JOURNAL OF ELECTRIC POWER AND ENERGY CONVERSION SYSTEMS (JEPECS)JEPECS VOL. 1, NO. 1, PP. 32-36, SPRING 2016

ISSN: 2345-4830 print/2345-4733 online

Simple Boost Control Method Optimized withGenetic Algorithm for Z-Source Inverter

Hamed Hosseinnia∗, Daryoush Nazarpour†and Saeed Sabernia‡

In this paper, Genetic Algorithm (GA) optimization is proposed in order to tune parameters of simple boost control methods for theZ-source inverter. One of the characteristics of a good source is its quality that is measured by its lower harmonic. Impedance sourceinverter (ZSI) that acts as an interface between source and power grid, has harmonics in its output. GA optimization is used in simpleboost control method to minimize these harmonics by setting control parameters on optimized values. The simulation has been carriedout using MATLAB/Simulink and results prove that simple boost control method with optimized values has a lower total harmonicdistortion (THD).

Keywords: Genetic Algorithm, Optimization Problem, Simple Boost Control Method, Z-source Inverter (ZSI).

Received March 2015; Revised August 2015; Accepted Feb. 2016.

I INTRODUCTION

The voltage-source inverter (VSI) and current-source inverter(CSI) are two types of traditional power inverter typologies(Fig. 1) [1]. These two types of converter have some limi-tations. In most inverters two switches on the same leg can-not conduct simultaneously, because this will cause a short cir-cuit across the source. One of the advantages of the impedancesource converter over other converters is its impedance networkthat is placed between the source and inverter bridge and pre-vents input source short circuit and introduces shoot-throughzero switching state. Also higher and better output can be ex-pected with equal source conditions. Fig. 2 shows the generalstructure of z-source inverter [2]. The DC source can be either avoltage or current source. When it is in series with an inductor,it acts as a current source and when it is in parallel with a capac-itor it becomes a voltage source. The diode is responsible forpreventing the capacitor discharge through the DC input volt-age. Z-source inverter performs as buck-boost converter. Whenthe input voltage is not high enough to generate required outputvoltage, the shoot-through zero state is used to boost the volt-age and when the input voltage is enough to make the desiredoutput, shoot-trough zero state is not used and z-source inverterperforms the buck conversion [3, 4].

II Z-SOURCE INVERTER

A Modified ZSIThe modified z-source inverter is shown in Fig. 3. As it can beseen the used element are exactly the same as the traditional type.It has been modified by changing diode and inverter bridge po-∗Department of Electrical Engineering, Urmia University Iran, E-mail:

Hhosseinnia66@yahoo.com.†Department of Electrical Engineering, Urmia University Iran E-mail:

d.nazarpour@urmia.ac.ir, (Crossponding Author)‡Department of Electrical Engineering, Urmia University Iran E-mail:

Saeed.Sabernia@gmail.com.

(a) Voltage source inverter

(b) Current source inverter

Figure 1: Traditional power inverter typologies.

c© 2016 Faculty Engineering, Shahed University, P.O.Box: 18151-159, Tehran, Iran.Published by Shahed University Publishing Center.

33 H. HOSSEINNIA et al. OPTIMIZED BOOST CONTROL FOR Z-SOURCE INVERTER

Figure 2: Z-source inverter.

Figure 3: Improved ZSI.

sition .This modified circuit limits inrush current because thereis no path for startup current [5, 6].

B Simple boost control method

In this section simple boost control method which is used to con-trol the shoot-through duty ratio is introduced .In this method,two straight lines are employed in the usual PWM controlmethod. These two straight lines are equal to the peak valueof the three phase reference signals [7, 8]. When the carrier tri-angular wave is greater than the positive straight line or smallerthan the negative line, the inverter works in shoot-through zerostate that is prevented in traditional method. From the sim-ple boost control method description, it can be concluded thatcarrier wave frequency and shoot-through are two main factorsin this method and the output waveform is mostly affected bythem. So by optimizing these two parameters, the output canbe improved compared to that obtained from estimated values.Fig. 4 illustrates the simple boost control method [9, 10].

In simple boost control method, the shoot-through duty ra-tio decreases with the increasing of modulation index M. Themaximum shoot-trough duty ratio of the simple boost controlmethod is limited to (1-M), reaching zero at a modulation indexof one. So to generate an output voltage with a high voltagegain, a small modulation index has to be used [11,12]. Becauseof the important role of M and carrier signal frequency, both ofthe parameters are used in the optimization problem. In fact,optimization is a trade-off between value of output voltage andtotal harmonic distortion (THD) [13–15]. In the next section,genetic algorithm is introduced.

Figure 4: Simple boost control method.

III GENTIC ALGORITHEM (GA)

Genetic algorithm (GA) is a search technique used in computingto find true or approximate solutions to optimization and searchproblems. Genetic algorithms are categorized as global searchheuristics. Genetic algorithms are a especial class of evolution-ary algorithms that use inspired techniques by evolutionary bi-ology such as inheritance, mutation, selection, and recombina-tion [16]. What do we mean by genetic algorithm? It starts witha set of randomly generated solutions and recombines pairs ofthem by chance to produce offspring. Only the best offspringand parents are kept to produce the next generation. In Fig. 5genetic algorithm flowchart has been illustrated.

A Simple boost control method with genetic algorithm

As described in the previous section, the shoot-trough has animportant role in the operation of ZSI. This is because it makesit possible to buck or boost the input voltage. In this paper thegenetic algorithm is selected to adjust parameters M and fre-quency of the carrier signal (fs) in simple boost control method.Because of the importance of minimizing THD of output volt-age [17, 18], the objective function is defined as below:

J =

tsim∫0

t. |THD|.dt (1)

JEPECS VOL. 1, NO. 1, PP. 32-36, SPRING 2016 34

Figure 5: GA flowchart.

Figure 6: Voltage gain and stress versus modulation index.

where, tsim is the simulation time. The output voltage gain (G)is denied as

G =vp

0.5vdc(2)

in which vp is the peak phase voltage of the inverter output.Based on the simple boost control of the z source inverter, thevoltage gain can be obtained as

G =M

2M − 1(3)

where M is peak of the reference voltage amplitude which isshown in Fig. 4 by VP or VN . The main aim of the optimizationis to minimize the objective function in (1) due to constrain:

Mmin ≤M ≤Mmax

fsmin ≤ fs ≤ fsmax(4)

Mmin and Mmax are lower and upper constraints for modula-tion index. Also, f)min and fmax are carrier frequency con-straints that are given 2000 and 3000, respectively.In reference [19], the voltage stress across of the inverterswitches was introduced for different modulation methods. Inthe boost control method, this voltage stress is (2G−1)vdc . Fig.6 shows the variation of voltage gain and stress versus the mod-ulation index. This figure illustrates that the modulation indexrange for having a gain voltage greater than one is 0.5 ≤ M ≤0.9. Also, upper and lower bounds of the carrier frequency are

Figure 7: System configuration.

Figure 8: Output voltage with(M = 0.8 and fs = 1800).

given fmin = 2 kHz and fmax = 3 kHz. The optimal modula-tion index and carrier frequency which are yielded from geneticalgorithm, are M = 0.831451 and fs = 2 kHz.

IV SIMULATION RESULTS

Simulation was conducted with the configuration shown in Fig.5. The simulation parameters are:vdc = 250 v, L1 = L2 =500 µH , C1 = C2 = 1.5 mf , Lf = 500µH , Cf = 5 µfand RL = 15 Ω. The simulation results with the arbitrar-ily chosen modulation index (M = 0.8) and carrier frequency(fs = 1.8 kHz) are shown in Fig. 8 and Fig. 9. Thesefigures show the output voltage and total harmonic distortion(THD) respectively [21]. Operation of genetic algorithm isshown in Fig. 10. In this figure, the minimum cost versus iter-ation is plotted. The minimum cost is the minimum objectivefunction in each iteration. The simulation results with modu-lation index (M = 0.831451) and frequency of carrier signal(fs = 2000 kHz) optimized by a genetic algorithm, for out-put voltage and THD analysis are presented in Fig. 11 and Fig.12, respectively [22, 23]. In order to compare the result of thevalues obtained from genetic algorithm optimization with de-fault values, output voltage harmonic is considered as the mainfactor and by using FFT diagrams and comparing them, out-put voltage harmonic reduction will be provided. As its namesuggests, the boost converter output voltage is higher than theinput voltage. By comparing the output voltage waveform that

35 H. HOSSEINNIA et al. OPTIMIZED BOOST CONTROL FOR Z-SOURCE INVERTER

Figure 9: THD = 5.48 (without optimization).

Figure 10: Operation of GA algorithm.

obtained from arbitrarily chosen values and optimized values, itcan be concluded that optimization has also had a positive effecton output voltage and it has a better waveform. Because of in-creased voltage in the optimized case, the voltage stress on theswitches (2vc − vo) is decreased compared with normal condi-tion considering optimized parameters (i.e. fs and M [24, 25].

V CONCLUSION

As simulation results showed, by tuning carrier signal frequencyand modulation index to optimized values, THD of the output isbetter than arbitrarily chosen values and the quality of ZSI out-put is improved. This verifies effectiveness of the simple boostcontrol method with GA. Genetic Algorithm can be applied toany problem where optimization is required. Therefore, it canbe applied in many usages in power electronics. The compari-son of the results in this paper to similar work in the literatureshows that the GA approach for the harmonic optimization ofZ-Source inverter works properly.

REFERENCES

[1] M. H. Rashid, “Power electronics,”2nd ed. Prentice Hall, 1993.

Figure 11: Output voltage with optimized parameters.

Figure 12: THD = 3.11 (with optimized parameters).

[2] F. Z. Peng, “Z-Source inverter,” IEEE Transactions on Industry Applica-tion, vol. 39, no. 2, pp. 504-510, 2003.

[3] I. Boldea, R. Antal and N. Muntean, “Modified Z-source Single PhaseInverter with Two Switches,” presented at the IEEE International Sympo-sium 2008, Industrial Electronics ISIE, Cambridge, 2008.

[4] P. Chiang and et al.“Pulse width modulation of Z-Source Inverters,”IEEE Trans. on Power Electronics, vol. 20, no. 5, pp. 1346-1355, 2005.

[5] P. F. Zheng, M. Shen and Z. Qian, “Maximum boost control of the currentZ-Source inverter,” IEEE Transactions on Power Electronics, vol. 20, no.4, pp. 833-838, 2005.

[6] F. Gao, L. Poh Chiang, R. Teodorescu and F. Blaabjerg, “Diode-AssistedBuck-Boost Voltage-Source Inverters,” IEEE Trans. on Power Electron-ics, vol. 24, no. 9, pp. 2057-2064, 2009.

[7] B. J. Rabi and R. Arumugam, “ Harmonic Study and Comparison of Z-Source Inverter with Traditional inverters,” ISSN American journal ofapplied Sciences, vol. 2, no. 10, pp.1418-1426, 2005.

[8] G. Durgasukumar and M. K. Pathak “THD Reduction Performance ofMulti-Level Inverter fed Induction Motor Drive,” presented at the IndiaInternational Conference on Power Electronics (IICPE), 2011.

[9] S. Kouro, P. Lezana, M. Angulo and J. Rodrguez, “Multicarrier PWMWith Dc-Link Ripple Feed forward Compensation For Multilevel Invert-ers,” IEEE Trans. on Power Electronics, vol. 23, no. 1, pp. 52-59, January2008.

[10] J. Holtz, “Puls Width Modulation for Electronic Power Convertion,” inproc IEEE, vol. 82, no. 10, pp. 1194-1214, 1994.

JEPECS VOL. 1, NO. 1, PP. 32-36, SPRING 2016 36

[11] J. Holtz and B. Beyer, “Optimal Pulse Width Modulation for ac Servosand Low- Cost Industrial Drive,” IEEE Trans. industry Application, vol.30, no. 4, pp. 1039-1047, 1994.

[12] C. J. Gajanayake, L. F. Lin, G. Hoay, S. P. Lam and S. L. Kian, “ExtendedBoost Z-Source Inverters,” IEEE Trans. on Power Electronics, vol. 25, no.10, pp. 2642-2652, Oct. 2010.

[13] M. S. Shen, J. Wang, A. Joseph, F. Z. Peng, L. M. Tolbert and D. J.Adams, “Constant boost control of Z-Source Inverter to minimize currentRipple and voltage stress,” IEEE Trans. on Industry Applications, vol.42, no. 3, pp. 770-778, May-Jun 2006.

[14] P. C. Loh, F. G.ao, F. Blaabjerg and S. weilim, “Operational analysisand modulation control of three-level Z-source inverters with enhancedoutput waveform quality,” IEEE Trans. on Power Electronics, vol. 24,no.7, pp. 1767-1777, July 2009.

[15] F. Gao, P. C. Loh, F. Blaabjerg and C. J. Gajanayake, “Operational anal-ysis and comparative evaluation of Embedded Z-source inverter,” pre-sented at the IEEE PESC, Rhodes, 2008.

[16] D. E. Goldberg, “Genetic Algorithms in search ,Optimization and ma-chine learning,” Addison Wesley ,1989.

[17] Z. Du, L. M. Tolbert and J. N. Chiasson, “Harmonic elimination for multilevel converter with programmed PWM method,” presented at the IEEEIndustry Applications Soc. Annual Meeting, 2004.

[18] M. J. Schutten and D. A. Torrey, “Genetic algorithm for control of powerconverters,” presented at IEEE Power electronics Speciali, 1995.

[19] M. Shen and F. Z. Peng, “Operation modes and characteristics of theZ-source inverter with small inductance,”presented at the Industry Ap-plications Conference, 2005.

[20] B. J. Rabi, “Minimization of harmonics in pwm inverters based on ge-netic algorithms,” Journal of Applied Sciences, vol. 6, no. 9, pp. 2056-2059, 2006.

[21] Y. Tang, S. Xie, C. Zhang and Z. Xu, “Improved Z-source inverterwith reduced Z-source capacitor voltage stress and soft-start capability,”IEEE Trans. on Power Electronics, vol. 24, no. 2, pp. 409-415, 2009.

[22] S. Rajakaruna and L. Jayawickrama, “Steady state analysis and design-ing impedance network of Z-source inverters ,” IEEE Trans. on IndustrialElectronics, vol. 57, no.7, pp.2483-2491, July 2010.

[23] S. Rajakaruna and L. Jayawickrama, “Steady-state analysis and design-ing imedance network of Z-source inverters,” IEEE Trans. on industrialelectronics, vol. 57, no.7, pp. 2483-2491, July 2010.

[24] S. M. Yousuf , P. Vijayadeepan and D. S. Latha, “The Comperative THDAnalysis of Neural Clamped Multilevel Z-Source Inverter using NovelPwm Control Technique,” International Journal of Modern EngineeringResearch (IJMER), vol.2, no. 3, pp. 1086-1091, May-June 2012.

[25] S. Yang, F. Z. Peng, Q. Lei, R. Inoshita and Z. Qian, “Current-fed quasi-Z-source inverter with voltage buck-boost and regeneration capability,”IEEE Trans. on Industry Applications, vol. 47, no. 2, pp.882-892, 2009.

Hamed Hosseinnia was born in Khoy,Iran in 1988. He received his B.S.c degreefrom Azerbaijan University of TarbiatMoallem (AUTM), Tabriz, Iran in 2010.He is currently M.S.c. student on Electricalengineering in Urmia University, Urmia,Iran. His research interests include powerelectronics, flexible AC transmissionsystem (FACTS) and Power system Relia-

bility.

Daryoush Nazarpour was born in Urmia,Iran in 1958. He received his B.Sc. de-gree from Iran University of Science andTechnology, Tehran, Iran in 1982 and theM.Sc. degree from Faculty of Engineering,University of Tabriz, Tabriz, Iran in 1988.He received the Ph.D. degree from TabrizUniversity, in 2005 in Electrical Power En-gineering. He is now an assistant professor

in Urmia University, Iran. His research interests include powerelectronics, and flexible AC transmission system (FACTS).

Saeed Sabernia was born in Ur-mia in 1989. He received his B.S.in Electrical Engineering from Ur-mia University, Iran in 2011. He iscurrently pursuing a M.S. in Power(Systems) Engineering. His cur-rent research interests are powerelectronics, FACTS and smartgrids.