Shunt Active Parallel Filter, Grid Photovoltaic System

ACHWAK ALAZRAG, MOHAMED HAJJEJ, LASSAD SBIT1

Process Laboratory, Energetic, Environment and Electrical System (PEESE),

National Engineering School of Gabes (ENIG), University of Gabes,

TUNISIA

Abstract: - This paper introduces a study focused on managing a photovoltaic system connected to the electrical

grid. The primary components of this system include solar arrays linked via a DC bus to an inverter situated on

Power Filter (SAPF) with a three-phase four-wire configuration, consisting of four legs and adopting a split

capacitor topology. Furthermore, the paper includes an exploration of the instantaneous power theory and the

utilization of hysteresis block control for the SAPF. The findings of this study are demonstrated and analyzed

using Matlab/Simulink software.

Key-Words: - Photovoltaic system, Boost converter, Bidirectional DC-DC converters, MPPT, SAPF, PWM.

Received: April 4, 2023. Revised: October 21, 2023. Accepted: November 12, 2023. Published: December 31, 2023.

1 Introduction

However, a significant drawback associated with

photovoltaic power generation is its susceptibility to

external factors, particularly fluctuations in solar

radiation and temperature. This variability leads to

an inherent instability in the generated power

output. To circumvent this challenge and unlock the

full potential of photovoltaic systems, the

imperative to accurately track and maintain the

maximum power point (MPP) of the photovoltaic

generator becomes evident. At the core of this

to devise methodologies for continuously

identifying and maintaining this optimal MPP.

In the pursuit of achieving this objective,

various algorithms and techniques have been

developed for what is known as Maximum

PowerPoint Tracking (MPPT). These algorithms

seek to dynamically adjust the operational

parameters of the photovoltaic system, such as the

duty cycle of a DC/DC converter or other control

variables, to constantly align the system’s operating

exploration of a grid-connected photovoltaic system.

The central focus of this study is the integration of

MPPT control mechanisms into the inverter control

system. By seamlessly incorporating MPPT

algorithms within the inverter’s operation, the study

aims to ensure that the active and reactive power

levels injected into the grid are optimized in

response to changing environmental conditions.

This dynamic control mechanism ultimately

enhances the energy-harvesting efficiency of the

particularly in terms of variable switching

frequency, [1]. To mitigate these limitations and

refine the control mechanism, the paper introduces a

modulated hysteresis control strategy, [2]. This

innovation involves overlaying a triangular signal

presence of high-order harmonic currents within the

load, enhancing power quality and mitigating

potential disruptions caused by harmonic

distortions. The widespread use of nonlinear loads

in industrial, commercial, and residential systems

has led to significant power quality issues in

contemporary power distribution systems. Among

these issues, harmonic currents and reactive power

stand out as major concerns. Shunt Active Power

Filters (SAPFs) have emerged as the most

loads and the electrical grid. The block diagram of

the SAPF is depicted in Figure 1, [1]. Unlike

passive power filters (PPFs), which mainly consist

of inductive and capacitive elements (L-C) and

provide fixed compensation, SAPFs offer dynamic

and adaptable solutions. They can actively respond

to power quality problems, delivering precise

control while minimizing the resonance effects often

associated with passive filters. SAPFs also offer

advantages in terms of sizing and flexibility, making

nonlinear loads, [4]. Various SAPF topologies have

emerged to address these challenges, including the

four-leg (4L) configuration, [5], as mentioned in

Figure 2, the split capacitor or two capacitors (2C)

approach, [6], presented in Figure 3, and the three

H-bridges (3-HB) topology (Figure 4), with each H-

bridge consisting of four switches arranged in an H

shape, [7].

In the split capacitor topology, the neutral wire

is positioned between two capacitors, necessitating

fourth leg (neutral wire) to balance the neutral

current, resulting in superior performance compared

to the 2C topology. This configuration has been

extensively explored in various research studies, [8],

optimizing and creating various parallel active filter

topologies, [11], others have delved into perfecting

the control aspect of these filters, which represents

their “heart.” The filtering efficiency is closely tied

to the effectiveness of the reference current

extraction algorithm. These methodologies

generally fall into two main categories: frequency

domain methods and time domain methods.

In the frequency domain approach, techniques

like the fast Fourier transform (FFT) are utilized to

due to the substantial computational load imposed

by FFT calculations, necessitating complex and

costly systems to operate the filter in real time.

This has led other researchers to focus on time

domain methods, which include the p-q theory, [14],

instantaneous reactive power theory, [15],

synchronous reference frame theory (SRF), [16] and

p-q-r theory, [17], [18]. Among these, the SRF

technique stands out for its use of a phase-locked

loop (PLL) system, ensuring undistorted

considered simple and practical techniques for

SAPF and power switch device control. HCA

operates with two predefined bands, ensuring that

the modulated currents remain within these

specified limits. This leads to the compensating

current closely tracking the reference current. HCA

determines the duration of the VSI (Voltage Source

Inverter) switches' “ON” and “OFF” states, resulting

in straightforward implementation, robustness, and

high performance, [19], [20]. However, HCA has its

pulse width modulation (PWM) control, which

produces a modulation signal to adjust the duty

cycle of power electronic switches, [21].

Fig. 2: Bloc diagram of the SAPF connected to a

distribution network with four legs topologies

Fig. 3: Bloc diagram of the SAPF connected to the

distribution network with four legs and mid-point

capacitor topologies

Fig. 4: Bloc diagram of the SAPF connected to the

distribution network with 3 half bridge topologies.

This study introduces a control system designed

for a photovoltaic (PV) system connected to the

electrical grid. The primary system components

consist of solar arrays connected to an inverter via a

(MPP), which corresponds to the point of the

highest power output.

There are various algorithms available for

MPPT, [22], [23] and in this research, we have

employed the Perturb & Observe (P&O) method,

[24]. The P&O algorithm takes an error signal as

input and produces the duty ratio of the DC/DC

converter (or its variation) as output, aiding in the

discovery of the MPP. This configuration enhances

of the APF depends on the efficiency of both the

reference current extraction technique and the

control of VSI switching devices. For reference

current extraction, a mathematical algorithm based

on the SRF technique was developed and

implemented. This algorithm allowed for the real-

time extraction of harmonics generated by CFLs

from the power source. These currents were sensed

using four sensors placed on the load side.

Additionally, the SRF technique utilized the PLL

IGBT switches. It’s important to note that while

HCA offers numerous advantages, it is susceptible

to chattering, resulting in variable switching

frequencies, which remain unresolved. To address

this concern, a Pulse Width Modulation (PWM)

controller was integrated into the system instead of

HCA. This PWM controller stabilized the switching

frequency and mitigated associated harmonic losses.

The paper is structured as follows: Section 2

elaborates on the Photovoltaic Generator (PVG) and

intricacies of a three-phase four-wire Shunt Active

Power Filter (SAPF) with a four-leg configuration

and split capacitor topology are presented in Section

4. Additionally, this section provides an overview of

mitigation strategies. Finally, Section 6 concludes

the paper by summarizing the findings and

contributions of this research endeavor.

standard conditions reference (Eref and Tref ).

Eref : Reference sunshine.

Kicc: Current short-circuit-temperature coefficient

(A/C).

Where:

q: Charge of an electron.

KB: Constant of Boltzmann.

Tj: temperature of the junction(C).

n: Ideal factor of the solar cell.

current flows through the cells, and the resulting

characteristic of the series arrangement is obtained

by adding the voltages at a given current, i.e., the

voltages add up, and the current remains constant)

and np cells connected in parallel (the cells are

subjected to the same voltage, and the characteristic

resulting from the grouping is obtained by adding

the currents at a given voltage: the currents add up,

and the voltage remains constant).

The total current delivered by the PV array, Ipvg

Incorporating series resistance and shunt

resistances provides an accurate modeling

opportunity for the PV cell. Rs corresponds to

internal losses due to current flow, and Rsh

corresponds to the leakage current to the ground.

The incorporation of series modules (cells) ns

increases the output voltage of the photovoltaic

array, and the incorporation of parallel modules np

increases the output current of the photovoltaic

array.

Manufacturers of PV modules provide reference

values for specified operating conditions, such as

The simulation was carried out for different

levels of irradiance and also for different

temperature levels. Irradiation levels were varied

from 0 W/m² to 1000 W/m², and the resultant P-V

and I-V curves can be seen in Figure 8 and Figure 9.

Fig. 9: V-I and P-V Characteristics of PV module at

different temperatures

The simulation results obtained from Figure 8

and Figure 9 shows that the voltage variation with

changes in irradiation is minimal, whereas, with an

increase in temperature, the voltage decreases.

Typically, the voltage will decrease. It can also be

observed that each curve has an operating point for

a certain operating voltage at which the module

produces the maximum power. This point is known

are almost the same, and at different levels of

temperatures, the short circuit currents are almost

the same. This, in turn, illustrates that at different

levels of solar irradiation, the voltage at which the

maximum power point is located is almost the same.

However, at different levels of temperatures, the

maximum power point is located at various

operating voltages that are far from each other. This

maximum power point varies at every instance, and

to have an efficient system, it is necessary to track

Pin: Incident power on the surface of the

photovoltaic cell (W),

Ea : Incident global illumination on the photovoltaic

Used to assess the quality of a photovoltaic cell, it is

The maximum power (MP) is obtained when the

solar panel is operated at the voltage where the

global maximum of the P-V characteristic lies. It

shows that for one specific operating point, the

maximum power output can be obtained from the

solar panel. This point on the P-V characteristic

curve is called the Maximum Power Point (MPP).

This point always lies on the knee of the I-V curve

that on the I-V curve of the solar panel, there is a

point called MPP (Maximum Power Point), which

always occurs on the knee of the curve where the

generated PV power is maximized. This MPP

voltage and current of the solar panel are inputs of

the controller. The algorithm performs its

calculations depending on these inputs. The

controller produces an output which is the adjusted

duty cycle of the PWM. It drives the DC-DC

converter’s switching device. For every different

operating point, the controller produces a different

duty cycle. To obtain the maximum power from the

solar panels, an efficient tracker algorithm is

required for the MPPT. The tracker algorithm’s task

specifications of the various and famous MPPT

algorithms previously presented. Was evaluated and

compared these algorithms in terms of complexity,

precision, speed, and technical knowledge of PV

panel settings, [25], [26].

less fluctuation in the steady state, [28]. The

11.

Fig. 11: Perturb and Observe algorithm

A modified version is obtained when the steps

are changed according to the distance of the MPP,

resulting in higher efficiency. A frequent trouble in

P&O methods is that the output terminal voltage of

Figure 12 shows the DC bus voltage regulation

circumstances guarantee synchronization of the

phase θs used in the Park transform with that of the

quantities of the network. As shown in the Figure 13

the PLL (Phase Locked Loop) technique is to

quadrature, of the fundamental voltage in the Park

reference frame from a phase θ synchronized with

the real phase θ of the grid voltage.

Fig. 13: PLL phase lock loop

The basic principle of the three-phase PLL in an

PLL are the three-phase voltages of the electrical

grid and the output is the detected phase angle. In

the case of a balanced system, the three simple

voltages of the grid are expressed as follows:

Where, Vgrid is the rms value of the network

Or, Pθ is the Park matrix. These leads:

By applying trigonometric relation, the previous

equation is simplified as follows:

Under the last hypothesis, the voltage Vqr can be

expressed as follows:

SRF theory for reference current calculation and

HCA for IGBT switches control. While HCA offers

numerous advantages, it’s important to highlight the

issue of chattering, which can lead to variable

using both electrical and mathematical modeling.

The Active Power Filter (APF) demonstrates the

capability to effectively address distortions

stemming from both current and voltage, offering

multidimensional flexibility, which makes it an

appealing area for research. APFs are primarily

employed for mitigating current distortions,

including current harmonics, reactive power, and

neutral current.

As depicted in Figure 3, the fundamental

provide the fundamental frequency for

synchronization purposes. Additionally, this method

requires a Proportional-Integral (PI) controller to