Solar Energy 198 (2020) 247–274

Contents lists available at ScienceDirect

Solar Energy
journal homepage: www.elsevier.com/locate/solener

Review on novel single-phase grid-connected solar inverters: Circuits and T

control methods
Ersan Kabalcı
Department of Electrical and Electronics Engineering, Faculty of Engineering and Architecture, Nevsehir Haci Bektas Veli University, 50300 Nevsehir, Turkey

A R T I C LE I N FO A B S T R A C T

Keywords: An ever-increasing interest on integrating solar power to utility grid exists due to wide use of renewable energy
Grid-connected inverter sources and distributed generation. The grid-connected solar inverters that are the key devices interfacing solar
Micro inverter power plant with utility play crucial role in this situation. Although three-phase inverters were industry standard
Solar inverter in large photovoltaic (PV) power plant applications, the microgrid regulations increased the use of single-phase
PLL
inverters in residential power plants and grid interconnection. This paper presents a detailed review on single-
MPPT
Evolutionary MPPT
phase grid-connected solar inverters in terms of their improvements in circuit topologies and control methods.
Even though there are many reviews have been proposed in the current literature, this study provides a dif-
ferentiating approach by focusing on novel circuit topologies and control methods of string and micro inverters.
The single and multi-stage solar inverters are reviewed in terms of emerging DC-DC converter and unfolding
inverter topologies while the novel control methods of both stages have been surveyed in a comprehensive
manner. The isolated and transformerless circuit topologies have been investigated by reviewing experimental
and commercial devices. The soft computing, evolutionary and swarm intelligence based algorithms have been
summarized in MPPT methods section while current injection and grid-connection control methods of unfolding
inverters stage have been presented with and without PLL architecture. There are many papers have been
compared and listed in each section to provide further outcomes which is followed by a summarizing discussion
section and conclusion.

1. Introduction conversion requirements especially in low-voltage range (Schweizer

and Kolar, 2013) where the wide variety of power converter studies on
The use of renewable energy sources (RESs) is increasing day by day DC-DC converters and cascaded multilevel inverters have been pro-
in electricity generation due to their variety and support to utility grids. posed in the literature. The PV power plants hold the biggest share by
Besides the opportunities, the wide integrations of RESs to utility grid increasing their installed capacity up to 400 GWp by the end of 2017
have brought several challenges in terms of planning, operating, which is estimated to tackle 1000 GWp capacity by 2022 (Jaeger-
maintenance and management issues. The distributed generation (DG) Waldau, 2017; Masson et al., 2018; Schmela, 2018). The total installed
policies set by governments and demand side management (DSM) re- capacity of global PV power plants has been increased more than 100
quirements of distribution system operators (DSOs) have leveraged the GWp in 2017 as seen in Fig. 1. The main contribution is supported by
interoperability plans for generation plants including solar photo- subsidies provided by US and European governments. The decreased
voltaics (PVs), wind turbines, fuel cells, and micro sources. A DG system installation and generation costs of PV technology promise that incre-
comprised by various type of energy sources requires appropriate ment of PV power plants will tackle other RESs. Although the installed
power electronic devices for power conversion for coupling at a single capacity of PV power plants is increasing, the overall efficiency which
bus bar. The grid-connected inverters which are required for RES and depends on efficiencies of PV module, power converters and control
DG integration to utility play crucial role in resource management. methods pursue their importance. The main sections of a power con-
Moreover, the DC-DC converters are also required for regulating the DC verter which is comprised by DC-DC converters, inverter stages and
power generated by PV or fuel cell sources while inverters are inter- control methods are reviewed in several studies (Çelik et al., 2018;
facing the entire power plant with grid (Kabalci, 2015; Kabalci and Meneses et al., 2013; Romero-Cadaval et al., 2013). The commercial
Kabalci, 2018, 2017). and industrial PV inverter topologies have been improved to obtain
The microgrid applications have promoted efficient power maximum efficiency, low cost, lower sizes in terms of weights and

E-mail address: [email protected].

https://doi.org/10.1016/j.solener.2020.01.063
Received 16 October 2019; Received in revised form 16 January 2020; Accepted 22 January 2020
Available online 29 January 2020
0038-092X/ © 2020 International Solar Energy Society. Published by Elsevier Ltd. All rights reserved.
E. Kabalcı Solar Energy 198 (2020) 247–274

Fig. 1. Increment of installed global PV capacity (Masson et al., 2018).

volumes comparing to isolated inverters. Furthermore, the solar array MOSFET and insulated gate bipolar transistor (IGBT). The increased
strings and configurations are effective on obtaining maximum effi- switching frequency have tackled some of aforementioned deficiencies
ciency under partial shading conditions (Shams El-Dein et al., 2013a, such as poor power quality and higher THD, but they are still far from
2013b). ensuring the desired overall efficiency (Ankit et al., 2018; Goroohi
The main target of a PV power converter is to harvest available Sardou et al., 2018; Jana et al., 2017).
maximum power from PV beam and transferring to utility grid. Thus, The most recent centralized inverter configurations are comprised
the control algorithms known as maximum power point tracking by using 3L-NPC or 3L T-type inverter topologies which are shown in
(MPPT) play crucial role on maximizing the yielded output power from Fig. 3a and b respectively (Kouro et al., 2015). The 3L-NPC provides
a PV array. The whole configuration defined by PV arrays determine the constant common mode voltage (CMV) and low THD ratio comparing to
arrangement of power converter and control methods. Although there conventional inverter topologies. However, it requires higher isolation
are several definitions exist on power converter types used in grid and high number of switching devices for inverter configuration. The
connection of PV power plants, three main categories are the most rated power of a single central inverter is mostly lower than 1 MW to
common topologies as centralized, string and multi-string inverter increase power efficiency. The constant CMV that is supplied by neutral
(Díez-Mediavilla et al., 2014; Romero-Cadaval et al., 2013; Shayestegan point of 3L-NPC DC bus increases its advantages against conventional
et al., 2018; Sridhar and Umashankar, 2017; Zeb et al., 2018). H4 inverter topology. Therefore, the transformerless operation does not
There have been numerous studies presenting single-phase and cause any leakage current problem or modulation deficiencies in 3L-
three-phase inverter topologies in the literature. The most common PV NPC topology. The T-type topology which is introduced by Conergy is
inverter configurations are illustrated in Fig. 2 where the centralized PV used to clamp the phase voltage of utility grid to comprise a zero-vol-
inverters are mainly used at high power solar plants with the PV tage level by using two-way switching semiconductors. Thus, 3L T-type
modules connected in series and parallel configurations to yield com- inverter can operate without any transformer at the output as 3L-NPC
bined output. The conventional centralized inverters have been used for topology does (Ahmad and Singh, 2018; Faraji et al., 2017; Kouro et al.,
long years in PV plants due to their power density and MW scale con- 2015).
version capabilities. The main configuration of central inverters is de- The main drawback of centralized type inverter is the use of single
signed to interface large PV power plants to the utility grid. The central MPPT algorithm which limits the overall efficiency regarding to the
inverter topologies are mostly based on two-level (2L) full bridges or mostly shaded solar array. The string inverter is an undersized type of
recent three-level (3L) configurations such as neutral point clamped centralized inverter since each string is directly connected to an in-
(NPC), conventional H-bridge (H4), T-type or emerging voltage source verter and, total power capacity is increased with coupled inverter ar-
inverter (VSI) topologies. rays. In string inverter configuration, an additional boost transformer or
The main idea behind using these inverter topologies is their single DC-DC boost converter is required if the string voltage is not high en-
DC bus requirement which can be easily met by a simple DC-coupling of ough. The multi string PV inverters based on boost DC-DC converters
PV power plants. The DC bus voltage is mostly kept lower than 1000 V with lower cost and lower size are improved to tackle this drawbacks
limit of PV module insulation range and typically up to 800 V to pre- since they associate the advantages of centralized and string inverters
vent overvoltage risks of large PV array strings. The output stage of an due to their flexible design features (Kabalci et al., 2015c, 2015b,
inverter is comprised by a line frequency (LF) three-phase transformer 2015a; Romero-Cadaval et al., 2013).
that decreases losses and increases the low voltage (LV) output for The string inverters are essential to integrate any PV string with
providing a connection to medium voltage (MV) distribution lines. The dedicated MPPT control to utility grid. The configuration of this in-
widespread topologies of central inverters provide efficiency around verter type is mostly implemented in single-stage or two-stage topolo-
85–90% and they are produced in bulky and heavy structures due to gies regarding to including a DC-DC stage before inverter section or
transformers and coolers. Besides, the 2L inverter topologies cause being comprised by just a single inverter stage. The two-stage inverters
higher total harmonic distortion (THD) ratio and lower power factor are equipped with MPPT control at DC-DC conversion stage, and sup-
comparing to 3L topologies (Jana et al., 2017; Kouro et al., 2015). plies regulated DC bus voltage to input terminals of inverter.
While the first central inverters were based on thyristors as On the other hand, unfolding inverter stage can be implemented
switching devices, they have been evolved to power converters using with or without galvanic isolation depending on the use of line

248
E. Kabalcı Solar Energy 198 (2020) 247–274

Fig. 2. PV inverter configurations.

transformer at the output stage. Besides heavy and bulky LF transfor- recent achievements in circuit topologies and control methods have
mers, it is possible to use high frequency (HF) transformer to enable leveraged the overall efficiency of string inverters up to 97%-98% ac-
galvanic isolation at the converter stage. Any architecture including cording to producer data provided by SMA, Danfoss, Delta and others
isolated or non-isolated topologies with single or two-stage has led to (Gotekar et al., 2015; Guo et al., 2016; Zeb et al., 2018). Fig. 4 illus-
implementation of different string inverter topologies as shown in trates a collection of the most widely used circuit topologies of in-
Fig. 4. The transformerless inverter topologies are preferred due to their dustrial and residential string inverters. The conventional H-bridge
reduced cost and higher efficiency opportunities comparing to isolated topology which is named as H4 is an industry standard in string in-
devices. On the other hand, the isolation requirements are met by using verters, but the modified and improved varieties of H4 are also used to
capacitive isolation providing to overcome voltage fluctuations and increase efficiency and power quality of inverter. The common objec-
grounding issues in transformerless topologies. tive of H4, H6, NPC and T-type topologies is to stabilize the CMV in
The most common inverter topologies used in string PV inverters operation. The second inverter category comprised by H5 and HERIC
are conventional H4 topology, improved H5 topology, highly efficient topologies are improved to prevent circulating leakage currents owing
and reliable inverter concept (HERIC), and H6 configurations. The to additional switching semiconductors (Rizzoli et al., 2016).

Fig. 3. Centralized PV inverter configurations, (a) 3L-NPC topology, (b) 3L T-type topology.

249
E. Kabalcı Solar Energy 198 (2020) 247–274

Fig. 4. String PV inverter topologies, (a) H4 topology, (b) H5 topology, (c) HERIC topology, (d) H6D1 topology, (e) H6D2 topology, (f) two-stage topology, (g)
isolated two-stage dual active bridge topology.

250
E. Kabalcı Solar Energy 198 (2020) 247–274

Table 1 comparing to other topologies using HF or LF transformers. In addition

Comparison of H5, H6, and HERIC topologies. to efficiency, the weight, cost, and size of transformerless topologies are
Switching Frequency (kHz) H5 (W) H6 (W) HERIC (W) obviously lower than others. The mid-point dc link capacitor and PV
capacitors provide successful solutions for replacing galvanic isolation
10 79.64 79.30 66.60 in terms of eliminating leakage current and parasitic capacitance
15 91.43 88.10 78.10
(Kerekes et al., 2011). The two-stage inverter topologies presented in
20 103.22 98.89 89.89
25 115.02 108.35 101.68
Fig. 4f and g are used as transformerless topologies as well as shown in
their illustrated configurations. The conventional boost topology, fly-
back and interleaved converter topologies are widely used in two-stage
Three fundamental modulation schemes used for controlling the H4 configurations. On the other hand, dual active bridge (DAB) topologies
and H5 topologies seen in Fig. 4a and b are unipolar, bipolar and hybrid seen in Fig. 4g are widely conducted in string and multi-string PV in-
methods. It is noted that the unipolar and hybrid modulation methods verters due to their three-level output voltage generation capability.
generate three-level output voltages with lower core losses but higher Although the string inverters improve MPPT efficiency replacing the
leakage current in isolated device configuration. On the other hand, the single MPPT deficiency of central inverters, it suffers efficiency loss
bipolar modulator causes two-level output voltage with higher core when any of PV module is shaded in a string. The shading and energy
losses and lower leakage current. Therefore, it is suggested to use bi- loss issues forced to shift solar inverter technologies from string inverter
polar sinusoidal pulse width modulation (SPWM) method to decrease to micro inverters that are located at each module in a PV string or
the leakage current magnitudes and filter requirement (Ahmad and array.
Singh, 2017a, 2017b; Islam and Mekhilef, 2014; Rizzoli et al., 2016). Thus, the shading problem is limited to effect just only the shaded
The HERIC topology which is implemented to eliminate CM leakage module instead of entire string. The micro inverters have been devel-
current improves the overall efficiency comparing to H4 topology due oped and widely used in low-power residential PV systems (Meneses
to its AC bypass networks as shown in Fig. 4c. The bypass switching et al., 2013). The micro inverter configurations are improved to provide
generates zero-level voltage output and interrupts the reactive power MPPT control for each PV module and the rated power is generally
flow between filter and DC bus capacitor (Islam et al., 2015; Kouro between 100 and 400 W for any inverter to handle maximum power
et al., 2015; Zeb et al., 2018). Xiao et al. (2011) have noted that the rate of a single PV module. The emerging micro inverter technology is
HERIC topology has the least switching loss comparing to H5 and H6 based on mimicking the device topologies presented in Fig. 4. The
topologies. However, HERIC itself represents the worst characteristics micro inverters are also categorized according to isolation and inverter
in terms of CM while H6 is noted as the best one among others. Xiao stage number as presented in string inverters. The recent two-stage
et al. have implemented a generic testbed to compare efficiency and micro inverters are implemented with HF transformers in order to de-
operation conditions of three topologies where the input voltage varies crease size and volume while non-isolated inverter are seen in single-
between 340 and 700 Vdc, and rated power at 1000 W. The filter in- stage or multi-stage configurations such as two-stage or three-stage
ductances have been defined as 4 mH while a 6.6 μF capacitor has been topologies (Çelik et al., 2018; Khan et al., 2017; Petreuş et al., 2013;
used as filter capacitance. The common mode capacitors are used as 2.2 Rajgor, 2013; Sher and Addoweesh, 2012).
nF for each device topology where the switching loss analysis results In this paper, the micro inverters are surveyed in a comprehensive
under various switching frequencies are shown in Table 1 (Xiao et al., way to present a reference on converter stages, inverter topologies,
2011). MPPT methods, isolation and operation conditions. Therefore, a de-
The H6 topology which is introduced by Ingeteam is an improved tailed literature survey is performed to specify current situation of grid-
structure with an additional switch on negative DC bus of H5 topology. connected single-phase solar inverters, research tendencies, and evol-
It has been developed in two different configurations as seen in Fig. 4d ving circuit topologies and control methods. On the other hand, the
and e that are named H6D1 and H6D2 respectively implying the design criteria of single-phase inverters are presented by considering
number of capacitive dividers and diode arrangements. Both config- international standards, safety regulations, and commercial products
urations prevent interaction of freewheeling diodes with passive com- that are used in the area.
ponents and thus, they enable to obtain unipolar output comparing to This paper is organized as follows. In Section 2, current situation of
H5 topology. Moreover, the leakage currents are eliminated due to single-phase inverters and state-of-the-art are presented as a collection
freewheeling diodes (Islam et al., 2015; Kouro et al., 2015). The sur- of device and controller configurations. The DC-DC converter stages of
veyed researches have shown that the transformerless inverter topolo- inverters are introduced in Section 3 with non-isolated and isolated
gies are the most advantageous among others owing to several ways topologies. The survey of MPPT methods that are assumed as PV side
known in the literature to minimize the leakage current without de- controller are analysed in Section 4 while the unfolding stage of single-
creasing the efficiency and eliminating the transformer requirements phase inverters, namely grid side device topologies are given in Section
(Dutta et al., 2018; Kerekes et al., 2011; Meneses et al., 2013; Romero- 5 which is followed by inverter control methods for power regulations
Cadaval et al., 2013; Serban et al., 2017). and grid connection requirements in Section 6, and discussions and
It can be noted that the most recent tendency for solar inverters is future research anticipations are drawn in Section 7. Finally, conclu-
transformerless, single-stage and single-phase configurations as they sions and obtained results are presented in Section 7.
have been surveyed in several papers (Dutta et al., 2018; Kerekes et al.,
2011; Meneses et al., 2013; Romero-Cadaval et al., 2013). The galvanic 2. Single-phase inverter technologies and state-of-the-art
isolation is built on the DC-DC side at the output of converter or on the
AC side at the output of inverter with bulky line transformer. Although The solar inverters usually include an HF or LF transformer for
any of these configurations are preferred for safety, the efficiency of the ensuring galvanic isolation and voltage transformation ratios as seen in
entire device is drastically limited due to required components. On the string inverters. Particularly, HF transformers are used to increase DC-
other hand, the transformerless inverters are widely accepted in com- link voltage by four or five times at the output of DC-DC converter
mercial and industrial applications owing to their higher efficiency, which is supplied by a single PV module with input voltage typically
smaller size, and lower weight and volumes. The efficiency, weight, and ranging between 40 and 60 V in micro inverters. In addition to robust
volume of more than 400 different commercially available inverters monitoring and metering capabilities, the solar micro inverters provide
have been compared in Kerekes et al. (2011) where results denote that dedicated MPPT for each PV module differing from string inverters in a
transformerless topologies increase the overall efficiency up to 3% solar array. They are designed to be operated at low-power applications
comparing to central and string inverters. The micro inverter

251
E. Kabalcı Solar Energy 198 (2020) 247–274

Fig. 5. Solar micro inverter configurations.

Fig. 6. Isolated micro inverter topologies, (a) LF transformer grid-tied, (b) HF transformer DC-link topology, (c) HF transformer single stage topology, (d) HF
transformer two-stage topology.

configurations are shown in Fig. 5 where a string PV inverter ar- comparing to HF transformer-based topologies seen in Fig. 6b–d. The
rangement is presented on the right hand-side of figure for comparison HF DC-link based topologies increase PV module voltage at DC-DC
in terms of inverter number for each PV module. Each power converter converter stage with its lower size and light weighing structure which
is operated with MPPT controller at DC side of connection arrangement are widely used in interleaved or flyback type device configurations
seen in Fig. 5 where each module has its own controller while several (Zhang, 2013). Recently, the non-isolated topologies requiring parti-
modules in a PV string is controlled by a single MPPT controller. Thus, cular interest on leakage and fault currents are paid much attention in
the dedicated MPPT controllers increase the overall efficiency of micro micro inverter designs. Although there is a growing interest on trans-
inverter comparing to centralized and string inverters. However, micro formerless micro inverters, they should comply with strict safety reg-
inverters should tackle a number of challenging issues including its ulations and standards in addition to grid-connection standards. The
lower efficiency against string inverters (Hasan et al., 2017; Siwakoti transformerless solar inverter researches are noted in Siwakoti and
and Blaabjerg, 2018; Zhang, 2013). The literature surveys have shown Blaabjerg (2018) for decreasing leakage current and control methods
that three types of micro inverters are found as HF DC-link in single such as using clamping CMV or using common ground configurations.
stage topology, HF two-stage topology, and single DC-link topology The device topologies that are efficient in eliminating the leakage
with two-stage configuration. Additionally, non-isolated topologies are current are listed as zero decoupled topologies including H5 and HERIC
emerging as an alternative to existing arrangements. The two-stage and zero-state mid-point clamped topologies as oH5, HB zero voltage
micro inverters are widely used due to distributed control at each stage, rectifier (HB-ZVR). The novel inverter structures especially the resonant
simple structure and low DC link capacitance (Hasan et al., 2017). The network configurations require increased number of switching devices
two-stage topology provides high voltage gain at DC-DC converter, easy and passive components increasing the cost and complexity. Moreover,
power conversion at inverter stage and MPPT control on converter high number of switching semiconductors increases switching losses
stage. The isolated topologies which are implemented by using trans- due to physical state of regular silicon devices. The multilevel inverter
formers ensure voltage gain due to transformation ratios in addition to topologies such as NPC and H4 are assumed as appropriate solutions for
galvanic isolation as shown in Fig. 6. decreasing leakage current and THD ratios while increasing overall
The grid integration of micro inverter topology shown in Fig. 6a is efficiency and power quality (Hasan et al., 2017; Siwakoti and
performed by a LF transformer that has higher volume and size Blaabjerg, 2018; Yahya et al., 2018).

252
E. Kabalcı Solar Energy 198 (2020) 247–274

The integration of a solar micro inverter to utility grid requires The leakage current causing several concerns such as safety issues,
compliance to several international standards in terms of overall THD electromagnetic interference (EMI) noises, increased THD rates, and
ratio, individual harmonic orders, power factor, DC current limitations, power losses is also related with control method and CMV value. The
voltage and frequency deviation limits of normal and islanded opera- leakage current is eliminated by using a few methods such as removing
tion modes, grounding and leakage current levels, and automatic syn- connection of PV module with utility grid during freewheeling mode,
chronization. The foremost standards are IEEE 1547.1 IEEE Standard using the clamping DC-link capacitors at the input side or creating a
Conformance Test Procedures for Equipment Interconnecting Distributed connection between neutral of grid and negative terminal of PV
Resources with Electric Power Systems, IEEE 929-2000 Recommended module. The H5 topology which is a commercially patented config-
Practice for Utility Interface of Photovoltaic (PV) Systems, IEC 60364 2005 uration of recent micro inverters forces H4 section to keep the CMV
Electrical Installations of Buildings, UL 1741 Standard for Safety Inverters, constant while an additional switching device increases on-state losses
Converters, Controllers and Interconnection System Equipment for Use with two times. An improved configuration of H5 which is titled as oH5
Distributed Energy Resources, and IEC 61000-3-2 (2001) EMC Limits for brings an additional switching device for clamping the CMV. It is re-
Harmonic Emissions. According to common concern of these standards, ported that the control method of this new topology requires dead-band
three types of currents should be controlled and limited which are for each switching queues to prevent short circuit of capacitors (Tofigh
ground fault current that occurs in case of insulation failure, fault current Azary et al., 2018; Xiao et al., 2011).
that denotes total utility current, and leakage ground current that is the A bidirectional single-stage PV inverter which is implemented
resultant of capacitive parasitic components. The current control is against drawbacks of aforementioned topologies is presented in Fig. 7a
performed by a monitoring unit measuring instant levels of fault and (Xia et al., 2017) where the dc link capacitor (Clink) acts as voltage
leakage currents (Kerekes, 2009). source for PV MPPT. The power conversion section of this topology is
Recently, zero voltage switching (ZVS) topologies have been pro- based on asymmetrical half-bridge inverter while the grid-tied opera-
posed in addition to flyback or interleaved micro inverters. The ZVS tion of this inverter is achieved by replacing load resistance and capa-
topologies have been adopted to micro inverters with LC or LLC re- citance with grid connection. The two-way boost stage is comprised by
sonant converters following the DC-DC conversion stage where it S1 and S2 switches, Lb inductor and Clink capacitor. The input voltage Vin
comprises phase shifted full-bridge converter configuration. The LLC is provided by PV module while the high voltage output is obtained at
resonance circuit increases efficiency and power density of isolated Vlink terminals. The inverter section of inverter topology is comprised by
inverter topology, and it enables use of hybrid control and modulation S3 and S4 switches, and Linv inductor comprising the filter section (Xia
methods for micro inverter. As being one of the emerging micro in- et al., 2017).
verter topologies, wide operation range and voltage increment oppor- Another recent single-stage topology proposed to decrease the
tunity are the main contribution of LCC converters. There are several leakage current is shown in Fig. 7b that is known as H-NPC where the
different configurations proposed in the literature by improving pri- H-bridge configuration is comprised by NPC phase branches. Each
mary and secondary side networks of HF transformer (Abdel-Rahman, branch generates three-level output voltage, and thus the H-bridge
2012; Jeong et al., 2017; Jovanovic and Irving, 2016, 2015; Xingkui output is obtained at five-level staircase waveform where the CMV is
et al., 2016; Sun et al., 2017). The resonance network is implemented to eliminated by the three-level pulse-width-modulation (PWM) (Rojas
replace conventional half or full-bridge configurations with DAB et al., 2017). The H-NPC topology which is used for rooftop string in-
topologies that are capable to increase operating voltage range at input verters is assumed as a reasonable solution for micro inverters since its
and at the output of DC-link. low voltage requirement at the input and higher voltage increasing
The reminder of this section is organized to present emerging string capability comparing to regular NPC topology. The dv/dt ratio over the
and micro inverter topologies in threefold as non-isolated, isolated and parasitic capacitance is also decreased regarding to NPC topology that
resonant configurations. In this section, the overall device structures are increases leakage current elimination ration of H-NPC topology.
presented in terms of device topology, number of active and passive A recent common mode transformerless inverter topology that fo-
devices, leakage current properties, filtering structures, and efficiency. cuses on leakage current elimination is given in Fig. 7c (Tofigh Azary
The DC-DC converter topologies, MPPT methods, inverter topologies et al., 2018; Vazquez et al., 2015). The proposed topology deals with
and control methods are presented in the following sections. creating a CM resonant network to eliminate leakage current circulating
through PV module to ground (Vazquez et al., 2015). This topology is
2.1. Non-isolated topologies based on connecting neutral of grid with negative terminal of PV
module. Therefore, a complex control method is implemented to con-
Although the isolated inverter topologies are accepted as an in- trol inverter switches S1 and S2 while the boost stage is controlled by
dustrial standard in terms of galvanic isolation and safety requirements, using switch SB. On the other hand, passive devices of inverter topology
the novel transformerless, namely non-isolated, inverter topologies are comprised by capacitors CDC and Cf, and inductors L1, L2, and Lf, and
have been proposed for decreasing implementation costs and leakage an input capacitor that requires a boost converter to obtain high voltage
currents, and to increase the efficiency by eliminating the transformer for MPP tracking. Despite its low number of active switches, it needs
losses. The single-phase transformerless PV inverters have become an high number of passive devices that seriously decrease the efficiency.
industrial technology for a long time in grid integration of solar plants. The two-stage PV inverters comprise a DC-link between DC-DC
In recent years, these string inverter topologies lower than 5 kW rated converter and inverter stages where the fluctuations of input voltage
power have been widely used in low power solar micro inverters. are compensated by the converter stage. This configuration provides
The most recent topologies such as H-bridge, NPC, H-NPC, HERIC, stable input voltage to inverter section where the oscillations caused by
T-type, H5 and H6 are being widely used in commercial micro in- MPPT are decreased. The most widely used commercial topology, two-
verters. These configurations are designed in single-stage or two-stage stage HERIC, is presented in Fig. 7d. The topology targets decoupling
structures regarding to topology and grounding requirements. A set of the PV module from utility grid due to control of switches S5 and S6
novel transformerless configurations for single-stage and two-stage in- while diodes D5 and D6 manage the freewheeling during CM operation.
verters are shown in Fig. 7 (Chen et al., 2015; Kerekes et al., 2011; A CM and differential mode (DM) voltage control method is proposed to
Meneses et al., 2013; Rojas et al., 2017; Sasidharan and Singh, 2017; eliminate high frequency leakage currents in Chen et al. (2015).
Tofigh Azary et al., 2018; Vazquez et al., 2015; Xia et al., 2017). Al- Moreover, a complex EMI filter is comprised by using filtering inductor
though there is a growing interest on transformerless micro inverter and capacitors to decrease parasitic components by splitting reactive
topologies, they suffer from eliminating the leakage current due to elements in positive and negative terminals of grid connections.
parasitic capacitance that comes up from PV module through ground. The HERIC and H5 topologies are accepted as the most efficient and

253
E. Kabalcı Solar Energy 198 (2020) 247–274

Fig. 7. Transformerless PV inverter configurations, (a) single-stage topology with minimized capacitance (Xia et al., 2017), (b) single-stage H-NPC (Rojas et al.,
2017), (c) common-mode transformerless (Vazquez et al., 2015), (d) two-stage HERIC inverter.

leading configurations decoupling solar arrays and utility grid. The 2.2. Isolated topologies
derivatives of HERIC topology are H5 and H6 that are comprised by five
or six switching devices as their name imply (Chen et al., 2015; L. As discussed earlier, the CMV and CM current cause to several issues
Zhang et al., 2014). HERIC topology provides the least switching loss in terms of safety, EMI noises, increased THD ratios and other power
among other topologies under hard switching conditions. In Chen et al. quality deficiencies in transformerless topologies. There is a wide in-
(2015), it is noted that a single-phase transformerless HERIC inverter terest on isolated topologies including novel DC-DC converter topolo-
with SiC JFETs can provide 99% peak efficiency. While it is expressed gies and inverter configurations in order to minimize parasitic and
that peak efficiencies of H5 and H6 inverters can be increased up to disturbing drawbacks seen in transformerless device topologies. It is
98%, these topologies are not still able to cope with some drawbacks also widely accepted that isolated topologies present higher reliability
such as higher volume of passive components, decreased efficiency at and galvanic isolation comparing to transformerless topologies by
high switching frequencies, and high frequency common-mode voltage eliminating the CM disturbances. The main contribution of isolated
(Chen et al., 2015; L. Zhang et al., 2014; Xiao et al., 2017, 2015). The micro inverter topologies is galvanic isolation between PV module and
H5 topology is comprised by an additional switch to conventional H4 utility grid that facilitates decreasing the leakage current level in CM
topology that decouples PV module from grid when the inverter output where special requirements exist in non-isolated topologies. The HF
voltage is at zero. Thus, the leakage current is prevented. The HERIC isolation transformer ensures the galvanic isolation at the DC-DC con-
topology which is improved with similar idea required two additional verter section (Öztürk et al., 2018).
switches at the output of H4 inverter while H6 employs additional The single-stage isolated topologies are comprised by regular fly-
switches at the input stage of inverter. back or interleaved flyback converter topologies while the multi-stages
Zhange et al. compared three recent topologies in terms of leakage include a wide variety of DC-DC converter and inverter topologies in a
current levels under same operation conditions in Chen et al. (2015) single device configuration. The main sections are comprised by a
that H5 topology draws the least leakage current at 6 mA while others booster converter and by a power conditioning inverter in multi-stage
draw around 9 mA. On the other hand, The European efficiencies of H5, micro inverter structure. The two-stage or multi-stage topologies pro-
H6, and HERIC topologies are determined at 96.78%, 97%, and vide power decoupling and MPPT control at the first stage while active
97.09%, respectively. Although the HERIC topology presents highest and reactive control with grid connection are performed at the output
efficiency among others, it suffers from leakage current level. stage. The list of recent isolated micro inverter topologies are presented
in Fig. 8.
The power conversion of PV module is performed by single-stage or
multi-stage configurations as shown in Fig. 6 where one or more stages

254
E. Kabalcı Solar Energy 198 (2020) 247–274

Fig. 8. Isolated micro inverter topologies.

Fig. 9. The emerging resonant micro inverter topologies, (a) two-stage topology with synchronous boost and series-resonant converter, (b) quasi-Z source series
resonant converter, (c) LLCC series-parallel resonant converter, (d) series resonant converter with bidirectional switch.

are employed to increase DC-link voltage in multi-stage topologies. The widely inherited from conventional H4 structure at inverter section.
HF transformers can be located at DC-DC converter section of single- The SEPIC, push-pull, flyback, resonant, active clamp converter topol-
stage topologies while it can be coupled at DC-DC converter stages or at ogies are commonly used in DC-DC converter sections of multi-stage
inverter section of multi-stage configurations. The single-stage isolated micro inverter topologies (Hasan et al., 2017).
converters are fundamentally improved by using flyback or interleaved The DC-link between converter and inverter stages accommodates
topologies as expressed on the left side of Fig. 8. Both configurations high-voltage by decreasing decoupling capacitor requirement. This
can be implemented with regular topologies, soft switched resonant voltage increasing stage is comprised by using center-tapped trans-
topologies or three-port topologies while the multi-stage topologies are former in single stage topologies. The recent isolated topologies are

255
E. Kabalcı Solar Energy 198 (2020) 247–274

developed with center-tapped transformer in single stage or flyback DC- topologies are presented in Hasan et al. (2017) where primary side is
DC converters that are followed by a half-bridge or full-bridge inverters. comprised as shown in Fig. 9d. The switching losses of single-stage
The detailed presentation of isolated single and multi stage topol- primary side are decreased with resonant network and bidirectional
ogies are presented in the next section under DC-DC converter topolo- switches that are located before VDR section of micro inverter config-
gies. Moreover, the resonant converter and micro inverter topologies uration. The switching frequency of such a resonant converter is in-
are discussed in the following part. creased up to 1 MHz range which decreases switching losses and in-
ductance leakages while increasing the overall efficiency up to 97.6%
2.3. Resonant topologies (Hasan et al., 2017). The LLC resonant converters are usually operated
at varying switching frequencies to compensate input voltage and load
The DC-DC converters can be classified into two categories as hard changes. It is noted that the duty cycles applied to switching devices are
switching or PWM converter and resonant converters. The recently kept around 50% and dead band is applied between on-off transitions to
presented topologies belongs to PWM converter configuration since prevent switching losses in ZVS control.
they do not require any inductor or capacitor tank circuit for managing The switching frequency range which is determined by magnetizing
the switching operation. inductance LM of transformer depends to variation of input voltage for
However, resonant converters are comprised with additional L-C ensuring more flexible operation. However, wide range of switching
networks controlling the charge and discharge cycles of energy transfer. frequency causes adverse effects on performance and efficiency of
The resonant converters are commonly configured with HF transfor- converter. The magnetizing inductance determines a path for resonant
mers providing the galvanic isolation. The leakage inductance of HF current during very low or zero load operations where the decreasing
transformer causes additional voltage stress on switching devices and LM provides decrement on switching frequency and thus, converter acts
increases switching power losses of PWM converters while this stress is like a parallel resonant device under no load operation (Jovanovic and
eliminated by tank circuit of resonant converter. In addition to de- Irving, 2016).
creasing switching losses, the resonant converters are implemented in
less size and weights due to increased switching frequency capability
3. Recent DC-DC converter topologies used in solar inverters
that decreases component sizes, and it creates a suitable intermediate
converter infrastructure between PV module and inverter stage.
A typical solar inverter is located between PV module and utility
The DC-DC converter stage of a PV inverter is fundamentally re-
grid where it converts the harvested energy to AC waveform. The iso-
quired to adapt output voltage of PV module to interface inverter stage
lated topologies are widely used to meet safety requirements and to
by the MPPT control algorithm. The galvanic isolation requirement
eliminate the effects of leakage currents in grid-connected infra-
prevents increasing the switching frequency due to resultant increased
structure. As expressed earlier, the isolation is achieved by using an HF
leakage inductance and increased EMI noises in hard switched con-
transformer that provides classification of topologies with the name of
verter topologies. Therefore, the hard switching topologies are com-
DC-link, pseudo-DC link, and high frequency. It is noted that pseudo-DC
bined with resonant tank circuits to overcome these drawbacks. One of
link topology provides higher efficiency comparing others due to its
the combinational circuit topologies is shown in Fig. 9a that is com-
single stage structure (Hu et al., 2012). The micro inverter which pro-
prised by a conventional synchronous boost converter and a series re-
vides power conversion through utility grid should comply with life-
sonant DC-DC converter (SRC). The SRC provides high efficiency with
span of a PV module that is around 25 years. Therefore, the devices and
eliminating switching losses due to ZVS or zero current switching (ZCS)
components of solar inverter are required to be long-life and equipped
methods. The bidirectional power flow and core excitation enables SRC
with control methods improving the reliability and efficiency of micro
to have high power density comparing to conventional isolated topol-
inverter. The DC-DC converter section plays crucial role in two-stage
ogies. A single stage DC-DC converter can be implemented by replacing
micro inverter topologies since it is the vital stage on regulating the
boost converter with an impedance networks as depicted in Fig. 9b
fluctuating PV voltage. Some technical challenges faced in DC-DC
which is known as quasi impedance source (qZS) SRC. The Z source
converter stage are related with wide operating voltage at input, wide
network is comprised with capacitors, inductors, and diodes in a fea-
input power rate due to fluctuating irradiation and temperature effects,
tured configuration. The Z circuit acts as a short circuit to boost the
increasing power density which is required due to increment of PV
input voltage along the input terminals of SRC, and the input voltage
module power, and high efficiency demand (Hu et al., 2012; Zhang,
supplied by PV module is preregulated at this stage. The operation
2013).
modes of qZS network are defined as shoot-through and non-shoot-
These requirements are met using an intermediate storage element
through regarding to charge and discharge cycles of inductors. The
which is mostly a capacitor coupled between DC-DC converter and in-
shoot-through mode defines the charging cycle of inductor where the
verter stages. The electrolytic capacitors are usually chosen for inter-
diode is open circuit while the load is supplied with the charge of in-
mediate dc bus due to their operating lifetime ranging between 1000
ductors and capacitors in non-shoot-through cycle (Kabalci, 2018;
and 7000 h. The operating voltage and required capacitance are de-
Kouro et al., 2015).
termined according to usage section as input or output capacitor as
The single-stage inverters are also implemented in multi-resonant
shown in Fig. 10. In case of input capacitor used for low voltage input
topologies in order to improve short-circuit immunity and operation
voltage range with high efficiency. Such an inverter topology that is
known as LLCC series-parallel resonant converter is shown in Fig. 9c.
The output network of LLCC converter is also followed by a voltage
doubler rectifier (VDR) circuit. On the contrary of its complex resonant
and output networks, it includes minimum number of active devices
where series and parallel resonance devices are essential in decreasing
leakage inductance. According to a comparison table of emerging re-
sonant converter topologies, the DC-DC converters with series re-
sonance networks provide highest efficiency with 97.4% while parallel
topologies provide 96.6% at peak efficiency (Vinnikov et al., 2017).
There are several resonant converter topologies have been proposed
by integrating regular boost and LLC resonant converters in literature. Fig. 10. Circuit diagram of a conventional flyback converter for solar micro
A number of dual-mode resonant converters with single stage H4 inverter.

256
E. Kabalcı Solar Energy 198 (2020) 247–274

which ranges around 25 V to 55–60 V for PV modules, the capacitance et al., 2014).
should be selected high enough to decrease input voltage ripples (Hu The flyback topology operates as a voltage source which is not af-
et al., 2012). fected by load changes in CCM mode due to its magnetizing inductance
There are a wide variety of micro inverters proposed in literature at HF transformer seen as Lm1 and Lm2 in Fig. 11. The center-tapped
which are implemented for supplying MPPT control, voltage amplifi- transformer and two-way switching devices have been conducted on
cation at DC stage, output current and voltage control, and galvanic the grid side of proposed ZVS controlled converter in Sukesh et al.
isolation. Although many DC-DC converter topologies such as half- (2014). Another flyback converter topology ensuring CCM operation
bridge, conventional boost, resonant converters and DABs have been with open-loop controller on primary side of HF transformer has been
proposed in literature, the flyback and interleaved converters come proposed in International Electrotechnical Commission and Oruganti
forward as the most recent ones among others. On the other hand, (2014). The micro inverters based on flyback topologies are also de-
combination of both topologies, namely interleaved flyback arrange- scribed as module integrated converters (MICs) and flyback MIC (FMIC)
ments are assessed as emerging novel topologies due to their improved which are proposed in different configurations and control methods in
circuit configuration and control methods. The higher step-up cap- Edwin et al. (2014), Thang et al. (2014).
ability of these topologies meets the robust voltage increasing re- The preliminary literature survey exhibits that the DC-DC converter
quirements of micro inverters since the input voltage yielded from PV topologies of micro inverters are much more sophisticated comparing to
modules are generally between 40 V and 60 V. Moreover, HF trans- conventional non-isolated topologies. The improvements of resonant
former configuration enables to operate at high switching frequency converters and increased availability of centre-tapped HF transformers
with the aid of resonant devices besides galvanic isolation (Rezaei et al., have leveraged development of recent converter topologies. Although
2016; Zapata et al., 2016). The conventional flyback converter that is they would be improved in ZVS, resonant or push-pull topologies, they
followed by an H4 inverter stage is one of the most common micro are based on two fundamental topologies as flyback and interleaved
inverter topologies in industry. The device configuration is shown in configurations (Baka et al., 2019; Feng et al., 2019; Hasan and
Fig. 10 where the HF transformer is used to create a galvanic isolation Mekhilef, 2017; Tayebi et al., 2019; F. Wang et al., 2018; Zeb et al.,
between input and output of converter stage while the conversion ratio 2018; Zhang et al., 2018, 2019).
of transformer supplies higher voltage elevation capability to converter.
The converter topology and conversion ratio of transformer enable
flyback converter to operate as a conventional buck-boost device with 3.1. Flyback DC converter
increased voltage gain that is defined as follows where D denotes duty
cycle ratio; One of the earliest sensorless flyback converter for PV power sys-
tems has been proposed by Kasa et al. with a centre-tapped secondary
n2·D winding transformer as shown in Fig. 12 (Hasan et al., 2017; Kasa et al.,
VO = VPV
n1·(1 − D) (1) 2005). The flyback converter was not widely preferred when it has been
first proposed in the literature since there had not been much famil-
A conventional flyback topology combined in interleaved structure iarity. However, the unique configuration of flyback topology was
to comprise a single-phase inverter is illustrated in Fig. 11 where the eliminating the additional converter requirement and was able to
decoupling capacitors are located at the input and output sections of convert any DC power level to ac at the output. The main device to-
converter, and an unfolding inverter stage is comprised in H4 topology. pology is quite simple to build and required number of semiconductors
The surveyed topology which is originally proposed in Rezaei et al. has been drastically decreased comparing to any other topology.
(2016) is improved by adding active clamping switches for analysing Moreover, flyback topology provides better electrical potential due to
discontinuous conduction mode (DCM) and boundary conduction mode its isolation feature between PV module and grid connection (Kasa
(BCM) operations of converter section. It is noted in International et al., 2005).
Electrotechnical Commission and Oruganti (2014) that flyback con- Shimizu et al. proposed a flyback-type PV converter to overcome
verter draws lower peak currents during continuous conduction mode partial shading and lifetime issues of electrolytic input capacitor
(CCM) that ensures higher efficiency for dc converter and inverter (Shimizu et al., 2006). The proposed device configuration was based on
stages. The flyback topology is assumed as a proven configuration in replacing electrolytic capacitors with film capacitors at small values,
terms of efficiency and reliability with its decreased switching device and transmitting the stored energy to output over centre-tapped sec-
numbers. Sukesh et al. proposed to decrease switching losses caused by ondary windings. Hu et al. proposed two different three-port flyback
hard-switching methods by operating the flyback converter with ZVS converter for PV micro inverters where the circuit configurations are
control methods particularly during high frequency operations (Sukesh illustrated in Fig. 13a and b (Hu et al., 2013, 2012). The first topology
seen in Fig. 13a is improved to decrease effects of large capacitance
requirements in power decoupling of PV micro inverter.
The proposed topology which is called three-port flyback converter
is comprised in a novel integrated configuration with single-stage

Fig. 12. Single stage flyback micro inverter with centre-tapped secondary
Fig. 11. Circuit diagram of an interleaved flyback single-phase inverter. transformer.

257
E. Kabalcı Solar Energy 198 (2020) 247–274

Fig. 13. Three-port flyback micro inverters (a) without using electrolytic capacitors (Hu et al., 2013) (b) with power pulsation decoupling capability (Hu et al.,
2012).

power conversion and power decoupling features. This results in re- improved interleaved converters have been proposed for increasing the
duced component number, compact power stage, low cost, and in- efficiency and power density with the name of MICs in literature. Edwin
creased efficiency (Hu et al., 2013). et al. presented a detailed review on MIC topologies and control
The three-port flyback converter topology is improved by using methods in Edwin et al. (2012) which is focusing on DCM, CCM and
conventional topology where an additional switch S2 and a diode D2 are BCM operation modes of converters in PV applications. The expressed
used to realize the power decoupling function at the primary side in study classifies MIC topologies according to DC-link availability,
Fig. 13a. The two-winding secondary is used to generate ac output switching device numbers, soft switching methods, and decoupling
voltage depending to line frequency operation of two switches S3 and S4 capacitances.
in series with diodes D3 and D4, respectively. The decoupling capacitor However, it is well known that flyback inverters behave as a current
CD is located at the primary side to prevent double line-frequency stress source in DCM operation and it will cause higher current stress on
at the PV module section. The difference of excessive input power primary side of HF transformer that drastically decreases efficiency of
which is higher than output power is charged to decoupling capacitor micro inverter. On the other hand, the flyback topology itself provide a
while the deficit power in case of higher output power required is met little higher efficiency in CCM operation mode due to its lower elec-
by discharging the decoupling capacitor. In this case, the discharge of tromagnetic interference (EMI), decreased current stress on primary
decoupling capacitor is ensured by turning on the switch S2 and re- side and lower operating voltage comparing to DCM mode. The CCM
quired additional power is inherited to increase output power. There- operation is ensured by precisely controlled duty cycle of switching
fore, there are two operation modes exist for this converter where the devices preventing the saturation of magnetizing inductor at trans-
first one is the charging mode while the second one is the discharging former. The transition between DCM and CCM is known as BCM which
mode (Hu et al., 2013). is proposed by some researchers to be taken into consideration to
Hu et al. proposed a novel topology seen in Fig. 13b that is based on overcome aforementioned drawbacks (Dong and Tian, 2017; Edwin
the previous one to improve power decoupling capability of single stage et al., 2014; Lai, 2014; Lodh et al., 2016; Zhang et al., 2013a).
flyback configuration. In addition to its power decoupling capacity, the There are some literature outcomes expressing that the flyback
capacitor CD is used as a snubber capacitor for recycling the leakage converters operate as current source regarding to peak-current control
energy generated by primary of transformer. The additional diode D2 is methods in BCM mode. It is also possible to increase overall efficiency
included to prevent reverse current flow of power decoupling capacitor by eliminating the diode losses occurred in reverse-recovery and ap-
to PV module in this topology while diode D3 provides a leakage current plying ZVS control to flyback converter. The main issue in flyback
discharge path in the lower loop (Hu et al., 2012). converter is the increment of primary current while the output current
In addition to decoupling circuit improvements on primary side, decreases which is caused since the DCM and BCM modes are used
some other studies have proposed novel topologies placing decoupling together in control. Eventually, the flyback micro inverter should be
circuits at secondary side of flyback converter. Such topologies are operated in DCM mode around zero-crossing points of grid voltage for
defined with the names of active clamp converters and HF ac link fly- decreasing the switching losses and should be operated in BCM mode
back converters as conducted in Hasan et al. (2017). These kinds of during the other points to increase overall efficiency (Dong and Tian,
novel topologies are focused on replacing electrolytic capacitors with 2017; Zhang et al., 2013a). The single-stage flyback converters which
long life film capacitors and providing an active decoupling or clamping are developed by adding an unfolding inverter at the grid side con-
network on primary or secondary side of HF transformer. In addition to nection are novel solutions owing to their simple control, high power
flyback converters, the interleaved and interleaved flyback topologies density and high voltage increment features. However, the limitations
are also proposed in the literature for providing long life and reliable on magnetic devices and HF transformers force researchers to discover
micro inverter topologies. novel topologies and device configurations with or without magnetic
devices. The ZCS is another control method which is applied for DCM
3.2. Interleaved flyback DC converter and BCM operation modes of micro inverters (Lai, 2014; Lodh et al.,
2016).
It is noted in Edwin et al. (2014) that inverter failures are related to The power range of a flyback micro inverter is increased by con-
reliability of components and switching devices in terms of lifetime and necting several flyback converters in parallel to comprise an interleaved
operational capabilities. The double-frequency decoupling capacitor flyback configuration. A sample device topology is shown in Fig. 14a
which is concluded earlier is one of the most important components where two-phase interleaved structure is followed by an unfolding in-
limiting the lifetime of entire micro inverter. Besides the flyback verter bridge at the utility grid side. This device topology is proposed by
topologies replacing electrolytic capacitors with film capacitors, Zhang et al. for investigating the interleaved flyback micro inverter in

258
E. Kabalcı Solar Energy 198 (2020) 247–274

Fig. 14. Two-phase interleaved flyback micro inverter topology, (a) circuit diagram, (b) current waveforms of the interleaved flyback inverter (Zhang et al., 2013b).

BCM and DCM operation modes at different load conditions and loss 4. Maximum power point tracking methods
situations. The performed analyses have proved that DCM control
strategy provides higher efficiency over BCM mode up to 200 W rated MPPT is an essential control algorithm to detect the maximum input
power. The two-phase interleaved flyback micro inverter brings the power corresponding to fluctuations on source voltage and current.
advantages on ensuring the current and power sharing, and decreasing Although the MPPT algorithm targets to maximize efficiency, it will not
the current stress on primary side by managing conduction orders of properly work if both stable and varying conditions are not tracked. The
each phase. The decreased losses of switching transitions, transformer overall efficiency of a power conversion system is depended to con-
core, and driving devices gradually increase the overall efficiency of sidering all changes on source and load variations. Therefore, an MPPT
converter. algorithm that is used to improve efficiency of a PV system should
The current waveforms of interleaved flyback converter are shown consider the changes on operating circumstances. There are a lot of
in Fig. 14b where each phase is shifted at 180° to ensure current sharing fundamental and derivative MPPT methods proposed in the literature
and isolated conduction control (Zhang et al., 2013b). The unfolding (Kabalci, 2017; Moon et al., 2015; Romero-Cadaval et al., 2013). The
inverter section is comprised by a CSI configuration where switches SA main difference on MPPT algorithms are classified into two main
and SB act as main switching devices at the primary side while diodes groups that are defined as direct MPPT and indirect MPPT methods.
D1 and D2 provides rectification on the secondary side of transformer. The direct methods are operated by detecting PV voltage, current, and/
The loss distribution analyses have been performed according to both or power while indirect methods are based on calculation of PV voltage
operation modes as shown in Table 2 where the converter has been or current at the maximum power point.
controlled under 36–60 V input voltage, 220 VAC output voltage and at The direct methods are faster in detection and response quickly to
100 kHz switching frequency. The highest share of losses has been in- fluctuations since they are based on measurements instead of calcula-
dicated with bold in the list. tions. The most widely known and used MPPT methods are perturb-
The test outcomes prove that DCM operation mode draws seriously and-observe (P&O) and incremental conductance (InCon) MPPT algo-
less losses in terms of turn-off power losses and gate driving circuit rithms among others. The commercial PV inverters are also based on
losses while core losses are slight lower than BCM mode. The presented these direct and perturbative algorithms due to their faster response
analyses have also shown that the two-phase interleaved flyback con- and reliable operation. Some MPPT algorithms have been improved by
verter provides higher efficiency under light loads comparing to single- using computational algorithms as fuzzy logic control (FLC), artificial
phase flyback micro inverter topology (Zhang et al., 2013b). neural network (ANN), particle swarm optimization (PSO), and genetic
There are several other topologies have been derived from combi- algorithm (GA). Although the computational methods provide better
nation of interleaved and flyback converters that are listed in Hasan analytical calculations, they are mostly very complicated and require
et al., 2017; Kim et al. (2013), Meneses et al. (2015), L. Zhang et al. higher system performance. On the other hand, direct algorithms can be
(2014), Z. Zhang et al. (2014). However, all the topologies are im- modified to obtain global MPPT operation in addition to local MPPT
proved regarding to developing control methods and active clamping detection (Kabalci, 2017; Romero-Cadaval et al., 2013).
features of the micro inverter configuration in Fig. 14. The control The constant voltage MPPT, open-circuit MPPT, and short-current
methods are related with MPPT controller on dc stage and synchroni- pulse-based MPPT algorithms are widely used indirect algorithms due
zation control on inverting stage. to their simplicity. These are based on calculations of the parameters
such as system voltage, open-circuit voltage, short-circuit current as
their name imply. The operating voltage is depended to PV tempera-
Table 2 ture, and seasonal changes affecting the reliability. Furthermore, these
Loss analysis of two-phase interleaved flyback micro inverter in DCM and BCM algorithms lack on reliability due to aging and pollution of PV modules.
modes (Zhang et al., 2013b). Many MPPT methods have been compared to detect design criteria as
DCM BCM shown in Table 3.
The indirect MPPT methods are not true MPPT and indeed, the
Loss (W) Percentage (%) Loss (W) Percentage (%)
measured values of open circuit voltage, constant voltage, and short-
Turn-off power loss 0.8 18.3 2.50 31.5 circuit pulse-based algorithms are based on calculated parameters in-
Gate driving loss 0.83 18.9 2.83 35.7 stead of measurement. Therefore, the response time of these MPPTs are
Mosfet power loss 0.60 13.7 0.40 5.04 lower than direct or computational MPPT algorithms. On the other
Core losses 0.84 19.2 0.99 12.5 hand, computational algorithms require complex mathematical opera-
Copper losses 0.71 16.2 0.89 11.2
Diode power losses 0.6 13.7 0.32 4.04
tions and high precision microprocessors to operate predefined itera-
tions. This kind of MPPT control is achieved by using high cost

259
E. Kabalcı Solar Energy 198 (2020) 247–274

Table 3
Comparison of widely known MPPT methods.
MPPT Method True MPPT Array Based Response Time Complexity Measured Values

Perturb-and-Observe (P&O) Yes No Variable Low Voltage, Current

Incremental Conductance (InCon) Yes No Variable Medium Voltage, Current
Sliding Mode Yes No Fast Medium Voltage, Current
Capacitor Control No No Medium Low Voltage
Load I-V Control No No Fast Low Voltage, Current
Open-Circuit Voltage (VOC) No Yes Medium Low Voltage
Constant Voltage (CV) No Yes Medium Low Voltage
Short-Circuit Pulse-Based (ISC) No Yes Medium Medium Current
dP/dV and dP/dI feedback Yes No Fast Medium Voltage, Current
Fuzzy Logic Control (FLC) Yes Yes Fast High Variable
Artificial Neural Network (ANN) Yes Yes Fast High Variable
Particle Swarm Optimization (PSO) Yes Yes Fast High Variable
Genetic Algorithm (GA) Yes Yes Fast High Variable

Fig. 15. MPPT algorithms, (a) duty cycle based (b) reference voltage based (c) output power based.

hardware. Feed forward or feedback control algorithms including ca- global MPP (GMPP) under partial shading circumstances that are
pacitor control, dP/dV, dP/dI, and load I-V MPPTs comprise another caused by effects of clouds, buildings, dust, and trees. Therefore, reg-
category. These methods mostly do not require voltage and current ular MPPT methods lack to response under partial shading and load
measurement of solar array and can provide better performance by variations. A number of GMPPT algorithms have been proposed to
integrating to existing methods (Furtado et al., 2018; Kabalci, 2017; Li address partial shading problems (de Brito et al., 2013; Furtado et al.,
et al., 2018; Montoya et al., 2016). 2018; Li et al., 2018). These methods require additional circuits and
The MPPT software can be classified into three groups regarding to complex mathematical operations and the most widely known ones are
input parameters detection as duty cycle based, voltage based, and based on soft computing and optimization algorithms.
output power-based algorithms as shown in Fig. 15. The MPPT algo- The requirements cause hindrances for commercial PV systems by
rithms are mostly based on solar array voltage due to their dependence increasing costs due to high performance microprocessors, additional
to irradiation. The regular direct MPPT algorithms perform control sensor and circuit devices, and increased convergence times. The im-
operations with duty cycle or voltage-based configurations as shown in proved MPPT methods have been enhanced by taking into considera-
Fig. 15a and b, respectively. Another MPPT configuration with DC-DC tion of DC-DC converter output voltage since it is drastically decreased
converter is performed by using output power or feedback based al- under partial shading conditions. This approach is proposed as an al-
gorithms as seen in Fig. 15c (Kabalci, 2017; Romero-Cadaval et al., ternative to GMPPT algorithms such as 0.8 VOC, reduced voltage range,
2013). Although the MPPT algorithms target to maximize yielded en- and PSO methods (Ahmed and Salam, 2017; Chen et al., 2014; Furtado
ergy, they are unable to detect the difference on local MPP (LMPP) and et al., 2018; Li et al., 2018). The MPPT methods are also adapted

Table 4
Widely used MPPT methods (Çelik et al., 2018).
Offline MPPT Methods Online MPPT Methods Hybrid MPPT Methods

Short Circuit Current Perturb-and-Observe (P&O) Analytical Calculation and P&O

Open-Circuit Voltage (VOC) Incremental Conductance (InCon) Particle Swarm Optimization (PSO)
Look-up Table Sliding Mode Genetic Algorithm
Fuzzy Logic Control (FLC) Capacitor Control Optimized FLC
Artificial Neural Network (ANN) Load I-V Control P&O and ANN
Genetic Algorithm (GA) Ripple Correlation Control Differential Evolution
Constant Voltage (CV) Forced Oscillation Direct Prediction Method (DPM)
Short-Circuit Pulse-Based (ISC) dP/dV and dP/dI feedback
Temperature Based Method

260
E. Kabalcı Solar Energy 198 (2020) 247–274

Fig. 16. Flowchart of two conventional MPPT algorithms, (a) P&O, (b) InCon (Kabalci, 2017).

according to requirements of micro inverters in terms of power de- 4.2. Soft computing based MPPT algorithms
coupling and control features.
The recent development of microprocessors and decreased cost of
4.1. Conventional MPPT algorithms microelectronics have facilitated the use of soft computing methods to
comprise embedded powerful controllers. The nonlinear power char-
Although the MPPT control could be applied to inverter stage, it is acteristics of PV modules are also handled with soft computing methods
mostly used for control of dc converter stage in micro inverters. Celik which are derived from FLC, ANN and various evolutional algorithms.
et al. have presented a comparison table in Çelik et al. (2018) where The most important contribution of soft computing based MPPT
MPPT methods are classified as offline, online and hybrid methods methods are their overcome capability on uncertainty, imprecision and
which are shown in Table 4. low efficiency under extreme fluctuations that are mostly met in con-
The most widely used conventional MPPT methods are based on P& ventional MPPT algorithms (Salam et al., 2013). Nabipour et al. pro-
O and InCon algorithms where PV voltage and current are instantly posed a new MPPT method based on fuzzy approach after surveying the
measured and processed in a comparison algorithm to detect derivation conventional methods and detecting the deficiencies in Nabipour et al.
of power. These algorithms are based on detecting whether the oper- (2017). The FLC and ANN based algorithms are the most widely known
ating point of the load is on the right-hand side or on the left-hand side and most flexible soft computing methods that are applied to improve
of MPP regarding to following equations. In case the deviation of power existing MPPT approaches in several studies (Bendib et al., 2015;
is greater than that of voltage, this means the power curve is hill- Karami et al., 2017; Messalti et al., 2017; Yilmaz et al., 2018).
climbing and thus, it is required to increase duty cycle rate as seen in The soft computing methods that are used to improve novel MPPT
(2) while vice versa for decreasing ratio of power to voltage deviation algorithms are listed by Salam et al. as shown in Fig. 17. In addition to
as given in (3). FLC and ANN, evolutionary algorithms such as GA, differential evolu-
tion, ant colony optimization, and particle swarm optimization are
∂PPV
< 0 ⇒ M = M − ΔM presented in the literature. On the other hand, there can be found hy-
∂VPV (2) brid algorithms that are comprised by combining two or more soft
∂PPV computing methods such as ANN with FLC, ANN with PSO, GA with
> 0 ⇒ M = M + ΔM conventional methods and ANN with conventional methods (Salam
∂VPV (3)
et al., 2013). Since detailed surveys and comparisons on soft computing
The flowchart of P&O and InCon MPPT algorithms are presented in MPPT methods can be seen in several sources such as (Bendib et al.,
Fig. 16a and b respectively. The linear control methods lack on har- 2015; Bizon, 2016; de Brito et al., 2013; Eltawil and Zhao, 2013; Jiang
vesting available maximum power from PV module since it draws a et al., 2017; Karami et al., 2017; Messalti et al., 2017; Yilmaz et al.,
nonlinear characteristic in power-voltage curve. There have been sev- 2018), the detailed presentation of this section is based on applications
eral studies presented in literature for conventional methods. The im- and recent advances in micro inverter controllers.
proved processor capabilities have overcome the convergence speed, Joshi et al. expresses that ANN, FLC, and GA are the three most
dismissed MPP tracking and partial shading deficiencies of conven- prominent soft computing methods that are used in MPPT applications.
tional methods. Since the conventional methods are far from providing In addition to pure use of these methods, an intersection of any type is
required high efficiency, they will not be focused on this section. used to generate hybrid MPPT algorithms as ANN FLC, ANN GA, and
However, further readings can be proposed on improved conventional GA FLC in order to benefit robustness of each soft computing method.
MPPT controllers (Çelik et al., 2018; Godoy et al., 2017; Jiang et al., The ANNs are based on statistical learning methods operated at
2012; Kasa et al., 2005; Petreuş et al., 2013).

261
E. Kabalcı Solar Energy 198 (2020) 247–274

Fig. 17. Soft computing methods used to improve MPPT algorithms (Salam et al., 2013).

modelled biological neuron systems while FLCs are based on calculation algorithm due to its higher accuracy proven by increased hidden layer
of membership grades in a particular rule base, and GAs are improved structure (Joshi and Arora, 2017; Salam et al., 2013).
by benefiting from biological systems and behaviours (Joshi and Arora, A summary of ANN based MPPT algorithms has been tabularized in
2017). Most of the recent algorithms improved regarding to GA are Table 5 where significant approaches and outcomes have been re-
known as evolutionary and swarm based algorithms such as particle marked. There are plenty of ANN based algorithms based on regular
swarm optimization (PSO) (Mohapatra et al., 2017; Zeddini et al., back-propagation methods while different researches based on other
2016), artificial bee colony (ABC) (Benyoucef et al., 2015; Fathy, 2015; training and optimization methods. On the other hand, ANN based
Pilakkat and Kanthalakshmi, 2019), ant colony optimization (ACO), MPPTs have also been implemented to decrease power losses in DC-DC
Cuckoo search, grey wolf optimization (GWO) and firefly algorithm converter switching devices.
(Bendib et al., 2015; Bizon, 2016; Jiang et al., 2017; Karami et al.,
2017; Messalti et al., 2017; Yilmaz et al., 2018).
4.2.2. Evolutionary and swarm intelligence algorithms
The evolutionary algorithms such as GA and differential evolution
4.2.1. ANN based MPPT algorithms or swarm intelligence based MPPT methods as ACO and PSO are being
A regular ANN system is comprised by many neurons which are extensively researched due to their robust prediction and estimation
used to determine the weighting rates of inputs, and the calculations in capabilities (Joshi and Arora, 2017; Mohapatra et al., 2017; Salam
hidden layer to generate outputs. A featured ANN for MPP tracking et al., 2013).
requirements is seen in Fig. 18 where the input neurons are located at The GA which is listed under evolutionary algorithms is a problem-
Layer h as solar insolation or irradiation, and temperature data. The solving method based on several biological evolution principles. The
Layer j accommodates the hidden layers and neurons determining the main target of a GA is to create better species comparing to their an-
tracking capability of MPPT algorithm regarding the accuracy of weight cestors. Therefore, GA tries to find best solution by combining and
calculation. Although there are different propagation methods as Le- random selection of existing genes. The GA algorithms are operated in
venberg-Marquardt, Back-Propagation, and Reinforcement Learning, the order of initialization, evaluation, genetic operations, and program
most of the ANN designs are improved with back-propagation training termination. The initialization starts with referring to a set of solutions

Fig. 18. Block diagram of an ANN based MPPT algorithm (Salam et al., 2013).

262
E. Kabalcı Solar Energy 198 (2020) 247–274

provided by known chromosomes which are represented with a fixed

An ANN model operated with different propagation methods has been improved for

RBFN based ANN MPPT has been improved and compared to P&O and Incremental
ANN based MPPT has been proposed for triple-junction PV module. The converter
array. The measurements and efficiencies have been compared to auto-tuning PID
three-level NPC inverter. The accuracy of MPPT is compared to FLC MPPT where

A multilayer NN has been proposed in this study for buck-boost converter of PV

code length. A fitness function is operated to evaluate each chromo-

reference point setting loop and second is P&O based fine tuning loop has been
The ANN MPPT is improved to control buck-boost converter which interfaces a

A global MPPT algorithms has been proposed under partial shading conditions.
Multiple input ANN has been implemented with 8 couples of input parameters
estimating the output parameters of the PV module. The highest correlation of

A neuro-fuzzy MPPT which is improved regarding to incremental conductance

A hybrid MPPT method consisting of two loops where first one is ANN based

conductance MPPT algorithms. Switching power losses has been significantly

some sets and a fitness value is defined for each chromosome. The
genetic operators as selection, crossover, and mutation are generated by

training set has been achieved with resilient back-propagation method

the fitness function for optimizing the population. The selection op-
eration is applied to a certain population for comprising a mating base
where crossover operators will be applied generating better genes by
using code sets. The string matching is performed randomly and mu-

efficiency compared in ANN and P&O MPPT operations

tation operators start to be changing between 0 and 1 in an iteration.
proposed. The controller provides 98.26% efficiency

The random valued iteration is sustained to operate until the stop cri-
terion achieved (Hadji et al., 2015; Harrag and Messalti, 2015; Kulaksız
and Akkaya, 2012; Salam et al., 2013).
ANN based MPPT responded more rapid

Besides GA, the differential evolution (DE) which has been proposed
by Storn and Price is one of the most widely used evolutionary algo-
rithm. It has attracted a significant interest as a global optimization
algorithm due to its efficiency and simplicity. It is also based on similar
MPPT has been proposed

decreased by ANN MPPT

operation orders as GA which starts with defining a population of n
that lacks in efficiency

particle to obtain the optimal or best solution. The initial population is

defined with random genes and then selection, crossover, and mutation
loops are operated to modify and optimize the population. The best
particles are selected and then mutation and crossover processes are
Remarks

repeated with a trial vector dui for each target vector (Zaki Diab and
Rezk, 2017). Tajuddin et al. (Tajuddin et al., 2013) define this D-di-
mensional parameter vector as X = (x1, x2,…, xD) ∈ RD where X is the
optimization criterion named objective function. The generated para-
Levenberg–Marquardt, Variousback-propagation (resilient,

meter vector is defined as follows,

X∗: f (X ) = min f (X ) (4)
Bayesian, Conjugate gradient, Gradient descent)

where the feasible solution space is restricted between lower and higher
design considerations as shown in (5),
Levenberg-Marquardt Back-Propagation
Radial Basis Function Network (RBFN)

L ⩽ X ⩽ H : L , H ∈ RD (5)
In a PV converter control, the DE is used to determine the best duty
NARMA Multilayer Perceptron

cycle Dbest by benefiting from mutation factor and duty cycles which
denotes the solution space as [dmin, dmax]. The mutation process is
Levenberg-Marquardt

presented as follows to determine the donor vector or generated mutant

Back-Propagation

Back-Propagation

particles;
Neuro-Fuzzy

dvi = Dbest + F ·(dr1 − dr 2) (6)

ANN Type

where r1 and r2 are the mutual random integers, and F is the scaling
factor. After the mutation stage, the crossover and selection operations
are get started to determine the best particles (Salam et al., 2013;
PV Voltage and Current (multiple

Tajuddin et al., 2013; Zaki Diab and Rezk, 2017). The block diagram of
a DE based MPPT tracker is given in Fig. 19a where the searching
mechanism have been shown in iterations (Zaki Diab and Rezk, 2017).
Open circuit voltage VOC

Irradiation, Temperature
Irradiation, Temperature

Irradiation, Temperature

Irradiation, Temperature
Short circuit current ISC,

PV Voltage and Current

PV Voltage and Current

inputs up to 8 couples)

Although the PSO algorithm is listed under evolutionary computing

algorithms, it differs from evolutionary algorithms since it is im-
Control Inputs

plemented regarding to cooperation and social behaviour determina-

Irradiation

tions which is defined as swarm intelligence. PSO is based on a popu-

lation as in other evolutionary algorithms where each particle provides
an optimized and best solution to the problem. Each particle has a
depended velocity that is determined by the update equation in search
Interleaved Boost

space. The equation is configured considering the recent experiences of

Summary of ANN based MPPT applications.

each particle and collective behaviours. The fundamental approach is

Buck-boost

Buck-boost

Full bridge
Converter

based on the idea that each particle has a certain velocity for dis-
SEPIC

placement and to find the best solution. At the initial stage, random
Boost

Boost
N/A

variables are located in search space and a cost function is produced to

evaluate them. Afterwards, all particles update their positions in the
Saravanan and Ramesh Babu
Harrag and Messalti (2019)
Rezk and Hasaneen (2015)

direction of the particle which provides the best solution with the de-
Rizzo and Scelba (2015)
Boumaaraf et al. (2015)

fined velocity. The update function is operated regarding to three

Çelik and Teke (2017)

Mekki et al. (2016)

parameters which are velocity of previous iteration, the particle best

solution (pbest), and the global best solution (gbest). The first velocity
Kassem (2012)

equation is presented by Eberhart and Kennedy as follows;

(2016)
Table 5

Vi (t + 1) = Vi (t ) + c1 R1 ⊗ (Pbest _i − Xi (t )) + c2 R2⊗(Gbest − Xi (t )) (7)

Ref.

where t denotes the iteration number, i is the number of particle, the

263
E. Kabalcı Solar Energy 198 (2020) 247–274

Fig. 19. Flowchart of evolutionary algorithms based MPPT, (a) DE, (b) PSO (Zaki Diab and Rezk, 2017).

vector Pbest_i is the best solution of particle i at the certain point, the vik + 1 = ωvik + c1 R1 (Pbest − dik ) + c2 R2 (Gbest − dik ) (9)
vector Gbest denotes the best solution of all particles at the certain point
while the c1 and c2 coefficients are the acceleration parameters, the Fig. 19b represents the flowchart of a PSO based MPPT algorithm
vectors R1 and R2 are random numbers uniformly distributed between 0 which is proposed by Zaki Diab and Rezk (2017). The comparison of DE
and 1, and the symbol ⊗ represents element-by-element vector multi- and PSO algorithms provides several similarities in terms of operation
plication (Fermeiro et al., 2017). Rezk et al. proposed a PSO based steps comprised by initialization, updating and optimization to obtain
MPPT algorithm regarding to (7) which is represented by the following best mutant or particle among others in the random search space. The
equation (Rezk et al., 2017); first generated samples are applied to power converter for determining
error rate of MPP, and then optimization is get started to achieve the
dik + 1 = dik + vik + 1 (8)
best solution in both algorithms.
where vik+1 represents the step size at the iteration k + 1 which is Table 6 represents a summary of evolutionary and swarm in-
depended to inertial weight ω, acceleration coefficients c1 and c2, and telligence based MPPT algorithms that have been selected due to their
random values R1 and R2 between [0,1] (Fermeiro et al., 2017; Rezk differentiating features from literature survey. The conducted survey
et al., 2017); has shown that the swarm intelligence, and particularly PSO based

264
 E. Kabalcı

Table 6
Summary of evolutionary and swarm intelligence based MPPT applications.
Algorithm Type Ref. Control Inputs Converter and Controller Remarks

Genetic Algorithm Harrag and Messalti PV power Boost Converter, Matlab Simulink GA based P&O and PID controlled GA MPPT algorithms have been proposed as
(2015) offline and online methods, and comparisons presented in terms of dynamic
response, overshoot, and ripple reduction where GA based P&O overcomes
Kulaksız and Akkaya Irradiation, Temperature, and Three-phase full bridge inverter, PSIM GA based ANN optimization has been proposed for MPPT control. The presented
(2012) PV voltage Simulation control method was promised to eliminate DC-DC converter requirement which
enables to direct connection of PV to inverter
Daraban et al. (2014) PV voltage and current Buck converter, DSP based control GA based P&O algorithms has been proposed to improve efficiency under partial
shading. The global MPPT control has been analysed with experimental testbed
Lasheen et al. (2016) PV voltage and current Boost converter, PSIM Simulation GA based constant voltage (CV) MPPT has been proposed for increasing the tracking
efficiency
Differential Evolution Zaki Diab and Rezk (2017) PV voltage and current Boost Converter, Matlab Simulink Flower pollination and differential evolution algorithms have been proposed to
decrease partial shading effect in building integrated PV systems
Tajuddin et al. (2013) PV voltage and current Buck-boost converter, PLECS Simulation A DE based MPPT has been proposed to analyse the performance comparing to hill
climbing MPPT in terms of the dynamic response under rapid and large irradiation
fluctuations. The performance of DE verified by authors
Ramli et al. (2015) PV voltage and current Buck-boost converter, PV solar array A new MPPT approach has been proposed named modified DE to overcome random
simulator based experimental system selection drawbacks of regular DE MPPT. The results have been compared and
evaluated under partial shading conditions
Xiong et al. (2018) PV voltage and current N/A, experimental An improved DE MPPT algorithm has been proposed with the name of whale
optimization algorithm and compared to conventional DE MPPT in terms of different
irradiation, temperature, and weather conditions
Zhao et al. (2018) PV voltage and current SGSADE and CEC2017 benchmark A novel hybrid algorithm based on self-adaptive Gravitational Search Algorithm
(GSA) and DE is proposed for solving single objective optimization. The self-adaptive
mechanism has been proposed

265
Muhsen et al. (2019) Weather data and component N/A, simulation A decision making system has been improved to determine optimal sizing for a
specifications standalone PV system. The algorithm has been improved with multi-objective DE
methods
Swarm Intelligence Based Algorithms Fermeiro et al. (2017) PV voltage and current Boost converter, TMS320F28027 DSP The conventional PSO algorithm has been improved to obtain more robust control in
(ACO, PSO, Cuckoo) global and local best particles
Rezk et al. (2017) PV voltage and current Boost Converter, Matlab Simulink PSO and Cuckoo Search (CS) MPPT algorithms have been proposed and compared to
incremental resistance MPPT
Titri et al. (2017) PV voltage and current Boost Converter, Matlab Simulink ACO MPPT has been improved and compared to conventional and soft computing
methods such as P&O, ANN, FLC, ANFIS, FL GA, and PSO based MPPT algorithms
Mosaad et al. (2019) Irradiation, Temperature Boost Converter, Matlab Simulink CS based MPPT has been proposed and compared to ANN and P&O MPPT methods
Sundareswaran et al. PV voltage and current Boost Converter, Simulation and PIC A new PSO method has been proposed and compared to regular PSO and P&O MPPT
(2015) microcontroller algorithms for evaluating the efficiency under partially shading conditions
Oshaba et al. (2015) Irradiation, Temperature Buck-boost Converter, Matlab Simulink Bat search algorithm has been proposed as an alternative to PSO MPPT and the
improved method has been compared for a motor drive application which was
supplied by PV
Yang et al. (2019) Irradiation, Temperature Boost converter, DS1104 HIL experimental A novel evolutionary algorithm named memetic salp swarm algorithm (MSSA) has
setup been proposed. The proposed MPPT compared to INC, GA, PSO, ABC, CS, grey wolf
optimizer (GWO), SSA, and teaching-learning-based optimization (TLBO) MPPTs
Mirza et al. (2019) PV voltage and current Boost Converter, Matlab Simulink Adaptive Cuckoo Search Optimization (ACO), General Regression Neural Network
GRNN), Fruit fly Optimization (FFO), and Dragonfly Optimization (DFO) MPPT
algorithms proposed
Douiri (2019) Irradiation, Temperature, SEPIC converter, Simulink, PSO optimized neuro-fuzzy MPPT improved and dynamic response of algorithms
output voltage compared to existing models
Saad et al. (2018) PV voltage and current Boost Converter, Matlab Improved PSO MPPT presented for dynamic response and steady-state performance
analysis of DC microgrid
Dileep and Singh (2017) PV voltage and current SEPIC Converter, OPAL-RT OP5060 Adaptive PSO MPPT has been proposed to improve overall efficiency for GMPP
tracking
Solar Energy 198 (2020) 247–274
E. Kabalcı Solar Energy 198 (2020) 247–274

algorithms are comprising most of the soft computing MPPT methods. converter stage that is followed by center-tapped HF transformer and a
The surveyed and summarized studies in Table 6 show that an in- full bridge converter as shown in Fig. 20. Sukesh et al. proposed a micro
creasing interest exist on swarm intelligence-based algorithms for im- inverter topology by using proven cost-effective flyback converter in
proving competitive MPPT methods. The recently proposed researches Fig. 20a (Sukesh et al., 2014). The proposed converter has been oper-
present more efficient outcomes comparing to conventional PSO based ated in BCM mode since the power transfer capability of flyback con-
MPPTs in terms of adaption to partial shading and rapidly changing verter is limited in DCM operation. The primary side MOSFET Sm is
weather conditions. On the other hand, the meta-heuristic methods switched on for charging the magnetizing inductance Lm to achieve
have been proposed mimicking the behaviours of many different reference current value. The charged current generates the required
swarms such as bats, wolves, cuckoos, fireflies, and others in addition to energy storage which will be transferred to utility grid by switching Sp1
ant or artificial bee colonies. and Sn1 devices regarding to positive and negative half-cycles of grid
voltage (Sukesh et al., 2014).
The proposed topology ensures reliable operation due to BCM mode
5. Device topologies in inverter section
and switching controls. An alternative topology shown in Fig. 20b is
titled flyback MIC which has drawn significant attention in recent years
Despite their low power application capabilities, the string and
inverter, and proposed as interleaved flyback MIC (IFMIC) in Edwin
micro inverters are being paid increasing attention in grid-connection
et al. (2014). The main contributions of IFMIC are noted as equal load
of PV systems and modules. The achieved technological improvements
sharing on dc converter stage, reduced input voltage and pseudo-DC
and decreasing costs of power electronics and device developments
link voltage ripples, and reducing the EMI noise by current sharing. The
have leveraged opportunities on use of single-phase inverters. The
solar inverter is operated in CCM mode at interleaved flyback stage, and
galvanic isolation is one of the most widely discussed title in circuit
fourth-order LC filter has been implemented to attenuate distortions
topologies since the HF transformer increases the cost but provides no
before transferring the converted power to utility grid.
special solution for isolation. However, the capacitive isolation enables
The inverter stages of Fig. 20b and c are operated at line-frequency
designers to build transformerless topologies with additional cares on
since they have been implemented with gate turn off (GTO) switches.
safety requirements (Siwakoti and Blaabjerg, 2018). If we recall the
Although the topology is based on fundamental H4 inverter, GTO based
inverter topologies given in figures Fig. 4, Fig. 7 and Figs. 9–14, it can
architecture eliminates LF transformer requirement on secondary side
be seen that the grid connections are performed by commonly using H4,
of HF transformer due to decreased voltage stress because of GTO
and novel topologies such as H5, oH5, H6 and HERIC type inverter
switches. Since the primary side converter is comprised by interleaved
section that follows the HF transformer or output of intermediate DC-
flyback, the BCM operation mode is the most appropriate method for
link stage.
increasing the transferred power rate in such topologies.
Hasan et al. have presented a few additional inverter topologies in a
The primary current of HF transformer is forced to track reference
review (Hasan et al., 2017) where most of them are based on flyback

Fig. 20. Various alternative inverter topologies used in solar inverters (Hasan et al., 2017), (a) Single flyback inverter with soft switching (Sukesh et al., 2014), (b)
CCM control interleaved flyback inverter (Edwin et al., 2014), (c) BCM control interleaved flyback inverter (Gao et al., 2014), (d) ILFI with soft switching (Kim et al.,
2013).

266
E. Kabalcı Solar Energy 198 (2020) 247–274

Table 7
Comparison of recent inverter stage topologies in solar micro inverters.
Ref. Component Numbers Active switch number at each half- Switching Method Peak Efficiency Galvanic Isolation
cycle

S D C L (+) (−)

(Islam et al., 2015) Flying Capacitor 4 1 2 0 2 2 Hard switching 99.2% No

(Xiao et al., 2011) oH5 6 0 2 0 3 3 Active Clamping 97.2% No
(Kerekes et al., 2011) H4-ZVR 5 5 2 0 2 2 Hard switching 94.88%0 No
(L. Zhang et al., 2014) H5 5 0 2 2 3 3 Hard switching 96.78% No
(L. Zhang et al., 2014) H6 6 0 2 2 3 2 Hard switching 97.09% No
(L. Zhang et al., 2014) HERIC 6 0 2 2 2 2 Hard switching 97% No
(Sukesh et al., 2014) Flyback Inverter 5 0 2 1 2 2 BCM soft switching 94% Yes
(Edwin et al., 2014) ILF Inverter 6 2 2 1 2 2 CCM soft switching 95.7% Yes
(Kim et al., 2013) ILF Inverter 8 2 5 1 2 2 DCM soft switching 95.11% Yes
(Gao et al., 2014) ILF Inverter 6 2 2 1 2 2 BCM soft switching 94% Yes
(Vinnikov et al., 2019) Optiverter 9 2 6 5 4 4 Quasi-resonant soft 95.3% Yes
switching
(H. Wang et al., 2018) Aalborg Inverter 6 6 1 1 3 2 CCM NA Yes
(Liu et al., 2017) Modified LLCL Aalborg 6 4 2 4 3 3 Soft switching 98.3% No
Inverter

S = Switch, D = Diode, C = Capacitor, L = Inductor, (+) = positive half-cycle, (−) = negative half-cycle.

Table 8
Grid connection standards for micro inverters (Hasan et al., 2017; DeBlasio, 2009; International Electrotechnical Commission and Oruganti, 2014; Meneses et al.,
2013).
Requirement IEC 61727 IEEE 1547-2008 EN61000-3-2 VDE

Nominal Power 10 kW 30 kW 16A 230 V –

Order (h) Limit Order (h) Limit Order (h) Limit (A) Order (h) Limit (A/MVA)

Harmonic Content 3–9 4.0% 3–9 4.0% 3 2.31 3 3

11–15 2.0% 11–15 2.0% 5 1.14 5 1.5
17–21 1.5% 17–21 1.5% 7 0.77 7 1
23–33 0.6% 23–33 0.6% 9 0.4 9 0.7
> 35 0.3% 11 0.33 11 0.5
13 0.21 13 0.4
Even harmonics are limited to 25% of the odd harmonic limits (15–39) 2.25/h 17 0.3
19 0.25
2 1.08 23 0.2
4 0.43 25 0.15
6 0.3 25–40 3.75/h
THD < 5% (8–40) 1.84/h Even 1.5/h
greater than40 4.5/h
DC Current Injection < 1% of rated output current < 0.5% of rated output current < 0.22A < 1A, max. trip time = 0.2 s
Voltage Deviations Range (%) Time (s) Range (%) Time (s) Range (%) Time (s) Range (%) Time (s)
V < 50 0.1 V < 50 0.16 – V < 85 0.2
50 ≤ V < 88 2 50 ≤ V < 88 2 V ≥ 110 0.2
110 ≤ V < 120 2 110 ≤ V < 120 1
V ≥ 120 0.05 V ≥ 120 0.16
Frequency Deviation Range (Hz) Time (s) Range (Hz) Time (s) Range (Hz) Time (s) Range (Hz) Time (s)
49 < f < 51 0.2 59.3 < f < 60.5 0.16 47.5 < f < 50.2 0.2

current, and magnetizing is maximized (Gao et al., 2014). Another ILF topologies. Although these topologies are paid more attention in
IFMIC topology which has been named as interleaved flyback inverter recent architectures, the peak efficiencies given in Edwin et al. (2014),
(ILFI) has been proposed by Kim et al. as seen in Fig. 20d (Kim et al., Gao et al. (2014), Kim et al. (2013), Sukesh et al. (2014) are lower
2013). The topology uses a decoupling capacitor for eliminating almost than all others.
double-frequency harmonics at 120-Hz of 60 Hz line frequency. Besides The efficiency and reliability of inverter stage are mostly ensured by
the converting unfolding inverter switches to IGBTs, this inverter control method in power transfer operation of micro inverter to utility
structure has been proposed with a novel control method based on grid. Therefore, the control methods and controllers implemented for
single MPPT for both interleaved phases and PLL for inverter stage (Kim grid connection are comprehensively introduced and discussed in the
et al., 2013). A detailed comparison of recent device topologies used in following section.
inverter stage of micro inverters has been summarized in Table 7 where
several particular features such as component numbers, active switch 6. Inverter controllers
numbers at each half-cycle, switching method, peak efficiency, and
galvanic isolation have been listed. The flying capacitor topology pre- Since the solar inverters are responsible for connecting the gener-
sented in Islam et al. (2015) provides the highest efficiency with the ated power at PV side to utility grid, two separate control infra-
least component number. On the other hand, it should be noted that the structures are required where the first one is located at PV side while
transformerless topologies are reported to provide higher efficiency the other one is operated at the output of inverter to interface the entire
than isolated architectures. Another noticeable issue from Table 7 is on device with utility grid. The PV side control represents converter side

267
E. Kabalcı Solar Energy 198 (2020) 247–274

control mechanism which is mostly MPPT as discussed earlier. On the phase inverter designs and proposed controllers for grid interconnec-
other hand, grid side control is requested to improve power quality and tion. The deadbeat control which is one of the widely known digital
efficiency of inverter to ensure reliable operation. Therefore, grid side control schemes has been proposed by Mattavelli (2005) for single-
controller of solar inverter should meet grid interconnection require- phase uninterruptible power supply (UPS) application. The controller
ments, provide secure grounding, and power decoupling features. The has been implemented by using output voltage and inductor current of
inverters improved for operating in single-phase grids should comply inverter where a state estimator has been proposed to overcome com-
with grid requirements described by several international and regional putational delay. Castilla (2008) proposed a linear current control
standards. The control architectures of inverters are classified into three scheme for single-phase grid-connected PV inverters. In spite of regular
categories due to device topology as two-stage, single stage without dc harmonic compensators that are parallel connected to tracking reg-
converter, and power control shifting phase (PCSP) approaches ulator, the serial connected harmonic compensator has been proposed
(Hassaine et al., 2014; Mahela and Shaik, 2017; Meneses et al., 2013). to improve synchronization accuracy and eliminating the PLL require-
All three methods are improved to achieve fast dynamic response, in- ment in Castilla (2008). In another study, Miret et al. (2009) proposed a
stant current control, and simple device architecture while PCSP pro- selective harmonic compensator to eliminate certain orders of voltage
vides additional reactive power control opportunity. and current harmonics at the output of inverter. The controller focuses
A micro inverter operating in grid-connected mode should satisfy on generating a reference current to detect the most influential har-
the grid connection standards in terms of power quality, THD ratios, monic orders and forces modulator to attenuate the detected harmo-
islanding detection, grid interfacing limits for voltage and frequency, nics. A proportional resonant (PR) current feedback control method has
and grounding. Menenes et al. have presented a list of standard re- been proposed in Shen et al. (2010) to achieve synchronization for a
quirements in Meneses et al. (2013) which are shown in Table 8 that grid-connected single-phase inverter. The output stage of inverter has
presents a summary of international standards which should be fol- been comprised by an LCL filter where the weighted average of the
lowed to ensure grid requirements in micro inverter interconnection to inverter current and the grid current have been used as feedback
grid. The most important concerns are related to THD ratios, voltage parameters in controller. Ho et al. (2009) proposed a constant-fre-
and frequency deviations, and DC current injection ratios. The injection quency hysteresis current controller to overcome interconnection defi-
ratio of DC current is a serious indicator for transformerless device ciencies while a rotating-sliding-line-based sliding mode controller
topologies and should be lower than 1% of rated output current of in- (SMC) for single-phase UPS inverters has been presented in Komurcugil
verter. However, such a low current rate requires a robust and precise (2012). Besides these control methods, it is seen that active damping,
controller for the micro inverter. Although the transformerless topolo- droop, and repetitive control based schemes are widely used in grid
gies bring increased efficiency, light weight, and decreased cost, they connection of single-phase inverters (Bao et al., 2014; Eren et al., 2015;
eliminate the galvanic isolation which causes leakage ground currents Guo et al., 2014; Xu et al., 2014; Yao and Xiao, 2013; Zhu et al., 2016).
with CMV deviations. Xu et al. have presented a summary of active damping based
The grounding is another important issue in micro inverter con- methods that are improved by using a single current feedback in Xu
trollers for decreasing the leakage ground currents. On the other hand, et al. (2014). Although the operation stability is guaranteed under
power demand is expected to be continuously met by a single PV certain current and filter parameters, it is not easy to ensure achieving a
module in micro inverter applications. Since a single PV module is high bandwidth in operation. Therefore, a number of improvements
capable to generate pure DC power at any operation instant, an internal have been done in general active damping controllers considering the
energy storage mechanism which is titled as power decoupling and LCL output filter is used in single-phase inverter as shown in Fig. 21
mostly supplied by a capacitor is required. The decoupling capacitor, where filter is comprised by inverter-side inductor L1, capacitor C1,
which is either electrolytic or film type stores the generated energy and grid-side inductor L2, equivalent parasitic resistors r1 and r2. The in-
eliminates double line-frequency deviations. The power decoupling is verter output voltage is denoted with uinv while the grid voltage is ug
also an important control procedure in micro inverter applications be- and injected grid current ig that is sampled for closed-loop current
sides grid interconnection, grounding and power quality controls control.
(Mahela and Shaik, 2017; Meneses et al., 2013). The equation denoting the relation between grid voltage and in-
Although it is easy to connect the three-phase inverters to utility jected grid current is represented in (10) while the natural resonance
grid by using regular PLL based controllers, the design of dc-bus voltage frequency that may cause to resonating current is seen in (11);
control scheme and current injection to grid are challenging issues in
i 1
single-phase inverters due to double line-frequency ripples and non- Guginv =
L1 L2 C1 s 3 + (r2 L1 + r1 L2) C1 s 2 + (L1 + L2 + r1 r2 C1) s + r1 + r2
trajectory nature of single-phase. Table 9 represents a list of single-
(10)

Table 9
Single-phase inverter controllers and features.
Ref. Topology and Rated Power Application Filter Controller Type Reference Parameter Feedback Loop

Mattavelli (2005) Full Bridge – 2 kVA UPS LC Deadbeat VO, IInv Multiple
Castilla (2008) Full Bridge – 1.5 kW PV LCL Current control IInv, IRef Single
Miret et al. (2009) H4 – 500 VA APF NA Selective compensator VO, IInv Multiple
Shen et al. (2010) Two-stage (H4 converter, Half Bridge Fuel Cell LCL PR current control Weigthed average of IInv, Single
inverter – 5 kW) IGrid
Ho et al. (2009) H4 – 300 W DG L Constant f hysteresis I IL, VGrid Single
Komurcugil (2012) H4 – NA UPS LC Sliding mode control VO, IO Multiple
Yao and Xiao (2013) Dual-Buck Full-Bridge – 1 kW VSI LC Hysteresis I IGrid, VGrid Multiple
Guo et al. (2014) H4 – NA VSI L Repetitive IGrid Single
Xiao et al. (2011), Xu et al. oH5 – 5 kW PV LCL Active damping (AD) based IGrid, VGrid Multiple
(2014)
Bao et al. (2014) H4 – 5 kW VSI LCL Capacitor current AD IGrid, VGrid Multiple
Eren et al. (2015) H4 – 1.8 kW PV, Wind LCL Droop control VBus, IGrid Multiple
Zhu et al. (2016) H4 – NA VSI LCL Fractional-order repetitive IGrid Single
control

268
E. Kabalcı Solar Energy 198 (2020) 247–274

denoting the injected grid current exist to reject the resonance (Xu
et al., 2014).
Although the active damping control schemes presented in Xu et al.
(2014) do not include PLL controller, Bao et al. have proposed a similar
active damping based controller that uses capacitor–current as feedback
parameter (Bao et al., 2014). The circuit schematic of single-phase grid-
connected VSI with LCL filter is seen in Fig. 23a while the transfer
Fig. 21. LCL filter of grid-connected single-phase inverter (Xu et al., 2014). function of multiple feedback controller is shown in Fig. 23b. In the
proposed application of (Bao et al., 2014), authors express that PLL is
1 L1 + L2 used to synchronize the the reference of injected grid current ig with the
fres = grid voltage vg. The feedback coefficient and gain of injected currents
2π L1 L2 C1 (11)
are represented by Hi1 and Hi1 blocks respectively. The injected grid
Fig. 22 represents improved active damping control methods that current regulator is described with Gi(s) while the transfer function of
are developed by using the single-phase grid current control method. inverter is denoted with Ginv block.
The controller using grid current injection method is seen with black Zeb et al. have presented a list of inverter controllers as non-linear
lines in Fig. 22 where kg is the grid voltage, iref is the reference current, group including SMC, partial feedback linearization (PFL), hysteresis,
and Gc(s) denotes the current regulator which is followed by LCL filter H∞, model predictive (MPC), ANN, repetitive, FLC, and autonomous
blocks. The controller seen in Fig. 22a is improved with active damping controllers in Zeb et al. (2018). However, the surveyed controllers and
method based on capacitor current control loop that is controlled by control schematics presented in Zeb et al. (2018) are mostly proposed
kAD proportional feedback parameter. The transfer function of this for three-phase inverters and medium power applications. Therefore,
closed-loop controller denoting the injected current is proposed as seen the PLL based control algorithms are listed in three-phase applications.
in (12) that is achieved by revising (10);

ig GC (s ) 7. Discussion and future studies

Giref =
L1 L2 C1 s 3 + kAD L2 C1 s 2 + (L1 + L2) s + GC (s ) (12)
It is seen that an increasing and extensive interest has been paid to
The transfer function of ug to ig is presented in this situation as string and micro inverters that are improved mostly to be used in single-
follows; phase applications. The main motivation behind this situation is highly
i −kAD C1 s − L1 C1 s 2 created by integration of distributed energy resources (DERs) to utility
Gugg = grid. Many countries are promoting the DG for individuals besides the
L1 L2 C1 s 3 + kAD L2 C1 s 2 + (L1 + L2) s + GC (s ) (13)
industrial plants, and a wide variety of regulations are available for
Since there is a correlation exist between capacitor current and in- subsiding residential users that are now defined as prosumer. Therefore,
jected grid current, the capacitor current feedback is inserted to de- the single-phase inverters are increasing their importance and wide-
veloped controllers seen in Fig. 22b–d. On the other hand, an additional spread use in RES integration in microgrid and nanogrid architectures.
closed-loop proportional controller which is based on grid-side inductor The reviewed studies and papers listed in current literature have
voltage is added to active damping controllers seen in Fig. 22c and d for shown that there are many aspects affecting the design and im-
decreasing the current harmonics. The revised transfer function of ug to plementation of a single-phase inverter exist. Most of the leading factors
ig is obtained as seen in (14); are related to isolation requirement, single or two-stage architecture,
−L1 C1 s 2 MPPT control approaches, inverter stage topology, and inverter con-
i
Gugg = trollers in grid-connection. Although the isolation was an industry
L1 L2 C1 s3 + kAD L2 C1 s 2 + (L1 + L2) s + GC (s ) (14)
standard and a strict requirement until a few years ago, it is seen that
It is seen that the first-order derivative parameter is removed from the novel circuit topologies tackle the leakage current elimination by
numerator, and a second-order derivative feedback (kADL2C1s2) several capacitive isolation methods. On the other hand, the input

Fig. 22. Active damping based control schemes (Xu et al., 2014) (a) capacitor current feedback, (b) equivalent capacitor voltage feedback, (c) equivalent grid-side
inductor voltage feedback, and (d) equivalent injected grid current feedback.

269
E. Kabalcı Solar Energy 198 (2020) 247–274

Fig. 23. PLL based active damping control scheme (Bao et al., 2014) (a) circuit schematic, (b) transfer function.

capacitor arrangements are also proposed to overcome leakage current As a summary of discussions, the multi-stage single-phase PV in-
and power decoupling requirements in literature. The multi-stage verters are required to be improved in terms of power decoupling, ef-
topologies are convenient to eliminate parasitic effects. Therefore, the ficiency under partial shading, operation mode control of converter
reliability, power density, overall cost, and performance of transfor- stage, grid-connection and islanding detection of unfolding stage, and
merless inverter topologies are optimized in recent designs. The im- device topologies to eliminate potential hazards of transformerless
proved power semiconductors such as SiC and GaN based mosfets are operation.
replacing the conventional Si switching devices, and thus the efficiency
of inverter is significantly increased by decreasing the switching and
cooling losses. It is foreseen that increased availability of SiC and GaN 8. Conclusion
devices will allow to replace IGBTs with MOSFETs in inverter topolo-
gies, and thus high frequency switching will be an industry standard. The integration of PV power plants to utility grid has been rapidly
On the other hand, the operating mode of switching device in a DC-DC increased in single-phase applications due to microgrid and nanogrid
converter affects the efficiency of inverter where DCM or BCM modes concepts that are brought by DG. Such a widespread application area
are remarkable in addition to regular CCM. which is integrated to utility grid should be managed in accurate, ef-
It is also seen that the efficiency of DC-DC converter stages are ficient, and reliable ways. There are numerous standards defining the
improved by novel topologies such as flyback, interleaved flyback, and interconnection and disconnection of single-phase inverters to utility
resonant topologies while the soft switching methods based MPPT grid available. The solar inverters are one of the most extensively re-
controllers are promising higher efficiency even under partial shading searched topics in emerging power electronics due to their variety in
operations in PV applications. The conventional MPPT methods such as circuit and control architectures. In this paper, a comprehensive lit-
P&O, InCon, and Hill-climbing are given up while the increased interest erature review has been presented considering string and micro in-
has been directed to ANN based, evolutionary and swarm intelligence verters which differentiate from central inverter in terms of power level
algorithms. The surveyed papers have shown that evolutionary and and application areas. The targeted survey group has been comprised
swarm intelligence based MPPT algorithms provide higher efficiency in by single-phase grid-connected inverters, and single and multi-stage
local and global MPPT processes. Therefore, it should be noted that inverters have been reviewed. The multi-stage topologies that are im-
regular and conventional MPPT methods will not be considered any plemented with DC-DC converter and unfolding inverter stages have
longer. been researched for novel circuit topologies and recent control
Another remark achieved from literature surveys is the circuit methods. The widely known and conventional topologies have been
topologies in inverter section of a single-phase solar inverter that are disregarded to provide further focus on emerging systems. The litera-
beyond the conventional H-bridge, namely H4, or two-level topologies. ture survey has shown that there is an increasing tendency on isolated
The most widely used innovative topologies are improved with H5, and resonant DC-DC converter topologies due to galvanic isolation re-
oH5, H6, H6D1, H6D2, HERIC, and resonant circuit architectures. It is quirements. However, non-isolated topologies have also been re-
anticipated that novel inverter topologies will be improved to overcome searched in order to eliminate HF transformer requirement and for
reliability, voltage stress on switches, current injection control, and removing the leakage current by capacitive coupling. It has also been
transformerless operation issues. Moreover, it is obvious that the im- noted that the unfolding stage topologies improved by using novel
provements seen in circuit topologies will also require novel control structures differing from conventional two-level and H-bridge circuit
methods and schemes. The PLL based control methods were industry schemes. In addition to circuit architectures, the control methods of
standard in grid-connection of three-phase inverters, and they have converter and unfolding stages have also been surveyed where novel
been adapted to single-phase interconnection. However, there are many MPPT methods, and grid-connection schemes have been presented in a
reliability issues and challenging situations have been reported for comprehensive way. The literature surveys have been enriched with
single-phase PLL applications in the literature. The literature review figures illustrating device topologies and control methods, and tabu-
have also shown that many novel interconnection methods have been larized summaries have been presented to provide further reading on
proposed without PLL in single-phase applications. The active damping, certain issues. As a final remark, this review is anticipated to be a
hysteresis band controllers, selective harmonic elimination compen- comprehensive source for researchers and professionals who are in-
sator, and PR controller are some of these controllers. It is anticipated vestigating solar string and micro inverter in terms of device topologies,
several alternative control methods replacing the PLL requirement will control methods, development and implementation directions.
be proposed for interconnection of PV inverters to single-phase grid.

270
E. Kabalcı Solar Energy 198 (2020) 247–274

Declaration of Competing Interest 03.098.

Dutta, S., Debnath, D., Chatterjee, K., 2018. A grid-connected single-phase transformer-
less inverter controlling two solar PV arrays operating under different atmospheric
The authors declare that they have no known competing financial conditions. IEEE Trans. Ind. Electron. 65, 374–385. https://doi.org/10.1109/TIE.
interests or personal relationships that could have appeared to influ- 2017.2711577.
ence the work reported in this paper. Edwin, F., Xiao, W., Khadkikar, V., 2012. Topology review of single phase grid-connected
module integrated converters for PV applications. In: IECON 2012 - 38th Annual
Conference on IEEE Industrial Electronics Society. Presented at the IECON 2012 -
References 38th Annual Conference of IEEE Industrial Electronics. IEEE, Montreal, QC, Canada,
pp. 821–827. https://doi.org/10.1109/IECON.2012.6388645.
Edwin, F.F., Xiao, W., Khadkikar, V., 2014. Dynamic modeling and control of interleaved
Abdel-Rahman, S., 2012. Resonant LLC Converter: Operation and Design. Infineon
flyback module-integrated converter for PV power applications. IEEE Trans. Ind.
Technologies North America (IFNA) Corp. 19.
Electron. 61, 1377–1388. https://doi.org/10.1109/TIE.2013.2258309.
Ahmad, Z., Singh, S.N., 2018. Improved modulation strategy for single phase grid con-
Eltawil, M.A., Zhao, Z., 2013. MPPT techniques for photovoltaic applications. Renew.
nected transformerless PV inverter topologies with reactive power generation cap-
Sustain. Energy Rev. 25, 793–813. https://doi.org/10.1016/j.rser.2013.05.022.
ability. Sol. Energy 163, 356–375. https://doi.org/10.1016/j.solener.2018.01.039.
Eren, S., Pahlevani, M., Bakhshai, A., Jain, P., 2015. An adaptive droop DC-bus voltage
Ahmad, Z., Singh, S.N., 2017a. Comparative analysis of single phase transformerless in-
controller for a grid-connected voltage source inverter with LCL filter. IEEE Trans.
verter topologies for grid connected PV system. Sol. Energy 149, 245–271. https://
Power Electron. 30, 547–560. https://doi.org/10.1109/TPEL.2014.2308251.
doi.org/10.1016/j.solener.2017.03.080.
Faraji, F., Mousavi, G.S.M., Hajirayat, A., Birjandi, A.A.M., Al-Haddad, K., 2017. Single-
Ahmad, Z., Singh, S.N., 2017b. An improved single phase transformerless inverter to-
stage single-phase three-level neutral-point-clamped transformerless grid-connected
pology for grid connected PV system with reduce leakage current and reactive power
photovoltaic inverters: topology review. Renew. Sustain. Energy Rev. 80, 197–214.
capability. Sol. Energy 157, 133–146. https://doi.org/10.1016/j.solener.2017.08.
https://doi.org/10.1016/j.rser.2017.05.181.
007.
Fathy, A., 2015. Reliable and efficient approach for mitigating the shading effect on
Ahmed, J., Salam, Z., 2017. An accurate method for MPPT to detect the partial shading
photovoltaic module based on Modified Artificial Bee Colony algorithm. Renew.
occurrence in a PV system. IEEE Trans. Ind. Inform. 13, 2151–2161. https://doi.org/
Energy 81, 78–88. https://doi.org/10.1016/j.renene.2015.03.017.
10.1109/TII.2017.2703079.
Feng, X., Wang, F., Wu, C., Luo, J., Zhang, L., 2019. Modeling and comparisons of ag-
Ankit, Sahoo, S.K., Sukchai, S., Yanine, F.F., 2018. Review and comparative study of
gregated flyback microinverters in aspect of harmonic resonances with the grid. IEEE
single-stage inverters for a PV system. Renew. Sustain. Energy Rev. 91, 962–986.
Trans. Ind. Electron. 66, 276–285. https://doi.org/10.1109/TIE.2018.2821634.
https://doi.org/10.1016/j.rser.2018.04.063.
Fermeiro, J.B.L., Pombo, J.A.N., Calado, M.R.A., Mariano, S.J.P.S., 2017. A new con-
Baka, M., Manganiello, P., Soudris, D., Catthoor, F., 2019. A cost-benefit analysis for
troller for DC-DC converters based on particle swarm optimization. Appl. Soft
reconfigurable PV modules under shading. Sol. Energy 178, 69–78. https://doi.org/
Comput. 52, 418–434. https://doi.org/10.1016/j.asoc.2016.10.025.
10.1016/j.solener.2018.11.063.
Furtado, A.M.S., Bradaschia, F., Cavalcanti, M.C., Limongi, L.R., 2018. A reduced voltage
Bendib, B., Belmili, H., Krim, F., 2015. A survey of the most used MPPT methods: con-
range global maximum power point tracking algorithm for photovoltaic systems
ventional and advanced algorithms applied for photovoltaic systems. Renew. Sustain.
under partial shading conditions. IEEE Trans. Ind. Electron. 65, 3252–3262. https://
Energy Rev. 45, 637–648. https://doi.org/10.1016/j.rser.2015.02.009.
doi.org/10.1109/TIE.2017.2750623.
Benyoucef, A., Soufyane, Chouder, A., Kara, K., Silvestre, S., Sahed, O.A., 2015. Artificial
Gao, M., Chen, M., Zhang, C., Qian, Z., 2014. Analysis and implementation of an im-
bee colony based algorithm for maximum power point tracking (MPPT) for PV sys-
proved flyback inverter for photovoltaic AC module applications. IEEE Trans. Power
tems operating under partial shaded conditions. Appl. Soft Comput. 32, 38–48.
Electron. 29, 3428–3444. https://doi.org/10.1109/TPEL.2013.2279266.
https://doi.org/10.1016/j.asoc.2015.03.047.
Godoy, R.B., Bizarro, D.B., de Andrade, E.T., de Oliveira Soares, J., Ribeiro, P.E.M.J.,
Bizon, N., 2016. Global maximum power point tracking (GMPPT) of photovoltaic array
Carniato, L.A., Kimpara, M.L.M., Pinto, J.O.P., Al-Haddad, K., Canesin, C.A., 2017.
using the extremum seeking control (ESC): a review and a new GMPPT ESC scheme.
Procedure to match the dynamic response of MPPT and droop-controlled micro-
Renew. Sustain. Energy Rev. 57, 524–539. https://doi.org/10.1016/j.rser.2015.12.
inverters. IEEE Trans. Ind. Appl. 53, 2358–2368. https://doi.org/10.1109/TIA.2016.
221.
2642883.
Boumaaraf, H., Talha, A., Bouhali, O., 2015. A three-phase NPC grid-connected inverter
Goroohi Sardou, I., Zare, M., Azad-Farsani, E., 2018. Robust energy management of a
for photovoltaic applications using neural network MPPT. Renew. Sustain. Energy
microgrid with photovoltaic inverters in VAR compensation mode. Int. J. Electr.
Rev. 49, 1171–1179. https://doi.org/10.1016/j.rser.2015.04.066.
Power Energy Syst. 98, 118–132. https://doi.org/10.1016/j.ijepes.2017.11.037.
Castilla, M., 2008. Linear current control scheme with series resonant harmonic com-
Gotekar, P.S., Muley, S.P., Kothari, D.P., Umre, B.S., 2015. Comparison of full bridge
pensator for single-phase grid-connected photovoltaic inverters. IEEE Trans. Ind.
bipolar, H5, H6 and HERIC inverter for single phase photovoltaic systems - a review.
Electron. 55, 2724–2733. https://doi.org/10.1109/TIE.2008.920585.
In: 2015 Annual IEEE India Conference (INDICON). Presented at the 2015 Annual
Çelik, Ö., Teke, A., 2017. A Hybrid MPPT method for grid connected photovoltaic systems
IEEE India Conference (INDICON). IEEE, New Delhi, India, pp. 1–6. https://doi.org/
under rapidly changing atmospheric conditions. Electr. Power Syst. Res. 152,
10.1109/INDICON.2015.7443837.
194–210. https://doi.org/10.1016/j.epsr.2017.07.011.
Guo, X., Jia, X., Lu, Z., Guerrero, J.M., 2016. Single phase cascaded H5 inverter with
Çelik, Ö., Teke, A., Tan, A., 2018. Overview of micro-inverters as a challenging tech-
leakage current elimination for transformerless photovoltaic system. In: 2016 IEEE
nology in photovoltaic applications. Renew. Sustain. Energy Rev. 82, 3191–3206.
Applied Power Electronics Conference and Exposition (APEC). Presented at the 2016
https://doi.org/10.1016/j.rser.2017.10.024.
IEEE Applied Power Electronics Conference and Exposition (APEC). IEEE, Long
Chen, B., Gu, B., Zhang, L., Zahid, Z.U., Lai, J.-S.J., Liao, Z., Hao, R., 2015. A high-
Beach, CA, USA, pp. 398–401. https://doi.org/10.1109/APEC.2016.7467903.
efficiency MOSFET transformerless inverter for nonisolated microinverter applica-
Shen, Guoqiao, Zhu, Xuancai, Zhang, Jun, Dehong, Xu., 2010. A New feedback method
tions. IEEE Trans. Power Electron. 30, 3610–3622. https://doi.org/10.1109/TPEL.
for PR current control of LCL-filter-based grid-connected inverter. IEEE Trans. Ind.
2014.2339320.
Electron. 57, 2033–2041. https://doi.org/10.1109/TIE.2010.2040552.
Chen, K., Tian, S., Cheng, Y., Bai, L., 2014. An improved MPPT controller for photovoltaic
Hadji, S., Gaubert, J.-P., Krim, F., 2015. Theoretical and experimental analysis of genetic
system under partial shading condition. IEEE Trans. Sustain. Energy 5, 978–985.
algorithms based MPPT for PV systems. Energy Proc. 74, 772–787. https://doi.org/
https://doi.org/10.1109/TSTE.2014.2315653.
10.1016/j.egypro.2015.07.813.
Bao, Chenlei, Ruan, Xinbo, Wang, Xuehua, Li, Weiwei, Pan, Donghua, Weng, Kailei, 2014.
Harrag, A., Messalti, S., 2019. IC-based variable step size neuro-fuzzy MPPT improving
Step-by-step controller design for LCL-type grid-connected inverter with capaci-
PV system performances. Energy Proc. 157, 362–374. https://doi.org/10.1016/j.
tor–current-feedback active-damping. IEEE Trans. Power Electron. 29, 1239–1253.
egypro.2018.11.201.
https://doi.org/10.1109/TPEL.2013.2262378.
Harrag, A., Messalti, S., 2015. Variable step size modified P&O MPPT algorithm using GA-
Daraban, S., Petreus, D., Morel, C., 2014. A novel MPPT (maximum power point tracking)
based hybrid offline/online PID controller. Renew. Sustain. Energy Rev. 49,
algorithm based on a modified genetic algorithm specialized on tracking the global
1247–1260. https://doi.org/10.1016/j.rser.2015.05.003.
maximum power point in photovoltaic systems affected by partial shading. Energy
Hasan, R., Mekhilef, S., 2017. Highly efficient flyback microinverter for grid-connected
74, 374–388. https://doi.org/10.1016/j.energy.2014.07.001.
rooftop PV system. Sol. Energy 146, 511–522. https://doi.org/10.1016/j.solener.
de Brito, M.A.G., Galotto, L., Sampaio, L.P., e Melo, G.De.A., Canesin, C.A., 2013.
2017.03.015.
Evaluation of the main MPPT techniques for photovoltaic applications. IEEE Trans.
Hasan, R., Mekhilef, S., Seyedmahmoudian, M., Horan, B., 2017. Grid-connected isolated
Ind. Electron. 60, 1156–1167. https://doi.org/10.1109/TIE.2012.2198036.
PV microinverters: a review. Renew. Sustain. Energy Rev. 67, 1065–1080. https://
Díez-Mediavilla, M., Dieste-Velasco, M.I., Rodríguez-Amigo, M.C., García-Calderón, T.,
doi.org/10.1016/j.rser.2016.09.082.
Alonso-Tristán, C., 2014. Performance of grid-tied PV facilities based on real data in
Hassaine, L., Olías, E., Quintero, J., Barrado, A., 2014. Power control for grid connected
Spain: Central inverter versus string system. Energy Convers. Manage. 86,
applications based on the phase shifting of the inverter output voltage with respect to
1128–1133. https://doi.org/10.1016/j.enconman.2014.06.087.
the grid voltage. Int. J. Electr. Power Energy Syst. 57, 250–260. https://doi.org/10.
Dileep, G., Singh, S.N., 2017. An improved particle swarm optimization based maximum
1016/j.ijepes.2013.12.009.
power point tracking algorithm for PV system operating under partial shading con-
Ho, C.N.-M., Cheung, V.S.P., Chung, H.S.-H., 2009. Constant-frequency hysteresis current
ditions. Sol. Energy 158, 1006–1015. https://doi.org/10.1016/j.solener.2017.10.
control of grid-connected VSI without bandwidth control. IEEE Trans. Power
027.
Electron. 24, 2484–2495. https://doi.org/10.1109/TPEL.2009.2031804.
Dong, M., Tian, X., 2017. Dual-Mode interleaved flyback micro-inverter. In: 2017 Chinese
Hu, H., Harb, S., Fang, X., Zhang, D., Zhang, Q., Shen, Z.J., Batarseh, I., 2012. A three-port
Automation Congress (CAC). Presented at the 2017 Chinese Automation Congress
flyback for PV microinverter applications with power pulsation decoupling cap-
(CAC). IEEE, Jinan, pp. 7719–7724. https://doi.org/10.1109/CAC.2017.8244175.
ability. IEEE Trans. Power Electron. 27, 3953–3964. https://doi.org/10.1109/TPEL.
Douiri, M.R., 2019. Particle swarm optimized neuro-fuzzy system for photovoltaic power
2012.2188840.
forecasting model. Sol. Energy 184, 91–104. https://doi.org/10.1016/j.solener.2019.
Hu, H., Harb, S., Kutkut, N.H., Shen, Z.J., Batarseh, I., 2013. A Single-stage microinverter

271
E. Kabalcı Solar Energy 198 (2020) 247–274

without using eletrolytic capacitors. IEEE Trans. Power Electron. 28, 2677–2687. Khan, A., Ben-Brahim, L., Gastli, A., Benammar, M., 2017. Review and simulation of
https://doi.org/10.1109/TPEL.2012.2224886. leakage current in transformerless microinverters for PV applications. Renew.
IEEE Standard 1547.2-2008, DeBlasio, R., 2009. IEEE Application Guide for IEEE Std Sustain. Energy Rev. 74, 1240–1256. https://doi.org/10.1016/j.rser.2017.02.053.
1547, IEEE Standard for Interconnecting Distributed Resources with Electric Power Kim, Y.-H., Ji, Y.-H., Kim, J.-G., Jung, Y.-C., Won, C.-Y., 2013. A New control strategy for
Systems, pp. 1–207. improving weighted efficiency in photovoltaic AC module-type interleaved flyback
International Electrotechnical Commission, Oruganti, R., 2014. IEC 61727-2004 inverters. IEEE Trans. Power Electron. 28, 2688–2699. https://doi.org/10.1109/
Photovoltaic (PV) systems – Characteristics of the utility interface. TPEL.2012.2226753.
Islam, M., Mekhilef, S., 2014. An improved transformerless grid connected photovoltaic Komurcugil, H., 2012. Rotating-sliding-line-based sliding-mode control for single-phase
inverter with reduced leakage current. Energy Conv. Manage. 88, 854–862. https:// UPS inverters. IEEE Trans. Ind. Electron. 59, 3719–3726. https://doi.org/10.1109/
doi.org/10.1016/j.enconman.2014.09.014. TIE.2011.2159354.
Islam, M., Mekhilef, S., Hasan, M., 2015. Single phase transformerless inverter topologies Kouro, S., Leon, J.I., Vinnikov, D., Franquelo, L.G., 2015. Grid-connected photovoltaic
for grid-tied photovoltaic system: a review. Renew. Sustain. Energy Rev. 45, 69–86. systems: an overview of recent research and emerging PV converter technology. IEEE
https://doi.org/10.1016/j.rser.2015.01.009. Ind. Electron. Mag. 9, 47–61. https://doi.org/10.1109/MIE.2014.2376976.
Jaeger-Waldau, A., 2017. PV Status Report 2017 (Science for Policy). Joint Research Kulaksız, A.A., Akkaya, R., 2012. A genetic algorithm optimized ANN-based MPPT al-
Centre (JRC), Luxembourg. gorithm for a stand-alone PV system with induction motor drive. Sol. Energy 86,
Jana, J., Saha, H., Das Bhattacharya, K., 2017. A review of inverter topologies for single- 2366–2375. https://doi.org/10.1016/j.solener.2012.05.006.
phase grid-connected photovoltaic systems. Renew. Sustain. Energy Rev. 72, Lai, C.M., 2014. A single-stage grid-connected PV micro-inverter based on interleaved
1256–1270. https://doi.org/10.1016/j.rser.2016.10.049. flyback converter topology. In: 2014 International Symposium on Computer,
Jeong, Y., Kim, J.-K., Lee, J.-B., Moon, G.-W., 2017. An asymmetric half-bridge resonant Consumer and Control. Presented at the 2014 International Symposium on Computer,
converter having a reduced conduction loss for DC/DC power applications with a Consumer and Control (IS3C). IEEE, Taichung, Taiwan, pp. 187–190. https://doi.
wide range of low input voltage. IEEE Trans. Power Electron. 32, 7795–7804. https:// org/10.1109/IS3C.2014.59.
doi.org/10.1109/TPEL.2016.2639069. Lasheen, M., Rahman, A.K.A., Abdel-Salam, M., Ookawara, S., 2016. Performance en-
Jiang, J.-A., Su, Y.-L., Kuo, K.-C., Wang, C.-H., Liao, M.-S., Wang, J.-C., Huang, C.-K., hancement of constant voltage based MPPT for photovoltaic applications using ge-
Chou, C.-Y., Lee, C.-H., Shieh, J.-C., 2017. On a hybrid MPPT control scheme to netic algorithm. Energy Proc. 100, 217–222. https://doi.org/10.1016/j.egypro.2016.
improve energy harvesting performance of traditional two-stage inverters used in 10.168.
photovoltaic systems. Renew. Sustain. Energy Rev. 69, 1113–1128. https://doi.org/ Li, X., Wen, H., Hu, Y., Jiang, L., Xiao, W., 2018. Modified beta algorithm for GMPPT and
10.1016/j.rser.2016.09.112. partial shading detection in photovoltaic systems. IEEE Trans. Power Electron. 33,
Jiang, S., Cao, D., Li, Y., Peng, F.Z., 2012. Grid-connected boost-half-bridge photovoltaic 2172–2186. https://doi.org/10.1109/TPEL.2017.2697459.
microinverter system using repetitive current control and maximum power point Zhang, Li, Sun, Kai, Yan Xing, Mu., Xing, 2014a. H6 Transformerless full-bridge PV grid-
tracking. IEEE Trans. Power Electron. 27, 4711–4722. https://doi.org/10.1109/ tied inverters. IEEE Trans. Power Electron. 29, 1229–1238. https://doi.org/10.1109/
TPEL.2012.2183389. TPEL.2013.2260178.
Joshi, P., Arora, S., 2017. Maximum power point tracking methodologies for solar PV Liu, Z., Wu, H., Liu, Y., Ji, J., Wu, W., Blaabjerg, F., 2017. Modelling of the modified-
systems – a review. Renew. Sustain. Energy Rev. 70, 1154–1177. https://doi.org/10. LLCL-filter-based single-phase grid-tied Aalborg inverter. IET Power Electron. 10,
1016/j.rser.2016.12.019. 151–155. https://doi.org/10.1049/iet-pel.2016.0044.
Jovanovic, M.M., Irving, B.T., 2016. On-the-fly topology-morphing control—efficiency Lodh, T., Pragallapati, N., Agarwal, V., 2016. An improved control scheme for interleaved
optimization method for LLC resonant converters operating in wide input- and/or flyback converter based micro-inverter to achieve high efficiency. In: 2016 IEEE 1st
output-voltage range. IEEE Trans. Power Electron. 31, 2596–2608. https://doi.org/ International Conference on Power Electronics, Intelligent Control and Energy
10.1109/TPEL.2015.2440099. Systems (ICPEICES). Presented at the 2016 IEEE 1st International Conference on
Jovanovic, M.M., Irving, B.T., 2015. Efficiency optimization of LLC resonant converters Power Electronics, Intelligent Control and Energy Systems (ICPEICES). IEEE, Delhi,
operating in wide input- and/or output-voltage range by on-the-fly topology- India, pp. 1–6. https://doi.org/10.1109/ICPEICES.2016.7853693.
morphing control. In: 2015 IEEE Applied Power Electronics Conference and Mahela, O.P., Shaik, A.G., 2017. Comprehensive overview of grid interfaced solar pho-
Exposition (APEC). Presented at the 2015 IEEE Applied Power Electronics Conference tovoltaic systems. Renew. Sustain. Energy Rev. 68, 316–332. https://doi.org/10.
and Exposition (APEC). IEEE, Charlotte, NC, USA, pp. 1420–1427. https://doi.org/ 1016/j.rser.2016.09.096.
10.1109/APEC.2015.7104534. Xingkui, Mao, Qisheng, Huang, Qingbo, Ke, Yudi, Xiao, Zhe, Zhang, Andersen, M.A.E.,
Kabalci, E., 2018. The design and analysis of a two-stage PV converter with quasi-Z source 2016. Grid-connected photovoltaic micro-inverter with new hybrid control LLC re-
inverter. In: 2018 IEEE 18th International Power Electronics and Motion Control sonant converter. In: IECON 2016 - 42nd Annual Conference of the IEEE Industrial
Conference (PEMC). Presented at the 2018 IEEE 18th International Power Electronics Electronics Society. Presented at the IECON 2016 - 42nd Annual Conference of the
and Motion Control Conference (PEMC). IEEE, Budapest, pp. 451–456. https://doi. IEEE Industrial Electronics Society. IEEE, Florence, Italy, pp. 2319–2324. https://doi.
org/10.1109/EPEPEMC.2018.8521953. org/10.1109/IECON.2016.7793632.
Kabalci, E., 2017. Maximum power point tracking (MPPT) algorithms for photovoltaic Masson, G., Kaizuka, I., Cambiè, C., 2018. Snapshot of Global Photovoltaıc Markets (PVPS
systems. In: Bizon, N., Mahdavi Tabatabaei, N., Blaabjerg, F., Kurt, E. (Eds.), Energy No. Report IEA PVPS T1-33:2018). IEA International Energy Agency.
Harvesting and Energy Efficiency. Springer International Publishing, Cham, pp. Mattavelli, P., 2005. An improved deadbeat control for UPS using disturbance observers.
205–234. https://doi.org/10.1007/978-3-319-49875-1_8. IEEE Trans. Ind. Electron. 52, 206–212. https://doi.org/10.1109/TIE.2004.837912.
Kabalci, E., 2015. A smart monitoring infrastructure design for distributed renewable Mekki, H., Mellit, A., Salhi, H., 2016. Artificial neural network-based modelling and fault
energy systems. Energy Convers. Manage. 90, 336–346. https://doi.org/10.1016/j. detection of partial shaded photovoltaic modules. Simul. Model. Pract. Theory 67,
enconman.2014.10.062. 1–13. https://doi.org/10.1016/j.simpat.2016.05.005.
Kabalci, Ersan, Bayindir, R., Gokkus, G., Kabalci, Y., 2015a. Dual DC-DC converter and Meneses, D., Blaabjerg, F., García, Ó., Cobos, J.A., 2013. Review and comparison of step-
monitoring interface for asymmetrical string inverters. In: 2015 International up transformerless topologies for photovoltaic AC-module application. IEEE Trans.
Conference on Renewable Energy Research and Applications (ICRERA). IEEE, pp. Power Electron. 28, 2649–2663. https://doi.org/10.1109/TPEL.2012.2227820.
1580–1585. Meneses, D., Garcia, O., Alou, P., Oliver, J.A., Cobos, J.A., 2015. Grid-connected forward
Kabalci, E., Kabalci, Y., 2018. A wireless metering and monitoring system for solar string microinverter with primary-parallel secondary-series transformer. IEEE Trans. Power
inverters. Electr. Power Energy Syst. 96, 282–295. https://doi.org/10.1016/j.ijepes. Electron. 30, 4819–4830. https://doi.org/10.1109/TPEL.2014.2365760.
2017.10.013. Messalti, S., Harrag, A., Loukriz, A., 2017. A new variable step size neural networks MPPT
Kabalci, E., Kabalci, Y., Canbaz, R., Gokkus, G., 2015b. Single phase multilevel string controller: review, simulation and hardware implementation. Renew. Sustain. Energy
inverter for solar applications. In: Presented at the 4th International Conference on Rev. 68, 221–233. https://doi.org/10.1016/j.rser.2016.09.131.
Renewable Energy Research and Applications, Palermo Italy, pp. 109–114. Miret, J., Castilla, M., Matas, J., Guerrero, J.M., Vasquez, J.C., 2009. Selective harmonic-
Kabalci, E., Kabalci, Y., Gokkus, G., 2015c. Dual DC-DC converter design for string in- compensation control for single-phase active power filter with high harmonic rejec-
verters used in solar plants. In: Presented at the 4th International Conference on tion. IEEE Trans. Ind. Electron. 56, 3117–3127. https://doi.org/10.1109/TIE.2009.
Renewable Energy Research and Applications, Palermo Italy, pp. 115–119. 2024662.
Kabalci, Y., Kabalci, E., 2017. Modeling and analysis of a smart grid monitoring system Mirza, A.F., Ling, Q., Javed, M.Y., Mansoor, M., 2019. Novel MPPT techniques for pho-
for renewable energy sources. Sol. Energy 153, 262–275. https://doi.org/10.1016/j. tovoltaic systems under uniform irradiance and partial shading. Sol. Energy 184,
solener.2017.05.063. 628–648. https://doi.org/10.1016/j.solener.2019.04.034.
Karami, N., Moubayed, N., Outbib, R., 2017. General review and classification of different Mohapatra, A., Nayak, B., Das, P., Mohanty, K.B., 2017. A review on MPPT techniques of
MPPT Techniques. Renew. Sustain. Energy Rev. 68, 1–18. https://doi.org/10.1016/j. PV system under partial shading condition. Renew. Sustain. Energy Rev. 80,
rser.2016.09.132. 854–867. https://doi.org/10.1016/j.rser.2017.05.083.
Kasa, N., Iida, T., Chen, L., 2005. Flyback inverter controlled by sensorless current MPPT Montoya, D.G., Ramos-Paja, C.A., Giral, R., 2016. Improved design of sliding-mode
for photovoltaic power system. IEEE Trans. Ind. Electron. 52, 1145–1152. https:// controllers based on the requirements of MPPT techniques. IEEE Trans. Power
doi.org/10.1109/TIE.2005.851602. Electron. 31, 235–247. https://doi.org/10.1109/TPEL.2015.2397831.
Kassem, A.M., 2012. MPPT control design and performance improvements of a PV gen- Moon, S., Yoon, S.-G., Park, J.-H., 2015. A new low-cost centralized MPPT controller
erator powered DC motor-pump system based on artificial neural networks. Int. J. system for multiply distributed photovoltaic power conditioning modules. IEEE
Electr. Power Energy Syst. 43, 90–98. https://doi.org/10.1016/j.ijepes.2012.04.047. Trans. Smart Grid 6, 2649–2658. https://doi.org/10.1109/TSG.2015.2439037.
Kerekes, T., 2009. Analysis and Modeling of Transformerless Photovoltaic Inverter Mosaad, M.I., Abed El raouf, M.O., Al-Ahmar, M.A., Banakher, F.A., 2019. Maximum
Systems. Aalborg University, Institute of Energy Technology, Aalborg. power point tracking of PV system based cuckoo search algorithm; review and
Kerekes, T., Teodorescu, R., Rodriguez, P., Vazquez, G., Aldabas, E., 2011. A new high- comparison. Energy Procedia 162, 117–126. https://doi.org/10.1016/j.egypro.2019.
efficiency single-phase transformerless PV inverter topology. IEEE Trans. Ind. 04.013.
Electron. 58, 184–191. https://doi.org/10.1109/TIE.2009.2024092. Muhsen, D.H., Nabil, M., Haider, H.T., Khatib, T., 2019. A novel method for sizing of

272
E. Kabalcı Solar Energy 198 (2020) 247–274

standalone photovoltaic system using multi-objective differential evolution algorithm single-phase transformerless inverters for grid-connected photovoltaic (PV) systems.
and hybrid multi-criteria decision making methods. Energy 174, 1158–1175. https:// Renew. Sustain. Energy Rev. 82, 515–530. https://doi.org/10.1016/j.rser.2017.09.
doi.org/10.1016/j.energy.2019.03.046. 055.
Nabipour, M., Razaz, M., Seifossadat, S.G., Mortazavi, S.S., 2017. A new MPPT scheme Sher, H.A., Addoweesh, K.E., 2012. Micro-inverters — promising solutions in solar pho-
based on a novel fuzzy approach. Renew. Sustain. Energy Rev. 74, 1147–1169. tovoltaics. Energy Sustain. Devel. 16, 389–400. https://doi.org/10.1016/j.esd.2012.
https://doi.org/10.1016/j.rser.2017.02.054. 10.002.
Oshaba, A.S., Ali, E.S., Abd Elazim, S.M., 2015. MPPT control design of PV system sup- Shimizu, T., Wada, K., Nakamura, N., 2006. Flyback-type single-phase utility interactive
plied SRM using BAT search algorithm. Sustain. Energy Grids Netw. 2, 51–60. inverter with power pulsation decoupling on the DC input for an AC photovoltaic
https://doi.org/10.1016/j.segan.2015.04.002. module system. IEEE Trans. Power Electron. 21, 1264–1272. https://doi.org/10.
Öztürk, S., Poşpoş, P., Utalay, V., Koç, A., Ermiş, M., Çadırcı, I., 2018. Operating prin- 1109/TPEL.2006.880247.
ciples and practical design aspects of all SiC DC/AC/DC converter for MPPT in grid- Siwakoti, Y.P., Blaabjerg, F., 2018. Common-ground-type transformerless inverters for
connected PV supplies. Sol. Energy 176, 380–394. https://doi.org/10.1016/j.solener. single-phase solar photovoltaic systems. IEEE Trans. Ind. Electron. 65, 2100–2111.
2018.10.049. https://doi.org/10.1109/TIE.2017.2740821.
Petreuş, D., Daraban, S., Ciocan, I., Patarau, T., Morel, C., Machmoum, M., 2013. Low cost Sridhar, V., Umashankar, S., 2017. A comprehensive review on CHB MLI based PV in-
single stage micro-inverter with MPPT for grid connected applications. Sol. Energy verter and feasibility study of CHB MLI based PV-STATCOM. Renew. Sustain. Energy
92, 241–255. https://doi.org/10.1016/j.solener.2013.03.016. Rev. 78, 138–156. https://doi.org/10.1016/j.rser.2017.04.111.
Pilakkat, D., Kanthalakshmi, S., 2019. An improved P&O algorithm integrated with ar- Sukesh, N., Pahlevaninezhad, M., Jain, P.K., 2014. Analysis and implementation of a
tificial bee colony for photovoltaic systems under partial shading conditions. Sol. single-stage flyback PV microinverter with soft switching. IEEE Trans. Ind. Electron.
Energy 178, 37–47. https://doi.org/10.1016/j.solener.2018.12.008. 61, 1819–1833. https://doi.org/10.1109/TIE.2013.2263778.
Guo, Qian, Wang, Jun, Ma, Hao, 2014. Frequency adaptive repetitive controller for grid- Sun, X., Li, X., Shen, Y., Wang, B., Guo, X., 2017. Dual-bridge LLC resonant converter with
connected inverter with an all-pass infinite impulse response (IIR) filter. In: 2014 fixed-frequency PWM control for wide input applications. IEEE Trans. Power
IEEE 23rd International Symposium on Industrial Electronics (ISIE). Presented at the Electron. 32, 69–80. https://doi.org/10.1109/TPEL.2016.2530748.
2014 IEEE 23rd International Symposium on Industrial Electronics (ISIE). IEEE, Sundareswaran, K., Vignesh Kumar, V., Palani, S., 2015. Application of a combined
Istanbul, Turkey, pp. 491–496. https://doi.org/10.1109/ISIE.2014.6864662. particle swarm optimization and perturb and observe method for MPPT in PV systems
Rajgor, G., 2013. Microinverters – a solar game changer. Renew. Energy Focus 14, 16–17. under partial shading conditions. Renew. Energy 75, 308–317. https://doi.org/10.
https://doi.org/10.1016/S1755-0084(13)70047-4. 1016/j.renene.2014.09.044.
Ramli, M.A.M., Ishaque, K., Jawaid, F., Al-Turki, Y.A., Salam, Z., 2015. A modified dif- Tajuddin, M.F.N., Ayob, S.M., Salam, Z., Saad, M.S., 2013. Evolutionary based maximum
ferential evolution based maximum power point tracker for photovoltaic system power point tracking technique using differential evolution algorithm. Energy Build.
under partial shading condition. Energy Build. 103, 175–184. https://doi.org/10. 67, 245–252. https://doi.org/10.1016/j.enbuild.2013.07.085.
1016/j.enbuild.2015.06.058. Tayebi, S.M., Hu, H., Batarseh, I., 2019. Advanced DC-link voltage regulation and ca-
Rezaei, M.A., Lee, K.-J., Huang, A.Q., 2016. A high-efficiency flyback micro-inverter with pacitor optimization for three-phase microinverters. IEEE Trans. Ind. Electron. 66,
a new adaptive snubber for photovoltaic applications. IEEE Trans. Power Electron. 307–317. https://doi.org/10.1109/TIE.2018.2823700.
31, 318–327. https://doi.org/10.1109/TPEL.2015.2407405. Thang, T.V., Thao, N.M., Jang, J.-H., Park, J.-H., 2014. Analysis and design of grid-
Rezk, H., Fathy, A., Abdelaziz, A.Y., 2017. A comparison of different global MPPT tech- connected photovoltaic systems with multiple-integrated converters and a pseudo-
niques based on meta-heuristic algorithms for photovoltaic system subjected to DC-link inverter. IEEE Trans. Ind. Electron. 61, 3377–3386. https://doi.org/10.1109/
partial shading conditions. Renew. Sustain. Energy Rev. 74, 377–386. https://doi. TIE.2013.2281153.
org/10.1016/j.rser.2017.02.051. Titri, S., Larbes, C., Toumi, K.Y., Benatchba, K., 2017. A new MPPT controller based on
Rezk, H., Hasaneen, E.-S., 2015. A new MATLAB/Simulink model of triple-junction solar the Ant colony optimization algorithm for Photovoltaic systems under partial shading
cell and MPPT based on artificial neural networks for photovoltaic energy systems. conditions. Appl. Soft Comput. 58, 465–479. https://doi.org/10.1016/j.asoc.2017.
Ain Shams Eng. J. 6, 873–881. https://doi.org/10.1016/j.asej.2015.03.001. 05.017.
Rizzo, S.A., Scelba, G., 2015. ANN based MPPT method for rapidly variable shading Tofigh Azary, M., Sabahi, M., Babaei, E., Abbasi Aghdam, Meinagh, F., 2018. Modified
conditions. Appl. Energy 145, 124–132. https://doi.org/10.1016/j.apenergy.2015. single-phase single-stage grid-tied flying inductor inverter with MPPT and suppressed
01.077. leakage current. IEEE Trans. Ind. Electron. 65, 221–231. https://doi.org/10.1109/
Rizzoli, G., Mengoni, M., Zarri, L., Tani, A., Serra, G., Casadei, D., 2016. Comparison of TIE.2017.2719610.
single-phase H4, H5, H6 inverters for transformerless photovoltaic applications. In: Vazquez, N., Rosas, M., Hernandez, C., Vazquez, E., Perez-Pinal, F.J., 2015. A new
IECON 2016 - 42nd Annual Conference of the IEEE Industrial Electronics Society. common-mode transformerless photovoltaic inverter. IEEE Trans. Ind. Electron. 62,
Presented at the IECON 2016 - 42nd Annual Conference of the IEEE Industrial 6381–6391. https://doi.org/10.1109/TIE.2015.2426146.
Electronics Society. IEEE, Florence, Italy, pp. 3038–3045. https://doi.org/10.1109/ Vinnikov, D., Chub, A., Liivik, E., Kosenko, R., Korkh, O., 2019. Solar optiverter—a novel
IECON.2016.7792984. hybrid approach to the photovoltaic module level power electronics. IEEE Trans. Ind.
Rojas, C.A., Aguirre, M., Kouro, S., Geyer, T., Gutierrez, E., 2017. Leakage current miti- Electron. 66, 3869–3880. https://doi.org/10.1109/TIE.2018.2850036.
gation in photovoltaic string inverter using predictive control with fixed average Vinnikov, D., Chub, A., Liivik, E., Roasto, I., 2017. High-performance quasi-Z-source
switching frequency. IEEE Trans. Ind. Electron. 64, 9344–9354. https://doi.org/10. series resonant DC–DC converter for photovoltaic module-level power electronics
1109/TIE.2017.2708003. applications. IEEE Trans. Power Electr. 32, 3634–3650. https://doi.org/10.1109/
Romero-Cadaval, E., Spagnuolo, G., Franquelo, L.G., Ramos-Paja, C.A., Suntio, T., Xiao, TPEL.2016.2591726.
W.M., 2013. Grid-connected photovoltaic generation plants: components and op- Wang, F., Feng, X., Zhang, L., Du, Y., Su, J., 2018a. Impedance-based analysis of grid
eration. IEEE Ind. Electron. Mag. 7, 6–20. https://doi.org/10.1109/MIE.2013. harmonic interactions between aggregated flyback micro-inverters and the grid. IET
2264540. Power Electron. 11, 453–459. https://doi.org/10.1049/iet-pel.2017.0356.
Saad, N.H., El-Sattar, A.A., Mansour, A.E.-A.M., 2018. A novel control strategy for grid Wang, H., Wu, W., Chung, H.S., Blaabjerg, F., 2018b. Coupled-inductor-based Aalborg
connected hybrid renewable energy systems using improved particle swarm optimi- inverter with input DC energy regulation. IEEE Trans. Ind. Electron. 65, 3826–3836.
zation. Ain Shams Eng. J. 9, 2195–2214. https://doi.org/10.1016/j.asej.2017.03. https://doi.org/10.1109/TIE.2017.2760849.
009. Xia, Y., Roy, J., Ayyanar, R., 2017. A capacitance-minimized, doubly grounded trans-
Salam, Z., Ahmed, J., Merugu, B.S., 2013. The application of soft computing methods for former less photovoltaic inverter with inherent active-power decoupling. IEEE Trans.
MPPT of PV system: a technological and status review. Appl. Energy 107, 135–148. Power Electron. 32, 5188–5201. https://doi.org/10.1109/TPEL.2016.2606344.
https://doi.org/10.1016/j.apenergy.2013.02.008. Xiao, H., Xie, S., Chen, Y., Huang, R., 2011. An optimized transformerless photovoltaic
Saravanan, S., Ramesh Babu, N., 2016. RBFN based MPPT algorithm for PV system with grid-connected inverter. IEEE Trans. Ind. Electron. 58, 1887–1895. https://doi.org/
high step up converter. Energy Conv. Manage. 122, 239–251. https://doi.org/10. 10.1109/TIE.2010.2054056.
1016/j.enconman.2016.05.076. Xiao, H.F., Lan, K., Zhang, L., 2015. A quasi-unipolar SPWM full-bridge transformerless
Sasidharan, N., Singh, J.G., 2017. A novel single-stage single-phase reconfigurable in- PV grid-connected inverter with constant common-mode voltage. IEEE Trans. Power
verter topology for a solar powered hybrid AC/DC home. IEEE Trans. Ind. Electron. Electron. 30, 3122–3132. https://doi.org/10.1109/TPEL.2014.2331367.
64, 2820–2828. https://doi.org/10.1109/TIE.2016.2643602. Xiao, H.F., Zhang, L., Li, Y., 2017. A zero-voltage-transition HERIC-type transformerless
Schmela, M., 2018. Global Market Outlook for Solar Power/2018–2022. SolarPower photovoltaic grid-connected inverter. IEEE Trans. Ind. Electron. 64, 1222–1232.
Europe, Brussels, Belgium. https://doi.org/10.1109/TIE.2016.2611574.
Schweizer, M., Kolar, J.W., 2013. Design and implementation of a highly efficient three- Xiong, G., Zhang, J., Yuan, X., Shi, D., He, Y., Yao, G., 2018. Parameter extraction of solar
level T-type converter for low-voltage applications. IEEE Trans. Power Electron. 28, photovoltaic models by means of a hybrid differential evolution with whale opti-
899–907. https://doi.org/10.1109/TPEL.2012.2203151. mization algorithm. Sol. Energy 176, 742–761. https://doi.org/10.1016/j.solener.
Serban, E., Paz, F., Ordonez, M., 2017. Improved PV inverter operating range using a 2018.10.050.
miniboost. IEEE Trans. Power Electron. 32, 8470–8485. https://doi.org/10.1109/ Xu, J., Xie, S., Tang, T., 2014. Active damping-based control for grid-connected $LCL$
TPEL.2016.2641478. -filtered inverter with injected grid current feedback only. IEEE Trans. Ind. Electron.
Shams El-Dein, M.Z., Kazerani, M., Salama, M.M.A., 2013a. An optimal total cross tied 61, 4746–4758. https://doi.org/10.1109/TIE.2013.2290771.
interconnection for reducing mismatch losses in photovoltaic arrays. IEEE Trans. Yahya, A., El Fadil, H., Oulcaid, M., Ammeh, L., Giri, F., Guerrero, J.M., 2018. Control of
Sustainable Energy 4, 99–107. https://doi.org/10.1109/TSTE.2012.2202325. grid connected photovoltaic systems with microinverters: new theoretical design and
Shams El-Dein, M.Z., Kazerani, M., Salama, M.M.A., 2013b. Optimal photovoltaic array numerical evaluation: advanced control of microinverters. Asian J. Control 20,
reconfiguration to reduce partial shading losses. IEEE Trans. Sustain. Energy 4, 906–918. https://doi.org/10.1002/asjc.1704.
145–153. https://doi.org/10.1109/TSTE.2012.2208128. Yang, B., Zhong, L., Zhang, X., Shu, H., Yu, T., Li, H., Jiang, L., Sun, L., 2019. Novel bio-
Shayestegan, M., Shakeri, M., Abunima, H., Reza, S.M.S., Akhtaruzzaman, M., Bais, B., inspired memetic salp swarm algorithm and application to MPPT for PV systems
Mat, S., Sopian, K., Amin, N., 2018. An overview on prospects of new generation considering partial shading condition. J. Cleaner Prod. 215, 1203–1222. https://doi.

273
E. Kabalcı Solar Energy 198 (2020) 247–274

org/10.1016/j.jclepro.2019.01.150. phase synchronization control strategies for BCM interleaved flyback microinverters.
Yao, Z., Xiao, L., 2013. Control of single-phase grid-connected inverters with nonlinear IEEE Trans. Power Electron. 29, 5921–5932. https://doi.org/10.1109/TPEL.2014.
loads. IEEE Trans. Ind. Electron. 60, 1384–1389. https://doi.org/10.1109/TIE.2011. 2300483.
2174535. Zhang, Zhe, Chen, M., Chen, W., Qian, Z., 2013a. Design and Analysis of the synchro-
Yilmaz, U., Kircay, A., Borekci, S., 2018. PV system fuzzy logic MPPT method and PI nization control method for BCM/DCM current-mode flyback micro-inverter. In:
control as a charge controller. Renew. Sustain. Energy Rev. 81, 994–1001. https:// 2013 Twenty-Eighth Annual IEEE Applied Power Electronics Conference and
doi.org/10.1016/j.rser.2017.08.048. Exposition (APEC). Presented at the 2013 IEEE Applied Power Electronics Conference
Zaki Diab, A.A., Rezk, H., 2017. Global MPPT based on flower pollination and differential and Exposition - APEC 2013. IEEE, Long Beach, CA, USA, pp. 68–75. https://doi.org/
evolution algorithms to mitigate partial shading in building integrated PV system. 10.1109/APEC.2013.6520187.
Sol. Energy 157, 171–186. https://doi.org/10.1016/j.solener.2017.08.024. Zhang, Zhiliang, He, X.-F., Liu, Y.-F., 2013b. An optimal control method for photovoltaic
Zapata, J.W., Renaudineau, H., Kouro, S., Perez, M.A., Meynard, T.A., 2016. Partial grid-tied-interleaved flyback microinverters to achieve high efficiency in wide load
power DC-DC converter for photovoltaic microinverters. In: IECON 2016 - 42nd range. IEEE Trans. Power Electron. 28, 5074–5087. https://doi.org/10.1109/TPEL.
Annual Conference of the IEEE Industrial Electronics Society. Presented at the IECON 2013.2245919.
2016 - 42nd Annual Conference of the IEEE Industrial Electronics Society. IEEE, Zhang, Z., Zhang, J., Shao, S., 2018. A variable off-time control method for a single-phase
Florence, Italy, pp. 6740–6745. https://doi.org/10.1109/IECON.2016.7793098. DCM microinverter. IEEE Trans. Power Electron. 33, 7229–7239. https://doi.org/10.
Zeb, K., Uddin, W., Khan, M.A., Ali, Z., Ali, M.U., Christofides, N., Kim, H.J., 2018. A 1109/TPEL.2017.2759227.
comprehensive review on inverter topologies and control strategies for grid con- Zhang, Z., Zhang, Junming, Shao, S., Zhang, Junjun, 2019. A High-efficiency single-phase
nected photovoltaic system. Renew. Sustain. Energy Rev. 94, 1120–1141. https://doi. T-type BCM microinverter. IEEE Trans. Power Electron. 34, 984–995. https://doi.
org/10.1016/j.rser.2018.06.053. org/10.1109/TPEL.2018.2824342.
Zeddini, Mohamed, A., Pusca, R., Sakly, A., Mimouni, M.F., 2016. PSO-based MPPT Zhao, F., Xue, F., Zhang, Y., Ma, W., Zhang, C., Song, H., 2018. A hybrid algorithm based
control of wind-driven self-excited induction generator for pumping system. Renew. on self-adaptive gravitational search algorithm and differential evolution. Expert
Energy 95, 162–177. https://doi.org/10.1016/j.renene.2016.04.008. Syst. Appl. 113, 515–530. https://doi.org/10.1016/j.eswa.2018.07.008.
Zhang, Q., 2013. Optimization and Design of Photovoltaic Micro-Inverter. University of Zhu, M., Zhao, Q., Ye, Y., Xu, G., 2016. Improved repetitive control scheme for grid-
Central Florida, Florida USA. connected inverter with frequency adaptation. IET Power Electron. 9, 883–890.
Zhang, Z., Chen, M., Chen, W., Jiang, C., Qian, Z., 2014b. Analysis and implementation of https://doi.org/10.1049/iet-pel.2015.0057.

274