Low-Voltage and High-Current DC Output Realized by ...

J Electr Eng Technol.2017; 12(4): 1575-1585 http://doi.org/10.5370/JEET.2017.12.4.1575

1575Copyright The Korean Institute of Electrical Engineers

This is an Open-Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/ licenses/by-nc/3.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Low-Voltage and High-Current DC Output Realized by Multiple Power Cells Based on Deadbeat and Automatic Current Sharing Control

Jinfeng Liu†, Yu Zhang*, Xudong Wang* and Herbert Ho-Ching IU**

Abstract – This paper presents a synchronous generator with a distributed system of multiple parallel three-phase power cells. This generator can immediately output high DC. Each power cell comprises three-phase windings and a three-phase synchronous rectification bridge with a deadbeat control of load power feedforward, which can improve the characteristics of dynamic response and reflect the load variance in real time. Furthermore, each power cell works well independently and modularly using the method of automatic maximum current sharing. The simulation and experimental results for the distributed controller of multiple power cells demonstrate that the deadbeat control method can respond quickly and optimize the quality of the energy. Meanwhile, automatic maximum current sharing can realize the validity of current sharing among power cells.

Keywords: Synchronous generator, Power cell, Deadbeat control method, Load power feedforward, Automatic maximum current sharing

1. Introduction Many types of DC or pulse power supplies are widely

used in various electrochemistry fields, including elec-trolysis, electroplate, metal colouring, and electrophoresis. The quality, cost and efficiency of such products are directly relevant to the characteristics of the power supplies [1-3]. Traditional high power DC power supplies consist of multiple parallel DC power modules in order to output high current, each sharing a portion of the high power output. With the improvement in the production processes and efficiency in many companies, the deficiencies of these kinds of power supplies have increased. Firstly, these power supplies have high energy consumption and low efficiency. Second, they have high volume and weight because a power frequency transformer is usually used in the devices and many power modules are paralleled. Lastly, the control apparatus has low precision and high energy consumption, as in [4, 5]. Parallel-connected switching devices are usually used to expand capacity, however, many driver circuits would increase the failure-rate of the apparatus and make it difficult to achieve modularity.

Many scholars decrease the volume and weight of power supplies to improve their quality by using the high-frequency switching power technique. Reference [6] optimized a low-voltage and high-current permanent

magnet generator and fitted a controller in the generator to decrease the loss of connecting wires. This assembly is complex and cannot improve the DC output. Reference [7, 8] designed a dual-winding induction generator, consisting of a control winding and a power winding. The rectifier is connected to the power winding to achieve DC output. The excitation capacitor is large when the generator outputs high current and it will produce high oscillation to AC harmonics.

This paper proposes a low voltage and high direct current power supply based on a mature permanent magnet synchronous generator (PMSG). The generator can output high DC because 18 three-phase rectification bridges are fixed on the edge of the generator. An effective control strategy must be considered the high number of control objects. We adopt a distributed control system to improve the modularity and reliability, since this large capacity power apparatus has a clearly distributed circuit, layout and control function [9].

Section 2 presents the structure of the distributed control system which includes the power cell. Section 3 presents the deadbeat control and simulation models accordingly. Section 4 presents the principle of automatic maximum current sharing. Section 5 and 6 present the simulation results of deadbeat and automatic maximum current sharing control. Section 7 presents the experimental results from the test platform.

2. Structure of Distributed Control System The multiple parallel-connected power cells in this paper

based on distributed control system are shown in Fig. 1. A DC power supply can provide current to the DC bus after

† Corresponding Author: School of Electrical and Electronic Engineering, Harbin University of Science and Technology, China. ([email protected])

* School of Electrical and Electronic Engineering, Harbin University of Science and Technology, China. ([email protected], [email protected])

** School of Electrical, Electronics and Computer Engineering, University of Western Australia, Australia. ([email protected])

Received: October 13, 2016; Accepted: May 16, 2017

ISSN(Print) 1975-0102 ISSN(Online) 2093-7423


1576 J Electr Eng Technol.2017; 12(4): 1575-1585

rectification and the number of parallel-connected power cells can be varied to meet the power needs of different loads. Each power cell has the advantage of high power density, efficiency and fast response. The most important factor is that each module is completely impendent.

The three-phase power supply of each power cell is comprised of Y-connected three-phase windings of PMSG which can generate induction electromotive force through cutting magnetic lines as shown in Fig. 2.

We can remove the neutral wire because the current of the three-phase rectification bridge is symmetrical, that is, the sum of three-phase current is zero [10]. The use of an MOSFET can greatly reduce the power converter’s volume and weight as well as increase the quality of the high frequency power supply [11].

The minimum cell of each power cell is shown in Fig. 3. There are 18×3=54 heat sinks which are arranged on the end cover of the generator along a circle. Fig. 3(a) describes the minimum cell which is composed of each

winding and the leg of the bridge and Fig. 3(b) describes both MOSFETs on each leg which are fixed on the heat sink.

The distributed control system is used to supply the three-phase synchronous rectification bridge and can achieve automatic current sharing according to the maximum current on the bus. This synchronous generator can output direct voltage of 5 V, rated power of 10kW and its rated revolution per minute is 5,000 RPM.

3. Deadbeat Control Strategy of Load Power

Feedforward for PWM The deadbeat control method can not only improve the

performance of dynamic response for the PWM rectifier but also reflect the variance of load power in real time through a test of the load power and feedforward compensation [12, 13]. The fundamental principle of this method is to calculate the pulse width of the next switch period according to the state equation of the DC system, then feedback the output signal and the next referenced output [14, 15].

The PI control and buffer circuit are used to achieve voltage-loop control to maintain the stability of the DC output voltage and compensate for the energy loss of the MOSFET and post circuit [16]. Meanwhile, it will produce a reference current Id once it cooperates with Pout. The model which produces regulation current Id is shown in Fig. 4, where Urout is the referenced output according to the feedback of the DC bus and the referenced voltage Uref, while kv is the adjusting coefficient of the maximum current sharing. Then, one can obtain Iout and Pout, as shown in Fig. 4, as follows.

( ) ( )out p rout v dc i rout v dcI K U k U K U k U dt= − + −∫ (1)

out dc oP U I= × (2)

where Udc and Io are the sample signal of the DC output voltage and current, respectively.

In Fig. 4, Ucsb is the control signal of the current sharing, Iout is the output of the PI controller in the outer-loop, while

Fig. 1. Distributed control system of multiple modules

Fig. 2. Power cells of PMSG with rectification bridge

(a) (b)

Fig. 3. Minimum cell of each power cell

Jinfeng Liu, Yu Zhang, Xudong Wang and Herbert Ho-Ching IU

http://www.jeet.or.kr 1577

Kp and Ki are the proportion and integration coefficient, respectively. Therefore, the input of the PI controller is keUref+kcUcsb-kvUdc. ke, kc and kv are proportion control coefficients of current sharing. Pout is feedforward output power and it can produce feedforward current signal Is through the gain module whose gain is 2/3U [17], U is steady DC 5 V output voltage. Iout can compensate energy on the DC side and maintain the stability of the DC output. Then, referenced current signal Id can be estimated as follows.

d out sI I I= + (3)

If each leg of the bridge took on the respective switch

loss, the adjusting voltage signal can be obtained by multiplying the synchronous signal of the respective phase [18], that is, ida, idb and idc, which are shown in Fig. 5. The duty cycle d of the PWM in terms of the KVL of each leg of the bridge in one period is as follows [19].

1( ) / ,0 12s dc

D

id u L U dTΔ= − + < < (4)

where TD is the switch period, L is the filter inductance and us is the input AC voltage.

Each phase current signal can be substituted into Eq. (4) to express the three-phase duty cycle of the PWM

rectification as Eq. (5) and accordingly, the control model is shown in Fig. 5.

1( ) /21( ) /21( ) /2

da saa s dc

D

db sbb s dc

D

dc scc s dc

D

i id K u L U

Ti i

d K u L UT

i id K u L U

T

⎧ −= × − +⎪

⎪⎪ −

= × − +⎨⎪⎪ −

= × − +⎪⎩

(5)

First, Id combines with the three-phase synchronous

signal. Second, they compare with each sample phase current signal. Lastly, the three-phase control output signal can be produced by Eq. (5) of the three-phase duty cycle.

4. Automatic Maximum Current Sharing Averaging the load for every module is known as the

current sharing method [20, 21]. Automatic current sharing is divided into mean current sharing and maximum current sharing. Each module connects to the current sharing bus through a resistance based on the mean current sharing and can obtain better characteristics of current sharing. Meanwhile, the general power system cannot work well once the average bus is shortened or anyone of power cells

Fig. 4. Output model of referenced current Id

Fig. 5. The deadbeat control model of the three-phase PWM rectification



could not work [22-24]. However, the method of maximum current can solve this shortcoming and improve the current sharing.

The control diagram based on automatic maximum current sharing is shown in Fig. 6. This method can automatically recognize the power cell which has the maximum output current due to the existence of a diode and set it as the current basis of the entire power system. Next, the other power cells will adjust their output current based on this maximum current [25]. A unidirectional buffer, shown in Fig. 6, can provide high input impedance and separate the post from the prime circuit. If one of the power cells malfunctions, the power system can redistribute the output current and the other power cells would work continually without being influenced by this faulty cell.

Fig. 7 shows the corresponding transfer function diagram of Fig. 6, where kv is the voltage feedback coefficient and kc and ke are additional coefficients of the current sharing control signal and voltage basis, respectively. Rs is the sample current coefficient. Rf is the proportional coefficient between the output voltage and current. The buffer G1 is selective and unidirectional, and its inner structure is shown in Fig. 7(b). Rg is the resistance of the current

sharing bus to the ground. Ubus is the sample voltage of the current sharing bus. Gev is the PI regulating function of the feedback voltage. Gfv is the comprehensive function of the drive signal and power output which is relevant to the output voltage, output load and drive circuit of the synchronous generator.

5. Simulation Verification of Single Power Cell

based on Deadbeat Control The simulation model for a power cell is built by

Saber2012 to verify and test the deadbeat control. The parameters of high frequency power for each power cell are as follows: the input of rectifier is a three-phase AC signal of 4 V, the frequency is 50 Hz, the DC output of the power supply is a current of 0.11kA and has a voltage of 5 V. Full-wave rectification is applied to the high frequency power supply and the expected DC voltage and current can be obtained through an LC filter. A real resistance, whose value is 5 V/110A=45 mΩ, is used to simulate the load of each power cell. At first, the time-domain analysis is performed for a controllable PWM with deadbeat control and contrasts with an uncontrollable rectification bridge. Then, frequency-domain analysis is performed to com-pensate the open-loop system and contrast with the total harmonic distortion(THD) with a different rectification bridge.

5.1 Time-domain simulation analysis

Fig. 8 shows the simulation results of the three-phase

output voltage and current before the typically uncon-trollable three-phase rectifier is realized by diodes. It is clear that the waves of the three-phase power supply seriously distort and have a high harmonic. Therefore, the power supply is critically polluted.

The deadbeat simulation model of the load power feedforward for the three-phase synchronous rectification is shown in Fig. 9.

The sample voltage and current of the three-phase power

Fig. 6. The control diagram based on automatic maximum

current sharing

(a) Main control diagram

(b) Inner structure of buffer G1

Fig. 7. Logic control diagram of maximum current modeFig. 8. Simulation waves of three-phase output voltage and

current before rectifier



Fig. 10. The inner model of Compare module

Fig. 11. Simulation waves of three-phase output voltage

and current before controllable rectifier

supply are transmitted to the core control module in this model, that is without_diff_beat which is built according to the deadbeat control of Fig. 5. It produces duty cycle control signals for the three half bridges. Then, the SPWM module outputs six gate control signals for the six MOSFETs of the half bridges. PI_Module, whose inner model is shown in Fig. 4, is the regulating module which combines the PI control with the load power sample and the

Fig. 12. Simulation waves of DC output voltage and

current after controllable rectifier

Fig. 13. The fluctuation of voltage output with the

variation of load resistor trace difference of the voltage sample. It then produces the referenced current Id which is used by the without_diff_beat module. The Compare module can output current sharing control signal Ucsb which is used by the PI_Modulation module to compare the output signal of each power cell with the signal of the current sharing bus. The switchover module is used to provide different control signals to the SPWM which can perform frequency-domain analysis, which will be explained in part 5.2.

Fig. 9. Simulation model of deadbeat control method



The inner model of the Compare module is shown in Fig. 10 according to the method of maximum current seen in Fig. 6. The unidirectional buffer cooperates with the diode to achieve automatic current sharing.

The simulation results of the three-phase output voltage and current before the controllable rectifier are shown in Fig. 11.

The voltage distortion of the three-phase power supply has disappeared and the output voltage is steady without pollution. The current wave has returned to a sinusoidal wave. Although there is amplitude distortion within the first 100 ms due to the existence of transient vibration on the start of the simulation, the three-phase current wave is exactly symmetrical afterwards.

DC output voltage and current after the controllable rectifier are shown in Fig. 12. The DC output voltage is a steady 5 V after the filter. The DC output current is 110A and no current ripple is observed in the meantime. The output of the system is perfectly steady.

To test its anti-disturbance, the load resistor is changed from 45 mΩ to 55 mΩ at the time of 20 ms. We can observe the voltage output from Fig. 13. The output remains steady after the 2 ms fluctuation under the deadbeat control.

5.2 Frequency-domain simulation analysis

Fig. 14 presents the inner simulation model of the

Switchover module that can switch different control signals to the SPWM and conduct frequency-domain analysis. The switch_vin model of Saber can switch one of three inputs during the process of simulation. The first input port(in1) is the v_pwl model of Saber which will provide a constant open-loop control voltage to SPWM model. The second input port(in2) is the v_sin model which is used to perform small-signal frequency analysis. The third input port(in3) is the closed-loop control voltage provided by the deadbeat control method which is combined with PI control.

At first, we can alter the input of switch_vin to in1 and perform transient (TR) analysis to get the end-point-file, then alter to in2 and perform small-signal frequency analysis whose initial-point-file is the end-point-file of the TR analysis. We then obtain the frequency-domain analysis

result of the open-loop, which is shown in Fig. 15. It is clear that the open-loop system is not stable because the slope is -40 dB/dec at the cutoff frequency, and thus the PI compensation network, which is equivalent to a type п compensation network when C2 in type п is 0, must be applied as it can provide a zero and a pole to adjust the frequency-domain characteristics.

Because the switch frequency of the MOSFET is 200 kHz, the cutoff frequency fc can be 200 kHz/5=40 kHz in order to satisfy the convergence condition of the open-loop gain. According to Fig. 15, |G(ωc )|=-29.12 dB and φ(ωc)= -91.3°. Therefore, PI must provide a 29.12 dB gain, then 20 lgKp=29.12 dB, i.e., Kp=28. Suppose the phase margin is 45°, the allowable coefficient K for fc/fz is four on the basis of φ(ωc )[26]. Therefore, a zero is ωz=2πfz=2π×fc/4= 62.8×103rad/s, and the integration time constant Ti is 0.7μs according to Ki /Kp =1/(Ti Kp ) =ωz. When Kp and Ti are applied in the PI module of Fig. 4, we alter the input of switch_vin to in3 and perform transient (TR) analysis to

Fig. 14. Inner model of switchover module

Fig. 15. Frequency-domain analysis of open-loop

Fig. 16. Frequency-domain analysis of closed-loop



obtain the end-point-file, then alter to in2 and perform small-signal frequency analysis whose initial-point-file is the end-point-file of the TR analysis. Thus, we obtain the frequency-domain analysis result of the closed-loop, which is shown in Fig. 16.

We can find a waveform with the slope of -20 dB/dec that cuts across 0 dB line at the point of the cutoff frequency, which is approximately 40 kHz. We also see that the phase margin is sufficient. Therefore, the PI compensation network has achieved zero-pole compensation for the open-loop system.

Because of the existence of nonlinear components such as an inductor, capacitor and switch in the system, high-order harmonics will mix with fundamental waves to affect both input and output. Therefore, the THD should be analysed. Saber simulation provides the function of the THD test for the signal after the Fourier transform. Although the test result is decibel value for the real THD of the signal, the variation tendency of the THD is still clear.

Fig. 17 is the Fourier transform results of the input phase current and output current from 0 Hz to 1 kHz for the uncontrollable rectifier bridge. The measurement tool of Saber can obtain the THD, which is shown in Fig. 17 with the signal noise ratio (SNR) and signal to noise and

distortion ratio (SINAD). In Fig.17, the magnitude of harmonics for the input

phase current is very high from 0 Hz to 350 kHz. The highest magnitude has reached over 1e+21. Thus, the THD of the input phase current and output current are both higher.

Fig. 18 is the Fourier transform results of the input phase current and the output current for the open-loop controllable rectifier bridge. The magnitude of harmonics for the input phase current has slumped from 0 Hz to 1 kHz. The THD of the input phase current and output current both are much lower than the uncontrollable rectifier.

Fig. 19 shows the Fourier transform results of the input phase current and output current from 0 Hz to 1 kHz for the closed-loop controllable rectifier bridge with deadbeat control. The THD of the input phase current and output current both reach a lower level than the aforementioned circumstances. The final THD of the input phase current for the entire closed-loop system with deadbeat control is log-1(-32.382/20)=0.024.

6. Simulation Verification of Multiple Power Cells Based on Automatic Maximum Current

Sharing Fig. 20 shows the simulation model of multiple modules

with power cells which is used to realize current sharing of maximum current.

For convenience, Fig. 20 only shows 3 modules, while there are 18 modules in the process of the real simulation. The sample current of each module will compare with the current on the current sharing bus. The difference is amplified to produce Ucsb. This current sharing control signal can adjust the value of the voltage loop and then the output voltage of each module would vary accordingly. Therefore, automatic maximum current sharing is achieved. In the process of the simulation, the output of one module

Fig. 17. Fourier transform of input phase current and output current for uncontrollable rectifier bridge

Fig. 18. Fourier transform of input phase current and output

current for open-loop controllable rectifier bridge

Fig. 19. Fourier transform of input phase current and

output current for closed-loop controllable rectifier bridge



Fig. 21. The simulation waves of transient analysis when

kv=0.4 kc=0.22 ke=0.02


kv=0.42 kc=0.63 ke=0.04 connects to a reverse voltage source which simulates the breakdown by increasing the voltage slowly. Then, other modules, whether maintaining current sharing or not, can be tested through this model.

kv, kc and ke are three key parameters of the transfer function diagram of Fig. 7. To analyse the parametric variance, the parameters can be assumed from 0.01 to 1 by the step of 0.01 in Saber. Fig. 21 shows the output waves when kv=0.4 kc=0.22 ke=0.0,2 which are the worst. With the

connection of the reverse voltage source, it is obvious that the DC output voltage Vout and the current sharing bus voltage Vbus decrease dramatically, and thus the system cannot achieve current sharing.

Fig. 22 presents the simulation waves when kv=0.42 kc=0.63 ke=0.04, which are improved. Under the influence of the reverse voltage source, Vbus still increases to a steady value and Vout can maintain 5 V.

Fig. 23 presents the simulation waves when kv=0.48

Fig. 20. Simulation model with maximum current sharing


kv=0.48 kc=0.86 ke=0.06

Fig. 24. The simulation waves based on mean current

sharing



kc=0.86 ke=0.06, which are optimum. Vbus and Vout are not influenced by the reverse voltage source. Both can output stably and achieve optimum current sharing.

The simulation results with mean current sharing are shown in Fig. 24.

With the connection of a reverse voltage source, it is obvious that output of the system shows a disturbance at the beginning. Furthermore, Vout and Vbus are lower than the output maximum current sharing when they reach stability. Hence, the mean current sharing cannot be achieved once any one of the modules malfunctions.

7. Experimental Results To verify the correction of simulation, we designed a suit

of Electronic Control Unit (ECU) which is shown in Fig. 25(a). DSP28335 is used to produce the control signal and

collect the voltage and current signal. The MOSFET driver is the N-channel TPS51601A which is capable of driving high-side and low-side with the highest speed and lowest switching loss. Fig. 25(b) is a synchronous generator which is fitted on the side of the engine.

ECU and PMSG can be developed simultaneously with the help of hardware in the loop (HIL). Hence, we use HIL to simulate a control object, that is, PMSG. It can connect with ECU through IO ports and finish comprehensive and scientific testing for the ECU. Because HIL permits fault insertion in the process of simulation, we set the reverse voltage source as interference and observe the wave variance before and after rectification.

One can see that the interference would impact the power supply from Fig. 26 (a). Although the current wave shows a small increase, it can maintain stabilization after 400 ms and the voltage wave remains unchanged. Fig. 26 (b) shows the variance of the two-phase current which still remains symmetrical. Thus, the deadbeat control strategy of the load power feedforward can stabilize the power supply.

Fig. 27 shows the dynamic response comparison of the PI controller and PI with the deadbeat control, respectively, when the load resistor is increased from 50 mΩ to 45 mΩ. From the experimental results, compared to only a PI controller, the PI controller with deadbeat control can improve the peak of oscillation and considerably reduce the time of oscillation.

We adjust the parameters of maximum current control

(a) ECU

(b) Power system

Fig. 25. Hardware test platform

Fig. 26. The waveform of line voltage and phase current wave before rectification

(a) PI controller (b) Deadbeat controllerFig. 27. Dynamic response comparison with different

control methods

t(50ms/div)

(a) (b) (c) Fig. 28. The experiment waves of DC output under

different parameters



algorithm by DSP28335 and observe the DC output. Fig. 28(a) shows the output waves of Vout and Vbus when kv=0.4, kc=0.22, and ke=0.02. They all drop swiftly and current sharing is not achieved. Fig. 28(b) presents the output waves of Vout and Vbus when kv=0.42, kc=0.63, and ke=0.04. The output voltage is maintained at 5 V and the voltage Vbus of the current sharing bus increases to a steady value. Fig. 28(c) presents the output waves of Vout and Vbus when kv=0.48, kc=0.86, and ke=0.06. Both can output stably and then achieve optimum current sharing.

Therefore, the maximum current method can achieve automatic current sharing and one can obtain optimum current sharing results by adjusting the parameters of the control algorithm.

8. Conclusions The distributed control system with 18 power cells in the

permanent magnet synchronous generator was introduced in this paper. To optimize the output performance of each power cell, the PWM deadbeat control strategy of the load power feedforward is used to stabilize the output of each power cell. These power cells cannot be parallel-connected directly because they have different properties. Therefore, the method of maximum current is adopted and optimum current sharing is realized by optimizing the parameters of the control structure.

Simulation and experiment have verified the validity of the control method and current sharing method. However, we should conduct repeatable experiments to explore how to smoothly control partial power cells to switch and maintain output DC stabilization. The synchronous generator with low voltage and high current output under the deadbeat control method will become a novel and efficient generator.

Acknowledgements This research was supported by the Major Scientific

and Technological Project of Guangdong Province. (No. 2015B010118003).

References

[1] Qamaruzzaman, A. Purwadi and P. A. Dahono, “A DC high-current low-voltage power generating system,” in Proceedings of IEEE Power System Technology Conference, Piscataway, USA, October 2002.

[2] P. A. Dahono, M. Firmansyah and D. Pramasti Y., “A high-current low-voltage DC power supply,” in Proceedings of IEEE PEDS, Piscataway, USA, October 2001.

[3] K. Morimoto, T. Doi, H. Manabe and Hyun-Woo Lee, “Next generation high efficiency high power dc-dc converter incorporating active switch and snubbing capacitor assisted full-bridge soft-switching PWM inverter with high frequency transformer for large current output,” in Proceedings of IEEE APEC, Austin, United States, March 2005.

[4] Sanjeevikumar P., Blaabjerg F. and Wheeler P., “Five-phase five-level open-windings/star-winding inverter drive for low-voltage/high-current applica-tions,” in Proceedings of IEEE International ITEC, Busan, Korea, May 2016.

[5] Rajesh, Geetha, Suja Malar and T. Ahilan, “Current sharing in parallel connected boost converters,” The Journal of Engineering, vol. 1, no. 1, pp. 1-9, Oct. 2016.

[6] Zhang Fei, Shang Dandan and Yang Huizhong, “Optimization design of structure of the low-voltage high-current permanent-magnet DC motor,” Machinery Design & Manufacture, vol.9, no.5, pp. 96-105, Dec. 2013.

[7] Liu Lingshun, Zhou Hongqing and Hu Yuwen, “Optimal design of dual stator-winding induction generator system with low voltage and large current output,” Electric Machine & Control Application, vol.36, no.11, pp.38-45, Jan. 2010.

[8] O. Ojo and I.E. Davidson, “PWM-VSI inverter-assisted stand-alone dual stator winding induction generator,” IEEE Transactions on Industry Appli-cation, vol. 36, no. 6, pp. 1604-1611, Apr. 2000.

[9] Yang Hejun, Xie Kaigui, Wai Rongjong and Li Chunyan, “Reliability evaluation of electrical distri-bution network containing distributed generation using directed-relation-graph,” J. Electr. Eng. Technol., vol. 9, no. 4, pp. 1188-1195, Jul. 2014.

[10] Alatawi, Khaled, Fahad Almasoudi and Mohammad Matin, “Switching performance and efficiency investi-gation of GaN based DC-DC Buck converter for low voltage and high current applications,” in Proceedings of SPIE Optical Engineering+ Applications. Sep-tember 2016.

[11] Sobrayen, Lingeshwaren and Akshay K. Rathore, “Three-phase soft-switching bi-directional DC-DC converter for low voltage high power applications,” in Proceedings of IEEE 8th IPEMC-ECCE Asia, March 2016.

[12] Zhang Xueguang, Zhang Wenjie, Chen Jiaming and Xu Dianguo, “Deadbeat control strategy of circulating currents in parallel connection system of three-phase pwm converter,” IEEE Trans. Energy Convers., vol. 29, no. 2, pp. 406-417, Jun. 2014.

[13] Chen Qi, Pengfei Tu, Peng Wang and Michael Zagrodnik, “Predictive current control for multilevel cascade H-bridge inverters based on a deadbeat solution,” Journal of Power Electronics, vol. 17, no.2, pp. 76-87, 2017.



[14] Yongchang Zhang, Wei Xie and Yingchao Zhang, “Deadbeat direct power control of three-phase pulse-width modulation rectifiers,” IET Power Electronics, vol. 7, no. 6, pp. 1340-1346, Jun. 2014.

[15] An Luo, Huagen Xiao and Zhikang Shuai, “Double deadbeat-loop control method for distribution static compensator,” IET Power Electronics, vol. 8, no. 7, pp. 1104-1110, Jul. 2015.

[16] Eskandari-Torbati. H. and Khaburi D. A., “Direct power control of three phase PWM rectifier using model predictive control and SVM switching,” in Proceedings of Power Electronics, Drive Systems and Technologies Conference, Tehran, Iran, February 2013.

[17] Li Mingshui, Wan Jianru and Li Guangye, “Research on power feedforward control strategy of PWM rectifier,” in Proceedings of PESA, Hong Kong, China, June 2011.

[18] Jun Tian, Zhengyu Liu and Jun Shu, “Based on the ultra-short power prediction and feed-forward control of energy management for microgrid system applied in industrial park,” IET Generation, Transmission & Distribution, vol.10, no.9, pp.2259-2266, Mar. 2016.

[19] Yongchang Zhang, Zhengxi Li and Changbin Hu, “Performance improvement of direct power control of pwm rectifier with simple calculation,” IEEE Trans. Power Electron., vol. 28, no. 7, pp. 3428-3437, Jul. 2013.

[20] Y. Yingyi, F. C. Lee and P. Mattavelli, “Analysis and design of average current mode control using describing function-based equivalent circuit model,” in Proceedings of IEEE ECCE, Raleigh, USA, September 2012.

[21] Z. Ye, D. Boroyevich, J.-Y. Choi and F.C.Lee, “Control of circulating current in two parallel three-phase boost rectifiers,” IEEE Tran. Power Electron., vol. 17, no. 5, pp. 609-615, Sep. 2002.

[22] Qingqing Yuan and Kun Xia, “Current decoupling control for the three-level PWM rectifier with a low switching frequency,” J. Electr. Eng. Technol., vol. 10, no. 1, pp. 280-287, Jan. 2015.

[23] C.-T. Pan and Y.-H. Liao, “Modeling and coordinate control of circulating currents in parallel three-phase boost rectifiers,” IEEE Trans. Ind. Electron., vol. 52, no. 2, pp. 825-838, Apr. 2007.

[24] Chuang, Chen-Feng, Ching-Tsai Pan and Hao-Chien Cheng, “A Novel Transformer-less Interleaved Four-Phase Step-Down DC Converter With Low Switch Voltage Stress and Automatic Uniform Current-Sharing Characteristics,” IEEE Tran. on Power Electron. vol. 31, no. 1, pp. 406-417, Jan. 2016.

[25] K. I. Hwu, W. Z. Jiang and P. Y. Wu, “An expandable four-phase interleaved high step-down converter with low switch voltage stress and automatic uniform current sharing,” IEEE Trans. Ind. Electron., vol.63, no. 10, pp. 6064-6072, Oct. 2007.

Jinfeng Liu She received Ph.D. degree in electrical engineering from Harbin university of science and technology in 2011. She is associate professor in the same school. Her research interests are motor control, distribution system and resonance compensation.

Yu Zhang He is studying for a master’s degree at Harbin University of science and technology. His research interests are switching power supply and digital signal processing.

Xudong Wang He received Ph.D. degree in mechanical electrical engin-eering from Harbin institute of technol-ogy in 2000. He is professor at Harbin University of science and technology and director of education department research center. His research interests are renewable energy vehicle, synchro-

nous motor control and design.

Herbert Ho-Ching IU He received Ph.D. degree from the Hong Kong polytechnic university in 2000. He is currently a professor at the University of Western Australia. His research interests include power electronics, renewable energy, nonlinear dynamics and memristor.

Low-Voltage and High-Current DC Output Realized by ...

Documents

Transcript of Low-Voltage and High-Current DC Output Realized by ...