International Journal of Applied Power Engineering (IJAPE)

Vol. 11, No. 2, June 2022, pp. 120~133

ISSN: 2252-8792, DOI: 10.11591/ijape.v11.i2.pp120-133  120

Power quality improvement of grid-connected photovoltaic

systems using PI-fuzzy controller

Marwan Saded Ahmed, Dhari Yousif Mahmood, Ali Hussein Numan

Department of Electrical Engineering, University of Technology, Baghdad, Iraq

Article Info ABSTRACT

Article history: To ensure enhanced reliability and availability of electricity to consumers,
grid-connected photovoltaic systems need to improve their power quality,
Received Sep 4, 2021 this paper uses a three-phase five levels cascaded H-bridge inverter in grid-
Revised Mar 5, 2022 connected mode to improve the flexibility and efficiency of the photovoltaic
Accepted Mar 24, 2022 system. Each photovoltaic (PV) array in the proposed system has
a maximum power point tracker (MPPT) to extract the PV array maximum
power point for a particular irradiance and temperature and reduce the
Keywords: mismatch that causes an imbalance in the power sent from the inverter to the
main grid. The fuzzy logic controller is used to tune the proportional-integral
Five levels H-bridge inverter (PI) controller to regulate the current and voltage of the grid-connected
Photovoltaic grid-connected inverter by changing the gain of the PI controller (Kp, Ki) to obtain a fast
PI-fuzzy controller response and improve the power quality of the system despite different load
Power quality disturbances and inputs. The system was simulated in MATLAB/Simulink,
THD and the results show the superiority of the proposed control unit, in which a
pure and stable sine output voltage and current waveforms. Finally, the total
harmonic distortion (THD) is improved to reach 3.81% based on the fuzzy
PI controller, while 7.77% is based on the PI controller.
This is an open access article under the CC BY-SA license.

Corresponding Author:
Marwan Saded Ahmed
Department of Electrical Engineering, University of Technology, Baghdad, Iraq
Email: [email protected]

1. INTRODUCTION
Many countries nowadays rely heavily on wind energy and solar panel energy as a source of
renewable energy (RES), especially with the spread and development of technology, being more reliable and
sustainable to form a microgrid integrated with the traditional grid and solve the problem of increasing
energy demand and reducing dependence on fuel Fossil and environmental conservation [1].
Energy management and the conversion of renewable energy generated (RES) into electricity and its
storage, requires the use of electronic devices as power converter systems and must be ideal in configuration
and size and meet the requirements for optimum operation of microgrid in terms of high efficiency,
performance, reliability, stability, low cost, protection, and effective energy regulation [2], [3]. One of the
most important electronic power converters, which is an inverter. An inverter can be defined as an electronic
device that converts DC power into AC power while maintaining the output voltage and frequency. Inverters
are divided into two types: the first type is a two-level inverter, and the second is a multilevel inverter with
three levels or more [3].
Because of the high switching losses, high switching frequency, harmonic distortion, and other
problems in the two-level inverter as it cannot be used in high and medium voltage transformations, the
engineers of electronic power and modern technology have tended to pay attention to and use the multilevel
inverter because of its many advantages and high capacity to deal with medium and high energy with its

Journal homepage: http://ijape.iaescore.com

Int J Appl Power Eng ISSN: 2252-8792  121

positive contribution in the field of renewable energy and electronic capacity. Multilevel inverters are
classified into three classes [3], [4]: i) flying capacitor multilevel inverter, ii) cascaded H-bridge multilevel
inverter, and iii) diode clamped multilevel inverter.
This paper presents an inverter topology of five levels H-bridge cascaded connected to the utility
grid with reliance on photovoltaic panels as a DC source. one of the most important problems facing this
system is the mismatch of the photovoltaic panel, as each photovoltaic (PV) module gives its power
(maximum power point (MPP)) due to the change of irradiance and temperature or consumption of the
photoelectric board. So, the powers supplied to the inverter are unbalanced, resulting imbalance in the
injection of grid current to achieve equilibrium; the modulation must be compensated, so a control scheme is
presented for each PV module in the system. Therefore, any module can be dispensed with when shading or
any defect occurs in the module on the other hand, the traditional control technique deals with linear systems
and is affected by disturbances [5]. Therefore, fuzzy logic control (FLC) techniques that are non-linear have
been proposed approach because it includes three main sources from nonlinearity, i.e. base rule, inference
engine and jamming units this fuzzy logic-based control model is proven to be a successful approach to the
control of many non-linear it has been suggested as an alternative approach for conventional control
techniques (PI) and compare the results between the two methods [6]. Many studies have relied on the
method of the fuzzy logic controller for use with the photovoltaic system as [6]–[9]. The fuzzy logic control
algorithm (PI-fuzzy) is used to control the current and voltage of the grid-connected inverter by changing the
gain (Ki, Kp) of the PI controller, improving the efficiency of the PV system [7]–[10].

2. DESCRIPTION OF THE SYSTEM

The cascaded H-bridge multilevel inverter connected to a three-phase grid, using solar panels as a
DC voltage source, the system connected to the utility grid via an L filter to filter unwanted signals as well as
transformers for isolation and protection. Figure 1 shows the components of the system.

Figure 1. Components of a proposed PV system

2.1. PV array
The solar cell is the basic unit for forming PV panels. A solar cell can be defined as a semiconductor
material that converts solar radiation into electrical energy, the solar cell consists of a thin layer of a
semiconductor material that is fabricated on a PN junction such as silicon, cadmium telluride, gallium
arsenide, etc, its operational properties are similar to that of a diode p-n as it relies on solar radiation and
surface temperature to generate electrical current. A photovoltaic cell can generate almost voltage 0.5 to 0.8
volt depending on the advanced technology and type of semiconductor material used [11].
The voltage generated by a single solar cell is insufficient for the applications, so the solar cells are
connected in one frame in series and parallel to their component called the solar panel. The equivalent circuit
of a solar cell can consist of a single or double diode, and a single diode is more efficient and easier to
analyse. Figure 2 shows the equivalent circuit of a solar cell with a single diode [10], [11] .The relationship
between output voltage (VPV) and output current (IPV) in the equivalent circuit of a single diode can be
analysed mathematically as (1) [12].

𝐼𝑃𝑉 = 𝐼𝑝ℎ – 𝐼𝐷 – 𝐼𝑠ℎ (1)

Where:
𝐼𝑃𝑉 : Current of PV cell (A)
𝐼𝑝ℎ : Photo-current (A); generated by the photoelectric effect
𝐼𝐷 : Diode current (A)
Power quality improvement of grid-connected photovoltaic … (Marwan Saded Ahmed)
122  ISSN: 2252-8792

Figure 2. Equivalent circuit of a single diode solar cell

The (2) shows the dependence of the voltage and current of the solar cell on radiation level,
temperature, and their changes.

𝐼𝑃ℎ = 𝐼𝑆𝐶 + 𝐾𝑖 (𝑇– 𝑇𝑟 )(𝐺/𝐺𝑛𝑜𝑚 ) (2)

Where:
𝐼𝑆𝐶 : short circuit current (A)
𝐾𝑖 : short circuit current of a cell at (STC) standard test condition 25 °C and 1000 W/m 2.
𝑇: operating temperature (K)
𝑇𝑟 : Reference temperature of the PV cell (298.15 K)
G: Solar radiation in W/m2
𝐺𝑛𝑜𝑚 : Nominal solar radiation 1000 W/m2
𝑇 1 1
𝐼o = 𝐼rs [( )]3 𝑒𝑥𝑝(𝑞 × 𝐸𝑔𝑜 )/(𝐴𝐾)( ) − ( ) (3)
𝑇𝑟 𝑇𝑟 T

𝐼rs = 𝐼scr /𝑒𝑥𝑝(𝑞𝑉𝑜𝑐 )/(𝑁𝑠 𝐾𝐴𝑇) − 1 (4)

Where:
𝐼o : Reverse saturation current
𝐼rs : Cell reverse saturation current at Tr.
𝑞: The charge of the electron is equal to 1.6×10^ (-19) coulomb
𝐸𝑔𝑜 : bandgap energy of the semiconductor = 1.1 e V
𝐴: The ideality factor of the diode
𝐾: Boltzmann’s constant equal 1.3805×10^ (-23) J/k
𝑉𝑜𝑐 : Open circuit voltage
𝑁𝑠 : number of cells connected in series
Now, the output current of the PV cell is in the (5) and (6).

𝐼PV = 𝐼Ph − 𝐼o [𝑒𝑥𝑝(𝑞 × 𝑉𝑃𝑉 + 𝐼𝑃𝑉 × 𝑅𝑆 )/(𝑁𝑠 𝐴𝐾𝑇) − 1] + 𝐼sh (5)

and

𝐼sh = (𝑉 + 𝐼PV 𝑅𝑆 )/𝑅𝑠ℎ (6)

Where:
𝑅𝑆 : series resistance (Ω)
𝑅𝑠ℎ : shunt resistance (Ω)
𝑉𝑃𝑉 : Output voltage from PV panel
The energy that is practically used to power the loads [8]. The (7) represents the relationship
between I-V in the PV array.
𝑁 𝑁
( 𝑉𝑃𝑉+𝑁 𝑠 𝐼𝑃𝑉 𝑉𝑆 𝑉 𝑁 𝑠 +𝐼𝑃𝑉 𝑅𝑠
𝑝 𝑝
𝐼𝑃𝑉 = 𝑁P × 𝐼Ph − 𝑁P × 𝐼𝑜 {𝑒𝑥𝑝 [ ]} − 1 + (7)
𝑁𝑠 𝐴𝐾𝑇 𝑅𝑠ℎ

Where:
𝑁P : number of PV modules connected in parallel
𝑁𝑠 : number of PV modules connected in series

Int J Appl Power Eng, Vol. 11, No. 2, June 2022: 120-133
Int J Appl Power Eng ISSN: 2252-8792  123

2.2. Boost converter

The boost converter is called a step-up converter because the output voltage is greater than the input
voltage, Figure 3 shows the structure of the boost converter which consists of two semiconductors (diode and
MOSFET), and the inductor which is the power storage source in addition to the capacitance such as the
filter, load impedance and DC voltage source (PV array or battery).

Figure 3. Circuit diagrams of the boost converter

The MOSFET is used as a trigger switch in the circuit, and the boost switch operates in two modes.
The first is when the switch is closed in the ON position, the current is stored in the inductor for a certain
period and the diode prevents the current from flowing. With the load current remaining constant due to
capacitor discharge. In the second position, when the switch is opened, current flows through the inductor
and diode, and the capacitor is charged and from there to the load while the load current also remains
constant [13]. Using Faraday's law, the relationship between the output and input voltage of the boost
converter can be found as shown in (8) [14].

𝑉𝑠. 𝐷𝑇 = (𝑉𝑜 − 𝑉𝑠 )(1 − 𝐷)𝑇 (8)

Where:
𝑉𝑠: input voltage
𝑉𝑜: output voltage of DC-DC boost converter
𝐷: Duty cycle

2.3. Incremental conductance algorithm (INC)

Figure 4 represents a flowchart of the INC algorithm, features of this algorithm include the speed of
processing steady-state errors, rapid response to variable solar radiation, as well as the speed of convergence,
as it combines accuracy and speed. The maximum power point in the PV matrix is traced from the derivation
of the relationship P*V concerning voltage or current. MPP can be calculated using the dp/dv and -I/v
relationship and as shown by the following derivation:

𝑃 =𝑉×𝐼 (9)
dP dI
= I + V. (10)
dv dv

or
𝑑𝑃 𝑑𝑣
= 𝑉 + 𝐼. (11)
𝑑𝐼 𝑑𝐼

From the calculation to depend on the values I and V, the MPP values can be inferred from the
following expressions:
∆𝑰 −𝐈
> The operating point is to the left of MPP
∆𝑽 𝐕
∆𝐈 −𝐈
= The operating point is exactly at MPP
∆𝐕 𝐕
∆𝐈 −𝐈
< The operating point is to the right of MPP
∆𝐕 𝐕
According to these comparisons, the algorithm increases or decreases the voltage in the PV array [15], [16].

Power quality improvement of grid-connected photovoltaic … (Marwan Saded Ahmed)

124  ISSN: 2252-8792

Figure 4. Flowchart of INC algorithm

2.4. Cascaded H-bridge inverter

Cascaded H-bridge are several single-phase inverters or called H-bridge inverters connected in
series, each bridge consists of an independent DC source in addition to switches, as shown in Figure 5. This
inverter generates a sinusoidal output voltage as each H-bridge provides three levels (zero, positive output
voltage, negative output voltage). The total output voltage is the sum of all the voltages produced by the
connected H-bridge cells in the chain. Where if the number of H-bridge cells is M the total output voltage is
equal to (2M+1). High power quality reduced harmonic distortion, low switching losses, high-quality output
wave are among the main advantages of this inverter [4], [10].

Figure 5. Topology of the modular cascaded H-bridge multilevel inverter for a three-phase grid-connected
PV system

Int J Appl Power Eng, Vol. 11, No. 2, June 2022: 120-133
Int J Appl Power Eng ISSN: 2252-8792  125

3. CONTROL SCHEME
One of the most important problems facing the PV system is the mismatch of the PV panels due to
the uneven radiation and temperature changes in addition to the lifetime of the solar panels, which means that
the MPP is different for each PV model. This reduces the efficiency of the photovoltaic system and the
occurrence of problems in the three-phase system at each stage due to the lack of current balance in the grid,
so there must be a maximum power point tracker (MPPT) for each PV array independently, control and
synchronization between the grid and the system [5], [17], [18].
The control process consists of two closed loops, an inner control loop that regulates the current and
an external control loop that controls the DC voltage regulation, maintains a constant DC link voltage, and
produces signals to the inner control loop as a reference for the current. Changes in conditions affecting the
PV array (radiation and temperature) result in a difference in the power produced from it which reaches the
utility grid via the DC link. Therefore, the power produced does not match the power that reaches the utility
grid, hence the voltage control unit provides a match between the power produced from the PV array and the
power delivered to the utility grid by changing the active power of the reference current [19], [20].
The inner control unit (current) can be designed with different frames of reference which are
synchronous (dq), fixed frames (αβ), and natural frames (abc). It should be capable of harmonic
compensation, grid synchronization, active and reactive power control with a sfast-dynamic response. Since
the inverter used is three-phase multilevel, a combination of variable AC signals is used, namely, grid current
(Iabc) and grid voltage (Vabc), as well as converting these variable quantities from natural frame (abc) to
synchronous frame (dq). Therefore, Clark or Park transformations are required to be used for this purpose,
depending on the phase-locked loop (PLL) to estimate the synchronization angle.
The control unit (dq) independently controls the active and reactive power by component current (d)
for active power and component current (q) for reactive power, to get the unit power factor, component (q) is
controlled and made equal to zero and thus the control process becomes easier after the disengagement between
the two components (P and Q) and relying on PI-fuzzy controller for its advantages to make the system more
efficient, Where the proposed PI-fuzzy controller is used in the external control loop to stabilize the incoming
voltage of the inverter and optimize the wave coming out from the inverter so that it is sine and pure, and is also
used in the inner loop of current to improve power quality and reduce THD [7], [19], [21]. The scheme shown
in Figure 6 represents the MPPT single control and control circuit for the proposed system.

Figure 6. Control scheme for three-phase five-level cascaded H-bridge multilevel inverter

Park transformation (13) and (14) are applied to convert voltage and current from nature frames into
a synchronous reference frame (dq) [22].
2π 2π
sin(wt) sin (wt − ) sin (wt + Va)
Vd 2 3 3
[ ]=√ [ 2π ] [Vb]
2π (13)
Vq 3 cos(wt) cos (wt − ) cos (wt + )
3 3 Vc

2𝜋 2𝜋
𝑠𝑖𝑛(𝑤𝑡) 𝑠𝑖𝑛 (𝑤𝑡 − ) 𝑠𝑖𝑛 (𝑤𝑡 + ) 𝐼𝑎
𝐼𝑑 2 3 3
[ ]=√ [ 2𝜋 ] [𝐼𝑏 ]
2𝜋 (14)
𝐼𝑞 3 𝑐𝑜𝑠 (𝑤𝑡) 𝑐𝑜𝑠 (𝑤𝑡 − ) 𝑠𝑖𝑛 (𝑤𝑡 + )
3 3 𝐼𝑐

Power quality improvement of grid-connected photovoltaic … (Marwan Saded Ahmed)

126  ISSN: 2252-8792

4. FUZZY LOGIC CONTROLLER

The non-linear behavior of most real systems and the difficulty in modelling them is one of the most
important problems facing these systems, as classical controllers (PID, PI, P) which are feedback controllers
deal with fixed parameters, so they are not adapt with the environment variables and changes in the
parameter system, as well as the setting of parameters for the controller is difficult since an accurate
mathematical model is usually hard to obtain to adjust these parameters and solve these problems, So it is
necessary suitable control unit that adapts to various change. Fuzzy logic control is an intelligent control
method that uses fuzzy rule sets and linguistic representations of a human's knowledge to control a plant. It is
an important branch of intelligent control. Positive control effects have been obtained using fuzzy logic
technology in a variety of control fields. The essence of FLC is converting expert knowledge of related fields
and the experience of skilled operators into language rules, and then using fuzzy reasoning and fuzzy
decision-making to implement complex system control. The proposed controller (fuzzy-PI) is applied to the
outer control loop of the dc link to reduce the ripples of the dc-link, as well as to the two inner double current
loops (Id and Iq); i) The design process for an FLC entails determining the inputs, ii) establishing the rules,
and iii) devising a method for demulsifying the fuzzy result of the rules into an output signal, known as
defuzzification. Figure 7 shows the Structure of the fuzzy PI controller [8], [9].

4.1. Design fuzzy-PI

4.1.1. Determine the input and output
The deviation e and the change of error are the FLC's inputs. The FLC constantly checks and
calculates the inputs, and then the PI controller's parameters are adjusted online using fuzzy logic rules to
achieve optimal parameters. Error e and error change rate can satisfy the requirements of PI parameter self-
tuning at different times as FLC inputs in the fuzzy adaptive PI controller. Each of the input signals has seven
membership functions, while each of the output signals has seven. From Figure 7 a block diagram of the
proposed FLC algorithm with the input and output changes, wherein the input signals are the error current in
the d-q frame and the derivation of the regulated voltage is as follows:

𝑒 𝐼𝑑 = 𝐺𝑖 (𝑖𝑑− 𝑟𝑒𝑓 − 𝑖𝑑 ) (15)

𝑒 𝐼𝑞 = 𝐺𝑖(𝑖𝑞𝑟𝑒𝑓 − 𝑖𝑞) (16)

e 𝑉d = 𝐺𝑣 (𝑉𝑑_ 𝑟𝑒𝑓 − 𝑉𝑑) (17)

Where:
𝐺𝑣 and Gi: gain factors to unify the input signal per-unit (pu)
𝑉d 𝑉q & Id Id: Voltage and current variables in rotating reference frame
Vd_ref & 𝑉𝑞_𝑟𝑒𝑓: Voltages reference at d-axis and q – axis

Figure 7. Structure of the PI-fuzzy controller

Figure 8(a) and 8(b) (see Appendix) depicts the design of the input membership functions of
variables and the proportional gain Kp and the integral gain Ki membership functions are designed as shown
in Figure 9(a) and 9(b) (see Appendix). Simple normalization calculations yield the fuzzy set domain ranges

Int J Appl Power Eng, Vol. 11, No. 2, June 2022: 120-133
Int J Appl Power Eng ISSN: 2252-8792  127

of input and output variables. The input variables' fuzzy set domain is set to e, 𝛥𝑒 = {−1, 0, 1} and the fuzzy
set domain of outputs 𝛥𝐾𝑃 , 𝛥𝐾𝑖 are set as (-1, 1) and (-1, 1) respectively. The input and output variables for
the corresponding fuzzy sets are set as follows: e, 𝛥𝑒, 𝛥𝐾𝑃 , 𝛥𝐾𝑖 = {𝑁𝐵, 𝑁𝑀, 𝑁𝑆, 𝑍𝐸, 𝑃𝑆, 𝑃𝑀, 𝑃𝐵}. The fuzzy
sets' means are referred to as negative big (NB), negative medium (NM), negative small (NS), zero (ZE),
positive small (PS), positive medium (PM), and positive big (PB).

4.1.2. Fuzzification and fuzzy reasoning

The normal distribution applies to all fuzzy variables. The Mamdani type is used in the fuzzy
inference system, as follows: If e is A and 𝛥𝑒 is B, then u is c. Tables 1 and 2 show the fuzzy logic rules for
the output variables, respectively.

Table 1. Rule-base fuzzy of 𝛥𝐾𝑖 parameter Table 2. Rule-base fuzzy of ΔKp parameter
𝛥𝐾𝑖 𝛥e 𝛥𝐾𝑃 𝛥e
NB NM NS ZE PS PM PB NB NM NS ZE PS PM PB
e NB NB NB NM NM NS ZE ZE e NB PB PB PM PM PS ZE ZE
NM NB NB NM NS NS ZE ZE NM PB PB PM PS PS ZE NS
NS NB PM NS NS ZE PS PS NS PM PM PM PS ZE NS NS
ZE NM NM NS ZE PS PM PM ZE PM PM PS ZE NS NM NM
PS NM NM ZE PS PS PM PB PS PS PS ZE NS NS NM NM
PM ZE ZE PS PS PM PB PB PM PS ZE NS NM NM NM NB
PB ZE ZE PS PM PM NB PB PB ZE ZE NM NM NM NB NB

4.1.3. Defuzzification
Defuzzification entails devising a method for converting the rules' fuzzy results into output signals.
To obtain accurate FLC output, the weighted average method is used. The following is a description of the
weighted average method:
∑𝑛
𝑖=1 𝑢𝑖𝜇𝑐(𝑢𝑖)
𝑢0 = ∑𝑛
(19)
𝑖=1 𝜇𝑐(𝑢𝑖)

Online self-tuning of PI controller parameters the proposed self-tuning of PI parameters aims to uncover the
hazy relationship between two PI parameters and FLC inputs. The DC-link voltage Vdc is sensed and
compared to the reference DC-link voltage 𝑉𝑑𝑐𝑟𝑒𝑓 to implement the FLC algorithm of the grid-connected in a
closed loop. The FLC continuously checks and calculates the deviations e and 𝛥𝑒 to meet the control
parameters of different requirements for different e and 𝛥𝑒. The PI controller's parameters are then adjusted
online using fuzzy rules to achieve an excellent dynamic and static performance for the PV system. The self-
tuning of parameters is based on the processing of fuzzification results, fuzzy reasoning, and defuzzification.
The PI controller's parameter regulation formula is used to achieve the goal of online regulation.

𝐾𝑝_𝑝𝑢𝑛𝑒𝑤 = 𝐾𝑝 + 𝛥𝐾𝑃 (20)

𝐾𝑖_ 𝑝𝑢𝑛𝑒𝑤 = 𝐾𝑖 + 𝛥𝐾𝑖 (21)

5. SIMULATION RESULTS AND DISCUSSION

The cascade multilevel grid-connected inverter simulation was done using the MATLAB/Simulink
platform. Via an inductor filter and interface transformer, the inverter is coupled to the grid. Table 3 presents
the parameters of the PV array. For both irradiation and temperature fluctuations MPP set-point monitoring,
detailed simulation investigations were conducted to objectively analyze the performances. For profiles
where differences in MPP set-point monitoring are such that the full range of irradiation and temperature
values are used, a good controller is needed. In this section, the simulation results are carried under shading
conditions with (500 W/m2, 1000 W/m2) irradiation and (25 °C, 32 °C) temperature. The PV system is
comprising of six groups and each group has 54 panels, two panels in series, and 28 panels in parallel, and
the output voltage of one PV array is 109.4 volts, as well as the output current of the PV group, is 156.24 A.
The photovoltaic power output is approximately 17 Kw. The input DC voltage for a single-phase of inverter
which is including two H-bridge is 500 V.
There are two periodic in the profile used for this test case. For the first periodic (t=0-1.2 s), the
irradiation is reduced to 500 W/m2 and the temperature is maintained at STC (25 °C). while in the second
periodic (t=1.2-2s), the irradiation level returns to the maximum irradiation 1000 W/m 2 and the temperature is
increased to 32 °C, as shown in Figures 10-16.

Power quality improvement of grid-connected photovoltaic … (Marwan Saded Ahmed)

128  ISSN: 2252-8792

Table 3. Electrical parameters of the PV panel

Parameters Values
Maximum power (Pmax) 305 w
Rated Voltage (Vmp) 54.7 V
Rated Current (Imp) 5.58 A
Open Circuit Voltage (Voc) 64.2 V
Short Circuit Current (Isc) 5.96 A
Number of cells in the panel 96

The output voltage of five levels injected into the main grid by the proposed converter during
controlled operation based on the fuzzy PI and PI controller are shown in Figures 10(a) and 10(b).
The proposed controller based on fuzzy with CHB has a five-level output voltage waveform that is very
stable and has no overshoot as compared with the PI controller which is depicted in Figure 10(b). While the
output voltage of the proposed system after the L filter based on the fuzzy PI and PI controller is shown in
Figures 11(a) and 11(b). It is obvious from the results in this figure, the output voltage is having high stability
and it has a pure sine wave without a ripple at the steady-state region. In Figure 12, the output of AC currents
of fuzzy PI and PI controller is showing. It is obvious from the waveforms in this figure, the waveform of
three-phase AC currents based on the fuzzy PI is a pure sine wave and no overshoot with a high steady-state
as compared with the PI controller which has high ripples and lower stability at the steady-state region as
shown in Figure 12(b).

(a) (b)

Figure 10. Waveforms of output voltage of the inverter, (a) output voltage of five-level CHB based on PI
controller, (b) output voltage of five-level CHB based on fuzzy PI

(a)

(b)

Figure 11. Waveforms of AC load side after L filter: (a) output voltage of AC bus based on PI controller with
zoom and (b) output voltage of AC bus based on the fuzzy PI controller with zoom

Int J Appl Power Eng, Vol. 11, No. 2, June 2022: 120-133
Int J Appl Power Eng ISSN: 2252-8792  129

(a)

(b)

Figure 12. Waveforms of AC load side after L filter: (a) output current of AC bus based on PI controller with
zoom and (b) output current of AC bus based on the fuzzy PI controller with zoom

While the output power of PV groups and output power of AC side load based on the fuzzy PI
controller is very tracking to the maximum power desired and no oscillation at the steady-state region as
compared with PI controller at (500 W/m2, 25 °C), as well as in (1000 W/m2, 32 °C) as shown in
Figures 13(a) and 13(b). The DC link voltage waveform has been depicted in Figure 14. It is obvious from
this figure; the voltage remains constant which implies the system is very stable and rejected the
disturbances.

(a) (b)

Figure 13. Responses of output power: (a) output power of PV array for one group and (b) output power of
AC side

Figure 14. Response of DC voltage after boost converter

Power quality improvement of grid-connected photovoltaic … (Marwan Saded Ahmed)

130  ISSN: 2252-8792

The values of the PI controller of the five-level inverter based on the fuzzy system for DC voltage
controller and current controller have been depicted in Figures 15(a) and 15(b). Finally, to show the
superiority of the proposed control based on a fuzzy system in improved the power quality of the proposed
system, the harmonics analysis based on fast Fourier analysis for THD is depicted in Figures 16(a) and 16(b).
It can be seen in this figure, the percentage THD for AC current is 3.81% while the THD is equal to 7.77%
which proved the superiority of the proposed PI controller in enhancement the power quality of the overall
system and increased the reliability of the system.

(a) (b)

Figure 15. Values of PI parameters based on fuzzy system: (a) values of PI parameters of DC voltage
controller and (b) values of PI parameters of the current controller

(a)

(b)

Figure 16. THD percentage for AC current: (a) THD percentage using PI controller and (b) THD percentage
using a PI-fuzzy controller

6. CONCLUSION
This paper describes a modified connection using a five-level cascade H-bridge inverter with six PV
groups system that avoids mismatch by separating each PV group with a DC-DC converter and MPPT to

Int J Appl Power Eng, Vol. 11, No. 2, June 2022: 120-133
Int J Appl Power Eng ISSN: 2252-8792  131

reach the maximum output power of six PV groups to achieve the design power of 100 Kw and PI-fuzzy
controller has been proposed to control the DC voltage controller and the Idq current controller. The results
showed that the output voltage waveform is stable and has a pure sine wave without a ripple in the steady-
state region, as well as the waveform for three-phase AC currents based on the fuzzy PI is a pure sine wave
and there is no overshoot with the high steady-state as compared to the PI controller. Finally, the total
harmonic distortion (THD) is reduced in the AC currents from 7.77% to 3.81 % in the proposed system,
which means the fuzzy PI controller has increased the reliability and stability of the system and its rejection
of disturbances, and achieved the desired expectations in improving power quality.

APPENDIX

(a)

(b)

Figure 8. Fuzzy subgroups of the input signals: (a) error and (b) change of error

Power quality improvement of grid-connected photovoltaic … (Marwan Saded Ahmed)

132  ISSN: 2252-8792

(a)

(b)

Figure 9. Fuzzy subgroups of the output signals: (a) linguistic output changes (Kp) and
(b) linguistic output changes (Ki)

REFERENCES
[1] P. Ray and M. Biswal, Eds., Microgrid: Operation, control, monitoring and protection, Lecture No. Springer, 2020. [Online].
Available: https://link.springer.com/book/10.1007/978-981-15-1781-5
[2] S. Bayhan, H. Abu-Rub, J. I. Leon, S. Vazquez, and L. G. Franquelo, “Power electronic converters and control techniques in AC
microgrids,” in IECON 2017 - 43rd Annual Conference of the IEEE Industrial Electronics Society, Oct. 2017, pp. 6179–6186.
doi: 10.1109/IECON.2017.8217073.
[3] L. Castro, S. Perez-Londono, and J. Mora-Florez, “Application of Fuzzy Logic based Control in Micro-grids,” in 2019 FISE-IEEE/CIGRE
Conference - Living the energy Transition (FISE/CIGRE), Dec. 2019, pp. 1–8. doi: 10.1109/FISECIGRE48012.2019.8985002.
[4] A. Krishna R and L. P. Suresh, “A brief review on multi level inverter topologies,” in 2016 International Conference on Circuit,
Power and Computing Technologies (ICCPCT), Mar. 2016, pp. 1–6. doi: 10.1109/ICCPCT.2016.7530373.
[5] I. H. Shanono, N. R. H. Abdullah, and A. Muhammad, “A survey of multilevel voltage source inverter tpologies, controls, and
applications,” Int. J. Power Electron. Drive Syst., vol. 9, no. 3, p. 1186, Sep. 2018, doi: 10.11591/ijpeds.v9.i3.pp1186-1201.
[6] D. Tamilarasi and T. S. Sivakumaran, “Fuzzy PI control of symmetrical and asymmetrical multilevel current source inverter,” Int.
J. Fuzzy Syst., vol. 20, no. 2, pp. 426–443, Feb. 2018, doi: 10.1007/s40815-017-0352-8.
[7] N.-T. Nguyen and H.-H. Lee, “Fuzzy PI Controller for Grid-Connected Inverters,” in Lecture Notes in Computer Science,
Springer, 2012, pp. 300–308. doi: 10.1007/978-3-642-25944-9_39.
[8] O. Z. Bakhoda, M. B. Menhaj, and G. B. Gharehpetian, “Fuzzy logic controller vs. PI controller for MPPT of three-phase grid-
connected PV system considering different irradiation conditions,” J. Intell. Fuzzy Syst., vol. 30, no. 3, pp. 1353–1366, Mar. 2016,
doi: 10.3233/IFS-152049.
[9] A. Kaysal and R. Bayindir, “Design and Analysis of Fuzzy Logic Controllers for Microgrid Voltage Control,” in 2018 2nd
International Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT), Oct. 2018, pp. 1–6. doi:

Int J Appl Power Eng, Vol. 11, No. 2, June 2022: 120-133
Int J Appl Power Eng ISSN: 2252-8792  133

10.1109/ISMSIT.2018.8567258.
[10] S. S. Katkamwar and V. R. Doifode, “Cascaded H-bridge multilevel PV inverter with MPPT for grid connected application,” in
2016 International Conference on Energy Efficient Technologies for Sustainability (ICEETS), Apr. 2016, pp. 641–646. doi:
10.1109/ICEETS.2016.7583832.
[11] T. S. Kishore, S. D. Kaushik, and Y. Venu Madhavi, “Modelling, Simulation and Analysis of PI and FL Controlled Microgrid
System,” in 2019 IEEE International Conference on Electrical, Computer and Communication Technologies (ICECCT), Feb.
2019, pp. 1–8. doi: 10.1109/ICECCT.2019.8869379.
[12] X. H. Nguyen and M. P. Nguyen, “Mathematical modeling of photovoltaic cell/module/arrays with tags in Matlab/Simulink,”
Environ. Syst. Res., vol. 4, no. 1, p. 24, Dec. 2015, doi: 10.1186/s40068-015-0047-9.
[13] N. Ramalingam et al., “Implementation of PI controller for boost converter in PV system,” Int. J. Adv. Res. Manag. Archit.
Technol. Eng., vol. II, no. April, pp. 6–10, 2016.
[14] M. H. Rashid, Power Electronics Handbook. Butterworth-Heinemann, Elsevier, 2017.
[15] F. L. Tofoli, D. de Castro Pereira, and W. J. de Paula, “Comparative study of maximum power point tracking techniques for
photovoltaic systems,” Int. J. Photoenergy, vol. 2015, pp. 1–10, 2015, doi: 10.1155/2015/812582.
[16] R. I. Putri, S. Wibowo, and M. Rifa’i, “Maximum power point tracking for photovoltaic using incremental conductance method,”
Energy Procedia, vol. 68, pp. 22–30, Apr. 2015, doi: 10.1016/j.egypro.2015.03.228.
[17] A. Lashab et al., “Cascaded multilevel PV inverter with improved harmonic performance during power imbalance between power
cells,” IEEE Trans. Ind. Appl., vol. 56, no. 3, pp. 2788–2798, May 2020, doi: 10.1109/TIA.2020.2978164.
[18] M. M. Paul Raj and S. S. Meenakshi Sundaram, “Cascaded H-bridge five-level inverter for grid-connected photovoltaic system
Using proportional-integral controller,” Meas. Control (United Kingdom), vol. 49, no. 1, pp. 33–41, 2016, doi:
10.1177/0020294016629175.
[19] N. M. Mohsin, “Design of a Grid Connected Photovoltaic Power Electronic Converter,” 2017. [Online]. Available:
http://www.bioone.org/doi/abs/10.3159/10-RA-011.1
[20] D. Y. Mahmood, A. H. Numan, and J. K. Ateih, “Simple and efficient control method for battery charging in high penetration
photovoltaic array,” no. May, pp. 1813–1819, 2019.
[21] Rajan J. Devi and Supriya S. Kadam, “Synchronization of three phase inverter with electrical grid,” Int. J. Eng. Res., vol. V4, no.
05, May 2015, doi: 10.17577/IJERTV4IS050938.
[22] G. J. May, A. Davidson, and B. Monahov, “Lead batteries for utility energy storage: A review,” J. Energy Storage, vol. 15, pp.
145–157, Feb. 2018, doi: 10.1016/j.est.2017.11.008.

BIOGRAPHIES OF AUTHORS

Marwan Saded Ahmed is an M.Sc. degree student in the research level at

Electrical Engineering Department, University of Technology, Baghdad, Iraq, and complete a
B.Sc. degree in 2009 in the Electrical Engineering Department, University of Dyala, Dyala, Iraq.
Mr. Marwan is interested in Power Electronic Converters, PV Systems, Microgrid systems, and
Storage Energy Systems. He can be contacted at email: [email protected].

Dhari Yousif Mahmood is a. Prof at the Department of Electrical Engineering at

the University of Technology, Baghdad, Iraq. He was awarded his M.Sc and Ph.D. in
Electrical Power Engineering from Sant.Peterburg Polytechnical Institute-Russa in 1986 and
1990 respectively. His BSc degree was received in 1981 from the University of Baghdad-
Collage of Engineering-Iraq. He works in the field of renewable energy and power system
analysis. He supervises a large number of postgraduate students in both masters and Ph.D.
degrees. He can be contacted at email: [email protected].

Ali Hussein Numan received the BSc, M.Sc, and Ph.D. in Electrical Engineering
from the University of Technology, in 1999, 2003, and 2009, respectively. Currently, he is an
Associate Professor of Electrical Engineering in the Electromechanical Engineering
Department of the University of Technology (UoT). He has co-authored the textbook Small
and Special Electric Motor and Their Control Technique, 2016, and published 37 journals and
conferences papers. His research interests include power electronics, variable speed drives,
renewable energy, and modeling and simulation. He can be contacted at email:
[email protected].

Power quality improvement of grid-connected photovoltaic … (Marwan Saded Ahmed)