PV wave making more J. Sci. Technol. Res., vol. 4, no. 8, pp. 189–192, 2015.
efficient use of the solar spectrum,’’ Sol. Energy Mater. Sol. Cells, vol. 87, [35] F. Rodrigues and T. H. Barton, ‘‘A McMurray inverter controller,’’ IEEE
pp. 467–479, May 2005. Trans. Ind. Electron. Control Instrum., vol. IECI-2, no. 3, pp. 156–160,
[13] E. Kymakis and G. A. J. Amaratunga, ‘‘Electrical properties of single- Aug. 1979.
wall carbon nanotube-polymer composite films,’’ J. Appl. Phys., vol. 99, [36] M. H. Rashid, Power Electronics: Circuits, Devices, and Applications,
no. 8, 2006, Art. no. 084302. 3rd ed. New Delhi, India: Prentice-Hall, 2007.
[14] T. Fangsuwannarak et al., ‘‘Conductivity of self-organized silicon quan-
[37] J. Faiz and G. Shahgholian, ‘‘Modeling and simulation of a three-phase
tum dots embedded in silicon dioxide,’’ Proc. SPIE, vol. 6037, Dec. 2006,
inverter with rectifier-type nonlinear loads,’’ Armenian J. Phys., vol. 2,
Art. no. 60370T.
no. 4, pp. 307–316, 2009.
[15] A. Luque and A. Martí, ‘‘Electron–phonon energy transfer in hot-carrier
[38] G.-Q. Zeng, J. Chen, Y.-X. Dai, L.-M. Li, C.-W. Zheng, and M.-R. Chen,
solar cells,’’ Sol. Energy Mater. Sol. Cells, vol. 94, no. 2, pp. 287–296,
‘‘Design of fractional order PID controller for automatic regulator voltage
2009.
system based on multi-objective extremal optimization,’’ Neurocomput-
[16] S. Vavilapalli, S. Padmanaban, U. Subramaniam, and L. Mihet-Popa,
ing, vol. 160, pp. 173–184, Jul. 2015.
‘‘Power balancing control for grid energy storage system in photovoltaic
[39] F. Blaabjerg and E. Koutroulis, ‘‘Methods for the optimal design of
applications—Real time digital simulation implementation,’’ Energies,
grid—Connected PV inverters,’’ Int. J. Renew. Energy Res., vol. 1, no. 2,
vol. 9, no. 7, p. 928, 2017. doi: 10.3390/en10070928.
pp. 54–64, 2011.
[17] D. C. Martins, ‘‘Analysis of a three-phase grid-connected PV power
system using a modified dual-stage inverter,’’ ISRN Renew. Energy, [40] W. Libo, Z. Zhengming, and L. Jianzheng, ‘‘A single-stage three-phase
vol. 2013, Sep. 2012, Art. no. 406312. grid-connected photovoltaic system with modified MPPT method and
reactive power compensation,’’ IEEE Trans. Energy Convers., vol. 22,
[18] M. M. Hoque, M. A. Hannan, and A. Mohamed, ‘‘Optimal algorithms for
no. 4, pp. 881–886, Dec. 2007.
the charge equalisation controller of series connected lithium-ion battery
cells in electric vehicle applications,’’ IET Elect. Syst. Transp., vol. 7, [41] A. Luo, Y. Chen, Z. Shuai, and C. Tu, ‘‘An improved reactive current
no. 4, pp. 267–277, 2017. detection and power control method for single-phase photovoltaic grid-
[19] P. Sanjeevikumar and K. Rajambal, ‘‘Extra-high-voltage DC-DC boost connected DG system,’’ IEEE Trans. Energy Convers., vol. 28, no. 4,
converters topology with simple control strategy,’’ Model. Simul. Eng., pp. 823–831, Dec. 2013.
vol. 2008, pp. 1–8, Jan. 2008. [42] E. Koutroulis and F. Blabbjerg, ‘‘Methodology for the optimal design
[20] R. A. Messenger and J. Ventre, Photovoltaic Systems Engineering, 2nd ed. of transformerless grid-connected PV inverters,’’ IET Power Electron.,
Boca Raton, FL, USA: CRC Press, 2004. vol. 5, no. 8, pp. 1491–1499, 2012.
[21] P. Karuppusamy, G. Vijayakumar, and S. Sathishkumar, ‘‘Certain inves- [43] A. Keyhani, M. N. Marwali, and M. Dai, Integration of Green and
tigation on multilevel inverters for photovoltaic grid connected system,’’ Renewable Energy in Electric Power Systems. New Jersey, NJ, USA:
J. Circuits, Syst. Comput., vol. 25, no. 9, 2016, Art. no. 1650108. Wiley, 2010.
[22] D. S. Wankhede and M. V. Aware, ‘‘A neoteric space vector PWM [44] S. Nema, R. K. Nema, and G. Agnihotri, ‘‘Inverter topologies and control
method for multilevel inverter using image processing,’’ in Proc. IEEE structure in photovoltaic applications: A review,’’ J. Renew. Sustain.
Int. Conf. Smart Technol. Manage. Comput., Commun., Controls, Energy Energy, vol. 3, no. 1, 2011, Art. no. 012701.
Mater. (ICSTM), Chennai, India, Aug. 2017, pp. 364–370. [45] P. Veena, V. Indragandhi, R. Jeyabharath, and V. Subramaniyaswamy,
[23] K. Prabakar, M. Shirazi, A. Singh, and S. Chakraborty, ‘‘Advanced ‘‘Review of grid integration schemes for renewable power generation
photovoltaic inverter control development and validation in a controller- system,’’ Renew. Sustain. Energy Rev., vol. 34, pp. 628–641, Jun. 2014.
hardware-in-the-loop test bed,’’ in Proc. IEEE Energy Convers. Congr. [46] T.-F. Wu, C.-H. Chang, L.-C. Lin, G. -R. Yu, and Y.-R. Chang,
Expo. (ECCE), Cincinnati, OH, USA, Oct. 2017, pp. 1673–1679. ‘‘A D-6 digital control for three-phase inverter to achieve active and
[24] R. Akkaya and A. A. Kulaksiz, ‘‘A microcontroller-based stand-alone reactive power injection,’’ IEEE Trans. Ind. Electron., vol. 61, no. 8,
photovoltaic power system for residential appliances,’’ Appl. Energy, pp. 3879–3890, Aug. 2014.
vol. 78, pp. 419–431, Aug. 2004. [47] M. R. Patel, Wind and Solar Power Systems. Boca Raton, FL, USA:
[25] A. H. Azman and M. Muhamad, ‘‘Design of transformer-less DC to CRC Press, 1999.
AC converter using PWM technique,’’ in Proc. 3rd Int. Conf. Electron. [48] P. Herschan and K. Zweibel, Basic Photovoltaic Principles and Methods.
Design (ICED), Phuket, Thailand, Aug. 2016, pp. 547–552. Colorado, CO, USA: Solar Energy Research Institute, 1982.
[26] P. C. Loh and D. G. Holmes, ‘‘Analysis of multiloop control strategies for [49] G. R. Walker and P. C. Sernia, ‘‘Cascaded DC-DC converter connection
LC/CL/LCL-filtered voltage-source and current-source inverters,’’ IEEE of photovoltaic modules,’’ IEEE Trans. Power Electron., vol. 19, no. 4,
Trans. Ind. Appl., vol. 41, no. 2, pp. 644–654, Mar. 2005. pp. 1130–1139, Jul. 2004.

24952 VOLUME 7, 2019

M. A. Hannan et al.: Fuzzy Logic Inverter Controller in PV Applications

[50] P. J. Pérez, G. Almonacid, J. Aguilera, and J. de la Casa, ‘‘RMS current [73] R.-J. Wai, W.-H. Wang, and C.-Y. Lin, ‘‘High-performance stand-alone
of a photovoltaic generator in grid-connected PV systems: Definition and photovoltaic generation system,’’ IEEE Trans. Ind. Electron., vol. 55,
application,’’ Int. J. Photoenergy, vol. 2008, Feb. 2008, Art. no. 356261. no. 1, pp. 240–250, Jan. 2008.
[51] A. El Amrani, A. Mahrane, F. Y. Moussa, and Y. Boukennous, [74] Y. Ounejjar, K. Al-Haddad, and L.-A. Gregoire, ‘‘Packed U cells multi-
‘‘Solar module fabrication,’’ Int. J. Photoenergy, vol. 2007, Feb. 2007, level converter topology: Theoretical study and experimental validation,’’
Art. no. 27610. IEEE Trans. Ind. Electron., vol. 58, no. 4, pp. 1294–1306, Apr. 2011.
[52] B. K. Bose, P. M. Szczesny, and R. L. Steigerwald, ‘‘Microcomputer [75] R. M. Kamel, ‘‘New inverter control for balancing standalone micro-grid
control of a residential photovoltaic power conditioning system,’’ IEEE phase voltages: A review on MG power quality improvement,’’ Renew.
Trans. Ind. Appl., vol. IA-21, no. 5, pp. 1182–1191, Sep. 1985. Sustain. Energy Rev., vol. 63, pp. 520–532, Sep. 2016.
[53] Y. Zhu and J. Fei, ‘‘Disturbance observer based fuzzy sliding [76] R. C. Dorf and R. H. Bishop, Modern Control Systems, 12th ed.
mode control of PV grid connected inverter,’’ IEEE Access, vol. 6, New Jersey, NJ, USA: Prentice Hall, 2011.
pp. 21202–21211, 2018. [77] K. Ogata, Modern Control Engineering, 5th ed. New Jersey, NJ, USA:
[54] G. Grandi, C. Rossi, D. Ostojic, and D. Casadei, ‘‘A new multilevel Pearson Education, 2010.
conversion structure for grid-connected PV applications,’’ IEEE Trans. [78] M. Gaballah and M. El-Bardini, ‘‘Low cost digital signal generation
Ind. Electron., vol. 56, no. 11, pp. 4416–4426, Nov. 2009. for driving space vector PWM inverter,’’ Ain Shams Eng. J., vol. 4,
[55] T. Shimizu, M. Hirakata, T. Kamezawa, and H. Watanabe, ‘‘Generation pp. 763–774, Dec. 2013.
control circuit for photovoltaic modules,’’ IEEE Trans. Power Electron., [79] L. Saribulut, A. Teke, and M. Tümay, ‘‘Vector-based reference location
vol. 16, no. 3, pp. 293–300, May 2001. estimating for space vector modulation technique,’’ Electr. Power Syst.
[56] H. Boumaaraf, A. Talha, and O. Bouhali, ‘‘A three-phase NPC grid- Res., vol. 86, pp. 51–60, May 2012.
connected inverter for photovoltaic applications using neural network [80] D. A. Dominic and T. R. Chelliah, ‘‘Analysis of field-oriented controlled
MPPT,’’ Renew. Sustain. Energy Rev., vol. 49, pp. 1171–1179, Sep. 2015. induction motor drives under sensor faults and an overview of sensorless
[57] T. V. Thang, A. Ahmed, C.-I. Kim, and J.-H. Park, ‘‘Flexible system schemes,’’ ISA Trans., vol. 53, pp. 1680–1694, Sep. 2014.
architecture of stand-alone PV power generation with energy storage [81] L. K. Letting, J. L. Munda, and Y. Hamam, ‘‘Optimization of a fuzzy logic
device,’’ IEEE Trans. Energy Convers., vol. 30, no. 4, pp. 1386–1396, controller for PV grid inverter control using S-function based PSO,’’ Solar
Dec. 2015. Energy, vol. 86, pp. 1689–1700, Jun. 2012.
[58] H. Qin and J. W. Kimball, ‘‘Closed-loop control of DC–DC dual-active- [82] A. R. Beig, ‘‘Constant v/f induction motor drive with synchronised
bridge converters driving single-phase inverters,’’ IEEE Trans. Power space vector pulse width modulation,’’ IET Power Electron., vol. 5,
Electron., vol. 29, no. 2, pp. 1006–1017, Feb. 2014. pp. 1446–1455, Sep. 2012.
[59] X. Li and A. K. S. Bhat, ‘‘A comparison study of high-frequency isolated [83] A. B. de Souza Júnior, E. de Castro Diniz, D. de Araújo Honório,
DC/AC converter employing an unfolding LCI for grid-connected alter- L. H. S. C. Barreto, and L. L. N. dos Reis, ‘‘Hybrid control robust using
native energy applications,’’ IEEE Trans. Power Electron., vol. 29, no. 8, logic fuzzy applied to the position loop for vector control to induction
pp. 3930–3941, Aug. 2014. motors,’’ Electr. Power Compon. Syst., vol. 42, no. 6, pp. 533–543, 2014.
[60] N. Mohan, T. M. Undeland, and W. P. Robbins, Power Electronics: [84] R. Abdelli, D. Rekioua, and T. Rekioua, ‘‘Performances improvements
Converters, Applications, and Design, 3rd ed. New Jersey, NJ, USA: and torque ripple minimization for VSI fed induction machine with direct
Wiley, 2003. control torque,’’ ISA Trans., vol. 50, pp. 213–219, Apr. 2011.
[61] S. Zhang and X. Yu, ‘‘Control strategy to achieve minimum/zero input [85] M. C. Paicu, I. Boldea, G.-D. Andreescu, and F. Blaabjerg, ‘‘Very low
current ripple for the interleaved boost converter in photovoltaic/fuel speed performance of active flux based sensorless control: Interior per-
cell power conditioning system,’’ in Proc. IEEE Energy Convers. Congr. manent magnet synchronous motor vector control versus direct torque
Expo., Sep. 2012, pp. 4301–4306. and flux control,’’ IET Electr. Power Appl., vol. 3, no. 6, pp. 551–561,
[62] A. A. Ahmad, A. Abrishamifar, and S. Samadi, ‘‘Low-frequency current 2009.
ripple reduction in front-end boost converter with single-phase inverter [86] S. Maiti, V. Verma, C. Chakraborty, and Y. Hori, ‘‘An adaptive speed sen-
load,’’ IET Power Electron., vol. 5, no. 9, pp. 1676–1683, 2012. sorless induction motor drive with artificial neural network for stability
[63] M. A. G. de Brito, L. Galotto, L. P. Sampaio, G. E. de Azevedo e Melo, enhancement,’’ IEEE Trans. Ind. Informat., vol. 8, no. 4, pp. 757–766,
and C. A. Canesin, ‘‘Evaluation of the main MPPT techniques for Nov. 2012.
photovoltaic applications,’’ IEEE Trans. Ind. Electron., vol. 60, no. 3, [87] N. Langer, A. H. Bhat, and P. Agarwal, ‘‘Neural-network-based space-
pp. 1156–1167, Mar. 2013. vector pulse-width modulation for capacitor voltage balancing of
[64] M. Arsalan, R. Iftikhar, I. Ahmad, A. Hasan, K. Sabahat, and A. Javeria, three-phase three-level improved power quality converter,’’ IET Power
‘‘MPPT for photovoltaic system using nonlinear backstepping controller Electron., vol. 7, pp. 973–983, Apr. 2014.
with integral action,’’ Sol. Energy, vol. 170, pp. 192–200, Aug. 2018. [88] A. H. Bhat and P. Agarwal, ‘‘Implementation of a neural-network-
[65] J. El Khazane and E. H. Tissir, ‘‘Achievement of MPPT by finite time based space-vector pulse-width modulation for a three-phase neutral-
convergence sliding mode control for photovoltaic pumping system,’’ Sol. point clamped high-power factor converter,’’ Electr. Power Compon.
Energy, vol. 166, pp. 13–20, May 2018. Syst., vol. 37, no. 2, pp. 210–233, 2009.
[66] M. A. Hannan, M. M. Hoque, A. Mohamed, and A. Ayob, ‘‘Review of [89] M. A. Hannan, Z. A. Ghani, A. Mohamed, and M. N. Uddin, ‘‘Real-time
energy storage systems for electric vehicle applications: Issues and chal- testing of a fuzzy-logic-controller-based grid-connected photovoltaic
lenges,’’ Renew. Sustain. Energy Rev., vol. 69, pp. 771–789, Mar. 2017. inverter system,’’ IEEE Trans. Ind. Appl., vol. 51, no. 6, pp. 4775–4784,
[67] T. Esram and P. L. Chapman, ‘‘Comparison of photovoltaic array max- Nov./Dec. 2015.
imum power point tracking techniques,’’ IEEE Trans. Energy Convers., [90] Y. Li, S. Tong, and T. Li, ‘‘Adaptive fuzzy output feedback dynamic sur-
vol. 22, no. 2, pp. 439–449, Jun. 2007. face control of interconnected nonlinear pure-feedback systems,’’ IEEE
[68] N. Femia, G. Petrone, G. Spagnuolo, and M. Vitelli, ‘‘Optimization of Trans. Cybern., vol. 45, no. 1, pp. 138–149, Jan. 2015.
perturb and observe maximum power point tracking method,’’ IEEE [91] P. Garcia, C. A. Garcia, L. M. Fernandez, F. Llorens, and F. Jurado,
Trans. Power Electron., vol. 20, no. 4, pp. 963–973, Jul. 2005. ‘‘ANFIS-based control of a grid-connected hybrid system integrating
[69] J. P. Lee, B. D. Min, T. J. Kim, D. W. Yoo, and J. Y. Yoo, ‘‘A novel renewable energies, hydrogen and batteries,’’ IEEE Trans. Ind. Informat.,
topology for photovoltaic DC/DC full-bridge converter with flat effi- vol. 10, no. 2, pp. 1107–1117, May 2014.
ciency under wide PV module voltage and load range,’’ IEEE Trans. Ind. [92] H. Xue and J. G. Jiang, ‘‘Fault detection and accommodation for nonlinear
Electron., vol. 55, no. 7, pp. 2655–2663, Jul. 2008. systems using fuzzy neural networks,’’ in Proc. CES/IEEE 5th Int. Power
[70] M. E. Başoğlu and B. Çakir, ‘‘Comparisons of MPPT performances of Electron. Motion Control Conf., vol. 3, Aug. 2006, pp. 1–5.
isolated and non-isolated DC–DC converters by using a new approach,’’ [93] C.-H. Lin, C.-Y. Wei, and M.-T. Wang, ‘‘The fuzzy neural network control
Renew. Sustain. Energy Rev., vol. 60, pp. 1100–1113, Jul. 2016. with adaptive algorithm for a PM synchronous motor drive,’’ in Proc. 6th
[71] M. A. Hannan et al., ‘‘Advanced control strategies of VSC based HVDC IEEE Conf. Ind. Electron. Appl., vol. 124, Jun. 2011, pp. 2518–2523.
transmission system: Issues and potential recommendations,’’ IEEE [94] J. Fei and T. Wang, ‘‘Adaptive fuzzy-neural-network based on RBFNN
Access, vol. 6, pp. 78352–78369, 2018. doi: 10.1109/ACCESS.2018. control for active power filter,’’ International Journal of Machine
2885010. Learning and Cybernetics. Berlin, Germany: Springer-Verlag, 2018,
[72] A. Mariaud, S. Acha, N. Ekins-Daukes, N. Shah, and C. N. Markides, pp. 1–12.
‘‘Integrated optimisation of photovoltaic and battery storage systems for [95] F. F. M. El-Sousy, ‘‘Adaptive hybrid control system using a recurrent rbfn-
UK commercial buildings,’’ Appl. Energy, vol. 199, no. 1, pp. 466–478, based self-evolving fuzzy-neural-network for PMSM servo drives,’’ Appl.
2017. Soft Comput., vol. 21, no. 8, pp. 509–532, 2014.

VOLUME 7, 2019 24953

 M. A. Hannan et al.: Fuzzy Logic Inverter Controller in PV Applications

[96] F.-J. Lin, K.-H. Tan, and D.-Y. Fang, ‘‘Squirrel-cage induction generator [120] K. Zdenko and B. Stjepan, Fuzzy Controller Design-Theory and Applica-
system using hybrid wavelet fuzzy neural network control for wind tions. Boca Raton, FL, USA: CRC Press, 2006.
power applications,’’ Neural Comput. Appl., vol. 26, no. 4, pp. 911–928, [121] C. Cecati, F. Ciancetta, and P. Siano, ‘‘A multilevel inverter for photo-
2015. voltaic systems with fuzzy logic control,’’ IEEE Trans. Ind. Electron.,
[97] F.-J. Lin and P.-H. Shen, ‘‘Adaptive fuzzy-neural-network control for vol. 57, no. 12, pp. 4115–4125, Dec. 2010.
a dsp-based permanent magnet linear synchronous motor servo drive,’’ [122] K. W. Kow, Y. W. Wong, R. K. Rajkumar, and R. K. Rajkumar, ‘‘A review
IEEE Trans. Fuzzy Syst., vol. 14, no. 4, pp. 481–495, Aug. 2006. on performance of artificial intelligence and conventional method in
[98] C.-F. Hsu, C.-M. Lin, and M.-C. Li, ‘‘Adaptive dynamic RBF fuzzy neural mitigating PV grid-tied related power quality events,’’ Renew. Sustain.
controller design with a constructive learning,’’ Int. J. Fuzzy Syst., vol. 13, Energy Rev., vol. 56, pp. 334–346, Apr. 2016.
no. 3, pp. 175–184, 2011. [123] G. J. Li, X. P. Zhang, S. S. Choi, T. T. Lie, and Y. Z. Sun ‘‘Control strat-
[99] M. A. Hannan, J. A. Ali, A. Mohamed, and A. Hussain, ‘‘Optimiza- egy for dynamic voltage restorers to achieve minimum power injection
tion techniques to enhance the performance of induction motor drives: without introducing sudden phase shift,’’ IET Gener., Transmiss. Distrib.,
A review,’’ Renew. Sustain. Energy Rev., vol. 81, no. 2, pp. 1611–1626, vol. 1, no. 5, pp. 847–853, 2007.
2018. [124] D. K. Chaturvedi, Modeling and Simulation of Systems Using MATLAB
[100] B. K. Sahu, T. K. Pati, J. R. Nayak, S. Panda, and S. K. Kar, ‘‘A novel and Simulink. Boca Raton, FL, USA: CRC Press, 2006.
hybrid LUS–TLBO optimized fuzzy-PID controller for load frequency [125] J. R. Nayaka, B. Shawa, and B. K. Sahu, ‘‘Application of adaptive-
control of multi-source power system,’’ Int. J. Elect. Power Energy Syst., SOS (ASOS) algorithm based interval type-2 fuzzy-PID controller
vol. 74, pp. 58–69, Jan. 2016. with derivative filter for automatic generation control of an intercon-
[101] V. R. Moorthi, Power Electronics: Devices, Circuits and Industrial Appli- nected power system,’’ Eng. Sci. Technol., Int. J., vol. 21, pp. 465–485,
cations, New Delhi, India: Oxford Univ. Press, 2005. Jun. 2018.
[102] D. Q. Dang, Y.-S. Choi, H. H. Choi, and J.-W. Jung, ‘‘Experimental [126] N. N. Karnik, J. M. Mendel, and Q. Liang, ‘‘Type-2 fuzzy logic systems,’’
validation of a fuzzy adaptive voltage controller for three-phase PWM IEEE Trans. Fuzzy Syst., vol. 7, no. 6, pp. 643–658, Dec. 1999.
inverter of a standalone DG unit,’’ IEEE Trans. Ind. Informat., vol. 11, [127] E. Hisdal, ‘‘The IF THEN ELSE statement and interval-valued fuzzy
no. 3, pp. 632–641, Jun. 2015. sets of higher type,’’ Int. J. Man-Mach. Stud., vol. 15, pp. 385–455,
[103] S. Sukumar, M. Marsadek, A. Ramasamy, H. Mokhlis, and S. Mekhilef, Nov. 1981.
‘‘A fuzzy-based PI controller for power management of a grid-connected [128] R. John, ‘‘Type 2 fuzzy sets: An appraisal of theory and applications,’’ Int.
PV-SOFC hybrid system,’’ Energies, vol. 10, no. 11, p. 1720, 2017. J. Uncertainty, Fuzziness Knowl.-Based Syst., vol. 6, no. 6, pp. 563–576,
doi: 10.3390/en10111720. 1998.
[104] A. Hmidet, R. Dhifaoui, and O. Hasnaoui, ‘‘Development, implementa- [129] O. P. Verma and H. Gupta, ‘‘Fuzzy logic based water bath temperature
tion and experimentation on a dSPACE DS1104 of a direct voltage control control system,’’ Int. J. Adv. Res. Comput. Sci. Softw. Eng., vol. 2, no. 4,
scheme,’’ J. Power Electron., vol. 10, no. 5, pp. 468–476, 2010. pp. 333–336, 2012.
[105] S. Alepuz, S. Busquets-Monge, J. Bordonau, J. Gago, D. Gonzalez, and [130] A. H. El Khateb, N. A. Rahim, and J. Selvaraj, ‘‘Type-2 fuzzy logic
J. Balcells, ‘‘Interfacing renewable energy sources to the utility grid approach of a maximum power point tracking employing SEPIC con-
using a three-level inverter,’’ IEEE Trans. Ind. Electron., vol. 53, no. 5, verter for photovoltaic system,’’ J. Clean Energy Technol., vol. 1, no. 1,
pp. 1504–1511, Oct. 2006. pp. 41–44, 2013.
[106] N. A. Rahim, J. Selvaraj, and C. Krismadinata, ‘‘Five-level inverter with [131] M. Sarvi and M. Keshmiri, ‘‘A fuzzy-PD controller to improve the
dual reference modulation technique for grid-connected PV system,’’ performance of HVDC system,’’ World Appl. Sci. J., vol. 22, no. 9,
Renew. Energy, vol. 35, no. 3, pp. 712–720, 2010. pp. 1210–1219, 2013.
[107] B. Parida, S. Iniyan, and R. Goic, ‘‘A review of solar photovoltaic tech- [132] Y. Sabzehmeidani, M. Mailah, M. Hussein, E. Gatavi, and M. M. Zain,
nologies,’’ Renew. Sustain. Energy Rev., vol. 15, no. 3, pp. 1625–1636, ‘‘A hybrid fuzzy-based robust controller for pneumatically actuated micro
2011. robot,’’ Int. Rev. Model. Simul., vol. 3, no. 6, pp. 1308–1316, 2010.
[108] M. Sreedevi and P. J. Paul, ‘‘Fuzzy PI controller based grid-connected PV [133] C. Guanrong and T. T. Pham, Introduction to Fuzzy Systems. Boca Raton,
system,’’ Int. J. Soft Comput., vol. 6, no. 1, pp. 11–15, 2011. FL, USA: CRC Press, 2006.
[109] P. Liutanakul, S. Pierfederici, and F. Meibody-Tabar, ‘‘Nonlinear control [134] N. Altin and I. Sefa, ‘‘dSPACE based adaptive neuro-fuzzy con-
techniques of a controllable rectifier/inverter-motor drive system with troller of grid interactive inverter,’’ Energy Convers. Manage., vol. 56,
a small DC-link capacitor,’’ Energy Convers. Manage., vol. 49, no. 12, pp. 130–139, Apr. 2012.
pp. 3541–3549, 2008. [135] D. S. Nimesh and M. P. Manisha, ‘‘Implementation of vector oriented con-
[110] S. Li, T. A. Haskew, D. Li, and F. Hu, ‘‘Integrating photovoltaic and power trol for induction motor using DS1104,’’ Int. J. Adv. Eng. Res. Develop.,
converter characteristics for energy extraction study of solar PV systems,’’ vol. 4, no. 1, pp. 132–141, 2017.
Renew. Energy, vol. 36, no. 12, pp. 3238–3245, 2011. [136] H. Saadat, Power System Analysis, 2nd ed. New York, NY, USA:
[111] S. Jayalath and M. Hanif, ‘‘Generalized LCL-filter design algorithm McGraw-Hill. 2004.
for grid-connected voltage-source inverter,’’ IEEE Trans. Ind. Electron., [137] J. M. Jacob, Power Electronics: Principles and Applications. Delmar Ltd.,
vol. 64, no. 3, pp. 1905–1915, Mar. 2017. 2002.
[112] S. Jain, M. B. Shadmand, and R. S. Balog ‘‘Decoupled active and reactive [138] I. Atacak and O. F. Bay, ‘‘A type-2 fuzzy logic controller design for buck
power predictive control for PV applications using a grid-tied quasi-Z- and boost DC–DC converters,’’ J. Intell. Manuf., vol. 23, pp. 1023–1034,
source inverter,’’ IEEE Trans. Emerg. Sel. Topics Power Electron., vol. 6, Aug. 2012.
no. 4, pp. 1769–1782, Dec. 2018. [139] O. F. Bay and M. O. Yatak, ‘‘Type-2 fuzzy logic control of a photovoltaic
[113] H. Armghan, I. Ahmad, A. Armghan, S. Khan, and M. Arsalan, ‘‘Back- sourced two stages converter,’’ J. Intell. Fuzzy Syst., vol. 35, no. 1,
stepping based non-linear control for maximum power point tracking in pp. 1103–1117, 2018.
photovoltaic system,’’ Sol. Energy, vol. 159, pp. 134–141, Jan. 2018. [140] T. Logeswaran, A. Senthilkumar, and P. Karuppusamy, ‘‘Adaptive neuro-
[114] Jaime Alonso-Martínez J. Eloy-Garcí, and S. Arnaltes,, ‘‘Direct power fuzzy model for grid-connected photovoltaic system,’’ Int. J. Fuzzy Syst.,
control of grid connected PV systems with three level NPC inverter,’’ Sol. vol. 17, no. 4, pp. 585–594, 2015.
Energy, vol. 84, pp. 1175–1186, 2010. [141] J. Fei and Y. Zhu, ‘‘Adaptive fuzzy sliding control of single-phase
[115] S. Subiyanto, A. Mohamed, and M. A. Hannan, ‘‘Intelligent maximum PV grid-connected inverter,’’ PLoS ONE, vol. 12, no. 8, 2017,
power point tracking for PV system using hopfield neural network opti- Art. no. e0182916.
mized fuzzy logic controller,’’ Energy Buildings, vol. 51, pp. 29–38, [142] B. Xu, D. Wang, Y. Zhang, and Z. Shi, ‘‘DOB-based neural control of
Aug. 2012. flexible hypersonic flight vehicle considering wind effects,’’ IEEE Trans.
[116] Z. Kovacic and S. Bogdan, Fuzzy Controller Design: Theory and Appli- Ind. Electron., vol. 64, no. 11, pp. 8676–8685, Nov. 2017.
cations, Boca Raton, FL, USA: CRC Press, 2006. [143] X. Wang, T. Zhang, and R. Wang, ‘‘Noniterative deep learning: Incor-
[117] M. Negnevitsky, Artificial Intelligence: A Guide to Intelligent Systemas, porating restricted Boltzmann machine into multilayer random weight
2nd ed. England, U.K.: Pearson Education Limited, 2005. neural networks,’’ IEEE Trans. Syst., Man, Cybern., Syst., to be published.
[118] L. Reznik, Fuzzy Controllers. New York, NY, USA: Oxford: Butterworth- doi: 10.1109/TSMC.2017.2701419.
Heinemann Linacre House, 1997. [144] I. Hamzaoui, F. Bouchafaa, and A. Hadjammar, ‘‘Investigation of the
[119] F. O. Karray and C. De Silva, Soft Computing and Intelligent Systems behavior of a three phase gridconnected photovoltaic system to control
Design: Theory, Tools and Applications. England, U.K.: Pearson Educa- active and reactive power with DPC,’’ Energy Procedia, vol. 6, no. 2011,
tion Limited, 2004. pp. 493–502, 2011.

24954 VOLUME 7, 2019

M. A. Hannan et al.: Fuzzy Logic Inverter Controller in PV Applications

[145] A. Kulkarni and V. John, ‘‘Mitigation of lower order harmonics in a ZAMRE ABD. GHANI received the B.Sc. degree
grid-connected single-phase PV inverter,’’ IEEE Trans. Power Electron., in electrical engineering from the University of the
vol. 28, no. 11, pp. 5024–5037, Nov. 2013. Pacific, Stockton, CA, USA, the M.Eng. degree in
[146] M. M. Haque and P. Wolfs, ‘‘A review of high PV penetrations in LV electrical engineering from Universiti Teknologi
distribution networks: Present status, impacts and mitigation measures,’’ Malaysia, Johor Bahru, Malaysia, and the Ph.D.
Renew. Sustain. Energy Rev., vol. 62, pp. 1195–1208, Sep. 2016. degree in electrical, electronic and systems engi-
[147] F. H. M. Rafi, M. J. Hossain, and J. Lu, ‘‘Hierarchical controls selection
based on PV penetrations for voltage rise mitigation in a LV distribu- neering from Universiti Kebangsaan Malaysia,
tion network,’’ Int. J. Elect. Power Energy Syst., vol. 81, pp. 123–139, Bangi, Malaysia, in 1987, 2007, and 2014, respec-
Oct. 2016. tively. He is currently a Senior Lecturer with the
[148] M. M. Hoque, M. A. Hannan, and A. Mohamed, ‘‘Charging and discharg- Faculty of Electronic and Computer Engineering,
ing model of lithium-ion battery for charge equalization control using Universiti Teknikal Malaysia, Melaka, Malaysia. His main research interest
particle swarm optimization algorithm,’’ J. Renew. Sustain. Energy, vol. 8, includes power electronic controllers for photovoltaic applications, such as
no. 6, 2016, Art. no. 065701. inverters and dc-dc converters.
[149] Y. Y. Ghadi, M. G. Rasul, and M. M. K. Khan, ‘‘Design and develop-
ment of advanced fuzzy logic controllers in smart buildings for institu-
tional buildings in subtropical Queensland,’’ Renew. Sustain. Energy Rev.,
vol. 54, pp. 738–744, Feb. 2016. MD. MURSHADUL HOQUE (S’16) received the
[150] R. Ortega, E. Figueres, G. Garcerá, C. L. Trujillo, and D. Velasco, ‘‘Con- B.Sc. degree in electrical and electronic engineer-
trol techniques for reduction of the total harmonic distortion in voltage ing from the Chittagong University of Engineering
applied to a single-phase inverter with nonlinear loads: Review,’’ Renew. and Technology, Chittagong, Bangladesh, in 2006,
Sustain. Energy Rev., vol. 16, no. 3, pp. 1754–1761, 2012.
and the M.Sc. degree in electrical, electronic and
[151] M. Farhat, O. Barambones, and L. Sbita, ‘‘Efficiency optimization of a
DSP-based standalone PV system using a stable single input fuzzy logic systems engineering from Universiti Kebangsaan
controller,’’ Renew. Sustain. Energy Rev., vol. 49, pp. 907–920, Sep. 2015. Malaysia, Malaysia, in 2016. He is currently an
[152] Q.-C. Zhong, ‘‘Harmonic droop controller to reduce the voltage harmon- Associate Professor with the Department of Elec-
ics of inverters,’’ IEEE Trans. Ind. Electron., vol. 60, no. 3, pp. 936–945, trical and Electronic Engineering, University of
Mar. 2013. Chittagong, Bangladesh. His research interests
[153] A. Houari, H. Renaudineau, J.-P. Martin, S. Pierfederici, and include power electronics, power conversion, BMS, power system control,
F. Meibody-Tabar, ‘‘Flatness-based control of three-phase inverter and automation,
with output LC filter,’’ IEEE Trans. Ind. Electron., vol. 59, no. 7,
pp. 2890–2897, Jul. 2012.
[154] A. Platakis, V. Bleizgys, A. Lucinskis, and A. Baskys, ‘‘Investigation of
photovoltaic inverter power quality,’’ Przeglad Elektrotechniczny, vol. 88, PIN JERN KER received the B.Eng. degree
no. 7b, pp. 180–182, 2012. (Hons.) in electrical and electronic engineer-
[155] G. F. Franklin, J. D. Powel, and A. Emami-Naeini, Feedback Control of ing from Universiti Tenaga Nasional (UNITEN),
Dynamic Systems. New Jersey, NJ, USA: Pearson Education, 2010.
[156] J. Yu, P. Liu, Q. Qian, W. Sun, and S. Lu ‘‘Gate-drive circuit with efficient Malaysia, in 2009, and the Ph.D. degree in elec-
energy recovery based on DC/DC converter,’’ Electron. Lett., vol. 52, tronic and electrical engineering from The Uni-
no. 11, pp. 952–953, 2016. versity of Sheffield, U.K. He is currently a Senior
Lecturer with the Department of Electrical Power
Engineering, UNITEN. He is also the Head of the
Unit (Electronics & IT) with the Institute of Power
Engineering, a research institute of UNITEN. His
research interests include the simulation and characterization of photode-
tectors, optical sensing, design of monitoring, and control system for
energy-related applications.

AINI HUSSAIN (M’98) received the B.Sc. degree

in electrical engineering from Louisiana State Uni-
versity, Baton Rouge, LA, USA, the M.Sc. degree
from UMIST, U.K., and the Ph.D. degree from
Universiti Kebangsaan Malaysia. She is currently
a Professor with the Centre for Integrated Systems
Engineering and Advanced Technologies, Faculty
of Engineering and Built Environment, Univer-
MAHAMMAD A. HANNAN (M’10–SM’17) siti Kebangsaan Malaysia. Her research interests
received the B.Sc. degree in electrical and elec- include decision support systems, machine learn-
tronic engineering from the Chittagong Univer- ing, pattern precognition, and signal and image processing.
sity of Engineering and Technology, Chittagong,
Bangladesh, in 1990, and the M.Sc. and Ph.D.
degrees in electrical, electronic, and systems engi-
neering from Universiti Kebangsaan Malaysia, AZAH MOHAMED (M’88–SM’05) received the
Bangi, Malaysia, in 2003 and 2007, respectively. B.Sc. degree from the University of London,
He is currently a Professor of intelligent systems in 1978, and the M.Sc. and Ph.D. degrees from
with the Department of Electrical Power Engineer- Universiti Malaya, in 1988 and 1995, respectively.
ing, College of Engineering, Universiti Tenaga Nasional, Malaysia. His She is currently a Professor with the Department
research interests include intelligent controllers, power electronics, hybrid of Electrical, Electronic and Systems Engineer-
vehicles, energy storage systems, image and signal processing, and artificial ing, Universiti Kebangsaan Malaysia. Her main
intelligence. He has received a number of gold awards for his innova- research interests include power system security,
tive research in ITEX, MTE, INNOFEST, SIIF, and PERINTIS. He is an power quality Azah, and artificial intelligence.
Associate Editor of the IEEE ACCESS.

VOLUME 7, 2019 24955