A Comparison of Control Methods for Z-Source Inverter

Transcription

1 Energy and Power Engneerng, 2012, 4, Publshed Onlne July 2012 ( A Comparson of Control ethods for Z-Source Inverter Cong-hanh Pham 1,2, Anwen Shen 1, Phan Quoc Dzung 3, Nguyen Bao Anh 3, Nguyen Xuan Phu 3 1 Department of Control Scence and Engneerng, Huazhong Unversty of Scence and echnology, Wuhan, Chna 2 Electrcal & Electroncs Engneerng Fulty, HCC Unversty of ransport, Ho Ch nh Cty, Vetnam 3 HCC Unversty of echnology, Ho Ch nh Cty, Vetnam Emal: phamcongthanh09@yahoo.com Receved ay 15, 2012; revsed June 18, 2012; cepted June 29, 2012 ABSRAC In recent years, Z-source nverters (ZSI) have been proposed as a replement power converson concept whch t has both voltage buck and boost abltes. In addton, ZSI doesn t requre dead-tme to protecton short crcut at two swtches any of the same phase leg n the nverter brdge and to heve optmal harmonc of current, voltage. hs paper presents two dfferent control methods (C) for ZSI. he am of ths study to compare between two modulaton methods, there are modfed spe vector pulse wdth modulaton method (SV) and the smple boost control (SBC) about the unque harmonc performance features, the total average and peak swtchng devce power of the nverter system. In addton, ths paper also analyzes about the ablty exceed modulaton ndex n lnear regon of two C usng ALAB/Smulnk. Keywords: Buck-Boost; Current Source Inverters (CSIs); Pulse-Wdth odulaton (PW); Voltage Source Inverters (VSIs); Z-Source Inverters (ZSI) 1. Introducton In a conventonal voltage source nverter, the two swtches of any phase leg cannot be gated at the same tme because ths may cause a short crcut stuaton and thus destroy the nverter. In addton, the maxmum output voltage cannot exceed the dc bus voltage. hese lmtatons n such conventonal voltage source nverter can be overcome by usng ZSI [1], Fgure 1. A comparson among conventonal pulse-wdth modulaton (PW) nverter, dc-dc boosted PW nverter and ZSI but ZSI shows that hgher effcency and lower cost. Because of these reasons, ZSI has appled n hybrd electrc vehcles [2,3]. In addton, n [4] show comparson among the tradtonal I-source nverter and (VSI) wth the Z-source network should requre less captance, nductance and sze smaller than tradtonal dc-dc boosted PW nverter. herefore, these results n ncreasng attenton on ZSI, especally for the applcaton where the dc nput source has a wde voltage varaton range such as fuel cell, battery, lead d. Also, the ZSI has appled for drect-drve wnd generaton system and photovolt power system base on trkng algorthm maxmum power and hybrd electrc vehcles [5-7]. Addtonally, the maxmum boost control s alke the conventonal carrer-based PW method base on all zero voltage vectors, shoot through state (the other zero vectors) and the sx tve states whch are mantaned unchanged but the man drawbks of ths modulaton method s bg nductor rpple and a large nductor when the output frequency s low. herefor, they wll ncrease the cost and sze of the crcut [8]. Gven ts many benefts, many studes have been researched nto C of ZSI. Cosequently, how chose C whch hghest effcency s very mportant n control systems of ZSI. hrough detaled analyss, the am of ths study to show how varous two C for ZSI, there are SV and the SBC whch they are analyzed and compared about the unque harmonc features of current, voltage and the total average swtchng devce power, the total peak swtchng devce power of the nverter system. In addton, ths paper also analyzes about the ablty exceed modulaton ndex n lnear regon of two C usng smulaton ALAB/Smulnk. 2. Equvalent Crcut, Operatng Prncple and otal Swtchng Devce Power of ZSI 2.1. Equvalent Crcut, Operatng Prncple of ZSI In Fgure 1 show ZSI voltage-type confguraton, whch nclude nductors (L 1 and L 2 ) are smlar nductance L,

2 188 C.-. PHA E AL. Fgure 1. ZSI voltage-type confguraton. smultaneously captors (C 1 and C 2 ) are smlar captance C. Where L and C are connected n X shape, the dc nput voltage source whch may be fuel cell, battery or dode rectfer. he general nverter have two zero voltage vector and sx tve voltage vectors but n the threephase ZSI has one addtonal zero voltage vector that s thrd zero state vector or the shoot-through zero state. he nfluence of ths phase-leg shoot-through on the nverter performance can be consdered the equvalent crcuts shown n Fgure 2. In operatng prncple of ZSI there are two operatng modes: non-shoot-through mode and shoot-through mode. In non-shoot-through mode Fgure 2(a) that the ZSI operates under the conventonal PW. he shoot-through state s forbdden n the conventonal nverter but n the ZSI the shoot-through mode are gven n Fgure 2(b). In Fgure 2(b) show that the load termnals of any phase legs are shorted both the upper and lower. At that tme, the dc captor voltage can be boosted to the desred value. Where β s the angle between the reference voltage vector Vref and voltage vector V 1. Because assume nductors (L 1 and L 2 ), captors (C 1 and C 2 ) have alke nductance (L) and captance (C) respectvely, we have V V V v v v C1 C2 C L1 L2 L Non shoot-through tme nternal of one perod non shoot-through zero state for a perod n n Fgure 2(a), we have: V d V0 v L V VC (2) v VC v L 2VC V0 Shoot-through tme nternal of one perod shootthrough zero state for a perod n Fgure 2(b), we have: v 0 vl VC (3) Vd 2VC (1) Fgure 2. Equvalent crcuts of ZSI. (a) Non-shoot-through mode; (b) Shoot-through mode. where V o s the dc source voltage and one perod sw- tchng cycle sf, where sf n and s sf the shoot-through tme duty rato. he average voltage of the nductors over one swtchng perod should be zero n steady state, from (2) and (3) thus captor voltage can expressed: 1 sf 1 VC1 VC2 VC Vd Vd (4) sf 0n 2Vc Vd 1 V v V V 12 d c sf 1 v V V 2 V V V 12 c L c d d 1 sf 0 0 (7) 2 2 From (6), f s ncrease then v s also ncrease, where v s the voltage stress ross swtchng devces otal Swtchng Devce Power (SDP) In an nverter system, choce swtchng devce must depend on the peak, average current and the maxmum voltage mpressed gong through t. he swtchng devce (5) (6)

3 C.-. PHA E AL. 189 power s ntroduced to quantfy the voltage and current stress of an nverter system [8]. SDP s a measure of the total semconductor devce requrement. he total average swtchng devce power of the nverter system SDP s gven n [8]: av P 2 2P v V πv cos (8) o o SDP 4 v 1 av op he total peak swtchng devce power of the nverter system SDP s also gven n [8]: pk SDP pk P 2 o Po v 4 v (9) V v cos op where N s the number of swtchng devces used, I L the nductor current, v the output peak phase voltage from the nverter and P o maxmum power output of the DC-source voltage. 3. Proposed C of SV and SBC 3.1. Proposed C of SV Nowadays, the spe vector PW (SV) method have wdely used at regulated PW nverter due to a hgher modulaton ndex and lower current harmoncs n [9]. he SV s sutable to control the shoot-through tme n ZSI. Where V 0, V 7 and shoot-through are zero vectors (n ZSI), where V 1 to V 6 are the sx tve vectors. If the ref- erence voltage vector V ref (the snusodal three-phase command voltage of output ZSI wth mnmum amount of harmonc dstorton) s located between the arbtrary vector V a and V b, the reference voltage vector s dvded nto the two adjent voltage vectors (V a, V b ) and zero vectors (V 0, V 7 and shoot-through) n Fgure 3. When Vref rotate around secton (1-6) of hexagon whle (a, b) are changed: (a, b) = (1, 2); (2, 3); (3, 4); (4, 5); (5, 6) n every sector, respectvely. In one samplng nterval, V a and V b are appled at tmes a and b, respectvely, and a b 0 the zero vector s appled at tme sf where 0 0. Consequently, from (10), the reference voltage vector can be gven by V ref V ref V a a V b b (10) 3 Vref sn π a s f v (11) 3 Vre f b 3 sf sn (12) v he conventonal SV n Fgure 4(a) and the SV n Fgure 4(b) where s shoot-through tme has an extra for boostng the dc lnk voltage of the nverter besde tme ntervals a, b and 0. Fgure 3. Voltage vector through conven-tonal SV of VSI. From Fgure 4 the shoot-through states (SS) are evenly assgned to eh phase wth 2 wthn zero 3 0 voltage perod and wthn tve voltage perod 4 3 a and wthn tve voltage perod b, where a and b are unchanged. So the SS does not affect the SV C of the nverter, and t s lmted to the zero state tme 0. Where are determned by (13). herefore, the SS should be nserted perod ntervals are mantaned at the start and end of the swtchng cycle to heve alke optmal harmonc performance (13) 3 4 From (7) we have sf 0 (14) 8 In the undermodulaton regon 0 max. Fgure 3, V ref always remans wthn hexagon n [10]. Let us defne a modfed modulaton ftor gven by V ref Vref (15) V 2 1sw v π where V ref = vector magntude (or phase peak value), n the lnear regon Vref s the output peak phase voltage from the nverter v Vref v and V 1sw = fundamen- tal peak value 2 v π of the square-phase voltage

4 190 C.-. PHA E AL. Fgure 4. SV (a) Swtchng patterns for tradtonal SV; (b) Swtchng patterns for SV. wave. he maxmum possble value of modulaton ftor max at the end of the under modulaton regon can derved. he radus Vrefm Fgure 3 of the nscrbed crcle can be gven as 2 π 1 V v cos v refm max (16) V 2 2 1sw v v π π hs means that 90.7 percent of the fundamental at the square wave s avalable n the lnear regon he Smple Boost Control (SBC) he SBC method strategy nserts SS n all the PW conventonal zero states durng one swtchng perod. he sx tve states are mantaned unchanged as n the conventonal carrer based PW [2]. When the trangular waveform s greater than the upper envelope, V, or po lower than the bottom envelope, V ne, the crcut turns nto shoot-through state, show Fgure 5. he output peak phase voltage from the nverter can be gven by v 1 Vd v 2 12 a a (17) 2 where a s the modulaton ndex of magntude magntude sne 0 a 1 n [10]. At a 1, magntude carrer = modulaton ftor, n the lnear regon (0 max ) Fgure 5. Smple boost control. where max = maxmum modulaton ftor determned by 1 v v 2 max (18) V 2 1sw v π At a 1, the maxmum value of fundamental peak 1 voltage s v, Whch s percent of peak voltage 2 2 v π of the square wave. From (16), (18) we can see that the maxmum value of fundamental peak voltage of the SV method s hgher than the SBC method.

5 C.-. PHA E AL. 191 herefore, the SV method modulaton ndex n lnear can exceed more than the SBC method. 4. Smulaton o verfy the valdty of the above analyss, smulaton models shown n Fgure 1 use atlab/smulnk for SV and SBC methods of z-source nverter are analyzed. he smulaton for two Cs wth parameters are the same, see able 1. In Fgure 6 shows the nverter s dc lnk voltage, the ZSI s captor voltage and comparson wth maxmum output voltage of DC-source. We observe from the fgure that the nverter s dc lnk voltage v boosted to (500 V) greater than maxmum output voltage of DC-source (400 V) s (100 V) see able 2. In Fgure 7 shows the swtchng output lne voltage and phases voltage waveform of ZSI wth SV methods. Fgures 8 and 9 show output lne voltage and current sne waveforms of ZSI wth SV and SBC methods, respectvely. Output lne peak voltage and peak current sne waveforms of ZSI wth SV method greater than output lne peak voltage and peak current sne waveforms of ZSI wth SBC method, respectvely. Fgures 10 and 11 show total harmonc dsturbance spectra of output lne voltage HD and total har- U monc dsturbance spectra of output current HD I whch HD I and HD of SV method less U than HD and HD of SBC method. hese re- I U sults are gven n able 2. Also n able 2 show calculate results of SV and SBC methods use Equatons (4)-(18). SDP, SDP av pk of SV method less than SDP av, SDP pk of SBC method. In the future, we wll apply SV for ZSI n many control systems (e.g. motor speed control system, hybrd electrc vehcles, photovolt, drect-drve wnd generaton system). In Fgure 12 shows the fgure experments for motor speed control system. 5. Acknowledgements I am extremely grateful to professors lecture speeches durng the course at Department of Control Scence and Engneerng, Huazhong Unversty of Scence and echno- able 1. Parameters are used for smulaton of SV and SBC. Parameter Z-source nductance (L 1 and L 2 ) Z-source captance (C 1 and C 2 ) Load resstance Load nductance Swtchng frequency (f sf ) Fundamental frequency (f 1 ) axmum power output of the DC-source voltage (P 0 ) Output DC-source voltage at maxmun power (V 0P ) axmum output voltage of DC-source (V 0 ) Command output voltage Value 1.5 mh 1000 μf 10 Ω mh 10 khz 60 Hz 3700 W 250 V 400 V 225 V he shoot-through tme duty rato ( ) 0.1 able 2. Calculate results of SV and SBC. Parameter SV SBC I D HD HD v V V V c I A L SDP VA av SDP VA pk Fgure 6. Smulaton SV. DC-lnk voltage (v ), Captor voltage (V c ) and DC-source voltage (V 0 ).

6 192 C.-. PHA E AL. Fgure 7. Swtchng output lne voltage and phases voltage waveform of ZSI (SV). (a) Swtchng output phase voltage; (b) Swtchng output lne voltage. Fgure 8. Output lne voltage and current sne wave-forms of ZSI (SV). (a) Output current sne wave-forms; (b) Output lne voltage sne waveforms.

7 C.-. PHA E AL. 193 Fgure 9. Output lne voltage and current wave-forms of ZSI (SBC). (a) Output current; (b) Output lne voltage. Fgure 10. Harmonc spectra of Z-source nverter (SV). (a) Harmonc spectra of output current; (b) Harmonc spectra of output lne voltage.

8 194 C.-. PHA E AL. Fgure 11. Harmonc spectra of Z-source nverter (SBC). (a) Harmonc spectra of output current; (b) Harmonc spectra of output lne voltage. harmonc spectra of output voltage and output current wth SV s less than harmonc spectra of output voltage and output current wth SBC, the total average and peak swtchng devce power of ZSI system wth SV method s less than the total average and peak swtchng devce power of ZSI system wth SBC, the SV modulaton ndex n lnear can exceed more than the SBC. All of these comparsons are used smulaton ALAB/Smulnk. REFERENCES Fgure 12. he fgure experments for motor speed control system. logy (HUS), Wuhan , Chna. Especally, professor Anwen Shen, assocate professor Phan Quoc Dzung, and master Nguyen Bao Anh who offers me advce and help me to fnsh ths paper. 6. Concluson hs paper presents a comparson of two dfferent there are SV and the SBC for ZSI. hey show the [1] I. P. C. Loh, D.. Vlathgamuwa, I. Senor, Y. S. La, G.. Chua and I. Y. W. L, Pulse-Wdth odulaton of Z-Source Inverters, IEEE ranstons on Power Electroncs, Vol. 20, No. 6, 2005, pp do: /pel [2]. Olzwesky, Z-Source Inverter for Fuel Cell Vehcles, US Department of Energy, Freedom CAR and Vehcles echnologes, EE-2G, Washngton, [3] K. Holland,. Shen and F. Z. Peng, Z-Source Inverter Control for rton Drve of Fuel Cell-Battery Hybrd Vehcles, Industry Applcatons Conference, 40th IAS Annual eetng, Vol. 3, No. 4, 2005, pp [4] I. Fang and Z. Peng, Z-Source Inverter, IEEE ranstons Industry Applcatons, Vol. 39, No. 2, 2003, pp [5] J. L, K. Q. Qu, X. L. Song and G. C. Chen, Study on

9 C.-. PHA E AL. 195 Control ethods of Drect-Drve Wnd Generaton System Based on hree-phase Z-Source Inverter, IPEC, Shangha, ay [6] S. A. K. H. ozafar Napoor, S. Danyal and. Sharfan, Photovolt Power System Based mppt Z-Source Inverter to Supply a Sensorless BLDC otor, 1st Power Electronc, Drve Systems and echnologes Conference, Unversty of abrz, abrz, February [7] U. Al and V. Kamaraj, A Novel Spe Vector pwm for Z-Source Inverter, Proceedngs of ICEEC, Nagercol, arch [8]. Shen, A. Joseph, J. Wang, F. Z. Peng and D. J. Adams, Comparson of radtonal Inverters and Z-Source Inverter for Fuel Cell Vehcles, IEEE Power Electroncs Socety, Vol. 22, No. 4, 2007, pp do: /pel [9]. Chun, Q. ran, J. Ahn and J. La, Ac Output Voltage Control wth nmzaton of Voltage Stress ross Devces n the Z-Source Inverter Usng odfed svpwm, PESC 37th IEEE, Jeju, June 2006, pp [10] B. K. Bose, odern Power Electroncs and AC Drvers, Pearson Educaton, Upper Saddle Rver, 2002.