A Comprehensive Review of Grid-Connected PV Systems Based on Impedance Source Inverter

The increase in linking photovoltaic (PV) power plants to utility grids are due to the world expansion in PV systems and its advantages of low running cost, renewable, etc. Although PV is now considered one of the main power sources in many countries, it has low efficiency. Therefore, the big issue to improve the efficiency is to enhance the interface inverters’ efficiency. Different inverter topologies have been proposed to relate to the PV panels; each has advantages and disadvantages. These topologies can be classified into two-stage and single-stage (impedance network) inverters. Impedance network inverters overcome the problems of traditional inverters and seek to realize the advantages of a two-stage system and reduce the number of power conversions. However, to the authors’ best knowledge, there is no comprehensive review of the applications of the impedance source inverter for the PV system, including the control techniques. Therefore, this paper reviewed the existing topologies by paying attention to four key research issues: 1) various impedance network inverters and discusses the main structures from the point of view of the PV application, 2) control techniques suitable for impedance source inverters, comparing them in terms of complexity and theoretical performance, 3) investigation into the most important control methods used to connect the inverter output current with the network voltage, and 4) Challenges and future direction. Finally, this paper is provided as a brief reference to help researchers choose the appropriate impedance source inverter topologies for their applications and the preferred control and modulation methods for this type.


I. INTRODUCTION
In the issue of climate change and environmental pollution is one of the issues of international attention.Therefore, the increase in investment shares in clean energy sources has become a strategic target globally.PV energy systems are the most used in both stand-alone and grid connected.

A. BACKGROUND AND CONTEXT
Research in the PV systems aims to raise the efficiency and quality of cells, operate the PV arrays to produce their maximum power, reduce the size and number of components in the The associate editor coordinating the review of this manuscript and approving it for publication was Inam Nutkani .system to limit cost and power losses, and raise the efficiency of the inverter used in power conversion, which is the most important component of the system.The impedance network inverters can be considered a promising means of converting power between the source and load in various electrical power conversion applications.Many structures of the impedance networks are provided with different control methods to use them in various applications, such as adjustable speed engines, uninterrupted power sources (UPS), distributed generation, and Electric vehicles [1].

C. RESEARCH MOTIVATIONS
PV applications are classified into stand-alone or gridconnected PV systems (GCPVSs), where a central grid system is designed to increase the capacity of the main grid.Further, the number of power conversion stages for both stand-alone and GCPVSs can be categorized into single-stage and two-stage inverters.These classifications rely on the voltage source inverter (VSI) or current source inverter (CSI) as a major component.However, VSI has some drawbacks, first, the DC input voltage must be greater than the AC output voltage (v ac = V pv 2 ).Second, the upper and lower switches of the bridge cannot be operated in one phase leg simultaneously.This may cause a short circuit to the switches and destroy the inverter [5].Hence, a time interval (dead time) must be used between the switching pulses of the upper and lower devices, which results in a distortion of the inverter's output waveform.Therefore, there are two ways to overcome the problem of the DC-link voltage value.Firstly, a line frequency transformer can be added on the AC side to raise and modify the value of the AC voltage to match the grid voltage.Secondly, a DC voltage boost converter can be added before the inverter.To achieve the required DC voltage, resulting in complicating the entire system, reducing efficiency, and increasing cost [2].However, the use of dead time is still necessary if any of the above two methods are used.Therefore, ZSI is proposed to eliminate these obstructions, using the shoot-through (ST) states to achieve boosted DC-link voltage.
Dead time can also be avoided since the ST state will become part of the inverter's operating states and the active and null states of conventional VSI.However, the traditional ZSI has some drawbacks, such as [1]: • In the case of ZSI, the DC-link voltage is less than the Z-source capacitor's voltage, which increases capacitor size.
• The ZSI input current is not continuous due to the diode's effect, which may be undesirable for PV.
• The inrush current in ZSI at start-up is very high.
To overcome the drawbacks of the original ZSI, much research has been done, and a set of modified ZSI topologies have been proposed [12], [15].These proposals aim to raise the efficiency of ZSI by increasing capacitance, decreasing component stresses, and decreasing the number of switches, switching frequency and passive components [20].Moreover, different structures have been suggested to increase the reinforcement capacity but did not solve other problems of traditional ZSI [21].Further, in [39], [43], [44], [45], [46], and [47], other topologies have been proposed to reduce the voltage of the Z-source capacitor, but they cannot fully increase the boosting capacity of the inverter.Generally, every structure has its successes and failures in solving various defects, so the optimal topology cannot be determined easily.

D. RESEARCH AIMS
As mentioned earlier, many ZSI topologies have been developed and tweaked where each topology has its unique features and applications that suit it best, and there is no one-size-fitsall solution.It is expected that new ZSI topologies will be constantly rolled out to meet and improve the performance of the inverter for different applications.Among them is the generation of renewable energy, such as PV, due to the unique ability of the ZSI to boost voltage with a minimal number of components and potentially low cost.Nevertheless, different modulation strategies are also needed to operate the inverter switches and adjust them to get the required outputs like frequency, the amplitude of voltage, and current.Different PWM schemes can control impedance source inverters with the same input and output relations, such as sinusoidal PWMs (SPWM) [83], [84], [85] and space vector PWM (SVPWM) [87], [88], [89], [90], [91], [92], [93], [94], [95], [96], to control the output voltage.There are limited numbers of review papers focusing on specific targets, for example, the work in [3] studying and developing the modification schemes suitable to this family of power inverters only.Also, other reviews have been concerned with studying different structures, advantages, and drawbacks [1], [3], [18], [42], [43], [44], [56].Further, some reviews focused on studying reaction energy only [55].Meanwhile, few reviews tried to study and cover different topologies, PWM techniques, and control strategies [5], [12], [57].In [12] there are two-part studies.The first part was concerned with all the topologies.The second part gave a brief study of the control strategies on the DC side, and another brief for the most important methods of PWM techniques.While they neglected the necessary control strategies on the AC grid side.Finally, [5] is a good review article, but it was a little old.It did not mention new types of impedance networks such as (Y-source, T-source, SL-ZSI, and SBI) and it did not cover the technique of 3 rd HI-CB-PWM.This paper presents a complete review of the most important studies and literature specializing in the parts of the grid-connected PV systems based on impedance source networks (ISNs) inverters.The second section of this paper is concerned with classifying the impedance source networks-based inverters, describing their different structures, and then comparing them in terms of characteristics, number of components and drawbacks.In the third section, the paper categorizes the different modulation techniques with detailed explanations and qualitative comparisons.In the fourth section, what has been published in the literature from the different control strategies that have been previously applied are reviewed.Thus, it is within the researcher's reach to choose the appropriate impedance source network topology for the target application and how to operate the inverter switches by controlling the pulse width.Finally, determining the control strategy that achieves the best performance for each system stage.

II. IMPEDANCE NETWORK CLASSIFICATIONS
The Z-source inverter (ZSI) was first introduced thirty years ago with a unique circuit where it can boost the input voltage.It has been used as a single-stage inverter in different applications, such as in the conversion systems of renewable energy sources.Using the Z-source impedance network (ZSN) features, the inverter can be controlled in both Shoot-through (ST) and Non-Shoot-Through (NST) states, as will be explained later in the following section.Although the ZSI can simultaneously perform the booting and inversion operations, it suffers from drawbacks, such as a discontinuous input current.This drawback is solved by deriving another topology called the quasi-Z-source network (QZSN), introduced in 2008 [5].This circuit is characterized by continuous DC input current.After that, many impedance circuits are developed by bypassing a specific defect in the original network to improve the inverter properties [7].Therefore, the main aim of this paper is to review the Impedance Source Networks (ISNs) coupled with VSI and clarify the advantages and disadvantages of each type [8].
The ISNs can be classified based on magnetic and transformer coupling into non-transformer-based and transformerbased types, as shown in Fig. 2. It can also be divided into the following groups [9]: a) ISNs including inductive components, b) ISNs including magnetically coupled inductive components, c) ISNs including a transformer, and d) ISNs including an active switch.
One advantage of ISN that includes magnetically coupled transformers is that they may have a high voltage boost factor FIGURE 2. Classifications of the different ISNs topologies [9].
according to turn ratio.Table 1 divides ISNs into two categories according to the type of input current; discontinuous input current (DIC) and continuous input current (CIC) [10].In general, the impedance source inverter operates in two modes of operation as follows: • Same as VSI when the input voltage is higher or equal to the required voltage level (modulation index is M ≤ 1).
• Like a step-up transformer, when the input voltage level is lower than the required voltage level; thus, the VSI operates in ST mode to boost the input voltage level.The ZSN is the original impedance network from which all other ISNs topologies were derived, either by modifying or rearranging the locations of inductors and capacitors [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20].Each ISN topology has specific features to meet the needs of a particular application.The new ISN topologies are being developed so far for the following reasons: • Reducing the quantity and rating of ISN components, • Increasing the voltage gain factor, • Realizing higher power density, and, • Increasing the inverter efficiency.Components such as inductances, capacitors, auxiliary diodes, or switches can be rearranged or increased to increase the network voltage.Fig. 3. shows the required components for each topology.Each topology has its advantages and disadvantages, as will be briefly discussed in the following subsections.

A. NON-TRANSFORMER BASED CONFIGURATIONS
These ISNs do not contain a transformer where they consist of only passive components.More details about these configurations are provided in the following sections.

1) Z-SOURCE AND Quasi-Z-SOURCE
As mentioned earlier, ZSI topology has appeared as an alternative to the traditional VSI or CSI.The ISN is considered a buffer between the source and the inverter bridge.It provides short circuits or open circuits at any time to suit the operating mode.Conventional ZSN, shown in Fig. 4(a), suffers from several problems, including that it draws discontinuous input current from the source, and it is not suitable for very low input DC voltages.Therefore, different voltage topologies derived from the ZSI have been proposed, such as QZSI.Two types of QZSI are shown in Fig. 4(b) and (c), where it is used to improve the performance of ZSI and overcome its drawbacks [5], [6].where V in equals the input DC voltage or PV voltage, V dc the boost voltage produced by the Z-source impedance, V C the voltage across the Z-source capacitor, Do the duty ratio of the shoot-through interval, and B the boost factor of the impedance network.2) EMBEDDED-Z-SOURCE NETWORK (E-ZSN) [6] An E-ZSN provides a continuous input current to the inverter, has a low capacitor voltage and is particularly suitable for PV applications [12].Figure 5(a) shows the topology of the E-ZSN circuit.This structure reduces the voltage and current ripple by using two independent DC voltage sources.But it increases the number of components and their cost.For this reason, E-ZSN is preferred in the case of PV arrays or fuel cells.The disadvantage of E-ZSN is that it operates asymmetrically at asymmetric input voltage.In the E-ZSN, inductors L 1 and L 2 are considered input current filters rather than external LC filters.

3) SWITCHED INDUCTOR/CAPACITOR
Following One of the most common methods is using a switched inductor instead of a permanent inductor which improves the boost ratio of the inverter, reduces the stress on the passive components and eliminates the star-up inrush current [39], [43], [44].The switched inductor ZSI is shown in Fig. 5(b).The switched inductor ZSI and QZSI provide a continuous input current and low voltage stress on the capacitor.The embedded ZSN with a switching inductor combines the advantages of both switched inductors ZSI (SL-ZSI) and (E-ZSI).Such as high boost ratio, low capacitor voltage stress, and low input current ripple.However, the switched inductor topology needs many passive devices, which increases the inverter cost.A switched capacitor can also be used instead of a permanent capacitor to achieve the same advantages.[14], [15] Some capacitors and diodes are added to the QZSI circuit to increase the voltage boost capability in this topology.Fig. 6(a) shows the Diode Assist Extended Boost (DAEB), in which one capacitor (C 3 ), one inductor (L3) and two diodes (D2 and D 3 ) are added to a conventional QZSI.Fig. 6(b) shows the modified diode assisted (MDAEB), which is derived from DAEB-QZSI by changing the contact points of the capacitor C 3 .Theoretically, such modifications can produce unlimited boosting capability, but this is impossible due to power losses.This structure is characterized by the capacitor's high voltage gain and low voltage stress.However, the passive components are increased.Therefore, Capacitor Assist Extended Boost (CAEB) topology, shown in Fig. 6(c), is preferred because it requires fewer semiconductor devices and has less power loss.

B. TRANSFORMER BASED OR MAGNETIC COUPLING CONFIGURATIONS [1], [2]
A general method was proposed to clarify the derivation of magnetically coupled networks in [1] and [2].It was found that magnetically coupled inductors and transformers improve the voltage boosting ratio and modulation index of the ISNs.It also decreases the passive components used in the network, increases the quality of power density, and reduces the system's cost.The following sections describe the ISNs topology based on the transformer or the coupled inductor.
1) TRANS-Z-SOURCE, TRANS-QZ-SOURCE, TRANS-QZ-SOURCE WITH CCM [8], [17], [20] Theoretically, the original ZSN, QZSN and embedded-ZSN have unlimited voltage gain.However, if a voltage gains higher than 2 or 3 will result in a higher stress voltage on the switches.Trans-ZSN has been proposed with a discontinuous input current [13].Trans-QZSN, and Trans-QZSN with continuous conduction mode (CCM) have been proposed [14].These proposals aimed to obtain higher voltage gains while keeping lower voltage stress and reducing the ZSN components to one coupled inductor, one capacitor, and one diode in Trans-QZSN.
Meanwhile, an additional capacitor is added to the Trans-QZSN-CCM.In Trans-ZSN, the boost factor is increased, and the coupled inductors are used as energy storage.The disadvantage of this topology is the presence of leakage inductance.Fig. 7., shows the Trans-ZSN, Trans-QZSN, and Trans-QZSN with CCM.
2) Y-SOURCE NETWORK [48] The Y-source network belongs to the networks with magnetically coupled components and consists of an inductor coupled with three coupled windings (N 1 , N 2 , and N 3 ), as shown in Fig. 8(a).Theoretically, the proposed inverter can obtain any voltage gain by adjusting the turns ratio and ST duty cycle of the switch.The disadvantage of this scheme is the presence of leakage inductance.The disadvantage of this scheme is the presence of leakage inductance.
3) THE -Z-SOURCE NETWORK [10] In Fig. 8(b), the -ZSI consists of one capacitor and a two-windings coupled transformer in the ISN to simultaneously increase the gain voltage and modulation index while decreasing the number of components.The Ã-ZSN has a discontinuous input current.The gain of the -ZSN increases with decreasing turns ratio compared to the T-SN, trans-ZSN, or LCCT-ZSN, whose gain increases with the 'turns ratio [16] The disadvantage of this topology is the presence of leakage inductance, which may affect the voltage and current on the semiconductor.[4], [15] The TZ-source achieves high voltage gain by setting the turns ratio of the transformer to greater than 1.The T-Z-source network contains magnetic components, which are a twowinding coupled inductor and one capacitor [17], as shown in Fig. 8(c).The voltage gain and boost factor of T-ZSI are higher than ZSI and QZSI.However, it has several drawbacks, including resonant effects.Because of that, transient currents can be high.Secondly, the substantial number of elements that make up this network and the fact that the input current is discontinuous.

5) THE LCCT-Z-SOURCE NETWORK AND LCCT-QZ-SOURCE NETWORK [9], [3]
The LCCT-Z-source is derived from Trans-Z-source.It consists of an inductor-capacitor-capacitor-transformer Z-source network, as shown in Fig. 9(a), while the LCCT-QZSI is shown in Fig. 9(b).The transformer used in these networks is ideal and of high frequency.The LCCT circuit prohibits saturation from occurring in the transformer core.For energy storage during the boosting process, only one inductor is used.With the combination of the inductor and the transformer in a common core, the LCCT-ZSN realizes high voltage gain and a higher modulation index but has a discontinuous input current.
Meanwhile, the LCCT-QZSN collects high gain voltage and continuous input current at any load level.It also prevents high-frequency fluctuations from the input current.The LCCT circuit can be improved by using high switching frequency diodes.

C. SWITCHED-BOOSTED INVERTER (SBI) TOPOLOGIES
The switched-boosted inverter (SBI) was introduced as an alternative to ZSI with the advantage of half number of the passive components.The inverse Watkins-Johnson topology, which is suitable for low power applications, is derived from the SBI topology, as shown in Fig. 10.The impedance network consists of one inductor, two diodes, one capacitor and an active switch where the capacitor connects to the negative terminal of the DC source, and the diode connects to the positive terminal.This type is illustrated in Fig. 10(a) and named SBI Type 1.The arrangement of the passive elements is changed to present a new type called SBI Type 2, where the diode connects between the capacitor and the positive side of the DC source, as clarified in Fig. 10(b).It is noticeable that SBI and ZSI in the ST state are similar in operation.Also, SBI could not overcome the most of ZSI drawbacks, such as the high voltage stress on the capacitor, the low boost capacitance, and discontinuous input current [26].

1) TRANS-SBI TOPOLOGIES
In the Trans-SBI type, the number of inductors is increased into two in series inductors instead of one as in the original SBI.In Trans-SBI Type1, the capacitor is connected between the boost circuit and the positive terminal of the inverter input.Meanwhile, in Trans-SBI Type 2, the capacitor is connected between the boost circuit and the negative terminal of the inverter input.Fig. 10(c) and Fig. 10(d) show these two topologies [27].

2) SWITCHED INDUCTOR SBI TOPOLOGIES
The switched inductor boost inverter (SL-SBI) is offered for is offered for achieving higher DC-link voltage.It is an upgraded model from SBI, but the inductor has been replaced by a circuit consisting of two inductors, three diodes and a DC source.Several substructures are derived from the underlying SL-SBI to bypass some defects by modifying the location of the DC sources.When the DC voltage source is placed in the place of the DC 4 , a constant and continuous input current can be obtained, and it is known as the SL-SBI Type-1.Whereas if the DC voltage source is placed in the position of the DC 2 , the input current will be discontinuous, and it is called the SL-SBI Type-2 because the diode is connected in series with the DC source.Further, placing the DC voltage source in the position of DC 3 leads to an input current ripple.Both structures are shown in Fig. 10(e) and Fig. 10(f).[28].

D. SUMMARY
Several modifications and improvements have been proposed in the original ZSI to overcome the drawbacks of the traditional ZSI and improve its performance [8], [9].Some succeeded in increasing the boosting capability, and others reduced the capacitor voltage and the start-up inrush current.Thus, each ISNs derived from the ZSI has advantages and disadvantages in solving the problems of the original configuration, so the user may not be able to select the appropriate network in practice quickly and accurately.Hence, to help the researcher determine the appropriate type required to achieve the objectives of the application, Tables 2 and 3 summaries the differences between the most famous types of ISNs in terms of components and characteristics, respectively.In contrast, Table 4 compares ISNs topologies in terms of advantages and disadvantages.

III. MODULATION TECHNIQUES
Impedance source inverters represent a different generation of power conditioning systems called single-stage inverters because they simultaneously adopt boost capability and reversal process.Therefore, it is an alternative to two-stage systems [1], [2], [3], [4], [5], [6], [7], [8].Thus, the research activities have increased to improve the performance of ZSI from different perspectives, such as voltage gain, voltage stresses across devices, input current continuity, and switching efficiency [5], [19], [20], [21], [22], [23], [24], [25].Among these are the various topological improvements already mentioned in the second section.It is noteworthy that all these improvements did not affect the basic operating principle of the original ZSI, as shown in Fig. 11.This refers to using an additional switching case compared to the standard cases of operating a three-phase voltage source inverter (VSI).This additional state is called the shootthrough state (ST) and is responsible for boosting the input voltage.
During this condition, the bridge appears as a short circuit by turning on at least two switches of the phase leg at the same time, as shown in Fig. 11(a).
On the other hand, during the non-ST states, the bridge is equivalent to a current source (Because the inverter bridge operates in one of the six active states, therefore acts as a current source that is displayed from the ZSI network.Due to the symmetric in the circuit, both inductors have identical current value.During this mode, the DC voltage source appears across the inductor and capacitor.The capacitor is charged and power flows to the load through the inductor), as shown in Fig. 11(b).In contrast to the traditional VSI, which does not allow the presence of the ST state.Because of the use of an impedance network, this state is allowed in impedance network inverters and must be inserted into any of the zero states so as not to affect the active states and, thus, the output AC voltage.This is the function of the pulse width modulation controllers.Three-phase impedance source inverters can use any of the conventional modulation schemes discussed in [5], [19], [20], [21], [22], [23], and [34], or the SVM schemes presented in [5], [23], and [24] and hybrid modulation (HM) proposed in [34], but the ST case may not be taken advantage of without modifying these methods to fit the ZSI principle of operation.
In general, three-phase ZSI modulation schemes can be classified into two categories: carrier-based modulation schemes and space-vector-based modulation schemes.This section provides a comprehensive review with a comparative evaluation of these different modulation schemes to measure performance and show how each scheme's properties influence ZSI parameters, such as voltage stress across devices, voltage gain value, and maximum modulation index.where SB is the simple-boost, MB is the maximum boost, CB is the constant boost, SM is the sinusoidal modulation, THM is the third-harmonic injected modulation, SVM is the space vector modulation, MSVM is the modified SVM, DCCM is the dc-clamped modulation, and HM is the hybrid modulation.,More details about each category will be discussed as follows:

A. CLASSIFICATION OF THREE-PHASE ZSI MODULATION SCHEMES
• Two ST insertion strategies: in which three-phase modulation schemes simultaneously operate through the three-phase legs as shown in Fig. 13(a).The ST states immediately replace the null states (111 and 000).This can be achieved using two straight DC lines, and the sinusoidal modulation signals of VSI.The ST duty cycle in this strategy is equally divided into two parts in one switching cycle.At each part, all three legs are in ST state, so that ST current is distributed evenly on three legs.On the other hand, the ST state inserts additional switch transitions.As a result, there are two more transitions in each leg than case of VSI in one switching cycle.[18] • Six ST insertion strategy: in which single-phase modulation schemes operate through one phase at a time.The principle of this method is that the ST states are inserted at every transition by overlapping the upper and lower driver signals as shown in Fig. 13(b).The transmission intervals in one switching cycle are the same as in the VSI.The ST state is divided into six segments, so that equivalent switching frequency of the impedance network equals six times the switching frequency.So, the size of the inductors can be greatly reduced.[18] FIGURE 13.PWM types based on ST insertion.
Each of these two categories can be divided into two subcategories as follows: [29] Continuous modulation: Two additional reference voltages (V p and V n ) are used with any of the reference signals (v a , v b , v c ) since the shape of V p and V n differs from one controller scheme to another.Discontinuous modulation: One additional reference voltage signal (V p or V n ) is used with DCCM reference signals.Any of the previous type also can be controlled using one of the following modulation techniques: • Simple Boost control (SB), • Maximum Boost control (MB), • The Maximum constant boost pulse width modulation (MCBC), • Constant Boost control with 3 rd Harmonic Injection (THM), • Modified Space Vector control (MSVPWM).
The literature discussed the use of the above methods in controlling pulse width modulation (PWM) in ZSI.In fact, each method is distinguished from others in several characteristics.
In case of using SBC or 3 rd -HI-CBC, the ST ratio is controlled by MPPT.Thus, it is suitable for PV systems.As for maximum boost control (MBC), it is not suitable for PV systems because the variable ST ratio with MBC will not be able to track the reference envelope.Also, MBC requires increasing the size of the inductors and capacitors to remove low frequency ripples [30].Therefore, the operating principle of each modulation method will be briefly explained.Moreover, a comparison of the most important differences is introduced in Tables 3.4 and 3.5.

1) SIMPLE BOOST CONTROL (SBC) [6]
This technique is the simplest PWM method to control the ZSI [20].Two values V p and V n are used as ST modulation signals.As shown in Fig. 14., the positive and negative peak values of these signals are greater than or equal to the peak values of the three-phase sinusoidal reference signals.ST only occurs when the carrier signal is greater than V p or less than V n .Otherwise, it works like conventional sinusoidal pulse width modulation (SPWM).Fig. 14 shows the simple boost pulse width modulation process and how to get the shoot-through interval.2) MAXIMUM BOOST CONTROL (MBC) [25] Fig. 15.shows the MBC technique.Compared with the SBC technique, MBC keeps the six active states of ZSI unchanged and converts all zero (null) states to ST (zero) states.Thus, the maximum gain G value for any modulation index M is achieved without distorting the resulting signal.The ST modulation signals are not two fixed lines but the positive and negative envelopes of the three-phase reference signals.ST occurs when the carrier signal is greater than the positive envelope or less than the negative envelope.Since the duty cycle of ST differs in each cycle.So, the mean value of ST during the cycle is considered to calculate the gain.
The ST state is repeated periodically every π/3.Assuming that the carrier signal frequency is much higher than the modulation signal frequency, the equation for the ST during one switching period in the interval (π/6 and π/2) can be formulated as follows: The average duty ratio of shoot-through can be formulated as follows: But in the MB method, due to the periodic variation of the operating ratio D o (t) per π 3 , the inductor current has a current ripple equal to six times the load frequency, which raises the required inductance value.
3) MAXIMUM CONSTANT BOOST CONTROL (MCBC) [26] Fig. 16.shows the MCB technique, which realizes the most voltage gain and keeps the ST duty ratio constant.Maintaining the ST constant reduces size and cost while increasing the voltage boost of any modulation index results in less voltage stress over the switches.
As in the earlier methods, there are five modulation signals: three sinusoidal references and two ST envelopes.ST occurs when the peak of the carrier wave is greater than V p or less than V n .Otherwise, the inverter is operated as in traditional SPWM control.
The variable B is figured out and kept constant by maintaining D o as constant.The main target of the MCBC is to achieve maximum boost while keeping it constant throughout, which helps to drop the current and voltage ripples that appear in the MBC technique.The frequency of the upper and lower envelops curves is three times the output frequency [8].There are two half-periods for both curves in a cycle.
The ST modulating signals (V p and V n ) are expressed as the following: The duty ratio of the ST is still constant since the distance between V p and V n curves is always constant and equal

4) CONSTANT BOOST WITH THIRD HARMONIC INJECTION
The MCBC method described above can be modified by a third harmonic injection [31].Where the three sinusoidal signals of the phase voltage are injected by a third harmonic component with a value equal to 1  6 of the fundamental signal, as in Fig. 17.At π 3 the value of V a reaches ( √ 3 2 m), which is the maximum possible value, while V b is at the lowest value, which is equal to (− √ 3 2 m).Therefore, to control the ST time using 3 rd HI, only two DC voltages (V p and V n ) are needed.Thus, it corresponds to the MCB control method in calculating the voltage gain.The difference is that the modulation index m is larger, which increased from one to ( 2√ 3 ).The voltage gain G can be varied from (0 to ∞) smoothly by changing the value of m.  [28] Space vector pulse width modulation (SVPWM) technology is commonly used because of the higher modulation index  ) in the linear region.Compared to SPWM, it supplies lower current harmonics.It has eight space vectors (six active vectors) and (seven null vectors).For ZSI, MSVPWM is used to obtain ST states as part of the switching pattern.The ST periods are included during the periods of null vectors of the total period T n , while the active state times T 1 , T 2 and T Z are unchanged, as seen in Fig. 18.

5) MODIFIED SPACE VECTOR CONTROL (MSVC)
Two types of MSVPWM can be used.The first pattern is based on the partial conversion of null states to ST states.In this pattern, the ST period of T o is limited to 3  4 of the period of the null state T n .That is, null states are not used as ST.This limits the achievement of higher values of gain.Therefore, it has been modified into the second pattern, where zero states are completely converted to ST.Thus, the gain of this pattern is higher than the gain of the first pattern.
Although MB-PWM has the highest voltage gain rate and boost ratio.However, it is not suitable for PV applications.Because the ST duty ratio is determined by the MPPT algorithm which targets the DC reference voltage at the maximum possible power output is achieved.CBPWM or MSVPWM is preferred because they achieve a relatively high gain and an appropriate voltage boost ratio and are suitable for PV applications.

IV. CONTROL STRATEGIES FOR GRID CONNECTED SYSTEMS
The task of a grid-connected PV system is to convert DC power produced by the PV array into AC power that can be fed into the grid.Thus, the control unit design aims to provide the maximum power from the PV arrays with the desired power factor and low harmonic distortion (THD) in the output.The designed controller must follow the reference and be able to reject disturbances.To implement a good ZSI behavior control strategy, it is necessary to formulate a good dynamic model of the inverter.Several small-signal analyses and mathematical models have been presented in the literature to study the overall system's performance.
• The AC side control converts the DC voltage into AC voltage and achieves the unity power factor.The inner current loop is adopted to ensure that the inverter current can follow the AC grid voltage [78].

A. THE DC CONTROL STRATEGIES
In As mentioned earlier, the PV array voltage is controlled by the ST duty cycle to track the maximum power points.There are four linear methods to control the DC-link voltage (V dc ) of ZSI, as shown in Fig. 20, which are:  1) CAPACITOR VOLTAGE (V c ) CONTROL [29], [32], [33], [34], [78], [79], [80], [81], [82], [83] capacitor voltage is controlled by adjusting the ST duty ratio where a PID controller is used and setting the modulation index to equal (1 − D o ). the gains of the PI controller can be obtained by try and error with saturation or by a neural network to achieve a wide-bandwidth control.In addition, a non-linear control method was used to control the V c , such as the sliding mode control method, the neural network controllers and so on.Fig. 21.sets out the different non-linear controllers.

2) INDIRECT CONTROL OF DC-LINK VOLTAGE [83], [84]
A PID-like fuzzy control unit is employed to control the mean value indirectly V dc .In [84], the peak of DC-link voltage (v dc (peak)) is indirectly controlled through controlling the peak of inverter output voltage (v ac ) using a PI control unit to adjust M, while Do is calculated by measuring the PV voltage (V in ) and comparing it to the desired peak of DC-link voltage (v dc ).
3) DIRECT CONTROL OF DC-LINK VOLTAGE [31], [85] The v dc is directly controlled through the adjusting value of the Do.The v dc value can be measured by an additional circuit due to its pulsating shape and hence complicating the control algorithm.On the other hand, it can be estimated by measuring V in and V c , while v dc = 2V c − V in .
4) UNIFIED CONTROL [86], [87] Unified control is used to simultaneously regulate the modulation index and the ST duty ratio.This is done by controlling the AC output voltage with a single PI control unit.The previous control methods share that they use the single-loop DC-link voltage control technique, where the DC-link voltage is directly affected by the changes in the PV input voltage, while the ST ratio is considered a command signal for the control method.It also suffers from two challenges in the case of high-power inverters.
• Inductor current is uncontrolled and can be overloaded during the transient mode.
• Finite stability limits.Fig. 22. and Fig. 23.show the scheme of the single control loop (SCL) and the double control loop (DCL) system, respectively [32].Therefore, double-loop voltage control in high power inverters is preferred to overcome these drawbacks.it is noted that the input inductor current (I L ) is the command signal for the control method.As in Fig. 23, the process divided into an inner and outer loop [26].The loop determines the DC-link voltage and ensures instantaneous response while controlling I L and V c using a PI controller.On the other hand, the PI control method enhances stability in the outer loop.Table 7 reviews the previous control methods and their characteristics.

B. GRID SIDE CONTROL STRATEGY
The control strategy for the grid side or the AC side is implemented in the dq-axis frame, where the three-phase quantities are transformed into a frame of reference with a shunt matrix given in Table 8 [37].A phase-locked loop (PLL) controller synchronizes the inverter's output current with the grid voltage, thus achieving the unity power factor.

1) PHASED LOCKED LOOP (PLL)
The PPL is needed to know the angle of the grid voltage and hence generate the reference signal with a similar angle.Therefore, the current can be in phase with the grid voltage, as shown in Fig. 24, and hence the unity power factor can be achieved.So, the PLL is used to generate this signal which is used as a reference signal and sent to the current controller.Similarly, for reactive power control, the PLL is used to generate a signal which is 90 • out of phase with the active voltage, as shown in Fig. 25.There are two ways to implement the PLL; open-loop and closed-loop.
First Method: The three-phase voltages (v a , v b , v c ) are converted to two-phase voltages v α , v β , and substitute the values for the two voltages v α , v β , in the equation β )), angle information (θ) can be obtained, which in turn is used to generate the reference signal for current in both the active and the reactive power, as shown in Fig. 26(a).Nevertheless, this method is not used in many situations due to some drawbacks such as: • The algebraic method with the simple mathematical operation.
• The open-loop system makes the system prone to go into unstable situations and critical network conditions.
• This PLL cannot withstand surges, harmonics, noises, and spikes.
• Drift the angle, which means that PLL output may produce wrong angle information.Second Method: Also like the previous method, starting by converting three-phase voltages in the normal frame (v a , v b , v c ) into a two-phase voltage in the stationary frame v α , v β , and then to a two-phase voltage in a synchronous frame (v d , v q ), as shown in Fig. 26(b), which shows the closed-loop PLL method, also Fig. 27(a), which shows the phasor diagram of the (α, β) axis, grid voltage, (d, q) axis, and the angle ωt, which is the angle between d-axis and (α) component.The phasor diagram shows that the d-axis does not completely align with the grid voltage, so non-zero values of the voltages of the d-axis and q-axis appear.But using some control mechanism, the q-axis voltage goes to equal zero, which will adjust the phasor diagram so that the grid voltage aligns perfectly with the d-axis.The new phasor diagram is shown in Fig. 27(b).
It is clear from the modified phasor diagram that the daxis voltage and the q-axis voltage are equal to zero, and the angle ( ωt) is changed to a new value.This new value of the angle ( ωt) can be used to generate the active and reactive components.
From the PLL scheme, the PI controller is used to keep the q-axis voltage equal to zero, so the reference value of the q-axis voltage must be zero.The output of the PI controller is sent to an integrator to find the angle ( ωt) and finally use the (sin and cos) functions to generate the active and reactive components, which are the final output of the PLL where the active component is aligned in phase with the d-axis, and the reactive component is aligned in phase with the q-axis.

2) TWO-LOOPS CONTROL STRATEGY
Usually, a double-loop or two-loops control strategy is used: [37] • The inner control loop in current that adjusts the grid current ensures power quality, current protection, and the function of harmonic compensation and dynamics.
• The outer control loop in voltage controls the DC-link voltage and works to balance the power flow through the inverter.It aims to maintain system stability with slow dynamics.However, other control strategies have been used in the different literature, such as: • The current injected into the grid is indirectly controlled in the outer control loop for DC-link voltage with an inner control loop in power.
• The outer control loop is in power with an inner control loop in the current.But all previous control strategies can be implemented with three frames of reference: • The synchronous frame, • The fixed frame, • The normal frame.The control carried out through these frames of reference is shown in Fig. 28(a, b, c).FIGURE Reference frames control.[38].

3) REQUIREMENTS
Generally, a PI controller with grid voltage feedforward is used for current-controlled PV [ 38] but this method has the following drawbacks: VOLUME 10, 2022 • A steady-state error occurs when the PI controller traces the sinusoidal reference.
• The system's ability to reject disturbance is poor due to a bad integrator performance.Therefore, two methods have been proposed to bypass these shortcomings, which are: • Generalized integration: This means using a double integrator that achieves infinite gain at a certain frequency, called the resonant frequency, and thus, can be used as a notch filter to compensate for harmonics in a highly selective manner.[12] • P + resonance controller: A stationery frame regulator is used, and the PI controller used as a DC compensator is converted into an equivalent AC compensator having the same frequency response characteristics at the desired bandwidth.
[39]  c) frame of reference can be formulated as follows [65]: where V is the peak voltage of the grid phase and ω is the angular frequency.The previous equations can be converted to dq synchronous frame of reference as follows: We can find that in the d-q frame, the current is coupled in the d-q axis, which brings difficulty in the design of the controller.Therefore, it is necessary to obtain a method for separating or decoupling the control [9].Suppose: Thus, we get: Using the PI controller, the equation can be formulated as follows: where K p is a proportional gain and K i is an integral gain of PI controller.Substituting ( 42) into (43) we get the following matrix: This realizes the process of controlling the current loop disconnection.As shown in Fig. 29-b.
FIGURE Decoupling control of the current loop.[38].

5) ACTIVE AND REACTIVE POWER CONTROL
Active and reactive power control is critical in a gridconnected PV system.Generally, all the radiation absorbed from the PV array is converted into active power by the grid-connected PV system and is taken away from reactive power generation.This gives designers and users the ability to control the system and adjust its performance easily.However, the widespread development in grid-connected PV systems, especially in places far from the main grids or those connected to sub-grids only, has made reactive power compensation and harmonic tuning another major consideration.The three-phase inverter circuit can be compatible with the three-phase reactive power compensator.It can output reactive power and reduce current distortion to the grid by appropriate control of the inverter currents [60], [61].
Some studies have combined ZSI with reactive power compensation or harmonic tuning.The firing time has been combined with sinusoidal pulse width modulation (SPWM) [62], [63].A standardized grid-connected photovoltaic system control with reactive power compensation and harmonic tuning based on ZSI has been proposed [64].

6) ANALYSIS OF ACTIVE AND REACTIVE POWER OF PV SYSTEM
Fig. 30.shows the Thevenin equivalent circuit for the gridconnected PV system where the Z g = X g + R g is the equivalent impedance of the grid side [60].When the unity power factor is adopted to control the PV system, the grid voltage V pv changes to V pv , so the identical current also changes to I pv as follow: where, P pv is the change in PV output active power, V pv is the rated voltage at the point of common coupling (PCC), S K is the short circuit capacity, Q is the equivalent impedance angle, and I p is the change in the active current of the PV system.According to the previous equation, when the PV output is greater than a certain value, the PCC voltage is higher than the normal values.Furthermore, when the PV system is connected to the reactive power voltage control mode, the change in the grid voltage can be expressed by: where, S pv is the change in active power generated by the PV system , θ is the power factor angle of the system connected to the PV grid, and I q is the change in the reactive current in the PV system, from (46), the path of change is shown in the Fig. 31.As per Fig.31, when the PV output active power is the V pv is at the top of the x-axis.This means that the grid-connected inverter's reactive power can be regulated and keep the PCC voltage constant.When V pv at point A in Fig. 34., the V pv will equal V pv * .Therefore, the grid-connected PV system must fully use the inverter's reactive power control to implement reactive power and voltage regulation [60] 7) ACTIVE AND REACTIVE POWER CONTROL OF ZSI Relying on the instantaneous power theory, the relation of instantaneous active power p and reactive power q with grid voltage V d and V q can be formulated in the synchronous frame of reference (d -q) as follow [63], [64]: where I d and I q does ZSI produce the active and reactive output currents.Since the grid voltage q is zero, (15) can be simplified to So, the active and reactive power of the ZSI can be controlled by adjusting the active current I d and reactive current I q produced by ZSI in a synchronous frame of reference.The ZSI mathematical model can be formulated as mentioned before in equations ( 9) and (10), which means that the inverter output currents I d and I q can be controlled separately from the reference current signals I * d and I * q as shown in Fig. 31.Finally, the active and reactive power of ZSI can be controlled.

V. CHALLENGES AND FUTURE DIRECTIONS
From the above it is clear that achieving all the advantages and overcoming all the disadvantages through one ISN design is impossible.Hence, several models have been proposed to complement the shortcomings of ZSI topologies depending on the intended application.The main challenges are increasing efficiency and reducing the system's overall cost.But the efficiency of the inverter requires a balance between increasing the advantages and reducing the disadvantages, for example, increasing the gain factor to the extent that the switches withstand voltage stress without damage.From the above, the passive components, such as capacitors and inductors, are the major contributors to the total cost of inverter design.The performance of the inverter is also closely related to the number of switches used in the design.Hence, a tradeoff must be ensured while forming the impedance network with active and passive components.
Moreover, the control techniques used in the system, will greatly affect the performance of the inverter in terms of losses.Therefore, in future work, researchers can use the active component based on GaN and SiC, which may reduce  9.A summary of the reference frames used for grid-side control.[37].
component losses and improve the efficiency, reliability, and power density of these inverters.Accordingly, the expansion of this work can be predicted by considering the following objectives: • Work on developing impedance source network models in terms of boost factor, the voltage stress on components, the total number of devices, and most importantly, inverter efficiency.
• Replacing two-stage inverters with ISNs reduces converter loss and improves the reliability and power quality of the whole system.

VI. CONCLUSION AND PERSPECTIVES
Since the first model of ISNs appeared, wide areas of research have opened up for many improvements and modifications to the inverter chassis models.In this review, an overview of several models of impedance source network inverters developed from the original ZSI or semi-ZSI is presented.
In fact, although it has not received enough attention, the EB-QZSI inverter has many advantages in the field of energy conversion, and they are summarized in Table No. four.This review also provided a compilation of the modified PWM techniques to suit impedance source network inverters that have been addressed in the literature with their applications and advantages for specific goals and objectives for their essential role in improving inverter performance.Hence, the study concluded that MCBC and MSVC may be the best performers in terms of ST work ratio, effort gain, and reinforcement ratio.
Finally, the study presented a proposal for an integrated inverter input and output control technology.Including DC-link voltage control methods, current control techniques, in addition to clarifying the concept of decoupling control in order to achieve the best performance of the grid connected system.

Fig. 12 .FIGURE 12 .
Fig.12.illustrates the various modulation schemes that are used with the three-phase ZSI, which can be classified into

FIGURE 24 .
FIGURE 24.Active current signal with the grid voltage.

FIGURE 25 .
FIGURE 25.Reactive current signal with the grid voltage.

4 )
DECOUPLING CONTROL OF THE CURRENT LOOP In Fig. 29(a), if v ao , v bo , v co , are the output voltages of the inverter and were V a , V b , V c , are the three-phase voltages of the grid, L a = L b = L c and they are the inductances of the line filter, r a = r b = r c are the internal resistances of the inductors, and finally i a , i b , i c are the inverter output currents.The voltage equations in the normal (a, b,

FIGURE 31 .
FIGURE 31.Vector analysis diagram for voltage change caused by the PV system.

TABLE 2 .
Comparison of the configuration of ISNs.

TABLE 3 .
Comparison of the characteristics of ISNs.

TABLE 4 .
Comparison of the optimization results of different approaches.