1008 CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 9, NO.

3, MAY 2023

Improved MPPT Control Strategy for PV Connected

to Grid Using IncCond-PSO-MPC Approach
Nora Kacimi, Abdelhakim Idir , Said Grouni, and Mohamed Seghir Boucherit

Abstract—This paper proposes a new hybrid maximum power Different topologies and control methodologies can be used
point tracking (MPPT) control strategy for grid-connected solar to implement these systems. However, they can be classified
systems based on Incremental conductance—Particle Swarm into two types: stand-alone systems and grid-connected sys-
Optimization and Model Predictive Controller (IncCond-PSO-
MPC). The purpose of the suggested method is to create as tems [4]. Standalone application produces very little electrical
much power as feasible from a PV system during environmental energy, it is only appropriate for low-power applications. PV
changes, then transfer it to the power grid. To accomplish this, energy sources must be connected to the grid system in order
a hybrid combination of incremental conductance (IncCond) to be used at medium and high-power levels, and a DC-
and particle swarm optimization (PSO) is proposed to locate AC inverter is typically used to regulate the DC-link voltage
maximum power, followed by model predictive control (MPC)
to track maximum power and control the boost converter to and manage active and reactive power flow between the PV
achieve high performance regardless of parameter variations. generator and the grid [5]. Furthermore, in order to be fully
A two-level inverter, likewise, controlled by Model Predictive operational, PV systems must deliver the maximum amount
Control, is employed to inject the PV power generated. In this of energy. However, because PV systems have nonlinear (P–
application, the MPC is based on minimizing the difference V) characteristics, a maximum power point tracking technique
between the reference and prediction powers, which is computed
to select the switching state of the inverter. The proposed system is utilized to extract maximum output power regardless of
is simulated and evaluated in a variety of dynamic conditions environmental changes.
using Matlab/Simulink. Results reveal that the proposed control Several MPPT techniques have been proposed in literature
mechanism is effective at tracking the maximum power point for better extraction of maximum power [6]–[14]. Under
(MPP) with fewer power oscillations. uniform conditions, most MPPT techniques perform well.
However, literature contains a multitude of other state-of-the-
Index Terms—DC/DC boost converter, grid-connected PV,
incremental conductance particle swarm optimization and model art partial shading MPPT methods. In [15], the authors pro-
predictive controller (IncCond-PSO-MPC), maximum power vide a comprehensive overview of 62 global MPPT methods.
point tracking (MPPT), photovoltaic (PV). The memetic salp swarm algorithm-based MPPT control for
partial shading conditions is discussed in [16], which is an
extension of the salp swarm algorithm. The authors of [17]
described a partial shading MPPT approach based on collective
I. I NTRODUCTION intelligence that employs five sub-optimizers. The authors
in [18] proposed a modified Fibonacci search technique for

E NERGY has become a critical factor in economic and

social development in our society over the last few
decades, and solar energy has gotten a lot of attention as a way
tracking the GMPP. In contrast, the authors in [19] discussed
a systematic modification of the P & O algorithm for tracing
the GMPP. The strategy based on the merging of P&O with a
to overcome increased energy consumption and environmental genetic algorithm is described in [20].
problems [1]. It is one of the most prominent sources for The Particle Swarm Optimization (PSO) method is the most
power generation. A solar photovoltaic module converts solar commonly used. The PSO method’s main drawbacks are its
energy directly into electricity [2]. Solar photovoltaic power slow operation and large oscillations. Several modifications
generating systems are gaining popularity because they are were made to the PSO method in order to overcome large
more abundant, efficient, and ecologically benign than conven- oscillation problems. Some changes were also made to the
tional power generation systems such as fossil fuel, nuclear, reinitiating of the PSO algorithm, and others used a hybrid of
or coal [3]. the PSO algorithm.
Manuscript received November 27, 2021; revised July 31, 2022; accepted- Furthermore, tracking delays occur with optimization-based
September 19, 2022. Date of online publication December 9, 2022; date of MPPT approaches due to the ultimate operating point, which is
current version December 17, 2022. calculated by comparing all population findings. Additionally,
N. Kacimi and M. S. Boucherit are with Laboratoire de Commande de
Processus, ENP Alger, Algiers, Algeria. when Partial Shading Conditions (PSC) are not present, it is
A. Idir (corresponding author, email: [email protected]; tracking performance dips behind that of standard methods. In-
ORCID: https://orcid.org/0000-0002-3510-5932) is with the Electrical En- cremental conductance (IncCond), on the other hand, is a well-
gineering Department, University of M’sila, BP 28000, M’sila, Algeria.
S. Grouni was with Boumerdes University, Boumerdes, 35000, Boumerdes, known technique with relatively good performance. However,
Algeria. He is now with the Tamaghasset University, Algeria. it has some drawbacks, such as not guaranteeing convergence
DOI: 10.17775/CSEEJPES.2021.08810 to the true maximum power point in some conditions, such as
2096-0042 © 2021 CSEE. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
KACIMI et al.: IMPROVED MPPT CONTROL STRATEGY FOR PV CONNECTED TO GRID USING INCCOND-PSO-MPC APPROACH 1009

in (PSC), and it is relatively slow, which reduces the efficiency Our paper’s main contribution is that it proposes a novel
of photovoltaic work. hybrid IncCond-PSO-MPC method for improving the quality
To address this issue, the authors proposed a Hybrid MPPT of the PV panel and extracting maximum power. The rest of
control strategy based on the IncCond–PSO-MPC approach. this paper is divided into two sections:
Under uniform conditions, the IncCond is used with the step 1) In the first stage, a novel method is used to extract the
size selected, which significantly reduces tracking time. In greatest amount of power. This method uses a hybrid IncCond–
order to detect MPP quickly under partial shading, the PSO is PSO to determine the optimal current of a PV system. This
used where the initial population is used by the last reference is then used in the MPC cost function to select the optimal
current. To increase tracking speed, a model predictive control switching of the DC/DC converter.
(MPC) is also employed to identify optimal switching of the 2) For the second stage, a predictive control coupled with
DC/DC converter. two-level three-phase voltage source inverter is proposed to
An MPC was employed at each sampling instant to solve a validate the control method. A PI controller collects reference
finite-horizon optimal control problem and get control actions power from the DC link voltage loop, which is dedicated to
for both the present and the future [21]–[24]. A predictive maintaining the DC link voltage at the goal value, and a cost
control strategy has numerous advantages, including ease function is used to select an ideal switching state.
of implementation and high performance in terms of time To evaluate the robustness of the suggested method in a
response. In this paper, a predictive current controller is used specified setup, simulation results are obtained in a Mat-
to replace the delay control. The control methodology consists lab/Simulink environment. According to the findings, the
of two steps. The first step is to calculate the reference current proposed IncCond-PSO-MPC technique optimizes efficiency
Impp using a hybrid (IncCond-PSO) that is related to the and extracts the ideal power that follows the power of the PV
MPP working current. The second step is a developed MPC panel.
regulator for the boost converter. The remainder of this work is structured as follows: Section
A two-level voltage source inverter controls the PV grid- II includes the researched system’s synoptic schematic, as well
connected inverter (2L-VSI). In literature, various control as a mathematical model of the PV source and boost converter.
techniques for grid-connected inverters have been presented, Section III goes through the framework of the proposed
including hysteresis and linear control with pulse width mod- hybrid MPPT approach in detail. Section IV addresses the
ulation, voltage-oriented control, direct power control (DPC), performance of the suggested technique and offers simulation
and so on. The most common controller for power converters results. Section V concludes with appropriate conclusions.
is the classical proportional-integral (PI) controller. However,
due to consideration of fixed gains, they are unable to handle
dynamic operating conditions [25]. II. PV S YSTEM D ESCRIPTION AND M ODELING
Because of its high accuracy and simple approach, as well as Figure 1 depicts overall system topology. It is made up of
the ease with which system nonlinearities and constraints can a PV system, a DC/DC boost converter output, and Rg , Lg
be integrated in the cost function, the predictive control model filters to decrease the ripple components induced by switching,
MPC has shown good performance [26]. This latter component as well as a voltage source inverter (VSI).
is crucial when building an MPC since it enables the selection The system is adapting in two steps. First, the MPPT-
of control objectives, which results in higher performance controlled boost converter harvests the maximum power from
for the PV grid-connected system in terms of operational the PV array and tracks it regardless of environmental fluctu-
efficiency, improved power quality, remote operations control, ations. The DC-DC boost power is subsequently fed into the
and environmental efficiency. MPC is recommended for grid- grid by the two-level inverter. The PI controller additionally
connected inverters due to the benefits mentioned previously. maintains DC-link voltage to ensure it is at the proper level.

DC-DC Boost Converter DC-AC Converter

L Idc

K1 K2 K3
eg
Rg Lg
d1 d2 d3 ia
a
Vdc • va0 b ib
vb0 c
vc0 ic

K1 K2 K3
d1 d2 d3
PV system

Fig. 1. Topology of the overall system under consideration.

1010 CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 9, NO. 3, MAY 2023

Mathematical representations for all system components are D

L
generated and supplied in the following sub-sections for con-
trol design considerations. + +
S
A. Array Model Vpv C Vc

A photovoltaic cell is an electrical device that converts − −

solar energy directly into electricity using the photovoltaic
effect [27]. The most recent modeling technique, seen in Fig. 3. Equivalent circuits of boost converter.
Fig. 2, consists of a current source comprised of a single diode
and resistors [28], [29]. of which consists of two parallel diode/transistor assemblies.
RS IPV
There are eight alternative switching voltage vectors that can
be obtained.
ID IP The following equations represent the mathematical model
of grid currents in the natural frame (abc):
Iph D RP VPV

dia
 Lg dt = −Rg Ia − ea + Van

Lg di
dt = −Rg Ib − eb + Vbn
b
(3)
 dic

Fig. 2. The equivalent electrical circuit of one diode. Lg dt = −Rg Ic − ec + Vcn
where Lg and Rg are the output filters’ inductance and
The solar cell model is a P-N semiconductor junction, resistance, respectively, ea is the grid voltage, and Van is the
similar to a diode; when exposed to light, it generates a DC output voltage of the two-level inverter.
current. The Clarke transformation is used to transfer grid voltage
As is generally known, solar irradiance, temperature, and and phase current components from the natural (a, b, c) ref-
load current all have an effect on the current generated by a erence frame to the stationary (α, β) frame, which is defined
solar cell. The following is the relationship between output as [32]:
current (IPV ) and output voltage (VPV ) [30].  
  r  1 1  xa
 q(VPV +Rs IPV )  V +R I
PV s PV xα 2 1 − √2
−√2
IPV = Iph − I0 e Ns T Ka −1 − (1) = 3
xb  (4)
RP xβ 3 0 2 − 23 xc
where IPV is the current supplied by the cell (A); I0 is
the diode saturation current; q is the Electron charge; K is The converter’s input current dynamic can therefore be
the Boltzmann constant; T is the cell temperature in Kelvin, expressed as:
a is the PV cell’s ideality coefficient; Iph the photocurrent (
Lg didtα = −Rg Iα − eα + Vα
generated by the PV cell in the presence of a particular amount di (5)
of light (A); Rs is the cell’s series resistance (Ω); RP is the Lg dtβ = −Rg Iβ − eβ + Vβ
cell’s parallel resistance (Ω); Ns is number of cells connected
in series. III. P ROPOSED C ONTROL D ESIGN

B. DC–DC Boost Converter Model The MPPT method is used to generate the switching signal
that maintains the PV system’s operation at the MPP for the
Electronic devices that convert one level of DC electrical
boost converter. This study considers a new combination to
power to another are known as DC–DC converters [31]. The
extract maximum power. The Incremental Conduction-Particle
boost converter architecture is utilized to provide high voltage ∗
Swarm Optimization technique is used to estimate current Iref
gain to the MPPT algorithm in order for it to attain maximum
at MPP, and then the latter is used to predict switching state
power. Fig. 3 depicts the basic circuit of a boost converter.
by minimizing a cost function of predictive control as shown
The boost converter’s dynamic model can be expressed as
in Fig. 4.
follows:
(
L di
dt (t) = −(1 − s)Vc (t) + Vpv (t)
L
S:Switching
dVc (2) Vpv
Hybrid MPPT I*ref
C dt (t) = (1 − s)IL (t) + R1 Vc (t) Using
Vdc
Ipv
IncCond-PSO DC/DC
where s is the switching state, which can be taken to 1 or 0; Predictive Boost
IL is the current inductor; Vc is the voltage capacitor; Vpv is IL Controller Converter
Equations
the PV array voltage. (9)
(9) and
and (10)
(10)

C. DC-AC Voltage Source Inverter Novel control strategy

(IncCond-PSO-MPC)
In a grid-connected PV system, a three-phase two-level
power converter is used. A three-phase two-level inverter’s Fig. 4. The proposed MPPT using hybrid IncCond –PSO-MPC approach.
conventional structure consists of three arms (a, b, c), each
KACIMI et al.: IMPROVED MPPT CONTROL STRATEGY FOR PV CONNECTED TO GRID USING INCCOND-PSO-MPC APPROACH 1011

A. Design Principle for a Conventional IncCond Technique. Step 5: Select Pbest ki = xki and Gkbest = xki ;
When the PV array is subjected to rapid changes in ex- Step 6: Updates velocity and position of particles;
ternal circumstances, the incremental conductance (IncCond) Step 7: Evaluates fitness Fik+1 = f (xk+1 i ), ∀i and deter-
technique outperforms other methods by tracking the MPP mines the best particle index.
more quickly [33]. The MPP is typically calculated by taking When PSC is present, the PV system has numerous local
a derivative of the PV power output with respect to voltage peaks (LP), one of which is the global peak (GP). A change
output. The photovoltaic panel’s output current and voltage are in irradiance causes the position of the GP to shift to another
utilized to compute conductance and incremental conductance. peak location, which the IncCond MPPT may not detect. As
This method’s fundamental equations are as follows: illustrated in Fig. 5, the proposed MPPT method employs
• Operation point is at (MPP) dI
dVpv Vpv
= − ipv hybrid MPPT algorithms. Under uniform conditions, anInc-
pv
dVpv Vpv Cond approach is used, while under partial conditions, the
• Operation point is at left side of MPP dIpv < − Ipv PSO technique is used.
dVpv Vpv
• Operation point is at right side of MPP dIpv > − Ipv To shorten search time and avoid particle convergence fail-
Figure 5 depicts the incremental algorithm’s flowchart. As ure, the PSO is re-initialized with the beginning values of the
shown, the method generates a reference current Iref , which latest reference current, and MPP is identified promptly. The
increases when the slope is positive and reduces when it shading pattern (SP) change can be easily checked using (8).
is negative [34]. This reference current will be fed into
Ppv,new− Ppv,old
the predictive control block, which will be explained in the > ∆P (%) (8)
following sections. Ppv,old
Under partial shading, PV systems have multiple peaks, and A predictive control is used to optimize power and generate
IncCond can get stuck in a local peak instead of a global a switching signal for the DC/DC converter. Using a first order
peak. To overcome this problem, a PSO is proposed in this approximation for the derivative in (2), the predicted current
paper. The block diagram of the hybrid technique is depicted and voltage in discrete time are given by:
in Fig. 5.
Ts Ts
IL (k + 1) = IL (k) − (1 − S) Vc (k) + Vpv (k) (9)
B. PSO Algorithm Lf Lf
PSO is a robust searching algorithm created in 1995 [35] by Ts Ts
Vc (k + 1) = Vc (k) + (1 − S) IL (k) + Vc (k) (10)
Eberhart and Kennedy based on patterns seen in bird flocking C RC
and fish schooling. A group of unique agents known as a where Ts is sample time.
swarm of particles is used in this procedure. The algorithm By minimizing a cost function, the anticipated variables will
discovers the fitness value corresponding to each particle in the be used to find the ideal switching state.
search space of the optimization problem, and the algorithm In this research, a cost function is evaluated for each
selects the optimal solution from among those. The most conceivable switching state and will be chosen to track the
effective solution will attract the other particles. This searching current Iref given by the hybrid (IncCond–PSO) controller, as
criterion will be repeated until the particles have converged on illustrated in Fig. 5.
the global optimal solution [36]. The next movement of the
particles will then be determined by information obtained in C. DC-DC Link Voltage Regulation
the swarm. The particle position can be calculated as follows:
This controller regulates DC link voltage, which is critical
xi (k + 1) = xi (k) + vi (k + 1) i = 1, 2, · · · , N (6) for power transfer and overall system stability. The goal of
this control algorithm is to keep the value of Vdc close to the
where Vi (k) denotes the particle’s velocity, k is the number reference value.
of perturbation iterations, and N denotes the total number of
A basic PI controller is implemented to avoid variations
particles. Speed is determined by:
and instability in DC link voltage. The controller works by
Vi (k + 1) = ωVi (k) + c1 r1 (Pbest,i − xi (k)) using a proportional gain to boost dynamics and an integrator
+ c2 r2 (Gbest − xi (k)) (7) quantity via the gain to improve steady-state tracking. The
transfer function of the system must first be specified in order
where ω is inertia weight, c1 is the cognitive coefficient, c2 to identify controller gains.
is the social coefficient, Pbest,i is the best solution of particle The following is the dynamic equation for DC link voltage:
i, and Gbest is the best solution of the entire swarm (global
dVdc
best), and r1 , r2 ∈ (0, 1). The following are the PSO steps: C = −iinv (11)
Step 1: Set the PSO parameters Wmin and Wmax , as well dt
as c1 and c2 ; where iinv is the inverter current.
Step 2: Generate a population of particles with locations x The voltage transfer function for a closed loop DC connec-
and velocities V ; tion is represented as follows:
Step 3: Put iteration k = 1;
Vdc Kp /C(Ki /Kp)
Step 4: Calculate particle fitness Fik = f (xki ), ∀i and = 2 (12)
determine the best particle index Pbest ; Vdc ref S + Kp /CS + Ki /C
1012 CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 9, NO. 3, MAY 2023

Start

Input (Ipv (k), Ipv (k − 1), Vpv (k), Vpv (k − 1), Iref , VC (k), ∆Ipv )

Calculate (Ppv (k), Ppv (k − 1), ∆Vpv , ∆Ppv , ∆Ipv )

Yes Ppv (k) − Ppv (k − 1) No

> 0.05
Ppv (k − 1)

No Yes
Reinitialization: ∆Ipv (k) = 0
I1 = 0.9 × Iref
I2 = Iref
I3 = 1.3 × Iref ∆Vpv (k) Vpv (k) Yes Yes
=− ∆Vpv (k) = 0
∆Ipv (k) Ipv (k)

Iref (k) = Iref (k − 1)

P SO Algorithm Iref (k) No
No

No
∆Vpv (k) > 0
∆Vpv (k) Vpv (k) Yes
>−
∆Ipv (k) Ipv (k) Yes
Iref (k) = Iref (k − 1) + ∆Iref
No
Iref (k) = Iref (k − 1) − ∆Iref
IncCond Algorithm

Ts Ts
IL (k + 1) = IL (k) − (1 − S) Vc (k) + Vpv (k)
Lf Lf

∗
gs=0,1 = IL (k + 1) − Iref

No Yes
gS=0 > gS=1

s=1 s=0

Finite Step Model Predictive Control

Return

Fig. 5. The proposed flowchart (IncCond-PSO-MPC).

From (12), the relation between Vdc and Vdc ref is the Fig. 6. Model predictive control is based on a discrete-time
transfer function of the second order system model that predicts behavior of all voltage source
Vdc ωn2 inverter switching states. By applying Euler forward to the
= (13) equation (5), the equations (14) and (15) below are obtained:
Vdc ref S2 + 2ζωn S + ωn  
Rf Ts Ts
By equaling (12) and (13). The PI parameters are chosen iα (k + 1) = 1 − iα (k) + (vα (k) − eα (k)) (14)
as follows: Kp = 2Cζωn and Ki = Cωn2 Lf Lf
 
Rf Ts Ts
iβ (k + 1) = 1 − iβ (k) + (vβ (k) − eβ (k)) (15)
D. DC-AC Control With MPC Lf Lf
The PV array is connected to the grid via a three-phase The grid’s active and reactive power at the next sampling
converter controlled by a conventional MPC, as shown in instant can be estimated in coordinates as follows:
KACIMI et al.: IMPROVED MPPT CONTROL STRATEGY FOR PV CONNECTED TO GRID USING INCCOND-PSO-MPC APPROACH 1013

PV Module DC-DC
Converter Idc DC-AC Converter
IPV

Rg Lg

Vdc

ec eb ea
!

S1 S6
IPV VPV Control signals
ia
Vdc_ref α,β ib
MPPT Function Cost Minimization Predictive
Model a,b,c i
c
equations α,β ea
Vdc_ref Qref Pref (14) and (15)
a,b,c eb
Vdc + ec
− Iαβ (k+1) Vαβ (k+1)
Pdc_ref PPV P (k+1)
+ active and reactive power
PI Controller − "! calculation using (16)
− Q (k+1) and (17).
PLoss

Fig. 6. The block diagram control of DC/AC of PV system connected to grid.

P (k + 1) = eα (k + 1)iα (k + 1) + eβ (k + 1)iβ (k + 1) (16) IV. S IMULATION R ESULTS

Q(k + 1) = eα (k + 1)iα (k + 1) − eβ (k + 1)iβ (k + 1) (17) Figure 1 is implemented in MATLAB/Simulink to vali-
date performance of the suggested system. The PV array’s
The required quantity of active and reactive power can be
effectiveness has been proven in a variety of atmospheric
fed into the grid by selecting the optimal voltage space vector.
situations. Every PV generator has a maximum power point
A cost function that should be defined in accordance with the
that is affected by solar irradiation and temperature. To achieve
system’s control requirements.
The following is the definition of a cost function: the desired PV power, the modules can be connected in series
or in parallel to form a string, resulting in a PV array of
g(k + 1) = |Pref (k + 1) − P (k + 1)| the desired capacity. The PV array in this study is composed
+ |Qref (k + 1) − Q(k + 1)| (18) of three panels (two SM110 modules connected in series
and two connected in parallel). Table I shows the electrical
The grid reference injected active power is computed as specifications of the system.
follows: Figure 8(b) shows the (P-V) output characteristic of a solar
Pgref ≃ PPV − Pdc ref − PLOss (19) panel under uniform changing irradiance and a temperature
of 25 ◦ C. The red circles represent the maximum power
PLoss = PJ = 3Rg Ig Ig∗ (20)
point positions at the corresponding irradiance. The (P-V)
where PPV is PV array power, Pdc ref is the reference active output characteristic of a PV array under partial shadowing
power required to compensate for inaccuracy in the DC is depicted in Fig. 8(b). It can detect several peaks caused
link voltage described earlier. PLoss is the amount of power by partial shading and a single MPP caused by uniform solar
dissipated as a result of output filter resistance. irradiation.
At time k, grid voltage and current vg(k) and ig(k) are To investigate the suggested controller’s performance and
measured. Predicted variables voltage and current are then dynamic response, we proposed using a resistance as load in
calculated at time (k + 1) using equations (14) and (15). the first part and a PV system connected to the grid in the
Predicted active and reactive powers, as well as their refer- second part. The proposed MPPT is tested in a variety of
ences, are then used to evaluate the cost function as follows. weather conditions. Table II lists the parameters for the PSO.
g(k + 1) A. Performance of PV Array Under Standard Test Conditions
As shown in the flowchart in Fig. 7, the switching state that We tested the proposed algorithm under uniform irradiation
minimizes the cost function will be chosen to be applied to to validate its performance. As shown in Table III, three PVs
the voltage source inverter at the next sampling period. are subjected to 1000 W/m2 (STC Case).
1014 CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 9, NO. 3, MAY 2023

1400
Start G =1000W/m2
1200
1000
Measure the grid current and voltage Ig (k), Ig (k)

Power (W)
800 G =800W/m2

600 G =600W/m2
Initialize the cost function (g)
400
G =400W/m2
200

For m = 0 to 7 0
0 50 100 150 200 250
Voltage (V)
(a)
Calculate the grid voltage vg (k + 1) 1000
Global Maxima
977.6 W
800 702.7W Local Maxima
549W
Predict the grid current ig (k + 1) for each 386.8W

Power (W)
voltage vector (V0 to V7 ) 600

400
Calculate the active and reactive power
Pg (k + 1) and Qg (k + 1) from (16) and (17) 200

0
Calculate the cost function from (18) 0 50 100 150 200 250
Voltage(V)
(b)
No
M >7
Fig. 8. Characteristic of the PV. (a) Under different irradiance level. (b) Under
Yes Partial shading.
Minimize the cost function and
select the predicted switching states TABLE III
S HADING PATTERNS TAKEN FOR T HIS S TUDY

Cases VMPP (V) IMPP (A) GMPP (W)

Apply selected switching state s(k + 1)
STC [1000, 1000, 1000] 212.4 6.21 1320
PSC1 [900, 780, 480] 143.3 4.90 702
PSC2 [1000, 600, 300] 143.9 3.81 549

Fig. 7. Flowchart diagram of the implemented MPC.

the power curve reveals the suggested MPPT can track MPP
TABLE I
S TUDY S YSTEM PARAMETERS effectively under standard conditions. The maximum power
point has been tracked before 0.05 s with minimal power
Electrical characteristics Value oscillation.
Maximum power 110 W
Open circuit voltage 43.5 V
Figure 10 shows the proposed output power for the four
Short circuit current 3.45 A different MPPT techniques (Perturb and Observe (P&O), in-
PV output voltage at MPP 35 V cremental conductance (IncCond), hybrid IncCond-PSO and
PV output current at MPP 3.15 A
Number of cells connected in series 72
IncCond-PSO-MPC.
Electrical parameters of Boost converter Value From Fig. 10, it can be seen that all techniques track the
Inductance (L) 45 mH GMPP. Amplitude of oscillation is very small with respect
Capacitance (C) 1110 µF
Load resistive (R) 80 Ω to the MPP for the proposed technique. On the other hand,
the P&O exhibits large oscillations, as shown by the zoomed
TABLE II portion (see Fig. 10).
PSO PARAMETERS
B. Performance of PV Array Under Step Change of Irradiance
Electrical characteristics Value
Particle numbers 03 We tested the proposed algorithm’s performance under
Maximum Number of iterations 50 varying irradiance conditions to validate its performance. The
Weight (w) 0.9 irradiance level for all PV is initially set to 600 W/m2 , then
C1 1.6
C2 1.2 gradually increased to 1000 W/m2 at 0.6 s, then decreased at
1.2 s to 800 W/m2 , and the temperature is kept constant at
25◦ C.
The results of the tracking voltage and current and power Figure 11 depicts the simulation outcome. As seen in
outputs are depicted in Fig. 9. The PV operated at 1320 W, Fig. 11, a step drop or increase in irradiance causes a change in
which corresponds to V = 213.4 V and I = 6.21 A. Analyzing PV power. As irradiance increases, the hybrid MPPT method
KACIMI et al.: IMPROVED MPPT CONTROL STRATEGY FOR PV CONNECTED TO GRID USING INCCOND-PSO-MPC APPROACH 1015

8
200

PV Current (A)
PV Voltage (V)
Vpv
150 Ipv
4
100

50

0 0
0 0.5 1 1.5 2 0 0.5 1 1.5 2
Time (s) Time (s)
(a) (b)

IncCond-PSO-MPC
Power (W)

1000 Pref

500

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Time (s)
(c)

Fig. 9. MPPT tracking under STC. (a) PV Voltage. (b) PV Current. (c) PV Power output.

C. Performance of PV Array Under Partial Shading

The suggested MPPT is evaluated first in STC, then under
1200
Pref step changes in irradiance, and as shown in Figs. 10 and 11,
Zoom
P&O it converges to the MPP.
Power (W)

IncCond To evaluate the performance of the proposed IncCond-PSO-

800 IncCond-PSO
IncCond-PSO-MPC [Proposed]
PC technique, a 2 s simulation is run, including two partial
shade pattern tests, P1 and P2.
1320 Table III shows the various irradiance patterns, as well as
400
the corresponding MPP.
1280 The partial shading pattern exposes the three PVs panels
0.5 0.51 0.52 to values of (900, 780, 480) W/m2 in the first test. The
0
0 0.5 1 1.5 2 highest output power is 702 W. The proposed MPPT, as shown
Time (s)
in Fig. 12, tracks the maximum power point with minimal
Fig. 10. PV array power output results under Standard condition. oscillations.

1500 800

Zoom 600
1000 1350
Zoom
Power (W)
Power (W)

1300 1050
400
1200 1000
0.88 0.9 700 702 W 675 W
Pref
Power (W)

Zoom
500 Pref 1.52 1.54 500 P&O
750 374.4 W IncCond
P&O 200 300
700 IncCond IncCond-PSO
0.14 0.15 0.16 IncCond-PSO 0 IncCond-PSO-MPC [Proposed]
IncCond-PSO-MPC [Proposed] 0 50 100 150 200 250
0 0 Voltage (V)
0 0.5 1 1.5 2 0 0.5 1 1.5 2
Time (s) Time (s)

Fig. 11. PV array power output results under step change irradiance. Fig. 12. PV power output under Partial shading P1.

quickly converges to its new operating point. Under uniform The PV characteristics corresponding to shading condition
change situations, it is found the suggested MPPT can track P1 are depicted in Fig. 12.
MPP efficiently and with reduced power oscillation. Table IV The proposed algorithm’s tracking result is represented in
displays the outcome of the proposed algorithm. the image above. Traditional IncCond is discovered to be stuck
1016 CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 9, NO. 3, MAY 2023

in LMPP at 374.4 W, while P&O are also found to be stuck 120

P&O
in LMPP at 675 W. 100 IncCond
However, as shown in Fig. 12, the proposed MPPT approach

Efficiency (%)
80 IncCond-PSO
accurately follows the corresponding 702 W. The tracking
60 Proposed
outcome obtained by the suggested algorithm is depicted in
Fig. 12. Classical IncCond is found to be stuck in LMPP at 40
374.4 W, also P&O stuck in LMPP at 675 W. 20
However, as shown in Fig. 12, the suggested MPPT ap-
0
proach accurately follows the Pref corresponding to 702 W. Test 1 test 2 Test 3 Test 4
Similarly, adopting the proposed MPPT, a second test, P2, (a)
0.6
was performed. The partial shade arrangement exposes the P&O

Tracking speed (s)

0.5
three PVs panels to values of (1000, 600, 300) W/m2 , with a 0.4 IncCond
maximum power of 549 W. Results are depicted in Fig. 13. 0.3 IncCond-PSO
The proposed MPPT approach (see Fig. 13), has clearly proved 0.2 Proposed
its usefulness and accuracy in tracking the GMPP. But IncCond 0.1
and P&O are stuck at 628.6 W. The suggested MPPT strategy 0
Test 1 test 2 Test 3 Test 4
is compared to existing techniques in Table IV. (b)

600 Fig. 14. Comparison various MPPT techniques. (a) Efficiency. (b) Tracking.

D. Dynamic Response of the Overall PV System With Hybrid

Power (W)

400 MPPT
600
Pref This section includes the injected PV power from PV
Power (W)

400
P&O systems into the grid under various irradiation changes and
200 200 IncCond
IncCond-PSO
illustrates the efficiency of the approach used. The MPC-
0 IncCond-PSO-MPC[Proposed] controlled two-level voltage source inverter is used. The PI
0 50 100 150 200 250
Voltage (V) controller is employed to keep the DC link voltage steady
0
0 0.5 1 1.5 2 enough to be referenced. Table V depicts the DC/AC converter
Time (s) characteristics used in simulation studies.
First, the global system’s performance is evaluated under
Fig. 13. PV power output under Partial shading P2.
uniform irradiation, as shown in Fig. 15. The irradiance level
for all PV is initially set at 600 W/m2 .
The proposed MPPT outperforms the IncCond-PSO and PV array output oscillates around the MPP, as seen in
normal IncCond approaches, as shown in Table IV. It can Fig. 15(a), but is completely maintained at its reference value
be concluded the proposed method improved convergence (700 V). Then, as shown in Fig. 15(c), a 0.6 s change in
speed significantly, and its total efficiency is over 99%, which irradiation from 600 to 1000 W/m2 leads to an increase in
is greater than existing methods. When using the proposed PV power production and a slight divergence in Vdc from its
method, tracking speed can also display higher precision reference.
tracking.
TABLE V
As can be observed, simulation results validate the sug- E LECTRICAL PARAMETERS OF DC/AC C ONVERTER
gested method’s improved performance in terms of tracking
speed and efficiency when compared to other existing ap- Electrical characteristics Value PI controller Value
proaches. As shown in Fig. 14(a), the suggested method has Grid voltage (e) 50 V Proportional gain (Kp ) 5
Voltage frequency (f ) 50 Hz Integral gain (Ki ) 40
an overall efficiency of almost 99 %, which is greater than Filter resistance (Rf ) 0.5 Ω
other approaches by more than 2 %. Filter inductance (Lf ) 10 mH
DC source voltage (Vdc ) 700 V
Tracking Speed is depicted in Fig. 14(b).

TABLE IV
C OMPARATIVE R ESPONSE P ERFORMANCE A NALYSIS R ESULTS

Tests Performances P&O IncCond IncCond-PSO IncCond-PSO-MPC[Proposed]

Test 1: STC Tracking speed (s) 0.1 0.06 0.09 0.05
Efficiency (%) 95 99 97 99
Test 2: Variable Tracking speed (s) 0.2 0.11 0.12 0.05
Efficiency (%) 97 98 96 98
Test 3 PSC 1 Tracking speed (s) – – 0.3 0.05
Efficiency (%) 88 54 97 97
Test 4 PSC 2 Tracking speed (s) – – 0.5 0.15
Efficiency (%) 76 78 95 99
KACIMI et al.: IMPROVED MPPT CONTROL STRATEGY FOR PV CONNECTED TO GRID USING INCCOND-PSO-MPC APPROACH 1017

1500
1000
Irradiance (W/m2)
1000

Power (W)
800

500
Pref
600
IncCond-PSO-MPC [Proposed]
0
0 0.6 1.2 2 0 0.6 1.2 2
Time (s) Time (s)
(a) (b)
1000
Active Power
Vdc Reactive Power
Vdc-ref
800 800
Voltage (V)

Power (W)
600 400

0
400
0 0.6 1.2 2 0 0.6 1.2 2
Time (s) Time (s)
(c) (d)
20

10
(ia,ib,ic) (A)

−10 Zoom Zoom

10 10
−20 0 0
−10 −10
0.58 0.6 0.62 0.64 1.3 1.32 1.34
−30
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Time (s)
(e)

Fig. 15. Performance of global system under uniform irradiation changes. (a) Solar irradiance profile; (b) PV power; (c) DC-link voltage; (d) Grid-injected
active and reactive power; and (e) Grid current of the three-phase system.

Increases in grid current are caused by increases in PV shading pattern exposes the three PV panels to values of (G1
power, as seen in Fig. 15(e), where the amplitude of the = 900, G2 = 780, G3 = 480) W/m2 , with a maximum power
grid current is proportional to PV power despite a modest of 702 W. The irradiance will then be increased to 1000 W/m2
overshoot from its reference (700 V). The MPC’s ability to at 0.6 s for all three PV panels, and then lowered to 600 W/m2
inject maximum available power is in response to irradiation at 1.2 s, as illustrated in Fig. 16(a).
variations. The proposed MPPT tracks maximum power for all irra-
As can be seen, the grid’s instantaneous active and reactive diation conditions. Vdc is maintained with little deviation to
power has a low ripple in the injected power. It can also be its reference, as seen in Fig. 16(c). Reactive power is nearly
seen that active power behaves similarly to the power, and 0 V, and grid current has preserved its sinusoidal shape, as
the reactive power stabilizes at zero and is unaffected by illustrated in Fig. 16(e).
variations in irradiation. Furthermore, despite the alteration,
the grid current maintained its sinusoidal structure. There is a V. C ONCLUSION AND F UTURE W ORK
shift in irradiance; there is a slight difference in current. The In this paper, a new hybrid MPPT controller for grid-
proposed control mechanism, as shown in Fig. 15, ensures connected solar systems combining Incremental conductance
rapid regulation of this variance. — Particle Swarm Optimization and Model Predictive Con-
Figure 16 depicts the second test. Initially, the partial trol (IncCond-PSO-MPC) is presented. The proposed solution
1018 CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 9, NO. 3, MAY 2023

1500
1000
Irradiance (W/m2) G1

G2 1000

Power (W)
800

600 500 Reference

IncCond-PSO-MPC [Proposed]
G3

0 0.6 1.2 2 0
0 0.5 1 1.5 2
Time (s) Time (s)
(a) (b)
1000
Vdc-ref 1200
Vdc Active Power
Reactive Power
800 800
Voltage (V)

Power (W)
400
600

0
400
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Time (s) Time (s)
(c) (d)
20

10
(ia,ib,ic) (A)

−10 Zoom
10
−20 0
−10
−30 0.58 0.6 0.62 0.64
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Time (s)
(e)

Fig. 16. Performance of global system under irradiation changes. (a) Solar irradiance profile; (b) PV power; (c) DC-link voltage; (d) Grid-injected active
and reactive power; and (e) Grid current of the three-phase system.

involves producing as much power as possible from a PV tion of the MPC controller with the IncCond-PSO not only
system under changing environmental conditions and then helped dramatically reduce steady-state oscillations in output
transferring it to the power grid. Maximum power is located power, but also reduced total maximum power point tracking
using a hybrid combination of Incremental conductance (Inc- time when compared to the IncCond-PSO and traditional
Cond) and Particle Swarm Optimization (PSO), and the Model IncCond. We can conclude the new proposed MPPT is a viable
Predictive Control (MPC) then records maximum power and technique for boosting efficiency of a PV system in partial
controls the boost converter to achieve high performance shade.
regardless of parameter variations. A two-level inverter, which As future work, further research could focus on experiments
is likewise controlled by MPC, is utilized to inject the PV comparing these techniques for mismatched situations such as
power generated. The proposed MPPT was tested using factors partial shadowing, non-uniformity of PV panel temperatures,
such as power tracking speed and efficiency. A number of damage to panel glass, and dust or shadow impacts for specific
experimental simulations on a variety of instances (tests) have applications.
demonstrated the proposed approach is highly effective.
In tracking time and power conversion efficiency in PVout- R EFERENCES
put power, the suggested method greatly outperforms tradi-
[1] M. Hosenuzzaman, N. A. Rahim, J. Selvaraj, M. Hasanuzzaman, A.
tional P&O, conventional IncCond and Incremental conduc- B. M. A. Malek, andA. Nahar, “Global prospects, progress, policies,
tance – Particle Swarm Optimization (IncCond-PSO). Integra- and environmental impact of solar photovoltaic power generation,”
KACIMI et al.: IMPROVED MPPT CONTROL STRATEGY FOR PV CONNECTED TO GRID USING INCCOND-PSO-MPC APPROACH 1019

Renewable and Sustainable Energy Reviews, vol. 41, pp. 284–297, Jan. systems affected by partial shading,” Energy, vol. 74, pp. 374–388, Sep.
2015. 2014.
[2] M. Ahmed, M. Abdelrahem, and R. Kennel, “Highly efficient and robust [21] P. E. Kakosimos andA. G. Kladas, “Implementation of photovoltaic
grid connected photovoltaic system based model predictive control with array MPPT through fixed step predictive control technique,” Renewable
Kalman filtering capability,” Sustainability, vol. 12, no. 11, pp. 4542, Energy, vol. 36, no. 9, pp. 2508–2514, Sep. 2011.
Jun. 2020, doi: 10.3390/su12114542. [22] K. Barra and D. Rahem, “Predictive direct power control for photovoltaic
[3] B. Bendib, H. Belmili, and F. Krim, “A survey of the most used MPPT grid connected system: An approach based on multilevel converters,”
methods: conventional and advanced algorithms applied for photovoltaic Energy Conversion and Management, vol. 78, pp. 825–834, Feb. 2014.
systems,” Renewable and Sustainable Energy Reviews, vol. 45, pp. 637– [23] S. Chen, Q. Yang, J. Zhou, and X. Chen, “A model predictive control
648, May 2015. method for hybrid energy storage systems,” CSEE Journal of Power and
[4] B. Parida, S. Iniyan, and R. Goic, “A review of solar photovoltaic Energy Systems, vol. 7, no. 2, pp. 329–338, Mar. 2020).
technologies,” Renewable and Sustainable Energy Reviews, vol. 15, [24] H. Chen, R. Xiong, C. Lin, and W. Shen, “Model predictive control
no. 3, pp. 1625–1636, Apr. 2011. based real-time energy management for hybrid energy storage system,”
[5] A. A. S. Mohamed, H. Metwally, A. El-Sayed, and S. I. Selem, CSEE Journal of Power and Energy Systems, vol. 7, no. 4, pp. 862–874,
“Predictive neural network based adaptive controller for grid-connected Jul. 2021.
PV systems supplying pulse-load,” Solar Energy, vol. 193, pp. 139–147, [25] E. Z. Bighash, S. M. Sadeghzadeh, E. Ebrahimzadeh, and F. Blaabjerg,
Nov. 2019. “High quality model predictive control for single phase grid-connected
[6] . R. B. Bollipo, S. Mikkili and P. K. Bonthagorla, “Hybrid, optimal, photovoltaic inverters,” Electric Power Systems Research, vol. 158,
intelligent and classical PV MPPT techniques: A review,” CSEE Journal pp. 115–125, May 2018.
of Power and Energy Systems, vol. 7, no. 1, pp. 9–33, Jan. 2021. [26] B. Boukezata, J. P. Gaubert, A. Chaoui, and M. Hachemi, “Predictive
[7] U. Yilmaz, A. Kircay, and S. Borekci, “PV system fuzzy logic MPPT current control in multifunctional grid connected inverter interfaced by
method and PI control as a charge controller,” Renewable and Sustain- PV system,” Solar Energy, vol. 139, pp. 130–141, Dec. 2016.
able Energy Reviews, vol. 81, pp. 994–1001, Jan. 2018 [27] S. Sumathi, L. A. Kumar, and P. Surekha, Solar PV and Wind Energy
[8] N. Kacimi, S. Grouni, A. Idir, and M. S Boucherit, “New im- Conversion Systems, Cham: Springer, 2015.
proved hybrid MPPT based on neural network-model predictive control- [28] A. Guichi, A. Talha, E. M. Berkouk, S. Mekhilef, and S. Gassab, “A new
Kalmanfilter for photovoltaic system,” Indonesian Journal of Electrical method for intermediate power point tracking for PV generator under
Engineering and Computer Science, vol. 20, no. 3, pp. 1230–1241,Dec. partially shaded conditions in hybrid system,” Solar Energy, vol. 170,
2020. pp. 974–987, Aug. 2018.
[9] C. Larbes, S. M. A. Cheikh, T. Obeidi, and A. Zerguerras, “Genetic [29] K. Rajani, and T. Ramesh, “Maximum power enhancement under partial
algorithms optimized fuzzy logic control for the maximum power point shadings using a modified Sudoku reconfiguration,” CSEE Journal of
tracking in photovoltaic system,” Renewable Energy, vol. 34, no. 10, Power and Energy Systems, vol. 7, no. 6, pp. 1187–1201, 2020.
pp. 2093–2100, Oct. 2009. [30] S. A. Tadjer, A. Idir, F. Chekired, “Comparative performance evalua-
[10] B. Talbi, F. Krim, T. Rekioua, A. Laib, and H. Feroura, “Design and tion of four photovoltaic technologies in saharan climates of Algeria:
hardware validation of modified P&O algorithm by fuzzy logic approach Ghardaı̈a pilot station,” Indonesian Journal of Electrical Engineering
based on model predictive control for MPPT of PV systems,” Journal and Computer Science (IJEECS), vol. 18, no. 2, pp. 586–598, May
of Renewable and SustainableEnergy, vol. 9, no. 4, pp. 043503, Aug. 2020.
2017. [31] U. K. Das, K. S. Tey, M. Seyedmahmoudian, S. Mekhilef, M. Y. I.
[11] M. Kermadi and E. M. Berkouk, “Artificial intelligence-based maximum Idris, W. Van Deventer, B. Horan, and A. Stojcevski, “Forecasting
power point tracking controllers for photovoltaic systems: Comparative of photovoltaic power generation and model optimization: a review,”
study,” Renewable and Sustainable Energy Reviews, vol. 69, pp. 369– Renewable and Sustainable Energy Reviews, vol. 81, pp. 912–928, Jan.
386, Mar. 2017. 2018.
[12] Naghmash, H. Armghan, I. Ahmad, A. Armghan, S. Khan, and M. [32] S. S. Lee and Y. E. Heng, “Predictive direct power control of multilevel
Arsalan, “Backstepping based non-linear control for maximum power direct current link converter for grid connected battery energy storage
point tracking in photovoltaic system,” Solar Energy, vol. 159, pp. 134– systems,” Journal of Renewable and Sustainable Energy, vol. 8, no. 3,
141, Jan. 2018. pp. 034104, May 2016.
[13] R. I. Putri, S. Wibowo, and M. Rifa, “Maximum power point tracking for [33] V. Salas, E. Olı́as, A. Barrado, and A. Lázaro, “Review of the maximum
photovoltaic using incremental conductance method,” Energy Procedia, power point tracking algorithms for stand-alone photovoltaic system,”
vol. 68, pp. 22–30, Apr. 2015. Solar Energy Materials and Solar Cells, vol. 90, no. 11, pp. 1555–1578,
[14] N. Kacimi, A. Idir. S. Grouni, M. S. Boucherit, “A new combined Jul. 2006.
method for tracking the global maximum power point of photovoltaic [34] A. Laib, F. Krim, B. Talbi, A. Kihal, and H. Feroura, “Improved
systems”, Rev. Roum. Sci. Techn.–Électrotechn. et Énerg, vol.67, 3, control for three phase dual-stage grid-connected PV systems based on
pp. 349–354, Sep, 2022. predictive control strategy,” Journal of Control Engineering and Applied
[15] B. Yang, T. J. Zhu, J. B. Wang, H. C. Shu, T. Yu, X. S. Zhang, and Informatics, vol. 20, no. 3 pp. 12–23, Oct. 2018.
W. Yao, “Comprehensive overview of maximum power point tracking [35] T. S. Babu, J. P. Ram, T. Dragicevic, M. Miyatake, F. Blaabjerg,and
algorithms of PV systems under partial shading condition,” Journal of N. Rajasekar, “Particle swarm optimization based solar PV array re-
Cleaner Production, vol. 268, pp. 121983, Sep. 2020. configuration of the maximum power extraction under partial shading
[16] B. Yang, L. E. Zhong, X. S. Zhang, H. C. Shu, T. Yu, H. F. LI, L. conditions,” IEEE Transactions on SustainableEnergy, vol. 9, no. 1, Jan.
Jiang, and L. M. Sun, “Novel bio-inspired memetic salp swarm algorithm 2018.
and application to MPPT for PV systems considering partial shading [36] A. M. Eltamaly, M. S. Al-Saud, A. G. Abokhalil, and H. M. H. Farh,
condition,” Journal of Cleaner Production, vol. 215, pp. 1203–1222, “Simulation and experimental validation of fast adaptive particle swarm
Apr. 2019. optimization strategy for photovoltaic global peak tracker under dynamic
[17] B. Yang, T. Yu, X. S. Zhang, H. F. Li, H. C. Shu, Y. Y. Sang, and partial shading,” Renewable and Sustainable Energy Reviews, vol. 124,
L. Jiang, “Dynamic leader based collective intelligence for maximum pp. 109719, May 2020.
power point tracking of PV systems affected by partial shading condi-
tion,” Energy Conversion and Management, vol. 179, pp. 286–303, Jan.
2019.
[18] R. Ramaprabha, M. Balaji, and B. L. Mathur, “Maximum power point
tracking of partially shaded solar PV system using modified Fibonacci Nora Kacimi received her Engineer degree in Auto-
search method with fuzzy controller,” International Journal of Electrical matic Control and Magister degree in Electrical En-
Power & Energy Systems, vol. 43, no. 1, pp. 754–765, Dec. 2012. gineering from M’Hamed Bougara University, Alge-
[19] H. Patel and V. Agarwal, “Maximum power point tracking scheme ria. She is a Ph.D. student in Electrical Engineering
for PV systems operating under partially shaded conditions,” IEEE at National Polytechnic School (ENP), Algeria. Her
Transactions on Industrial Electronics, vol. 55, no. 4, pp. 1689–1698, current interests include Diagnosis research team
Apr. 2008. and renewable energy.
[20] S. Daraban, D. Petreus, and C. Morel, “A novel MPPT (maximum
power point tracking) algorithm based on a modified genetic algorithm
specialized on tracking the global maximum power point in photovoltaic
1020 CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 9, NO. 3, MAY 2023

Abdelhakim Idir received his B.S. and M.S. de- Mohamed Seghir Boucherit graduated from Al-
grees in Control from Bejaia University and Setif geria’s National Polytechnic School (ENP) with an
University, respectively, in 2003 and 2006. He re- Engineer degree in Electrotechnics, a Master degree,
ceived his doctorate and HDR degrees in Electrical and a Ph.D. degree in Electrical Engineering in 1980,
Engineering from University M’Hamed Bougara in 1988, and 1995, respectively. Following graduation,
Boumerdes, Algeria, in 2015 and 2018, respectively. he enrolled in the Electrical Engineering Department
From November 2008 to July 2019, he worked as a of the National Polytechnic School of Algiers (ENP).
Lecturer in the Department of Electrical Engineering He is a Professor and the Director of the Process
at Boumerdes University in Algeria. In 2019, he Control Laboratory’s Industrial Systems and Diag-
began working as an associate professor in the nosis research team.
department of Electrical Engineering at Mohamed
Boudiaf University in M’Sila, Algeria. His current research interests include
fractional system modeling, simulation, and control, fractional PID control,
AC drives, and renewable energy.

Said Grouni graduated from the National Polytech-

nic School of Algiers (ENP) in 1988 with an En-
gineer degree in Electrotechnics, a Magister degree
in Electrical Engineering, and a Doctorate degree
in Electrical Engineering. He is currently a Profes-
sor at Algeria’s M’hamed Bougara of Boumerdes
University (UMBB). At the Laboratory of Applied
Automatics, he directs the control system and fault
diagnosis research teams (LAA). Since 2017, he
has served as the dean of the electrical engineering
systems department at Boumerdes University, where
he has also taught electrical machine courses. His current research interests
include electrical drives, process control applications, electrical machines, and
fault diagnosis.