Three-Phase Quasi-Z-Source Inverter with Constant Common-Mode Voltage for Photovoltaic Application

Transcription

1 Three-Phase Quasi-Z-Source Inverer wih Consan Common-Mode Volage for Phoovolaic Applicaion Negar Noroozi, Suden Member, IEEE, Mohammad Reza Zolghadri, Senior Member, IEEE Absrac In rasformerless grid-conneced phoovolaic (PV) sysems, common-mode volage (CMV) flucuaions cause leakage curren flow hrough he sray capaciance of he PV panels. Shoo-hrough (SH) saes in a quasi-zsource inverer (q-zsi), increase he ampliude of high order harmonics of CMV. In his paper, by using he modulaion echnique based on odd PWM (OPWM) and minor change in he Z nework of he hree-phase q-zsi, he leakage curren is blocked. No exra semiconducor elemen is added. By he proposed echnique, CMV is kep nearly consan during swiching cycles. The experimenal resuls for CMV analysis in a prooype are presened o verify he heoreical analysis. Index Terms common-mode volage, leakage curren, Quasi-Z-Source inverer, phoovolaic power sysem, odd PWM. M I. INTRODUCTION os of he insalled PV sysems are grid-conneced. According o Inernaional Energy Agency Phoovolaic Power Sysem repor (IEA-PVPS Annual Repor 2016), a he end of 2015, more han 99 % of he globally insalled PV capaciy (228 GW) were grid-conneced. A ransformerless PV sysem has higher efficiency and lower weigh and size; bu, by removing he galvanic isolaion beween he DC source and he grid, he leakage curren issue is appeared [1]. The leakage curren is originaed from CMV flucuaions. In a ransformerless sysem, he leakage curren flows hrough he Manuscrip received Mar. 18, 2017; revised Jun. 17, 2017and Aug. 29, 2017; acceped Sep. 30, The Auhors are wih he Deparmen of Elecrical Engineering, Sharif Universiy of Technology, Av. Azadi, Tehran, Iran, ( nenoroozi@gmail.com; zolghadr@sharif.ir). parasiic capaciance, mainly he sray capaciance beween he PV array and he grounded PV frame. Safey issues may arise from his curren [2]. Increasing radiaed elecromagneic emissions, grid curren disorions and losses are he oher disadvanages of a ransformerless srucure [3]. Several mehods are proposed for CMV and leakage curren reducion in lieraures [4]-[15]. Some mehods are based on he srucure modificaion adding some exra semiconducor elemens o he main srucure [4]-[7]. In [5], H7 opology, using seven swiches, is proposed for CMV reducion in hreephase grid-conneced sysem. Four-leg inverer wih modified modulaion echnique is proposed in [6]. Srucure-based mehods usually increase he complexiy and he cos of he sysem. Some oher reduced CMV mehods (RCM) are based on modulaion modificaion which are less expensive and have simpler implemenaion. For example, acive zero sae PWM (AZSPWM), remoe sae PWM (RSPWM) and near sae PWM (NSPWM) in [8]-[13] are all based on modulaion modificaion. Inerleaved carrier-based CMV reducion mehod inroduced in [14], is anoher mehod from his ype. RSPWM is he only mehod which can keep CMV consan in a classic volage source inverer (VSI); however, i enails some limiaions in he sysem. The amoun of CMV reducion and he oal performance of hese mehods are compared in [11] and [14]-[15]. Z-source inverer (ZSI) is a beneficial srucure in PV applicaions wih buck/boos characerisic in a single sage [16]-[18].Various srucures are derived from he original ZSI srucure and named as q-zsis [19]-[24]. q-zsi wih he srucure shown in Fig. 1, is applied widely in disribued generaion [21]-[26]. The hybrid srucure of q-zsi wih baery sorage is presened in [27]-[30]. q-zsi shown in Fig. 1, has some advanages comparing wih he radiional ZSI. _ V C s C2 C 2L2 + L 1 D grid C 2 inverer L 1 D L 2 C + + v L1 1 + v L2 + C s capacior filer/ V a b c qz nework in local load V C1 C 1 i leakage Z e Fig. 1. Three-phase q-zsi. he main opology; grid-conneced q-zsi wih PV connecion a he inpu: including he sray capaciors and he ground modeling impedance.

2 Is inpu curren is coninuous and he volage across is much lower han ha on. The laer leads o reduced passive componen raing and lower manufacuring cos [31]. None of he modulaion-based RCM mehods is ideal for Z- source family inverers. Because of SH saes, high-frequency harmonics of CMV are larger han hose in a simple VSI and leads o higher disruping leakage curren. The leakage curren pah and he sray capaciance ( ) in a q-zsi are shown in Fig. 1. In [3], he hree-phase ZSI wih exra fas recovery diode (ZSI_D) is presened o decrease he leakage curren in ransformerless PV sysems. ZSI_D applies odd and even PWM mehod (OEPWM). In ZSI_D, an exra fas recovery diode is conneced o he negaive erminal of he PV. The inpu diodes in ZSI_D provide isolaion in erminals of he PV array from he inverer swiches during SH saes. In ZSI_D, he CMV sill conains high-frequency harmonics. Alhough he leakage curren is reduced in ZSI_D, bu in pracice, he leakage curren could flow in boh urning-on and off momens of diodes. In addiion, using exra semiconducor elemen increases he cos of he sysem. In [32] he modified near sae mehod is presened for CMV reducion in ZSI. Similarly, he CMV is no cancelled compleely by using his mehod. In his paper, by a minor change in he q-zsi inpu inducor design, while using OPWM mehod (also called RSCMV1 [8]- [13]), he CMV is kep consan and he leakage curren is compleely blocked. In secion II, he main opology of q-zsi is reviewed. The CMV is analyzed in a hree-phase q-zsi in secion III. In secion IV, OPWM is described and he implemenaion of he proposed CMV reducion mehod is presened. The deailed experimenal resuls presened in secion V show ha he CMV and he leakage curren are canceled in hree-phase q-zsi by using he proposed mehod. II. SVM-BASED QUASI-Z-SOURCE INVERTER The q-zsi shown in Fig. 1, has wo operaing modes [33]: 1) Inverer mode: Similar o classic inverers, in his mode, q-zsi has six acive vecor saes and wo zero saes. This mode is named as non-shoo-hrough mode or inverer mode. In his mode, he inpu curren flows hrough he inpu side diode ( in Fig. 1) and he q-zsi impedance nework capaciors ( and ) are charged during his mode. 2) Shoo-hrough mode: In his mode, he inverer is shored by is one, wo or all hree legs. In his mode, he nework diode (D in Fig. 1) is urned off via he reveres-bias volage and he impedance nework capaciors charge he wo inducors ( ) a he same ime. The volage across he leg erminals during SH mode is approximaely zero and during inverer mode is equal o: (1) shown in Fig. 1, is he inpu DC source. In (1), B is he boos facor of a q-zsi and is defined as, (2) in (2) is he relaive duraion of SH sae in a swiching period which is obained by he conroller. The value of is calculaed based on source volage or as an oupu variable of he maximum power poin racking (MPPT) module, or as a conrol oupu which is compaible wih he baery siuaion in a hybrid q-zsi [34]. In a simple q-zsi, is usually regulaed by a volage conrol loop hrough adjusmen. In seady sae, he impedance nework capaciors volage are obained as ([33]): 1 (3) (4) In SVM mehod, here are six acive swiching saes and wo zero saes. Fig. 2, shows he volage space vecors and A- ype secors in αβ plane. In addiion o he eigh radiional swiching saes of a VSI, in a q-zsi, here are seven SH saes; when boh he upper and lower swiches of one or muliple phase legs are on [3]. The duraion of he vecors in a swiching cycle is calculaed like ha of a simple SVM inverer [35]. For example, if he reference vecor is in he firs secor ( in Fig. 2), he duraion of he acive and passive vecors are obained as below [35]: sin, 0 (5) sin (6) 0.5T (7) In (7), refers o he duraion of he passive sae in which he hree lower swiches are on ( vecor); while is relaed o he inerval in which he hree upper swiches are on ( vecor). is he swiching period and is he line frequency. is modulaion index in q-zsi, where: The modulaion index definiion in q-zsi is similar o a radiional VSI. V 4 (011) V 3 (010) A 3 V 5 (001) V 6 (101) Fig. 2. The space vecors and A-ype secors in SVM mehod. A 2 TABLE I THE CMV VALUES IN Q-ZSI Vecor,, odd vecors,, Even vecors V 2 (110) A 1 A 5 θ=ωm V 1 (100) A 4 A 6 CMV * Seven vecors for he shoo hrough saes. (8)

3 1 2/3 1/3 III. COMMON MODE VOLTAGE IN QUASI-Z-SOURCE INVERTER The CMV for hree-phase inverer is defined as [3]: (9) Table I, shows he swiching saes and he CMV values in a q-zsi. The CMV waveform in q-zsi is no uniform wihin a swiching period and i includes sep variaions. For example as i is seen in Table I, applying odd and even saes cause 1/3 and 2/3 levels in he CMV waveform respecively. Therefore, he arrangemen of odd, even and zero saes affecs he CMV waveform and i can be concluded ha, he CMV waveform is changed by applying differen modulaion echniques. The arrangemen of he sae vecors and SH inervals during a swiching period (in secor) for a SVM-based q-zsi is shown in Fig. 3, in which he SH inervals are deermined by he filled recangles. The CMV waveform for his swiching cycle is also depiced in Fig. 3. The flucuaions in he CMV waveform leads o leakage curren flow in he grid-conneced PV sysem. In [36]-[37], he equivalen circui for he common mode curren flow in hree-phase inverers is presened. The equivalen circui is shown in Fig.4. and in Fig. 4 are he line inducance filer elemens, scaled by one-hird in he equivalen circui. The CMV obained from (9) is used o esimae he leakage curren. The sray capaciance is abou / [36]. The ground impedance,, has an uncerain value and is mosly resisive. I migh be around a few Ohms. The equivalen impedance hrough he pah (for he frequency of ) equals o: (10) If he CMV includes he harmonic of, he effec of his harmonic appears in he leakage curren due o he pah impedance. According o (10), he pah impedance for zero frequency is infinie; herefore, heoreically here is no DC C s T sh /4 N T sh /4 V cm L f /3 i leakage V CM /V DC T sh /4 R f /3 T sh / T 0 /2 T a /2 T b /2 T 7 T b /2 T a /2 T 0 /2 Fig. 3. The swiching saes and he CMV waveform for SVM-based q-zsi during a swiching period in A 1 secor. Z e Fig. 4. The equivalen circui for leakage curren flow [37]. curren flow in he circui. IV. ODD VECTOR PWM METHOD Using only he odd vecors in OPWM mehod or he even vecors in even PWM (EPWM) are wo similar ways o CMV cancellaion in VSIs, which are presened in lieraure as [3] and [32]. In OPWM mehod, only he odd vecors (,, ) are used o compose he oupu reference vecor. As shown in Fig. 5, in his mehod, he αβ plane is divided in hree secors. According o [3], in a classic VSI, he relaive ime for each odd vecor in a swiching period (, 1,3,5) is calculaed as: cos sin sin (11) In (11), is he DC volage across he leg erminals, and θ equals o he reference phase angle. OPWM mehod is applicable in q-zsi. From vecor calculaions, he relaive duraion for odd vecors in a q-zsi are obained as: V 3 (010) V 5 (001) V ref θ cos V 1 (100) Fig. 5. The odd vecors and secors in OPWM mehod. sin sin (12) One-hird of SH duraion could be applied beween each wo odd vecors. A. Scalar Implemenaion of OPWM in DSP For scalar implemenaion of OPWM in DSP-conrolled inverers, imer phase shif mehod is proposed. In Fig. 6, he imer signals and he swiching commands are depiced for OPWM mehod. Timer_a, Timer_b, and Timer_c are relaed o he phases a, b and c respecively. The period of imers, PRD, is seleced due o he swiching period. Timer_a is assumed he reference imer. Using (12), in each swiching cycle he phase of imers are se as: 0 (13) The compare signals values (,,,, ) for and swiches equal o:

4 PRD Timer _a phs a CMPR ah CMPR al T sw phs b Timer _b phs c T 1 +D sh /3 T 1 +T 3 +2D sh /3 CMPR bh CMPR bl Timer _c CMPR ch CMPR cl S ah S al S bh S bl S ch T 1 D sh /3 T 3 D sh /3 T 5 D sh /3 S cl Fig. 6. Timer phase shif mehod for scalar implemenaion of OPWM in DSP-conrolled q-zsi. Fig. 8. The CMV waveform of he q-zsi. using SVM modulaion; using OPWM modulaion. 1/3 V CM/V DC T sh /6 T sh /3 T sh/3 T sh/6 T 1 T 3 T 5 Fig. 7. The CMV in q-zsi using OPWM modulaion echnique.. ;.. ;.. ;. In Fig. 6, he compare signals are shown. (14) B. Common-Mode-Volage in OPWM q-zsi Due o SH saes beween odd vecors, applying OPWM mehod in q-zsi doesn lead o a consan CMV (unlike radiional VSI). In Fig. 7, he ypical CMV waveform for OPWM q-zsi is depiced. Comparing wih CMV in SVM q- ZSI in Fig. 3, he sep variaions of CMV are reduced in OPWM mehod, bu he CMV sill conains high-frequency harmonics. A q-zsi wih specificaion shown in Table II is considered as an example. The CMV for SVM and OPWM mehods are shown in Fig. 8 and Fig. 8 respecively., and equals 380v, 0.15 and 0.6. The CMV harmonics are measured hrough a high-frequency specrum analyzer (Agilen HP 8561B) and shown in Fig. 9- for SVM and OPWM mehods respecively. From Fig. 9 -, i is seen ha, for he same leg volage and he same modulaion index, using OPWM mehod in q-zsi resuls in CMV harmonic reducion, bu he issue of high-frequency harmonics is no Fig. 9. The CMV harmonics. Using SVM mehod. Using OPWM mehod. disappeared compleely. Considering ha he leg volage in OPWM mehod is larger han in SVM, he harmonics are even larger han shown in Fig. 9.

5 C. Modified quasi Z-Source using Odd-PWM According o (3)-(4), he volage across he Z nework inducor (wih he polariy shown in Fig. 1) during he inverer and SH modes is calculaed as: (15) 1 The CMV waveform in OPWM mehod and ypical volage across he inducor L 1 ( ), are depiced in Fig. 10. In despie of he CMV waveform,, equals is maximum value in SH inervals (1. ) and during he inverer mode i has is minimum value (. ). As shown in Fig. 11, he inpu inducor is divided o wo series inducors, and, wih he raio of, where: 1 (16) 1 CMV in his case is obained as: (17) The node,, is shown in Fig. 11. From (15) and Table I, CMV is equal o: /3 (18) 1 0 The CMV waveform becomes consan, if equals 1/3. In oher words, if and are seleced as 1/3 and 2/3, he CMV is equal o a consan value: (19) q-zsi wih wo-piece inpu inducance is named as modified q-zsi (mq-zsi). CMV in mq-zsi using OPWM is a DC volage, wihou high-frequency harmonics. D. OPWM Comparing wih SVM As presened in [3], he modulaion index in OPWM mehod is limied o 0.61, while in SVM i is limied o This means ha, in OPWM mehod, higher DC volage is required a he leg erminals and he inpu source. However, in Z-source family inverers, SH inervals booss he leg erminals volage and he problem of need o a higher inpu volage could be raher relieved. In SVM mehod, he passive sae is implemened hrough zero vecors ( and ), while in OPWM mehod, i is implemened hrough applying equally he hree odd vecors. For example, if he relaive passive ime is in OPWM mehod, each odd vecor is applied for /3. The relaive passive duraion in OPWM mehod is equal o: 3,, (20) In Z-source family, SH duraion shouldn exceed he passive duraion; herefore, for proper operaion of he inverer using OPWM, i is concluded: (21) From (12) and (20)-(21), i can be said: 1 (22) Using SVM mehod, he permied range for is obained, as: D sh D sh The permied work zone Fig. 12. The permied work zones in q-zsi using SVM or OPWM. TABLE II THE TOTAL SPECIFICATION OF THE DESIGNED PROTOTYPE. line o neural volage M qz (OPWM) The permied work zone M qz (SVM) 110 line frequency 50 V CM /V DC 1/3 v L1/V DC (1-D sh ) -D sh non-sh non-sh non-sh Fig. 10. The ypical volage across he nework inducor comparing wih he CMV. V in P N L 1 + =2/3L 1 _ 2/3V L1 Fig. 11. mq-zsi main srucure. non-sh non-sh non-sh C 2 D _ 1/3V L1 + L 1 N L 2 T sw C 1 a b c Z nework capaciance 220 Z nework inducance 1 TABLE III SVM AND OPWM METHODS PARAMETERS FOR THE Q-ZSI. SVM OPWM permied inpu volage ( ) 160V-357V 342V-554V permied SH duraion ( ) he peak of leg volage ( ) 380v 590v modulaion index ( ) TABLE IV SIMULATION RESULTS FOR SVM Q-ZSI AND OPWM MQ-ZSI curren THD volage THD leakage curren SVM q-zsi OPWM mq-zsi SVM q-zsi OPWM mq- ZSI SVM q-zsi OPWM mq-zsi 0 1.8% 4.8% 128% 195% % 4.8% 130% 195% 15.3A % 4.8% 128% 195% 1.74A % 4.8% 129% 195% 1.52A % 4.8% 128% 195% 0.84A 1 *Resonan frequency of he leakage impedance is equal o 1/

6 1 1 (23) The permied range of due o he modulaion index is shown in Fig. 12 for boh SVM and OPWM mehods. To have 0.25), he modulaion index should a boos facor of 2 ( be designed less han 0.5 and.87 in OPWM and SVM mehods respecively. The oupu volage oal harmonic disorion (THD) in OPWM is larger han ha in SVM mehod. In [11], a oal comparison is done for THD levels in differen RCM mehods. Flowing leakage curren increases THD level in he oupu curren, while THD of he oupu volage is no significanly affeced. Two 3 sysems wih specificaion shown in Table II-III are designed and simulaed. Due o (22)-(23), he ranges for he seleced permied inpu volage and are shown in Table III. The THD levels of he oupu currens and volages and also he amoun of leakage curren for differen values of sray capaciance are shown in Table. IV. The blocking volage of he IGBTs is 600. The inverer oupu filer inducance is 6 and he swiching frequency is equal o 9.2. For he sray capaciance of 150nF, he resonan frequency of he leakage curren pah is equal o he swiching frequency, which leads o he wors case THD and equals is leakage curren in SVM q-zsi. In Table IV, is less han 600 in boh SVM q-zsi maximum value and and OPWM mq-zsi. The volage raing of he IGBTs limis ( 0.51) and also due o pracical he lower value of consideraions, is no seleced higher han 0.3. While in is achievable SVM q-zsi he maximum pracical level of, maximum value is lower in by selecing 0.82 for OPWM q-zsi because of lower value limiaion (Table III). In he simulaion, is zero. The curren THD in SVM q-zsi, is affeced by he value of sray capaciance, while his parameer is consan in OPWM mq-zsi. As shown in Table IV, for zero leakage, he curren THD in SVM q-zsi and OPWM mq-zsi are 1.8% and 4.8% respecively. The curren THD is increased o 62% for 150nF sray capaciance in SVM q-zsi while i is consan in OPWM mq-zsi. Due o higher volage THD in OPWM mq-zsi comparing o SVM mq-zsi, larger oupu inducance is required o have he same curren THD level. However, higher leakage curren increases significanly curren THD in SVM q-zsi. In OPWM q-zsi, and olerances affec heir raio, and consequenly he leakage curren; bu, for a small olerance his effec is negligible. For example, 5% and Fig. 13. The designed prooype for q-zsi. Fig. 14. The hree-phase oupu curren; RMS value is 3.5 A (he sensor gain is 200mv/A). using SWM; using OPWM. 10% olerance in he inducors values (for he case of 150nF sray capacior), leads o he leakage currens of 7mA and 20mA respecively for he wors case. V. THE EXPERIMENTAL RESULTS A low-volage q-zsi prooype is designed and boh SVM and OPWM conrol mehods are implemened. The oal specificaion of he sysem is shown in Table II-III. The inverer oupu inducance and he swiching frequency are 3 and 21. The implemened prooype is shown in Fig. 13. The digial signal processor, TMS320F2808 is used as he main conroller, providing he swiching and he proecion commands. The phase locked loop (PLL) based on dq ransformaion is implemened for being synchronized o he grid when he sysem is grid-conneced. Isolaed volage and curren measuremens are provided by LEM sensors, he measured values from he sensors are ransferred o he Fig. 15. The upper swiches commands by imer phase shif mehod.

7 (c) Fig. 16. The phase erminals oupu volage and he CMV volage SVM q-zsi;. OPWM q-zsi; (c) OPWM mq-zsi. processor by he analog cable (shown in Fig. 13). The PV inpu volage is modeled wih a DC supply volage. The module FSBS15CH60 from FAIRCHILD is used as he power converer, a he boom of he main board in Fig. 13. and in Fig. 13 is nearly equal o 333 and 666 respecively. The oupu phase currens (wihou leakage curren) for SVM and OPWM mehods are shown in Fig The RMS value of he oupu curren equals 3.5A. The prooype sysem provides 1.2kW. The upper swiches commands for phase shif mehod in OPWM scalar implemenaion are shown in Fig. 15. The odd vecors (, and ) are implemened by he pulses shown in Fig. 15, wih he same order depiced in Fig. 6. For example, pulse duraion in Fig. 15, represens duraion and a par of SH sae. The odd vecor ransiions occur in SH inervals. The CMV and he oupu erminals volage of he q-zsi are measured for SVM and OPWM mehods. In Fig. 16-, he CMV of he q-zsi is shown for SVM and OPWM mehods respecively. The CMV waveforms in Fig. 16-, have he (c) Fig. 17. The leakage curren. SVM q-zsi;. OPWM q-zsi; (c) OPWM mq-zsi. ypical forms of he CMV in a SVM q-zsi (Fig. 3) and OPWM q-zsi (Fig. 7). Alhough he leg erminal volage is increased by he facor of 1.5 in Fig. 16, he CMV sep variaions are decreased compared o ha shown in Fig. 16. The CMV flucuaions in Fig. 16 is due o SH saes which can be idenified by zero-volage inervals. The oupu erminals volages (, and ) in Fig. 16 conain uniform sep variaions wih he ampliude of 590. In Fig. 16(c), he CMV for OPWM mq-zsi is depiced. Using a wopiece inducor a he inpu of q-zsi, non-uniform flucuaions wih he ampliude of approximaely 200 are added o he oupu erminals volages. The CMV in Fig. 16 (c) is nearly consan and is equal o 170. A lile glich is seen on he CMV a SH momens. Considering 150 sray capacior and zero ground impedance, he leakage curren is measured for SVM q-zsi, OPWM q-zsi and OPWM mq-zsi. The resuls are shown in Fig. 17-(c). SVM q-zsi has he maximum leakage curren, equals o 1.7A rms. The mos dominan harmonic in Fig. 17 is 21 (swiching frequency). In OPWM q-zsi, he leakage curren is reduced significanly and

8 is equal o 0.3A; his curren is shown in Fig. 17. In Fig. 17, he higher order swiching harmonics are also seen in he waveform. In mq-zsi, by redesigning he inpu inducor, he leakage curren is reduced o is smalles value of abou 15mA. Higher ampliude leakage curren (as in Fig. 17) affecs srongly he oupu curren THD. In SVM q-zsi, OPWM q-zsi and mq-zsi (which heir leakage currens are shown in Fig. 17-(c)) he oupu currens THD are 22.2%, 9.6% and 8.4% respecively. VI. CONCLUSION In his paper, by using OPWM modulaion echnique and minor change in q-zsi srucure, he leakage curren is blocked. Applying exising CMV reducion mehods as OPWM mehod in he radiional q-zsi doesn resul in consan CMV, because of SH inervals. In he example case, by applying OPWM, he leakage curren is reduced significanly. By a minor change in he Z nework filer design and applying OPWM, i is possible o keep he CMV as a nearly consan value. No exra semiconducor elemen is added o he converer; only he inpu inducor is splied ino wo pars. The volage variaions of CMV is compensaed wih he inducor connecing o he negaive erminal. REFERENCES [1] J. M. Shen, H. L. Jou, and J. C. Wu, Novel ransformerless gridconneced power converer wih negaive grounding for phoovolaic generaion sysem, IEEE Trans. on Power Elecron., vol. 27, DOI 10/1109/TPEL/2011/ , no. 4, pp , Apr [2] B. Yang, W. Li, Y. Gu, W. Cui, and X. He, Improved ransformerless inverer wih common-mode leakage curren eliminaion for a phoovolaic grid-conneced power sysem, IEEE Trans. on Power Elecron., vol. 27, DOI 10/1109/TPEL/2011/ , no. 2, pp , Feb [3] F. Bradaschia, M. C. Cavalcani, P. E. P. Ferraz, and F. A. S. Neves, Modulaion for hree-phase ransformerless Z-source inverer o reduce leakage currens in phoovolaic sysems, IEEE Trans. on Ind. Elecron., vol. 58, DOI 10/1109/TIE/2011/ , no. 12, pp , Dec [4] C. T. Morris, D. Han and B. Sarlioglu, Reducion of common mode volage and conduced EMI hrough hree-phase inverer opology, IEEE Trans. on Power Elecron., vol. 32, DOI 10/1109/TPEL/2016/ , no. 3, pp , Mar [5] T. K. S. Freddy, N. A. Rahim, W. P. Hew, and H. S. Che, Modulaion echniques o reduce leakage curren in hree-phase ransformerless H7 phoovolaic inverer, IEEE Trans. on Ind. Elecron., vol. 62, DOI 10/1109/TIE/2014/ , no. 1, pp , Jan [6] P. Garg, S. Essakiappan, H. S. Krishnamoorhy, and P. N. Enjei, A faul-oleran hree-phase adjusable speed drive opology wih acive common-mode volage suppression, IEEE Trans. on Power Elecron., vol. 30, DOI 10/1109/TPEL/2014/ , no. 5, pp , May [7] A. L. Julian, G. Orii, and T. A. Lipo, Eliminaion of common-mode volage in hree-phase sinusoidal power converers, IEEE Trans. on Power Elecron., vol. 14, DOI 10/1109/63/788504, no. 5, pp , Sep [8] C. C. Hou, C. C. Shih, P. T. Cheng, and A. M. Hava, Common-mode volage reducion pulsewidh modulaion echniques for hree-phase grid-conneced converers, IEEE Trans. on Power Elecron., vol. 28, DOI 10/1109/TPEL/2012/ , no. 4, pp , Apr [9] X. Wu, G. Tan, Z. Ye, Y. Liu, and S. Xu, Opimized common-mode volage reducion PWM for hree-phase volage-source inverers, IEEE Trans. on Power Elecron., vol. 31, DOI 10/1109/TPEL/2015/ , no. 4, pp , Apr [10] L. Guo, X. Zhang, S. Yang, Z. Xie, and R. Cao, A model predicive conrol-based common-mode volage suppression sraegy for volagesource inverer, IEEE Trans. on Ind. Elecron., vol. 63, DOI 10/1109/TIE/2016/ , no. 10, pp , Oc [11] A. M. Hava, and E. Un, Performance analysis of reduced commonmode volage PWM mehods and comparison wih sandard PWM mehods for hree-phase volage-source inverers, IEEE Trans. on Power Elecron., vol. 24, DOI 10/1109/TPEL/2008/ , no. 1, pp , Jan [12] A. M. Hava, and E. Un, A high-performance PWM algorihm for common-mode volage reducion in hree-phase volage source inverers, IEEE Trans. on Power Elecron., vol. 26, DOI 10/1109/TPEL/2010/ , no. 7, pp , Jul [13] E. Un, and A. M. Hava, A near-sae PWM mehod wih reduced swiching losses and reduced common-mode volage for hree-phase volage source inverers, IEEE Trans. on Ind. Appl., vol. 45, DOI 10/1109/TIA/2009/ , no. 2, pp , Mar.- Apr [14] J. W. Kimball, and M. Zawodniok, Reducing common-mode volage in hree-phase sine-riangle PWM wih inerleaved carriers, IEEE Trans. on Power Elecron., vol. 26, DOI 10/1109/TPEL/2010/ , no. 8, pp , Aug [15] H. Chen, and H. Zhao, Review on pulse-widh modulaion sraegies for common-mode volage reducion in hree-phase volage-source inverers, IET Power Elecron., vol. 9, DOI: 10/1049/ie-pel/2015/1019, no. 14, pp , Nov [16] F. Z. Peng, Z-Source Inverer, IEEE Trans. on Ind. Appl., vol. 39, DOI: 10/1109/TIA/2003/808920, no. 2, pp , Mar./Apr [17] N. Sabeur, S. Mekhilef, and A. Masaoud, Exended maximum boos conrol scheme based on single-phase modulaor for hree-phase Z- source inverer, IET Power Elecron., vol. 9, DOI: 10/1049/iepel/2015/0124, no. 4, pp , Mar [18] Y. Tang, S. Xie, and J. Ding, Pulsewidh Modulaion of Z-Source Inverers Wih Minimum Inducor Curren Ripple, IEEE Trans. on Ind. Elecron., vol. 61, DOI: 10/1109/TIE/2013/ , no. 1, pp , Jan [19] J. Anderson, and F. Z. Peng, A Class of Quasi-Z-Source Inverers, in IEEE Indusry Applicaions Sociey Annual Meeing, DOI 10/1109/08IAS/2008/301, pp. 1-7, Oc. 2008,. [20] D. Cao, and F. Z. Peng, A Family of Z-source and Quasi-Z-source DC- DC Converers, in IEEE Applied Power Elecronics Conference and Exposiion, DOI 10/1109/APEC/2009/ , pp , Mar [21] Y. Liu, B. Ge, H. Abu-Rub, and D. Sun, Comprehensive modeling of single-phase quasi-z-source phoovolaic inverer o invesigae lowfrequency volage and curren ripple, IEEE Trans. on Ind. Elecron., vol. 62, DOI 10/1109/TIE/2014/ , no. 7, pp , Jul [22] Y. Zhou, H. Li, and H. Li, A single-phase PV quasi-z-source inverer wih reduced capaciance using modified modulaion and doublefrequency ripple suppression conrol, IEEE Trans. on Power Elecron., vol. 31, DOI 10/1109/TPEL/2015/ , no. 3, pp , Mar [23] B. G. e. al, An acive filer mehod o eliminae DC-side low-frequency power for a single-phase quasi-z-source inverer, IEEE Trans. on Ind. Elecron., vol. 63, DOI 10/1109/TIE/2016/ , no. 8, pp , Aug [24] S. Bayhan, H. Abu-Rub, and R. S. Balog, Model predicive conrol of quasi-z-source four-leg inverer, IEEE Trans. on Ind. Elecron., vol. 63, DOI 10/1109/TIE/2016/ , no. 7, pp , Jul [25] M. S. Diab, A. A. Elserougi, A. M. Massoud, A. S. Abdel-Khalik, and S. Ahmed, A pulsewidh modulaion echnique for high-volage gain operaion of hree-phase Z-source inverers, IEEE Journal of Emerging and Seleced Topics in Power Elecron., vol. 4, DOI /JESTPE/2015/ , no. 2, pp , Jun [26] H. Abu-Rub, A. Iqbal, Sk. Moin Ahmed, F. Z. Peng, Y. Li, and G. Baoming, Quasi-Z-Source inverer-based phoovolaic generaion sysem wih maximum power racking conrol using ANFIS, IEEE Trans. on Susainable Energy, vol. 4, DOI 10/1109/TSTE/2012/ , no. 1, pp , Jan [27] Y. Liu, B. Ge, H. Abu-Rub and H. Sun, Hybrid pulsewidh modulaed single-phase quasi-z-source grid-ie phoovolaic power sysem, IEEE Trans. on Ind. Inf., vol. 12, DOI 10/1109/TII/2016/ , no. 2, pp , Apr [28] J. Khajesalehi, K. Sheshyekani, M. Hamzeh and, E. Afjei, Highperformance hybrid phoovolaic -baery sysem based on quasi-zsource inverer applicaion in microgrids, IET Generaion,Transmission

9 & Disribuion, vol. 9, DOI 10/1049/ie-gd/2014/0336, no. 10, pp , Jul [29] S. Hu, Z. Liang, D. Fan and X. He, Hybrid ulracapacior baery energy sorage sysem based on quasi-z-source opology and enhanced frequency dividing coordinaed conrol for EV, IEEE Trans. on Power Elecron., vol. 31, DOI 10/1109/TPEL/2016/ , no. 11, pp , Nov [30] J. Khajesalehi, K. Sheshyekani, M. Hamzeh, and E. Afjei, Maximum consan boos approach for conrolling quasi-z-source-based inerlinking converers in hybrid AC DC microgrids, IET Generaion, Transmission & Disribuion, vol. 10, DOI 10/1049/ie-gd/2015/0607, no. 4, pp , Mar [31] Y. Li, S. Jiang, J. G. Cinron-Rivera, F. Z. Peng, "Modeling and conrol of quasi-z-source inverer for disribued generaion applicaions, IEEE Trans. on Ind. Elecron., vol. 60, DOI 10/1109/TIE/2012/ , no. 4, pp , Apr [32] V. Erginer, and M. H. Sarul, A novel reduced leakage curren modulaion echnique for Z-source inverer used in phoovolaic sysems, IET Power Elecron., vol. 7, DOI 10/1049/ie-pel/2013/0187, no. 3, pp , Mar [33] B. Ge e al. An energy-sored quasi-z-source inverer for applicaion o phoovolaic power sysem, IEEE Trans. on Ind. Elecron., vol. 60, DOI 10/1109/TIE/2012/ , no. 10, pp , Oc [34] Y. Liu, B. Ge, H. Abu-Rub, and F. Z. Peng, Modelling and conroller design of quasi-z-source inverer wih baery-based phoovolaic power sysem, IET Power Elecron., vol. 7, DOI 10/1049/ie-pel/2013/0389, no. 7, p , Jul [35] J. Liu, S. Jiang, D. Cao, and F. Z. Peng, A digial curren conrol of quasi-z-source inverer wih baery, IEEE Trans. on Ind. Inf., vol. 9, DOI 10/1109/TII/2012/ , no. 2, pp , May [36] B. Sahan, A. N. Vergara, N. Henze, Alfred Engler, and P. Zacharias, A single-sage PV module inegraed converer based on a low-power curren-source inverer, IEEE Trans. on Ind. Elecron., vol. 55, DOI 10/1109/TIE/2008/924160, no. 7, pp , Jul [37] Y. Bae, and R. Y. Kim, Suppression of common-mode volage using a mulicenral phoovolaic inverer opology wih synchronized PWM, IEEE Trans. on Ind. Elecron., vol. 61, DOI 10/1109/TIE/2013/ , no. 9, pp , Sep Negar Noroozi (S 17) was born in Tabriz, Iran, in She received her B.S. degree in elecrical engineering from Universiy of Tehran, Iran, in 2007, and her M.S. degree in elecrical engineering from Sharif Universiy of Technology, Tehran, Iran, in She is currenly a Ph.D. candidae a he Deparmen of Elecrical and elecronics engineering of Sharif Universiy of Technology, Tehran, Iran. Her research ineress are renewable energy, AC/DC and DC/AC converers. Mohammad Reza Zolghadri (SM 13) is an associae professor and head of power sysem group a he deparmen of Elecrical Engineering, Sharif Universiy of Technology, Tehran, IRAN. He received his B.S. and M.S. degrees from Sharif Universiy of Technology, in 1989 and 1992, respecively, and he Ph.D. degree from Insiue Naional Polyechnique de Grenoble (INPG), Grenoble, France, in 1997, all in elecrical engineering. In 1997 he joined he deparmen of Elecrical Engineering of Sharif Universiy of Technology. From 2000 o 2003, he was a Senior Researcher in he Elecronics Laboraory of SAM Elecronics Company, Tehran. From 2003 o 2005, he was a Visiing Professor in Norh Carolina A&T Sae Universiy, USA. He is he founder and head of Elecric Drives and Power Elecronics Lab (EDPEL) a Sharif Universiy of Technology. He is a member of founding board of Power Elecronics Sociey of Iran (PESI). He is he auhor of more han 100 publicaions in power elecronics and variable speed drives. His fields of ineres are applicaion of power elecronics in energy sysems, modeling and conrol of power elecronic converers and variable speed drives.