Article

Stability Analysis of Chaotic Grey-Wolf Optimized Grid-Tied

PV-Hybrid Storage System during Dynamic Conditions
Mukul Chankaya 1, Ikhlaq Hussain 2, Aijaz Ahmad 1, Hasmat Malik 3,* and Majed A. Alotaibi 4,*

1 Department of Electrical Engineering, NIT Srinagar, Srinagar 190006, India;

[email protected] (M.C.); [email protected] (A.A.)
2 Department of Electrical Engineering, University of Kashmir, Srinagar 190006, India;

[email protected]
3 BEARS, University Town, NUS Campus, Singapore 138602, Singapore

4 Department of Electrical Engineering, College of Engineering, King Saud University, Riyadh 11421,

Saudi Arabia
* Correspondence: [email protected] (H.M.); [email protected] (M.A.A.)

Abstract: This paper presents the stability improvement of the three-phase four-wire (3P-4W) grid-
tied PV-hybrid energy storage system (HESS) using chaotic grew wolf optimization (CGWO) for
DC bus voltage (𝑉 ) and AC bus voltage (𝑉 ) control. The CGWO tuned fractional order pro-
portional–integral (FOPI) controllers reduce the 𝑉 and 𝑉 variations during diverse, dynamic
conditions, i.e., sudden irradiation variations, deep voltage sag/swell, etc. The DC bus is responsible
for the current injection/extraction control, maximum PV power extraction, bi-directional power
flow, dc second-harmonics component elimination, and active power balance. At the point of com-
mon coupling (PCC), the AC bus is accountable for bi-directional power flow and active and reac-
tive power control. The two-level voltage source converter (VSC) is controlled by a novel variable
Citation: Chankaya, M.; Hussain, I.; step-size incremental least mean square (VSS-ILMS) in zero voltage regulation (ZVR) mode. Due to
Ahmad, A.; Malik, H.; Alotaibi, M.A. its varying step size, VSC control is less prone to noise signals offers better stability, improved con-
Stability Analysis of Chaotic vergence rate, dc offset rejection, and tracking speed during dynamics, i.e., large oscillations. A bat-
Grey-Wolf Optimized Grid-Tied tery and ultracapacitor are coupled to the DC link by buck-boost converters in the HESS. To regulate
PV-Hybrid Storage System during
power transit between the DC bus and the grid, the HESS current control technique is designed to
Dynamic Conditions. Electronics
shift frequently from charging to discharging stage and vice versa. The novelty of the PV-HESS
2022, 11, 567. https://doi.org/10.3390/
system lies in CGWO tuned VSS-ILMS control of VSC, which effectively and efficiently filter out
electronics11040567
the active fundamental constituents of load current and eliminate dc offset from VSC output. The
Academic Editor: Zbigniew HESS control maintains the DC bus voltage profile by absorbing and delivering energy (during
Leonowicz dynamic conditions) rather than curtailing it. The presented system is simulated in a MATLAB/SIM-
Received: 30 January 2022 ULINK environment. The simulation results in graphical and numerical forms verify the stable and
Accepted: 11 February 2022 satisfactory operation of the proposed system as per IEEE519 standard.
Published: 13 February 2022
Keywords: PV power generation; DC bus stability; power electronics; power quality; adaptive con-
Publisher’s Note: MDPI stays neu-
trol; hybrid energy storage system; chaotic grey wolf optimization; optimization techniques
tral with regard to jurisdictional
claims in published maps and institu-
tional affiliations.

1. Introduction
Nowadays, photovoltaic (PV) power generation is widely acclaimed due to its scal-
Copyright: © 2022 by the authors. Li- able nature and less initial investment than wind energy systems [1]. The grid-tied PV
censee MDPI, Basel, Switzerland. systems offer generation diversification, peak shaving capability, improved reliability, re-
This article is an open access article
duced ancillary reserves, improved voltage profile, and reduced transmission losses, etc.
distributed under the terms and con-
Moreover, grid-tied PV systems as distributed generators (DG) lessen the impact of cyber-
ditions of the Creative Commons At-
attack improve power utility security while delivering the critical load [2,3]. In India, the
tribution (CC BY) license (https://cre-
total installed capacity of renewable energy sources (RES) is 103 GW by 2021 (excluding
ativecommons.org/licenses/by/4.0/).
hydropower), in which PV installation is 40%. The total RES installed capacity is expected

Electronics 2022, 11, 567. https://doi.org/10.3390/electronics11040567 www.mdpi.com/journal/electronics

Electronics 2022, 11, 567 2 of 23

to grow up to 450 GW by 2030 with a 60% share of PV power (280 GW) [4]. Due to the
cost-effectiveness, fewer space constraints, compact size, and ready availability make the
grid-tied PV systems suitable for both utility-scale and rooftop plants.
The penetration of RES is increasing day by day in the modern grid to match the
energy requirements, which raises many technical issues to be managed before delivering
PV power to the grid [5,6]. Considering the cost of PV arrays in PV power generation
systems, it is highly desirable to operate them at the highest efficiency. Maximum power
point tracking (MPPT) algorithms are utilized for the same. In [7,8] many conventional
and artificial intelligence (AI) based MPPT algorithms are given. The transformer-less PV-
HESS grid integration is more economical and efficient. However, dual-stage configura-
tion with DC–DC boost converter is best suited for enhanced DC bus stability rather than
using single-stage configuration [9,10]. Several technical challenges related to grid inte-
gration of PV systems involves designing and control of voltage source converters (VSC),
voltage and current harmonics, active and reactive power control, voltage and frequency
deviations, stability, power balancing at DC and AC bus, weak grid conditions, PV power
fluctuations, bi-directional power flow, lower efficiency, reliability issues, etc. [11,12]. For
the satisfactory and healthy operation of grid integration of PV system and to deliver PV
power to the grid, technical complications, i.e., stability and power quality issues, must
be dealt with.
The stability of the grid-tied PV system significantly relies on DC bus and AC bus
stability. The DC bus plays an essential role in managing bi-directional power flow, active
power balance, maximum PV power extraction, reducing PV power fluctuations, gener-
ating active loss component of current, etc. [13,14]. At the point of common coupling
(PCC), the AC bus manages bi-directional active and reactive power flow, islanding de-
tection, low voltage ride through, reactive power compensation, etc. In the case of grid-
following converters, the PCC voltage is very much dependent on the grid, which reduces
the burden of the AC bus and leaves it with mainly reactive power control [15].
The PI controllers generally control the dc and ac buses due to their simple design
procedure, good performance, less computational complexity, and ease of tuning [16,17].
Much artificial intelligence (AI), i.e., fuzzy logic and neural network-based controls, have
also been utilized for proportional–integral (PI) controller gain tuning, which increases
the system’s complexity [18,19]. Many metaheuristic optimization techniques (MOTs)
have been implemented in an offline mode to deliver optimal gains of the PI controller
[14,20–23]. The fractional-order PI controller offers enhanced performance than the PI con-
troller due to its non-integer integral gain. The gains of fractional-order PI (FOPI) can be
obtained by using MOTs, i.e., salp swarm optimization (SSO), chaotic grey wolf optimi-
zation (CGWO), and many more [20,23]. The increased tuning parameter as non-integer
integral gain increases the system’s robustness to deliver optimal control, better system
response, and power quality [24–26]. The FOPI controlled DC bus and AC bus reduces
the initial transients and dynamic variations of the DC bus (𝑉 ) and AC bus voltage
magnitude (𝑉 ) to a greater extent and performs significantly better than the PI control-
ler. Table 1 presents the optimization technique for parameter optimization in a grid-tied
system
The HESS consists of static storage devices (i.e., battery, ultracapacitor (UC), fuel
cells), which have a shorter response time to any system dynamics than the compressed
air storage system (flywheel storage, etc.) [14,20]. Energy storage systems with faster re-
sponse times are suitable for handling sudden, sustained, and prolonged disturbances,
i.e., fault, load variations, and voltage sag/swell.
Electronics 2022, 11, 567 3 of 23

Table 1. MOTs for parameter estimation.

Operational Operating
Ref Optimization Technique Objective DG Used
Mode Conditions
𝑉
regulation, filter pa-
[13] Jaya optimization Solar PV Grid-Tied Load unbalancing
rameter estimation
[14] GNDO 𝑉 regulation Solar PV-HESS Grid-Tied Weak grid weak
[15] Fuzzy logic 𝑉 , 𝑉 regulation Solar PV Grid-Tied grid
[18] Fuzzy parameter optimization 𝑉 regulation Solar PV Grid-Tied PF correction
[20] SSO 𝑉 regulation Solar PV-HESS Grid-Tied Weak grid
[21] MRFO, GWO, WO, GHO, ASO 𝑉 , 𝑉 , MPPT regulation Solar PV Grid-Tied Insolation change
[22] Taylor approximation 𝑉 regulation Solar PV Grid-Tied Unstable control

The power quality issues, i.e., voltage and current harmonics suppression, should be
seriously considered while designing the VSC controls. Various advance control algo-
rithms have been used to reduce the total harmonics distortion (THD) level well below
5% as per IEEE 519 standard and perform multi-functional operations. Time-domain con-
trols, i.e., synchronous reference frame (SRF), power balance theory (PBT), etc. [27,28],
have been widely used due to their ease of implementation, fairly adequate performance,
and less computational burden. The time-domain control does not perform well under
transients, delivers a low convergence rate during dynamic conditions, and achieves
multi-functional operations. The frequency-domain controls, i.e., wavelet and Stockwell
transform algorithms, offer better filtering capability but at the cost of higher computa-
tional burden and slow convergence rate [29,30]. The adaptive controls, i.e., least mean
square (LMS), least mean fourth (LMF), etc., come up with ease of implementation and
with slightly higher complexity in comparison to time-domain controls [16,17]. The LMS
algorithm uses a fixed step size for adaptation to find the balance between the conver-
gence equation and maladjustment error results at low-convergence speed and more sig-
nificant steady-state oscillations. The LMF algorithm acts as a higher-order adaptive filter
for better MSE reduction with a fixed step size. Due to the fixed adaptation size, efficiency
reduces during abrupt dynamic changes. The above-mentioned adaptive algorithm has
inspired many derivatives of these controls, i.e., variable step-size LMS (VSS-LMS), de-
layed LMS (DLMS), median LMS (MLMS), hyperbolic cosine LMS (HCLMS), combined
LMS-LMF, etc. VSS-LMS [31] offers reduced steady-state error, but its performance re-
duces during significant dynamic changes. DLMS [32] is similar to LMS, introducing a
delay factor in the unit template for better weight generation. MLMS [33], with an addi-
tional scaling factor other than step-size, improves the noise cancellation. HC-LMS [34]
offers a better convergence rate during dynamic conditions than LMS. Combined LMS-
LMF [35] shifts control from LMS to LMF (higher-order filter) depending on the magni-
tude of the error. Variable step-size incremental LMS (VSS-ILMS) [36] is a new addition
to the derivatives of the LMS algorithm, which offers better filtering capability against
noise signals in the sparse environment due to its incremental nature of adaptation step.
The main objective is to present a stable grid-tied PV-HESS with low voltage ride-
through capability during fragile grid conditions. The DC and AC bus voltages are stable
even during deep voltage sag/swell conditions. The power quality of the system remains
intact with CGWO tuned FOPI controllers. The novelty of the presented work lies in VSS-
ILMS based VSC control that performs multi-functional operations, including power and
load balancing, active and reactive power control, power quality enhancement, dc offset
removal, etc. The CGWO tuned FOPI controlled DC, and the AC bus also stabilizes the
system during deep voltage sag/swell conditions.
This paper is organized in the following manner: Section 2 presents the general de-
scription of the system and topology implemented; Section 3: illustrates the implemented
research methodology; Section 4: gives the detailed explanation of implemented control
algorithms, including, 𝑉 and HESS control and VSS-ILMS based VSC control. Section
Electronics 2022, 11, 567 4 of 23

5: Simulation results are presented and discussed in detail to show the efficacy of the pro-
posed work; Section 6: Conclusion of the presented work is given.

2. System Description
Figure 1 shows the topology of the proposed system. The presented system is a three-
phase four-wire (3P-4W) grid-tied PV-HESS. The PV array of 32 kW is connected to the
DC bus via a DC-DC boost converter to stabilize PV voltage (𝑉 ) during irradiation var-
iation. The MPPT is achieved by using the incremental conductance (InC) method. The
HESS involving battery and UC is also attached to the DC bus via a bidirectional DC-DC
buck-boost converter to control the charging and discharging operation.

PPV to VSC Ripple Filter

Control Gating signal Gating Signals
for Boost Non-Linear
InC based Converter Load
DC Bus
MPPT Interfacing
at 700V S1 S3 S5 S7
Inductors
VPV IPV iLa iLb iLc iLn
iCa
T Vdc
iCb
iCc
G
iCn PCC
PV Array Boost converter S2 S4 S6 S8

RS & LS
IBat
Rs Lb Sb1
Multi-Objective Chaotic Grey
Vdc Wolf Optimization VPCC iSa iSb iSc
Cp Rp Sb2 Optimized Optimized Vterr
Verr
B - gains gains - iSn
FOPI FOPI
Buck-Boost Converter and Battery + Control Control +
Three-Phase Four-Wire AC
SUC1 Grid 415V at 50Hz
RUC LUC IUC
to VSC Control
SUC2 PCC Voltage Control
CUC
Sb1 & Sb2 +
Buck-Boost Converter and UC Control PI IBat
PWM -
Control &
SUC1 & SUC2
Control
IUC
Bidirectional Buck-Boost Converter Control

Figure 1. Proposed Topology.

The 𝑉 and 𝑉 are controlled by multi-objective CGWO (MOCGWO) tuned FOPI

controller to minimize their variations during dynamic conditions. The FOPI managed
DC bus generates precise 𝑖 , which acts as a reference signal for battery current control,
suppresses the second-harmonics component at DC bus, and improves the power quality
of the system. The two-level VSC is attached to the PCC via interfacing inductors to reduce
the voltage and current ripples. The grid at 415 V, 50 Hz is also linked to the PCC. The
VSC control is provided by the VSS-ILMS algorithm, which generates the direct and quad-
rature fundamental component of load current to generate the required reference currents
and eliminates the dc offset at PCC. The proposed system is termed smart due to its capa-
bility in operating in three modes, i.e., (1). The total power is delivered to the grid during
PV to grid mode, (2). During PV-DSTATCOM mode, the load is satisfied with both the
PV and grid (3). during DSTATCOM mode, the load is satisfied by the grid only due to
the unavailability of PV power. The proposed system with low voltage ride-through ca-
pability is also analyzed under load unbalancing and deep voltage sag/swell conditions.
Electronics 2022, 11, 567 5 of 23

The two-level VSC is attached to the PCC via interfacing inductors and R–C filters to re-
duce the current and voltage ripples produced due to VSC switching. The grid at 415 V,
50 Hz is also attached to the PCC.
The primary characteristics of the presented work are as follows.
a. Smart grid-tied PV operation: The grid-tied PV system performs satisfactorily in three
modes: (1) PV to grid mode, (2) PV-DSTATCOM mode, (3) PV-DSTATCOM to
DSTATCOM, and vice versa mode.
b. Dynamic state operations: The presented system is observed under weak grid condi-
tions like load unbalancing and deep voltage sag/sell condition, i.e., 20%, 40%, and
60%, respectively.
c. Multi-functional operation: The proposed system performs multiple operations, i.e.,
load balancing, harmonics elimination, active and reactive power control, etc.
d. DC and AC bus stability performance: The DC and AC bus control is provided by
CGWO tuned FOPI controller to stabilize the system during dynamic conditions.
e. HESS: The HESS ensures continuous supply to critical load and enhances the system’s
power quality by suppressing the second-harmonics content at the DC bus.

3. Implemented Research Methodology

Figure 2 shows the implemented research methodology of the MOCGWO technique.
Various signals, i.e., source voltage (𝑣 ), source current (𝑖 ), 𝑉 , 𝑉 , PV voltage,
current, power (𝑉 , 𝐼 , 𝑃 ). The 𝑉 and reference 𝑉 , i.e., (𝑉 ∗ ) are compared to gen-
erate the first objective function (Obj-1), which is an integral square error (ISE-1) as per
(1), to be minimized by MOCGWO for delivering optimal gains of FOPI controller, i.e.,
𝑘 , 𝑘 , 𝜆 for 𝑉 control. The direct loss component of current (𝑖 ) is generated from
the 𝑉 control contributes to generating the fundamental elements of load current by
VSC control. The 𝑉 is also compared with the reference 𝑉 , i.e., (𝑉 ∗ ) to generate a
second objective function (Obj-2) with ISE-2 as its error indices as per (2). The Obj-2 min-
imization also delivers, 𝑘 , 𝑘 , 𝜆 for the 𝑉 control. The quadrature loss component of
current (𝑖 ) is further included in the quadrature fundamental component of load
current by VSC control. With optimized AC and DC buses, VSC control generates more
accurate weight signals and reference currents, which further improves the performance
of the proposed system.

Obj– 1 = (𝑉 ∗ – 𝑉 ) = (𝑉 ) (1)

Obj– 2 = 𝑉∗ – 𝑉 = (𝑉 ) (2)

Multi-Objective Chaotic Grey Wolf Optimization Technique

Figure 3 shows the flowchart of the MOCGWO algorithm, which is inspired by the
social hunting behavior of grey wolves. Chaos theory is introduced into the GWO algo-
rithm for accelerating its global convergence speed. Despite having a reasonable conver-
gence rate, GWO has low solving accuracy and poor local searchability. To overcome the
shortcomings of GWO, the CGWO algorithm is developed by introducing a piecewise
chaotic map in the GWO algorithm to produce a diverse initial population with uniform
distribution as per (3). The reader may refer to more information and paradigm [37–41].
The local minima stagnation is avoided by chaos, randomness, and ergodicity.
Electronics 2022, 11, 567 6 of 23

Sense VPV,IPV,Vdc,Vpcc, vSabc, and iSabc

VPV & IPV

Boost converter control by InC

based MPPT
Vpcc PPV Vdc
Feed-forward

Chaotic Grey Wolf Optimization

ISE1
term
Verr

Vdc regulation by optimized

FOPI controller
idLoss
WPV
Direct axis fundamental load
current component generation
Vpcc
ISE2

Vpcc regulation by optimized

FOPI controller
iqLoss
Quadrature axis fundamental load
current component generation

Comparison of & iSabc delivered

to the Hysteresis current controller

VSC Switching sequence generation

Figure 2. Research methodology.

Initialize random population for MOCGWO

and provide termination criteria
M odelling Phase

Increment iterations as i=i+1

Provide design variables X with

constraints and design objectives

Specify upper bound ,lower bound

E xploration phase

Initialize MOCGWO parameters ('a','A' and 'C')

and initial position by piecewise chaotic map
M ulti-O bjective C haotic G rey Wolf

Calculate the fitness function as per equation 4

O ptim ization
E xploitation phase

Evaluate first three best fit wolves, i.e.,

Update chaotic sequence of the piecewise chaotic

map along with wolves positions

Evaluate fitness again and replace the worst wolf

with the best grey wolf

NO
Termination criteria
D ecision Phase

satisfied
YES

Output the best solution

Stop

Figure 3. GNDO flow chart.

Electronics 2022, 11, 567 7 of 23

𝑃 0≤𝑝 ≤𝑎
⎧ ⎫
⎪ 𝑎 ≤ 𝑝 ≤ 0.5 ⎪
.
𝑃 = (3)
⎨ 0.5 ≤ 𝑝 ≤ 1 − 𝑎⎬
.
⎪ ⎪
⎩ 1−𝑎 ≤𝑝 ≤1 ⎭

The exact hierarchy of wolves in a wolf pack is considered as GWO, which is 𝛼, 𝛽,

𝛾, and 𝜁. The roles and responsibilities of wolves in the group reduce with the descending
order. The 𝛼 wolf is a pack leader, responsible for giving instructions and decision mak-
ing. 𝛽 wolf is second in command and the best candidate to replace 𝛼. 𝛾 wolves take
order from 𝛼, 𝛽 and dominate 𝜁, which is fourth and the last level. 𝜁 wolves maintain
safety and integrity in the wolf pack.
This first step of MOCGWO is to generate a diverse population with uniform distri-
bution initialization of wolf pack. The chaotic map described in (3) is mapped with the
iterative algorithm to initialize the first chaotic numbers. Various parameters involved in
the CGWO algorithm for conducting exploration-exploitation mechanism viz. ‘a’, ‘A’ and
‘C’ are generated, the same as in GWO (4)–(6). The distances from 𝛼, 𝛽, and 𝛾 wolves,
i.e., 𝐷 , 𝐷 , and 𝐷 to each of the remaining wolves (𝑝⃗) as a fitness function, using
which the effect of 𝛼, 𝛽 and 𝛾 wolves on the prey viz. 𝑝⃗, 𝑝 ⃗, and 𝑝 ⃗ can be calculated
further. The fitness of an initialized population of the wolf in search space is calculated
and sorted using the introduced chaotic map and are sorted according to their fitness. The
fittest wolf position will be updated in the next iteration (7) and may get the wolf’s posi-
tion as the optimal solution in case termination criteria are fulfilled. In each iteration, as-
sociated parameter values are also updated (6). With termination, criteria satisfied the
leader wolf’s fitness or position will be considered the most optimal solution or PI con-
troller gains for 𝑉 and 𝑉 regulation by the MOCGWO algorithm.

𝐷 ⃗ = 𝐶⃗ × 𝑝 ⃗ − 𝑝⃗ , 𝐷 ⃗ = 𝐶⃗ × 𝑝 ⃗ − 𝑝⃗ , 𝐷 ⃗ = 𝐶⃗ × 𝑝 ⃗ − 𝑝⃗ (4)

𝑝 ⃗ = 𝑝 ⃗ − 𝐴 ⃗ × 𝐷 ⃗, 𝑝 ⃗ = 𝑝 ⃗ − 𝐴 ⃗ × 𝐷 ⃗, 𝑝 ⃗ = 𝑝 ⃗ − 𝐴 ⃗ × 𝐷 ⃗ (5)

𝐴⃗ = 2𝑎⃗ × 𝑟⃗ − 𝑎⃗, 𝐶⃗ = 2 × 𝑟⃗ (6)

𝑝⃗(𝑡 + 1) = (𝑝⃗ + 𝑝 ⃗ + 𝑝 ⃗)/3 (7)

4. Controlling Strategies
The proposed system is controlled by mainly three controls, which are (1). Incremen-
tal conductance (InC) based MPPT control. (2). DC bus and HESS current control and (3)
VSS-ILMS based VSC control.

4.1. DC Bus Voltage and HESS Current Control

Figure 4 shows 𝑉 control along with HESS current control. The DC bus is con-
trolled by a FOPI controller, which receives its optimized gains from the MOCGWO tech-
nique to minimize the 𝑉 variation during dynamic conditions. The 𝑖 acts as a ref-
erence signal for battery and UC current control or charging/discharging control. The op-
timized 𝑖 will be smoother and free from sharp edges, which reduces the stress on
the HESS, increasing the storage system’s longevity.
Electronics 2022, 11, 567 8 of 23

Figure 4. DC bus voltage and HESS current control.

4.2. Voltage Source Converter Control

Figure 5 shows the VSS-ILMS based VSC control. The proposed control with incre-
mental variable step-size reduces steady-state error and faster convergence speed. With
every iteration, the step-size increases, which can filter a wide variety of noises without
increasing the computational burden. The network approximation with the proposed con-
trol is easy, performing better with the sparse network. This gradient-based VSC control
reduces the mean square error (MSE) out of the load current component. The optimized
DC and AC bus components help VSC control generate the accurate weight components
and reference current signals, which further improves the VSC performance.
For the direct weight components generation (𝑊 , where x = a, b, c), error signals
(𝑒 , where x = a, b, c) for each phase are produced as per (8) as a function of the 𝑖 and
the in-phase component (𝜇 ). The 𝜇 , x = a, b, c, are calculated as per (9), 𝑉 is voltage
magnitude and 𝑣 , where x = a, b, c, are the phase to ground source voltages [15].
𝑒 = 𝑖 (𝑛) − 𝜇 × 𝑊 (𝑛), where x = a, b, c (8)

𝜇 = = , where x = a, b, c
(9)

The 𝑊 , where x = a, b, c are calculated for each phase as a function of the error
signal and incremental step-size (𝜇 (𝑖 + 1)) as per (10) and (11).
𝑊 (𝑛 + 1) = 𝑊 (𝑛) + 𝜇 (𝑖 + 1) × 𝜇 ×𝑒 ; where x = a, b, c (10)
𝜇 (𝑖 + 1) = 𝛼 × 𝜇 (𝑖) + 𝛾 × 𝑒 (𝑖) (11)
The overall direct weight signal (𝑊 ) is calculated as a function of average weight
components (𝑊 ) and feed-forward term (𝑊 ) as per (12)–(15).

𝑊 = = (12)

𝑊 = 𝑊 +𝑊 +𝑊 (13)

𝑊 =𝑊 −𝑊 (14)
𝑖∗ =𝑊 ×𝜇 where x = a, b, c (15)
For the quadrature weight component generation (𝑊 , where x = a, b, c), error signals
(𝑒 , where x = a, b, c) are calculated as per (16). The quadrature-phase components
(𝜇 , where, x = a, b, c) is calculated as a function of in-phase components as per (17).
𝑒 = 𝑖 (𝑛) − 𝜇 × 𝑊 (𝑛), where x = a, b, c (16)
Electronics 2022, 11, 567 9 of 23

√ ∗ √ ∗
𝜇 = ; 𝜇 = + ;𝜇 =− + (17)
√ √ √

The 𝑊 , where x = a, b, c are calculated for each phase as a function of the error signal
and incremental step-size (𝜇 (𝑖 + 1)) as per (18) and (19).
𝑊 (𝑛 + 1) = 𝑊 (𝑛) + 𝜇 (𝑖 + 1) × 𝜇 ×𝑒 ; where x = a, b, c (18)
𝜇 (𝑖 + 1) = 𝛼 × 𝜇 (𝑖) + 𝛾 × 𝑒 (𝑖) (19)
The overall direct weight signal (𝑊 ) is calculated as a function of average weight
components (𝑊 ) and 𝑖 as per (20)–(22).
1
𝑊 = 𝑊 +𝑊 +𝑊 (20)
3
𝑊 =𝑊 −𝑖 (21)
𝑖∗ =𝑊 ×𝜇 where x = a, b, c (22)
∗
The overall reference current (𝑖 , where x = a, b, c) is generated by doing the sum of
∗
𝑖 and 𝑖 ∗ as per (23).
𝑖∗ = 𝑖∗ + 𝑖∗ where x = a, b, c (23)
∗
The 𝑖 , where x = a, b, c are delivered to the hysteresis current controller (HCC) for
VSC switching signal generation.

Figure 5. VSC Control.

Electronics 2022, 11, 567 10 of 23

5. Results and Discussions

The presented system is observed and analyzed under various dynamic conditions,
including three modes of smart PV operation, shallow to deep voltage sag/swell, and load
unbalancing. The following results and discussion validate the performance of the system.

5.1. Steady-State Analysis

Figure 6a–c presents the steady-state analysis of phase ‘a’ of source voltage (𝑣 ),
source current (𝑖 ) and load current (𝑖 ). During steady-state, the solar irradiation level
was kept fixed at 1000 W/m2. The THD level of grid voltage and current were maintained
well within the 5% limit as per IEEE 519 standards. In the proposed system, AC and DC
bus control was provided by PI and FOPI controllers. The PI and FOPI controllers were
supplied with GA and CGWO optimized for better performance. Table 2 shows the grid
THD levels with diverse controllers implemented. The CGWO tuned FOPI controller de-
livered the lowest THD level during steady-state.
g Signal

300 40
Signal

10

iL a (A )
iS a (A )
v S a (V )

0 0 0
-40 -10
-300
0.12 0.13 0.14 0.15 0.16 0.17 0.18 0.19 0.2 0.21 0.1 0.12 0.14 0.16 0.18 0.1 0.12 0.14 0.16 0.18
FFT analysis
Time (s)
FFT analysis
Time (s) FFT analysis
Time (s)
100 100 100
Grid Voltage (V) = 339V, Load Current(A) = 18.16A,
Source Current (A) = 43.36A,
THD= 1.13% at 50Hz THD= 39.62% at 50Hz
v S a(V )

iL a(A )
iS a (A )

50 50 THD= 2.94% at 50Hz 50

0 0 0
0 200 400 600 800 1000 0 200 400 600 800 1000 0 200 400 600 800 1000
Frequency (Hz) Frequency (Hz) Frequency (Hz)
(a) (b) (c)
Figure 6. (a–c). THD analysis of 𝑣 , 𝑖 , and 𝑖 .

Table 2. THD comparison of 𝑣 , 𝑖 , 𝑖 with the PI and FOPI controllers tuned by GA and
CGWO during steady-state.

Without
Quantity FOPI-CGWO FOPI-GA PI-CGWO PI-GA
Optimization
𝑣 1.13% 1.42% 1.33% 1.6% 1.62%
𝑖 2.94% 2.96% 3.12% 3.38% 3.56%
𝑖 39.62% 39.62% 39.62% 39.62% 39.62%

The VSC control of the above-described system was also compared with the other
adaptive controls to prove the better filtering capability of the implemented VSS-ILMS
control. Table 3 shows the THD performance of the grid with LMS, LMF, and VSS-ILMS
control. For this analysis, AC and DC buses were regulated by CGWO tuned FOPI con-
troller.

Table 3. THD comparison of 𝑣 , 𝑖 , 𝑖 with VSS-ILMS, LMF, and LMS-based VSC controls.

Quantity VSS_ILMS LMF LMS

vSa 1.13% 1.6% 1.62%
iSa 2.94% 4.38% 4.71%
iLa 39.61% 39.61% 39.61%
Electronics 2022, 11, 567 11 of 23

5.2. Smart Grid-Tied PV System Performance Analysis

The presented system is termed a smart grid-tied PV system as it can operate in three
different modes and effortlessly shifts from one mode to another.
During mode-1, the grid-tied PV system does not deliver any power to the local load;
in contrast, the grid itself absorbs the full PV power.
During mode-2, the grid and PV system both satisfy the load requirements; if the
load demand is less than PV power generation, the remaining power will be delivered to
the grid. If the irradiation level reduces, the HESS and grid provide the remaining re-
quired power.
During mode-3, in sunshine conditions, the grid-tied PV system operates as a PV-
DSTATCOM, whereas during bad weather conditions or nighttime, the proposed system
shifts to DSTATCOM mode. In DSTATCOM mode, load requirements are fulfilled by the
grid only. When PV power becomes available, the system shifts to PV-DSTATCOM mode
effortlessly without any transients.
The presented system was analyzed under each mode of PV operation. Figure 7
shows the three smart PV operational mode analysis by associated AC quantities, i.e.,
𝑣 ,𝑖 , phase ‘a’ compensator and load current (𝑖 and 𝑖 ), neutral current of source,
compensator, and load (𝑖 , 𝑖 , 𝑖 ) and 𝑉 .
Mode-1 operation was executed from 0.2 s to 0.3 s, where no load was connected to
the system, and whole power was delivered to the grid. The 𝑣 and 𝑖 maintained
the out of phase relation, as power was provided to the grid. The 𝑖 magnitude in-
creased as more power was available without the local load. The 𝑖 and 𝑖 also re-
mained zero in the absence of the load. The compensator neutral current remained in
phase opposition of 𝑖 to maintain the 𝑖 = 0, in this case 𝑖 is already zero, hence 𝑖
is also remained zero as well 𝑖 . The 𝑉 was maintained at the desired level.
v Sabc (V)

300
0
-300
50
(A) i C n (A) i (A) i La (A) i C a (A) i Sabc (A)

0
-50
50
0
-50
20
0
-20
2
0
Sn

-2
20
0
-20
20
0
Ln

-20
V pcc (V) i

360
340
320
0.2 0.3 0.4 0.5 0.6 0.7
Mode1: PV to Grid Mode2: PV and Grid both Time(s) Mode3: PV-DSTATCOM to DSTATCOM
and vice versa
Figure 7. Smart PV mode analysis of 𝑣 ,𝑖 , 𝑖 , 𝑖 , 𝑖 , 𝑖 , 𝑖 , and 𝑉 .
Electronics 2022, 11, 567 12 of 23

Mode-2 operation was executed from 0.4 s to 0.5 s, where the irradiation level was
reduced from 1000 W/m2 to 600 W/m2 to confirm the combined operation of the grid and
PV system. With reduced irradiation level, the power delivered to the grid reduces, hence,
𝑖 also. Though 𝑣 and 𝑖 remained in phase opposition as PV power at the re-
duced insolation level was sufficient for load requirements. The 𝑖 and 𝑖 were in
phase opposition, canceling out each other at three times the fundamental frequency. The
𝑖 was maintained near zero, hence confirming the neutral current compensation.
Mode-3 is executed from 0.6 s to 0.7 s, during which the irradiation level is reduced
to zero. In the absence of PV power system shifts from PV-DSTATCOM to DSTATCOM
operation. At 0.6 s, the phase reversal of 𝑖 was observed, earlier 𝑣 and 𝑖 were
in phase opposition, but after 0.6 s, both came in phase with each other, confirming the
power supply from the grid to load. After 0.7 s, the irradiation level was again increased
to 1000 W/m2, due to which the system shifted from DSTATCOM to PV-DSTATCOM op-
eration. The 𝑉 magnitude was maintained at the desired level with the help of AC bus
voltage control.
Figure 8 shows the behavior of DC quantities during smart PV operations. The PV
voltage, current, and power (𝑉 , 𝐼 and 𝑃 ), battery and UC voltages and currents
(𝑉 , 𝐼 , 𝑉 , 𝐼 ) and DC bus voltage 𝑉 are shown as DC quantities of the system.
During Mode-1: The irradiation level is maintained at 1000 W/m2, where whole
power was delivered to the grid. The battery and UC start charging during mode-1, 𝐼
and 𝐼 became slightly negative and, 𝑉 and 𝑉 started increasing. The 𝑉 was
maintained at the desired of 700 V during PV to grid operation.
During mode-2: the irradiation level was reduced from 1000 W/m2 to 600 W/m2. The
𝑉 was kept at the fixed voltage level of 600 V, whereas 𝐼 and 𝑃 changed with the
reduced irradiation level. During this mode, both PV and the grid manage the load de-
mand. The 𝑉 was maintained at the desired level.
(kW) I (A) V PV (V)

600
0
80
40
PV

0
40
20
PV

0
(A)V Bat(V) I UC (A) V UC (V)P

400
390
380
10
0
-15
400
399
398
50
0
Bat

-50
V dc (V) I

720
700
680
0.2 0.3 0.4 0.5 0.6 0.7
Mode 1: PV to Grid Mode 2: PV and Grid both Time(s) Mode 3: PV-DSTATCOM to DSTATCOM
and vice versa
Figure 8. Smart PV mode analysis of 𝑉 , 𝐼 , 𝑃 , 𝑉 , 𝐼 , 𝑉 ,𝐼 , and 𝑉 .
Electronics 2022, 11, 567 13 of 23

During mode-3, the solar irradiation level reduced from 1000 W/m2 to 0 W/m2 result-
ing in the proposed system shifting from PV-DSTATCOM to DSTATCOM operation. The
𝑉 , 𝐼 , and 𝑃 became zero to simulate the absence of sunshine. The grid itself and
HESS charges managed the load slowly from the grid power. After 0.7 s system shifts from
DSTATCOM to PV-DSTATCOM operation without any significant transient on load and
grid side of the system. The 𝑉 was maintained at the desired level of 700 V, as shown
in Figure 8.
Figure 9 shows the active and reactive power flow of the grid side (𝑃 , 𝑄 ), load
side (𝑃 ,𝑄 ) and compensator side (𝑃 , 𝑄 ) during three modes of smart PV op-
erations. The reactive power delivered by the grid (𝑄 ) remained almost zero through-
out the smart PV operation as VSC gives the necessary reactive power to load, keeping
the reactive power exchange with grid minimum. The grid active power exchange (𝑃 )
changed with the mode of operation. During PV to grid mode, 𝑃 became negative as
the grid absorbed the active power. The load demand of active and reactive power became
zero as the load was disconnected during this mode. The compensator was converting DC
power to AC power and delivering to the grid, so the 𝑃 , 𝑄 was the same as
𝑃 ,𝑄 during mode-1.
(kW)

20
PGrid
Grid

0 Q Grid
/Q

-20
Grid

-40
P
(kW)

8
6
Load

4
/Q

2 PLoad
Load

0 Q Grid
P
(kW)

40
VSC

20
/Q

0
VSC

PVSC Q VSC
-20
P

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Mode 2: PV and Mode 3: PV-DSTATCOM to Load Unbalancing
Mode 1: PV to grid mode
grid both mode
Time(s)
DSTATCOM and vice versa mode
Figure 9. Smart PV mode analysis of 𝑃 ,𝑄 ,𝑃 ,𝑄 ,𝑃 , and 𝑄 .

During mode two operation, the 𝑃 ,𝑄 remained fixed. The 𝑃 changes

with the change in irradiation level and the same power were provided to the grid as
𝑃 .
During mode three operation, the PV power became zero in the DSTATCOM opera-
tion; the load demand remained fixed during this mode also. The 𝑃 became negative
to support the changing of HESS and 𝑃 was delivered to the load.
Figure 10 shows the power factor (PF) variations during smart PV operation. The
VSC control operated in zero voltage regulation mode, where it aimed to manage voltage
Electronics 2022, 11, 567 14 of 23

at PCC and compensate reactive power. The load PF remained around 0.8 p.u. during
smart PV operation and load unbalancing, which changed slightly with the connection
and disconnection of load during PV to grid mode. Both VSC and grid PFs have the same
behavior. During PV-DSTATCOM mode, the grid PF reached –1 due to the reversal of
flow of power, and the earlier grid was absorbing, and then the grid was supplying.

0.5
Power Factor

Power Factor during

Power Factor during PV and grid both mode VSC PF
PV to grid mode Load PF
0 Grid PF

Power Factor during

-0.5 PV-DSTATCOM to DSTATCOM
and vice-versa mode

-1
0 0.2 0.4 0.6 0.8 1.0
Time(s)
Figure 10. Power factor variations analysis during smart PV mode.

5.3. Load Unbalanced Analysis

Figure 11 shows the load unbalancing analysis created by disconnecting the phase ‘a’
of the load from the system. During load unbalancing, 𝑖 became zero, 𝑖 magnitude
increased slightly as only two phases of load, around 66%, needed to be managed, and
the remaining was delivered to the grid. The 𝑖 and 𝑖 cancelled each other at twice
the fundamental frequency, hence maintaining the 𝑖 = 0. The 𝑉 was kept at 339 V.
V p c c (V ) i L n (A ) i C n (A ) i S n (A ) i L a (A ) i C a (A ) i S a b c (A ) v S a b c (V )

300
0
-300
50
0
-50
50
0
-50
20
0
-20
10
0
-10
20
0
-20
20
0
-20
360
340
320
0.75 0.8 0.85 0.9 0.95
Time(s)
Electronics 2022, 11, 567 15 of 23

Figure 11. Load unbalancing analysis of 𝑣 ,𝑖 , 𝑖 , 𝑖 , 𝑖 , 𝑖 , 𝑖 , and V .

On the DC side, 𝑉 , 𝐼 and 𝑃 showed no change and kept following their refer-
ences as irradiation level was kept fixed at 1000 W/m2. The 𝐼 and 𝐼 charging rate
increased slightly. The DC bus stability was preserved, and 𝑉 was maintained at 700 V
while following the 𝑉 ∗ . As shown in Figure 12. The PF and grid, load, and VSC power
are demonstrated in Figures 9 and 10.
V P V (V )

600
300
0
60
P V (k W ) I P V (A )

50
40
35

25
V U C (V ) P

400
395
390
10
U C (A )

0
V d c (V ) I B a t (A )V B a t (V ) I

400
399
398
50
0
-50
720
700
680
0.75 0.8 0.85 0.9 0.95
Time(s)
Figure 12. Load unbalancing analysis of 𝑉 , 𝐼 , 𝑃 , 𝑉 , 𝐼 , 𝑉 ,𝐼 , and 𝑉 .

5.4. Abnormal Grid Voltage Analysis

The grid voltage sag and swell are created by varying the three-phase voltage mag-
nitude. The system was analyzed from shallow sag/swell of 20% to deep sag/swell of 60%.
Figure 13 shows the performance of AC quantities during abnormal grid voltages.
The grid voltage sag/swell of 20% was created from 1.0 to 1.2 s, the 𝑖 remained
sinusoidal, and its magnitude increased with voltage sag and reduced with the swell. The
magnitude of 𝑖 , 𝑖 , 𝑖 , 𝑖 also changed with the variation in grid voltage. The 𝑉
waveform also confirms the occurred sag/swell.
During 40% of voltage sag/swell, the variations in currents magnitude are higher, but
the 𝑖 maintained its sinusoidal nature. The 𝑖 was maintained at zero, confirming
the neutral current compensation.
During 60% grid voltage sag swell, the magnitude of various currents signals rose
while maintaining the system’s stability. During deep voltage sag/swell, PV remained
connected to the system with low ride-through capability. The VSC control retained the
sinusoidal nature of source current during these weak grid conditions, vouching for the
efficient control algorithm.
Figure 14 shows the performance of DC quantities during abnormal grid voltages.
The 𝑉 , 𝐼 , 𝑃 remained almost the same during voltage sag/swell conditions, other
than swell variations in deep voltage sag/swell conditions. During the transition from sag
to swell mode, battery and UC currents show significant variations but quickly adjusted
to new conditions. The battery system was supported with the current control mechanism,
Electronics 2022, 11, 567 16 of 23

whereas UC does not require high current protection, so the variation in 𝐼 was higher
than 𝐼 . The DC bus is crucial for the system’s stability; during deep voltage sag/swell
the 𝑉 variations were not very significant. The FOPI-based 𝑉 control along with VSC
control stabilized the DC bus quickly and did not allow a huge deviation from the refer-
ence point during extremely weak grid conditions.
Figure 15 shows the grid, load, and VSC active and reactive power performance dur-
ing abnormal grid voltage conditions (extreme weak grid conditions). With grid voltage
abnormalities, load active and reactive power requirements increased sharply. During
deep voltage, sag/swells reactive power demand increases rapidly, collectively managed
by the grid and VSC to maintain the healthy operation.
Figure 16 shows the PF variations of load, grid, and VSC side. The load weak grid PF
remained the same throughout the operation. With voltage sag and swell conditions, the
PF adjusted itself to new values depending on the phase variation with source currents.
The VSC PF complements the grid PF, but during deep voltage, sag/swell conditions PF
decreased to a shallow level confirming the growing demand for reactive power during
extreme weak grid conditions.
(A ) i C n (A ) i S n (A ) i L a (A ) i C a (A ) S abc v S abc (V )

300
0
-300
200
(A )

0
-200
i

200
0
-200
20
0
-20
10
0
-10
20
0
-20
20
0
Ln

-20
V p cc (V ) i

600
400
200
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7
20% Voltage Sag/Swell Time(s) 40% Voltage Sag/Swell 60% Voltage Sag/Swell

Figure 13. Abnormal grid voltage analysis of 𝑣 ,𝑖 , 𝑖 , 𝑖 , 𝑖 , 𝑖 , 𝑖 , and 𝑉 .

Electronics 2022, 11, 567 17 of 23

700
(A) V PV (V)

600
500
60
PV

50
(kW) I

35
PV

25
(A)V Bat(V) I UC (A) V UC (V) P

450
400
360
150
0
-150
400
399
398
50
0
Bat

-50
V dc (V) I

750
700
650
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7
20% Voltage Sag/Swell 40% Voltage Sag/Swell Time(s) 60% Voltage Sag/Swell

Figure 14. Abnormal grid voltage analysis of 𝑉 , 𝐼 , 𝑃 , 𝑉 , 𝐼 , 𝑉 ,𝐼 , and 𝑉 .

/ Q Load (kW) P Grid / Q Grid (kW)

20
PGrid
10 Q Grid

-10

15
10
PLoad
5
Q Grid
Load

0
P VSC / Q VSC (kW) P

10

0
PVSC
-10
Q VSC
-20
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9
40% Voltage Time(s) 60% Voltage Sag/Swell
20% Voltage Sag/Swell
Sag/Swell
Figure 15. Abnormal grid voltage analysis of 𝑃 ,𝑄 ,𝑃 ,𝑄 ,𝑃 , and 𝑄 .
Electronics 2022, 11, 567 18 of 23

0.5
Power Factor

Power factor during

0 20% Voltage Sag/Swell

VSC PF Power Factor during Power Factor during

40% Voltage Sag/Swell 60% Voltage Sag/Swell
Load PF
-0.5 Grid PF

-1
1.0 1.2 1.4 1.6 1.8 1.9
Time(s)
Figure 16. Power factor variations analysis during abnormal grid voltage conditions.

5.5. DC and AC Bus Stability Analysis

The DC bus stability is crucial for the stability of the grid-tied PV system. As DC bus
controls the flow of power from PV and HESS. The grid voltage itself assists the AC bus
voltage control in the grid following converters. The optimized DC and AC bus generate
an accurate direct and quadrature loss component of currents (𝑖 &𝑖 ). These loss
components help the VSC control generate accurate weights and reference current signals.
The VSC control with optimized AC and DC busses performs better in harmonics elimi-
nation and overall losses reduction.
Figure 17 shows the 𝑉 variations during diverse, dynamic conditions considered.
The CGWO tuned FOPI controller performed effectively in reducing the 𝑉 variations.
With the implemented DC bus control, the presented system maintains its stability during
extreme weak grid conditions, i.e., PV-DSTATCOM to STATCOM and vice-versa, in deep
voltage sag swell conditions.
Figure 18 shows the variations of 𝑖 during induced dynamic conditions. In ideal
conditions 𝑖 should be zero, with the help of CGWO tuned FOPI controller, pre-
sented system minimizes its variations. Even during the sag/swell conditions with FOPI
control the 𝑖 is maintained near-zero without significant sharp edges.
The AC bus voltage control in a grid-following converter is partially controlled by
the grid voltage also. The AC bus voltage magnitude control with CGWO and GA tuned
PI, and FOPI controller is observed. All of the AC bus controls maintain the 𝑉 at the
desired level, and CGWO based FOPI control maintains the 𝑉 at the desired level as
well maintains the 𝑖 component near zero. Table 4 presents the analytical perfor-
mance analysis of 𝑉 with CGWO and GA optimized FOPI and PI controllers.
Electronics 2022, 11, 567 19 of 23

V without
800 V PI-GA
dc
dc V PI-CGWO
dc
V FOPI-GA
dc
V FOPI-CGWO
dc V*
dc
optimization
780 710
705
710
700 700
760 700 690
695
0.6 0.7 0.8 0.9 1.0 1.1
740 Load
20% voltage
Mode3: PV-STATCOM Unbalancing sag/swell
V dc(V)

720 to STATCOM

700
680 Mode1: PV to Grid Mode2: PV and
Grid both
660 705
705

700 700 60% Voltage

640 40% Voltage
sag/swell sag/swell
695
0.2 0.3 695
620 0.4 0.5
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8
Time(s)
Figure 17. 𝑉 variations with GA, CGWO optimized PI and FOPI controller.

idLoss PI-CGWO idLoss PI-GA i dLosswithout optimization idLoss FOPI-CGWO idLossFOPI-GA

100 IdLoss during
4 0
2 20% Voltage sag/swell
0 -10
50 -2
-20
0.2 0.3 0.2 0.3

0
(A)

IdLoss during IdLoss during

-50 PV mode 1
I
PV mode 2 IdLoss during dLoss
during
dLoss

PV mode 3 Unbalanced
10 Loading
-100
i

5
4 IdLoss during
0 2
40% Voltage sag/swell
-150 -5 0
0.2 0.3 -2
0.2 0.3
IdLoss during
-200 60% Voltage sag/swell

0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8

Time(s)
Figure 18. 𝑖 variations with GA, CGWO optimized PI and FOPI controller.
Electronics 2022, 11, 567 20 of 23

Table 4. DC bus performance analysis.

Without
Parameters CGWO-FOPI GA-FOPI CGWO-PI GA-PI
Optimization
Steady-state error 0.57% 1.26% 1% 1.14% 1.71%
Convergence speed 5 ms 7 ms 7.8 ms 11 ms 16 ms
Dynamic state transients (PV to grid mode) 1.12% 1.87% 2.12% 2.18% 2.42%
Dynamic state transients (Irradiation change) 0.10% 0.17% 0.14% 1.12% 1.12%
Dynamic state transients (PV-DSTATCOM-STATCOM) 0.13% 1.14% 1.29% 1.71% 2%
Dynamic state transients (Unbalanced load) 0.29% 0.43% 0.57% 0.85% 0.13%
Dynamic state transients (Voltage sag/swell 20%) 1.71% 2% 2.57% 3.14% 3.57%
Dynamic state transients (Voltage sag/swell 40%) 4.98% 5.42% 8.29% 8.56% 10.20%
Dynamic state transients (Voltage sag/swell 60%) 6.68% 6.56% 7.62% 9.52% 11.45%

5.6. THD Performance Analysis during Dynamic Conditions

Table 5 describes the THD analysis of 𝑣 , 𝑖 , and 𝑖 during diverse conditions.
The table also presents the THD comparison of CGWO-FOPI control with other imple-
mented controls for AC and DC bus voltage control.

Table 5. THD performance analysis during dynamic conditions.

Without
Quantity FOPI-CGWO FOPI-GA PI-CGWO PI-GA
Optimization
Mode1: PV to Grid mode
𝑣 1.35% 1.44% 1.49% 1.66% 1.66%
𝑖 1.94% 2.20% 2.12% 2.38% 2.53%
𝑖 39.62% 39.62% 39.62% 39.62% 39.62%
Mode2: PV and Grid mode during Irradiation Change from 1000 W/m2 to 600 W/m2
𝑣 0.97% 1.22% 1.35% 1.62% 1.65%
𝑖 6.23% 6.51% 6.70% 6.99% 7.22%
𝑖 39.62% 39.62% 39.62% 39.62% 39.62%
Mode3: PV-DSTATCOM to DSTATCOM and vice versa mode
𝑣 0.97% 1.22% 1.35% 1.62% 1.62%
𝑖 6.23% 6.22% 6.25% 6.37% 7.29%
𝑖 39.62% 39.62% 39.62% 39.62% 39.62%
Load Unbalancing
𝑣 1.2% 1.49% 1.35% 1.62% 1.62%
𝑖 4.01% 5.32% 4.80% 4.91% 5.10%
𝑖 39.62% 39.62% 39.62% 39.62% 39.62%
Voltage Sag of 0.8 p.u.
𝑣 1.61% 1.62% 1.61% 1.62% 1.62%
𝑖 0.83% 0.92% 1.12% 1.72% 2.12%
𝑖 39.62% 39.62% 39.62% 39.62% 39.62%
Voltage Swell 1.2 p.u.
𝑣 1.61% 1.61% 1.61% 1.62% 1.62%
𝑖 1.28% 2.16% 2.76% 3.15% 4.29%
𝑖 39.62% 39.62% 39.62% 39.62% 39.62%
Voltage Sag of 0.6 p.u.
𝑣 1.61% 1.62% 1.61% 1.62% 1.62%
𝑖 0.56% 0.77% 1.78% 3.36% 4.17%
𝑖 39.62% 39.62% 39.62% 39.62% 39.62%
Electronics 2022, 11, 567 21 of 23

Voltage Swell 1.4 p.u.

𝑣 1.61% 1.61% 1.61% 1.62% 1.62%
𝑖 0.84% 1.46% 1.96% 2.39% 3.69%
𝑖 39.62% 39.62% 39.62% 39.62% 39.62%
Voltage Sag of 0.4 p.u.
𝑣 1.61% 1.62% 1.61% 1.62% 1.62%
𝑖 1.74% 2.77% 2.65% 5.86% 4.95%
𝑖 39.62% 39.62% 39.62% 39.62% 39.62%
Voltage Swell 1.6 p.u.
𝑣 1.61% 1.61% 1.61% 1.62% 1.62%
𝑖 0.82% 1.5% 1.95% 1.95% 3.63%
𝑖 39.62% 39.62% 39.62% 39.62% 39.62%

6. Conclusions
This paper describes a FOPI VSS-ILMS based VSC control of a three-phase four-wire
grid-tied PV-HESS system. The DC and AC buses are regulated by the chaotic grey wolf
optimized FOPI controller to reduce 𝑉 and 𝑉 voltage variation during induced dy-
namic conditions. The DC bus controls the active power flow from PV and HESS to the
grid and load, along with the flow of power, which makes the DC bus control crucial for
maintaining the overall system’s stability during steady and dynamic states. The opti-
mized DC bus eliminates the second-harmonics component on the DC coupling capacitor,
resulting in enhanced power quality. On the other hand, in the case of grid-following in-
verters, the AC bus voltage is partially controlled by the grid itself. Implementing differ-
ent controls for the AC bus regulation will have a minute impact on overall system per-
formance. The AC bus control helps to achieve desired reactive power flow while operat-
ing under zero voltage regulation mode. Both optimized AC and DC buses produce direct
and quadrature loss components of currents, which further enhance the VSC control per-
formance by generating accurate weight signals and reference current signals. The VSC
control suppresses the dc offset at PCC, which improves the system’s performance. The
VSS-ILMS algorithm controls the VSC, which increases the step size in each iteration to
accommodate a wide variety of noise signals to be passed through the filtering window.
The network approximation with proposed gradient-based adaptive control is easy, per-
forming better with the sparse network. The proposed control provides VSC with the ca-
pability to perform a multi-functional operation while ensuring the system’s stability. The
proposed system performs active and reactive power control, neutral current compensa-
tion, load balancing, harmonics elimination, etc., during the diverse, dynamic, and ex-
tremely weak grid conditions. Due to the offset rejection capability of the VSC and opti-
mized DC and AC buses, the system remains stable up to 60% voltage sag/swell condi-
tions. The THD of the attached distribution grid voltage is merely 1.13% and the current
is 2.94% during steady-state. Even during dynamic conditions, the THD level is well main-
tained with 5%, even during dynamic conditions, i.e., PV to grid mode, load unbalancing
and deep voltage sag/swell conditions. The CGWO tuned DC bus minimize the deviation
of 𝑉 from its reference. The steady-state deviation of 𝑉 is hardly 0.57% (≅3 V), how-
ever, during 60% voltage sag/swell condition, it varies maximum up to 6.68% (≅45 V),
which surely reduces the stress on DC coupling capacitor and VSC switches. The overall
performance of the proposed system is found satisfactory as per the IEEE519 standard.

Author Contributions: Conceptualization, M.C., I.H., A.A., H.M. and M.A.A.; methodology, M.C.,
I.H., A.A., H.M. and M.A.A.; software, M.C., I.H., A.A., H.M., and M.A.A.; validation, M.C.,
I.H., A.A., H.M., and M.A.A.; formal analysis, M.C., I.H., A.A., H.M. and M.A.A.; investiga-
tion, M.C., I.H., A.A., H.M. and M.A.A.; resources, M.C., I.H., A.A., H.M., and M.A.A.; data
curation, M.C., I.H., A.A., H.M., and M.A.A.; writing—original draft preparation, M.C., I.H.,
A.A., H.M. and M.A.A.; writing—review and editing, M.C., I.H., A.A., H.M., and M.A.A.;
Electronics 2022, 11, 567 22 of 23

visualization, M.C., I.H., A.A., H.M., and M.A.A.; supervision, M.A.A.; project administration,
H.M. and M.A.A.; funding acquisition, H.M. and M.A.A. All authors have read and agreed to the
published version of the manuscript.
Funding: The authors extend their appreciation to the Researchers Supporting Project at King Saud
University, Riyadh, Saudi Arabia, for funding this research work through the project number RSP-
2021/278.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Acknowledgments: The authors would like to acknowledge the support from King Saud Univer-
sity, Saudi Arabia. The authors would like to acknowledge the support from Intelligent Prognostic
Private Limited Delhi, India Researcher’s Supporting Project (XX-02/2021).
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Das, B.K.; Alotaibi, M.A.; Das, P.; Islam, M.S.; Das, S.K.; Hossain, M.A. Feasibility and Techno-Economic Analysis of Stand-
Alone and Grid-Connected PV/Wind/Diesel/Batt Hybrid Energy System: A Case Study. Energy Strategy Rev. 2021, 37, 100673.
https://doi.org/10.1016/j.esr.2021.100673.
2. Shah, F.A.; Khan, I.A.; Khan, H.A. Performance Evaluation of Two Similar 100 MW Solar PV Plants Located in Environmentally
Homogeneous Conditions. IEEE Access 2019, 7, 161697–161707. https://doi.org/10.1109/ACCESS.2019.2951688.
3. Rehman, S.; Ahmed, M.A.; Mohamed, M.H.; Al-Sulaiman, F.A. Feasibility Study of the Grid Connected 10MW Installed
Capacity PV Power Plants in Saudi Arabia. Renew. Sustain. Energy Rev. 2017, 80, 319–329.
https://doi.org/10.1016/j.rser.2017.05.218.
4. IRENA. Renewable Power Generation Costs in 2020; International Renewable Energy Agency: Abu Dhabi, United Arab Emirates,
2021. ISBN 978-92-9260-348-9 (accessed on 10 January 2022).
5. Gandhi, O.; Kumar, D.S.; Rodríguez-Gallegos, C.D.; Srinivasan, D. Review of Power System Impacts at High PV Penetration
Part I: Factors Limiting PV Penetration. Sol. Energy 2020, 210, 1–21. https://doi.org/10.1016/j.solener.2020.06.097.
6. Kumar, D.S.; Gandhi, O.; Rodríguez-Gallegos, C.D.; Srinivasan, D. Review of Power System Impacts at High PV Penetration
Part II: Potential Solutions and the Way Forward. Sol. Energy 2020, 210, 202–221. https://doi.org/10.1016/j.solener.2020.08.047.
7. Karami, N.; Moubayed, N.; Outbib, R. General Review and Classification of Different MPPT Techniques. Renew. Sustain. Energy
Rev. 2017, 68, 1–18. https://doi.org/10.1016/j.rser.2016.09.132.
8. Villegas-Mier, C.G.; Rodriguez-Resendiz, J.; Álvarez-Alvarado, J.M.; Rodriguez-Resendiz, H.; Herrera-Navarro, A.M.;
Rodríguez-Abreo, O. Artificial Neural Networks in Mppt Algorithms for Optimization of Photovoltaic Power Systems: A
Review. Micromachines 2021, 12, 1–19. https://doi.org/10.3390/mi12101260.
9. Estévez-Bén, A.A.; Alvarez-Diazcomas, A.; Rodríguez-Reséndiz, J. Transformerless Multilevel Voltage-Source Inverter
Topology Comparative Study for PV Systems. Energies 2020, 13, 1–26. https://doi.org/10.3390/en13123261.
10. López, H.; Rodríguez-Reséndiz, J.; Guo, X.; Vázquez, N.; Carrillo-Serrano, R.V. Transformerless Common-Mode Current-Source
Inverter Grid-Connected for PV Applications. IEEE Access 2018, 6, 62944–62953. https://doi.org/10.1109/ACCESS.2018.2873504.
11. Singh, B.; Chandra, A.; Al-haddad, K. Power Quality Problems and Mitigation Techniques, 1st ed; John Wiley Online Library.; West
Sussex,UK, 2015; ISBN 9781118922057.
12. Sevilla-Camacho, P.Y.; Zuniga-Reyes, M.A.; Robles-Ocampo, J.B.; Castillo-Palomera, R.; Muniz, J.; Rodriguez-Resendiz, J. A
Novel Fault Detection and Location Method for PV Arrays Based on Frequency Analysis. IEEE Access 2019, 7, 72050–72061.
https://doi.org/10.1109/ACCESS.2019.2920053.
13. Mishra, S.; Ray, P.K. Power Quality Improvement Using Photovoltaic Fed DSTATCOM Based on JAYA Optimization. IEEE
Trans. Sustain. Energy 2016, 7, 1672–1680. https://doi.org/10.1109/TSTE.2016.2570256.
14. Chankaya, M.; Hussain, I.; Ahmad, A.; Malik, H.; García Márquez, F.P.; Márquez, F.P.G. Generalized Normal Distribution
Algorithm-Based Control of 3-Phase 4-Wire Grid-Tied PV-Hybrid Energy Storage System. Energies 2021, 14, 4355.
https://doi.org/10.3390/en14144355.
15. Mosalam, H.A.; Amer, R.A.; Morsy, G.A. Fuzzy Logic Control for a Grid-Connected PV Array through Z-Source-Inverter Using
Maximum Constant Boost Control Method. Ain Shams Eng. J. 2018, 9, 2931–2941. https://doi.org/10.1016/j.asej.2018.10.001.
16. Agarwal, R.K.; Hussain, I.; Singh, B. Application of LMS-Based NN Structure for Power Quality Enhancement in a Distribution
Network under Abnormal Conditions. IEEE Trans. Neural Netw. Learn. Syst. 2018, 29, 1598–1607.
https://doi.org/10.1109/TNNLS.2017.2677961.
17. Agarwal, R.K.; Hussain, I.; Singh, B. LMF-Based Control Algorithm for Single Stage Three-Phase Grid Integrated Solar PV
System. IEEE Trans. Sustain. Energy 2016, 7, 1379–1387. https://doi.org/10.1109/TSTE.2016.2553181.
Electronics 2022, 11, 567 23 of 23

18. Zheng, Z.; Zhang, T.; Xue, J. Application of Fuzzy Control in a Photovoltaic Grid-Connected Inverter. J. Electr. Comput. Eng.
2018, 2018. https://doi.org/10.1155/2018/3806372.
19. Bhattacharyya, S.; Singh, B.; Chandra, A.; Al-Haddad, K. Artificial Neural Network Based Advanced Current Control for Grid-
Tied Photovoltaic System. In Proceedings of the 2019 National Power Electronics Conference, NPEC, Tiruchirappalli, India, 13–
15 December 2019. https://doi.org/10.1109/NPEC47332.2019.9034761.
20. Chankaya, M.; Hussain, I.; Ahmad, A.; Khan, I.; Muyeen, S.M. Nyström Minimum Kernel Risk-Sensitive Loss Based Seamless
Control of Grid-Tied PV-Hybrid Energy Storage System. Energies 2021, 14, 1365. https://doi.org/10.3390/en14051365.
21. Alturki, F.A.; Omotoso, H.O.; Al-Shamma’a, A.A.; Farh, H.M.H.; Alsharabi, K. Novel Manta Rays Foraging Optimization
Algorithm Based Optimal Control for Grid-Connected PV Energy System. IEEE Access 2020, 8, 187276–187290.
https://doi.org/10.1109/access.2020.3030874.
22. Tayebi, S.M.; Hu, H.; Batarseh, I. Advanced DC-Link Voltage Regulation and Capacitor Optimization for Three-Phase
Microinverters. IEEE Trans. Ind. Electron. 2019, 66, 307–317. https://doi.org/10.1109/TIE.2018.2823700.
23. Kohli, M.; Arora, S. Chaotic Grey Wolf Optimization Algorithm for Constrained Optimization Problems. J. Comput. Des. Eng.
2018, 5, 458–472. https://doi.org/10.1016/j.jcde.2017.02.005.
24. Thangam, T.; Muthuvel, K. Hybrid Algorithm Based PFoPID Control Design of a Grid-Connected PV Inverter for MPPT. In
Proceedings of the 2nd International Conference on Inventive Research in Computing Applications, ICIRCA, Coimbatore, India,
15–17 July 2020; pp. 992–998. https://doi.org/10.1109/ICIRCA48905.2020.9182806.
25. Lakshmi, M.; Hemamalini, S. Decoupled Control of Grid Connected Photovoltaic System Using Fractional Order Controller.
Ain Shams Eng. J. 2018, 9, 927–937. https://doi.org/10.1016/j.asej.2016.06.002.
26. Nasir, A.; Rasool, I.; Sibtain, D.; Kamran, R. Adaptive Fractional Order PID Controller Based MPPT for PV Connected Grid
System Under Changing Weather Conditions. J. Electr. Eng. Technol. 2021, 16, 2599–2610. https://doi.org/10.1007/s42835-021-
00782-w.
27. Rezkallah, M.; Chandra, A.; Singh, B.; Singh, S. Microgrid: Configurations, Control and Applications. IEEE Trans. Smart Grid
2019, 10, 1290–1302. https://doi.org/10.1109/TSG.2017.2762349.
28. Singh, B.; Solanki, J. A Comparison of Control Algorithms for DSTATCOM. IEEE Trans. Ind. Electron. 2009, 56, 2738–2745.
https://doi.org/10.1109/TIE.2009.2021596.
29. Mus-Ab, A.; Mishra, M.K. Wavelet Transform Based Algorithms for Load Compensation Using DSTATCOM. In Proceedings of
the 2017 IEEE PES Asia-Pacific Power and Energy Engineering Conference (APPEEC), Bangalore, India, 8–10 November 2017.
30. Kumar, R.; Singh, B.; Kumar, R.; Marwaha, S. Recognition of Underlying Causes of Power Quality Disturbances Using Stockwell
Transform. IEEE Trans. Instrum. Meas. 2020, 69, 2798–2807. https://doi.org/10.1109/TIM.2019.2926876.
31. Pradhan, S.; Hussain, I.; Singh, B.; Panigrahi, B.K. Modified VSS-LMS-Based Adaptive Control for Improving the Performance
of a Single-Stage PV-Integrated Grid System. IET Sci. Meas. Technol. 2017, 11, 388–399. https://doi.org/10.1049/iet-smt.2016.0351.
32. Hussain, I.; Agarwal, R.K.; Singh, B. Delayed LMS Based Adaptive Control of PV-DSTATCOM System. In Proceedings of the
2020 3rd International Conference on Energy, Power and Environment: Towards Clean Energy Technologies, Shillong, India,
5–7 March 2021; pp. 1–6.
33. Seema, S.S.; Singh, B.; Panigrahi, B.K. Median LMS Control Approach for SPV Grid Integrated Distribution System. In
Proceedings of the 2021 International Conference on Sustainable Energy and Future Electric Transportation (SEFET),
Hyderabad, India, 21–23 January 2021; pp. 1–5.
34. Kumar, A.; Seema; Singh, B.; Jain, R. Double Stage Grid-Tied Solar PV System Using HC-LMS Control. In Proceedings of the
PIICON 2020-9th IEEE Power India International Conference, Sonepat, India, 28 February–1 March 2020; pp. 3–8.
https://doi.org/10.1109/PIICON49524.2020.9112906.
35. Srinivas, M.; Hussain, I.; Singh, B. Combined LMS-LMF-Based Control Algorithm of DSTATCOM for Power Quality
Enhancement in Distribution System. IEEE Trans. Ind. Electron. 2016, 63, 4160–4169. https://doi.org/10.1109/TIE.2016.2532278.
36. Saeed, M.O.B.; Pasha, S.A.; Zerguine, A. A Variable Step-Size Incremental LMS Solution for Low SNR Applications. Signal
Processing 2021, 178, 107730. https://doi.org/10.1016/j.sigpro.2020.107730.
37. Malik, H.; Iqbal, A.; Yadav, A.K. Soft computing in condition monitoring and diagnostics of electrical and mechanical systems.
In Novel Methods for Condition Monitoring and Diagnostics; Springer: Berlin/Heidelberg, Germany, 2020; Volume 1096, p. 499.
https://doi.org/10.1007/978-981-15-1532-3.
38. Malik, H.; Iqbal, A.; Joshi, P.; Agrawal, S.; Bakhsh, F.I. (Eds.) Metaheuristic and Evolutionary Computation: Algorithms and
Applications; Springer Nature: Singapore, 2021; Volume 916, p. 820. https://doi.org/10.1007/978-981-15-7571-6.
39. Malik, H.; Fatema, N.; Alzubi, J.A. AI and Machine Learning Paradigms for Health Monitoring System: Intelligent Data Analytics;
Springer Nature: Singapore, 2021; Volume 86, p. 513. https://doi.org/10.1007/978-981-33-4412-9.
40. Iqbal, A.; Malik, H.; Riyaz, A.; Abdellah, K.; Bayhan, S. Renewable Power for Sustainable Growth; Springer Nature: Singapore, 2021;
Volume 723, p. 805. https://doi.org/10.1007/978-981-33-4080-0.
41. Tomar, A.; Malik, H.; Kumar, P.; Iqbal, A. Machine Learning, Advances in Computing, Renewable Energy and Communication;
Springer Nature: Singapore, 2022; Volume 768, p. 659. https://doi.org/10.1007/978-981-16-2354-7.