Soft-Switching Single-Phase Grid-Connecting Converter ... - IEEE Xplore

2922

IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 29, NO. 6, JUNE 2014

Soft-Switching Single-Phase Grid-Connecting Converter Using DCM Operation and a Turn-Off Snubber Capacitor Takanori Isobe, Member, IEEE, Kyohei Kato, Naoto Kojima, and Ryuichi Shimada, Member, IEEE

Abstract—This paper proposes a new single-phase dc/ac converter topology with soft-switching operation. The proposed converter achieves soft-switching by using DCM (discontinuous current mode) operation and a common turn-off snubber capacitor connected to the dc terminals of a full-bridge configuration. One additional semiconductor switch is used to disconnect the snubber capacitor from the dc side, and allows the snubber capacitor to be discharged. The DCM operation realizes zero-current turn-on and the snubber capacitor achieves zero-voltage turn-off. The proposed converter can be applied to a grid-connected converter for domestic renewable energy and energy storage, including Photovoltaic and battery energy storage devices. This paper describes operation principles, control and modulation techniques as a grid-connecting converter. Experimental verification including loss analysis with a 1-kW pilot device is provided. Index Terms—Discontinuous current mode, soft switching.

I. INTRODUCTION ITH increase of renewable energy power generation installation, the power conversion technology for these systems and energy storage devices along with the renewable energy sources are paid attention. For example, photovoltaic (PV) power generations are widely installed in domestic area and recently low-power grid-connected converters, which offer individual dc/ac conversion for each solar panel, are studied well [1]. By doing this, maximum power point tracking operation can be applied to each solar panel independently and higher energy harvest can be achieved, compared with doing by one centralized converter. For these low-power converters, compactness is especially needed.

W

Manuscript received November 29, 2012; revised March 20, 2013 and June 21, 2013; accepted July 9, 2013. Date of current version January 29, 2014. This paper was presented at 15th International Power Electronics and Motion Control Conference, EPE-PEMC 2012 ECCE Europe, Novi Sad, Serbia, Sep. 4–6, 2012. The original title is “Single-Phase Grid-Connecting Converter Based on a MERS Soft-Switching Concept for Renewable Energy and Energy Storage.” Recommended for publication by Associate Editor J. Liu. T. Isobe and R. Shimada are with the Reserach Laboratory for Nuclear Reactors, Tokyo Institute of Technology, Tokyo 152-8550 Japan (e-mail: tisobe@ nr.titech.ac.jp; [email protected]). K. Kato and N. Kojima are with the MERSTech Co., Ltd, Tokyo 107-0052, Japan (e-mail: [email protected]; [email protected]). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TPEL.2013.2274390

For the low-power grid-connected converter, usually voltage source type converter is used and an inductor is needed to connect to a voltage source type grid; on the other hand, converters can be directly connected to their load in motor drive applications. Therefore, the inductor can be one of space taking components in the grid-connected converter. The grid-connected converter can be designed with both continuous current mode and discontinuous current mode (DCM) operations by selecting the inductance. The DCM operations are well reported in dc/dc converters and power factor correction (PFC) rectifiers for a single phase [2], [3] and a three phase [4]; however, not popular in dc/ac inverters. It is said that the DCM operation has advantages in low power applications; therefore, can be also advantageous in the low-power grid-connected converter. Increase of converter switching frequency can reduce the size of passive components including the inductor and a filter, and contribute to downsizing. However, increase of switching losses and electromagnetic interference (EMI) emission due to the high frequency switching is one of challenges. High dv/dt generates conducted EMI with various parasitics in the converter; for example, parasitic capacitance of switching devices, gate drives, and sensors to the ground. Additionally in PV applications, a large parasitic capacitance of PV panels to the ground has a large impact to the EMI problems if there is no galvanic isolation [5]. Use of an EMI filter can reduce the conducted EMI emission; however, additional components and loss are introduced. Therefore, reduction in generating EMI is attractive from the point of compactness and efficiency of the converter. The soft-switching technology can improve these high frequency switching problems. This paper proposes a new singlephase dc/ac converter topology with soft-switching operation, which can reduce the switching losses and EMI generation. The proposed converter achieves the soft-switching by using DCM operation and a common turn-off snubber capacitor. The DCM operation realizes zero-current turn-on and the common turn-off snubber capacitor achieves zero-voltage turn-off. This turn-off snubber topology comes from a series reactive power compensator for the ac circuit proposed in [6]–[8]. It consists of full-bridge configured semiconductor switches and a capacitor, whose capacitance is relatively low. By operating with line frequency switching, the capacitor can have a zero voltage period in a half cycle. By applying this mechanism to converters with high frequency switching, zero-voltage turn-off can be achieved. A three-phase PFC rectifier with this soft-switching concept has been proposed in [9].

0885-8993 © 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications standards/publications/rights/index.html for more information.

ISOBE et al.: SOFT-SWITCHING SINGLE-PHASE GRID-CONNECTING CONVERTER USING DCM OPERATION

Fig. 1.

Circuit configuration of the proposed converter.

Fig. 2.

Schematic view of the proposed switching pattern.

II. PROPOSED CONVERTER A. Configuration and Switching Principles The configuration of the proposed converter is shown in Fig. 1. The configuration consists of a full bridge with a small snubber capacitor Cs and an additional disconnecting switch D. It is similar to usual single-phase converters; however, the capacitor connected to the dc-side of the bridge is just snubber use and the main dc-link capacitor is connected via the disconnecting switch D. As same√ as for usual dc/ac converters, output voltage range is defined as 2Vac < Vdc . In the ac side, an inductor Lac is connected; however, its inductance is relatively low since this converter operates with DCM. As a drawback of using DCM, ac-side low-pass filter is essentially needed. Schematic switching patterns are shown in Fig. 2. U and Y are controlled simultaneously and V and X are kept OFF during positive ac current period, and V and X are controlled during negative ac current period. φ is defined as a phase angle difference between the ac voltage and current. A reactive power control can be achieved by controlling φ. Operations within 90◦ < φ < 270◦ are rectifier operations. The disconnecting switch D is controlled to achieve zerovoltage turn-off. For rectifier operation without reactive power flow (φ = 180◦ ), the disconnecting switch D can be kept OFF all the time as proposed in [9]. This operation is attractive for unidirectional rectifier application since the disconnecting switch can be replaced by a diode. However, this paper proposes another control method using the disconnecting switch for all the operation.

2923

Simulated switching waveforms and current paths are shown in Figs. 3 and 4, respectively. In a switching cycle, ac-side voltage can be assumed as a constant; therefore, is represented as vac in the following discussion. vac is always lower than Vdc . Before the current path (a), D must be turned ON. Then by turning ON two bridge switches (U and Y in case of the period shown in Fig. 3) simultaneously, Lac current, ilac , starts to flow with the current path shown in Fig. 4(a). This is one main mode of this operation and the current is linearly increased by Vdc − vac . Then, the current starts to discharge Cs by turn-off of D, as shown in Fig. 4(b). After the capacitor voltage, vcs , reaches at zero, the current flows in the current path shown in Fig. 4(c). Then, the current starts to charge Cs by turn-off of the bridge switches, as shown in Fig. 4(d). Current paths (b), (c), and (d) are transient modes to achieve soft-switching turn-off, and has relatively short period compared to other main modes. After vcs reaches at Vdc , the current flows back to the dc side as shown in Fig. 4(e). This mode is another main mode and the current is linearly decreased by vac + Vdc . The discontinuous switch D must be turned ON before the next current path (a) will starts. In the current path (e), the discontinuous switch D is conducting and the voltage difference between Vdc and vcs is zero; therefore, D can be turned ON at any time in this period. D can be turned ON after ilac becomes zero; however, possible voltage difference between Vdc and vcs causes short-circuit current. ilac becomes zero in each switching cycle, in other words, circuit parameters must be designed to achieve this DCM operation. Turn-on is performed within the zero current period and di/dt is mitigated by Lac ; therefore, zero-current switching (ZCS) turn-on of the bridge switches and D is achieved. Prior turn-off of switch D results in complete discharge of Cs . Turnoff is performed within the zero voltage period and dv/dt is mitigated by Cs ; therefore, zero-voltage switching (ZVS) turnoff of the bridge switches is achieved. Switch D itself is also operated with soft switching as shown in Fig. 3(c). B. Modulation Technique PFC rectifiers with DCM operation are known to have good harmonic characteristics without modulation, especially when the step-up ratio is higher [4]. As same as for usual PFC converters with DCM, the three-phase PFC converter with the same soft-switching concept proposed in [9] generates good ac current waveform without modulation. The same technique can be applied to the proposed converter in rectifier operation; however, the principle is not valid for the inverter operation. Therefore, some active control methods must be applied. Additionally, the active control can improve the harmonic characteristics also in the rectifier operation. In PFC rectifier, some active control methods to achieve much better current waveform in DCM operation has already been proposed [3], [10]. Fig. 5 shows schematic diagram of the proposed modulation method. In this discussion, transient periods of charging and discharging of Cs , (b) and (d), and zero voltage period (c) are neglected. Those come from relatively small Cs , whose purpose is just dv/dt mitigation, and shortest Td to achieve zero voltage. With those assumption, ilac becomes a triangular waveform as

2924


(a)

(b)

(c)

Fig. 3. Schematic waveforms of several switching cycles during positive ac current period. Switch voltages, vsw , and currents, isw , of (a) switch U, which is controlled during this period, (b) switch X, which is kept OFF and operated as a diode, and (c) switch D. Negative isw means diode operation.

Fig. 5.

Fig. 4.

Possible current paths during positive ac current period.

shown in Fig. 5. The average of ilac in a switching cycle, ilac , can be easily calculated as ⎧ d2 Vdc (Vdc − vac ) ⎪ ⎪ ⎨ ilac = L f (V + v ) ac sw dc ac 2 ⎪ V (V + v −d ⎪ dc dc ac ) ⎩ ilac = Lac fsw (Vdc − vac )

Schematic diagram of the modulation technique.

for both current polarity in ac, where fsw is switching frequency, d is duty ratio which can be expressed as TON /Tsw . ilac is function of the ac line voltage phase angle, θ. Instantaneous value of iac around the switching cycle can be assumed to be same as ilac ; therefore, ilac should be controlled to be a sinusoidal waveform in a line cycle. To obtain sinusoidal ac current of Iac ∗ (in rms) with phase angle difference of φ

(sin (θ − φ) > 0) (1) (sin (θ − φ) < 0)

ilac (inv) =

√ 2Iac ∗ sin (θ − φ)

(2)


(a) Fig. 6.

2925

(b)

(c)

Simulation waveforms of ac currents with modulation. (a) φ = 0◦ . (b) φ = 30 ◦ . (c) φ = 180 ◦ (rectifier operation).

gives d as function of θ. From these equations, d can be expressed as √ ⎧ √ 2Lac fsw (Vdc + vac ) sin (θ − φ) ⎪ ∗ ⎪ ⎪ d = Iac · ⎪ ⎪ Vdc (Vdc − vac ) ⎪ ⎪ ⎪ ⎪ ⎨ (sin (θ − φ) > 0) . (3) √ ⎪ ⎪ √ 2L f (V − v ) sin (θ − φ) ⎪ ac sw dc ac ⎪ ⎪ d = Iac ∗ · ⎪ ⎪ −Vdc (Vdc + vac ) ⎪ ⎪ ⎩ (sin (θ − φ) < 0) Usually√ac line voltage is practically pure sinusoidal; therefore, vac = 2Vac sin θ. √ If Vdc , Vac , and φ are constant, d/ Iac ∗ depends on only θ; therefore, it can be obtained by off-line calculation and stored as a table as function of θ. d can be easily calculated from the table and given Iac ∗ by the controller. It is an attractive feature from the point of control implementation. To derive d, effects of Cs were neglected; however, should be considered for ZVS turn-off achievement. The mechanism to achieve the ZVS is contained in current paths (b), (c), and (d), and those paths are caused by Td . Required time for discharge and charge Cs depends on θ; therefore, optimum Td , which realizes zero “zero voltage period,” is also depends on θ. The optimum Td , which is equal to the capacitor discharging time, can be expressed as Td =

Cs · Vdc Ip (θ)

(4)

where Ip (θ) is the peak current of ilac with the assumption that Cs is relatively small and the current does not changed by the discharge and charge. The equation indicates that the required Td becomes large when Ip is low; therefore, the complete discharge can be difficult around zero-cross point of the ac current with the fixed switching frequency. This paper proposes using a fixed Td , which achieves the complete discharge for a certain wide period in line cycle. This results in not complete of the discharge and hard-switching turn-off in the rest of period; however, generated switching loss around the current zero cross is considered to be not large.

TABLE I SIMULATION PARAMETERS

TABLE II STATISTICS OF RESULTING WAVEFORMS IN SIMULATION

Simulation waveforms with line cycle scale are shown in Fig. 6. Parameters of the simulation circuit and resulting statistics, including fundamental ac current in rms and total harmonic distortion under 49th order, are shown in Tables I and II, respectively. Almost sinusoidal current waveforms were obtained. The same Iac ∗ (= 7 A) is given for all conditions; however, inverter operations generate larger current compared to one of rectifier operation. The reason for that is Td . Effective period of ilac increasing is shorter than the period which was used for the modulation calculation. This has different influence between inverter and rectifier operations.

C. Design Circuit Parameters The proposed modulation assumes DCM operation; therefore, zero current period must remain in every switching period in whole line phase. This determines the maximum ac current achieved by given circuit parameters. The following discussion also neglects periods (b), (c), and (d), and assumes fixed switching frequency, fsw . Critical d to

2926


Fig. 8. Overview of the fabricated converter. (a) Semiconductors and set of a snubber capacitors on PCB. (b) One of divided stepup inductors (L a c /2).

Fig. 7. Schematic diagram of the maximum instantaneous ac current, im a x , based on the parameters shown in Table I and various ac current waveforms.

TABLE III PARAMETERS OF THE FABRICATED CONVERTER

achieve the DCM operation, dm ax , can be easily calculated as √ Vdc + 2Vac sin θ dm ax = . (5) 2Vdc From equation (5), the maximum instantaneous ac current, im ax , is given by im ax =

Vdc 2 − 2Vac 2 sin2 θ 4Lac fsw Vdc

(6)

and the curve based on the simulation parameters listed in Table I are shown in Fig. 7 with various ac current waveforms. The instantaneous current must be within the range of ±im ax , and the maximum ac current amplitude depends on φ. From equation (6), the maximum Iac ∗ in φ = 0◦ operation, Iac.m ax(φ=0 ◦ ) , is obviously given by Vdc 2 − 2Vac 2 . Iac.m ax(φ=0 ◦ ) = √ 4 2Lac fsw Vdc

(7)

To realize the converter which satisfies given specifications of rated power and voltages, Lac is the main selectable parameter and fsw is usually given by implementation related conditions. A low-pass filter which consists of an inductor and a capacitor is used for removing high-frequency components from ilac and making iac a smooth waveform. Almost all the harmonic components can be assumed to flow in the filter capacitor, and its capacitance, Cf , can be determined by ripple voltage limitation of the capacitor. The modulation proposed in this paper assumes this voltage to be constant within a switching cycle; therefore, a finite capacitance introduces some error. By specifying acceptable amount of the voltage ripple from the point of modulation, Cf can be determined. Then, the filter inductance, Lf , can be determined to satisfy given amount of acceptable ripple in the output ac current. III. EXPERIMENTAL VERIFICATION A. Setup To verify the proposed concept and modulation technique, a pilot device with 1-kW rating was fabricated. Overview of the fabricated converter is shown in Fig. 8, and parameters are listed

in Table III. To reduce unwanted oscillation caused by stray inductance, a PCB for the proposed configuration was designed and fabricated as shown in Fig. 8(a). Snubber capacitors and switch D are connected to the bridge circuit as close as possible. IGBTs with 600 V and 22 A rating were used for this converter. Inductors shown in Fig. 8(b) were used for Lac , which use the Litz wire as winding since the flowing current includes relatively high high-frequency components. The inductance was determined to achieve 1-kW with given voltages, switching frequency and unity power factor by equation √ (7). The peak current of iac is expected to be 7.07 A (5 A × 2); however, double of the peak current flows in Lac at least, due to its discontinuous current waveform. Inductors used for Lac in experiments have peak current rating of 17 A including some margin. On the other hand, Lf uses usual copper wire and has peak current rating of 8.1 A. Schematic diagram of the experimental setup including controller is shown in Fig. 9. A small field-programmable gate array (FPGA) board was used for control with zero-cross detect circuit connected to the ac side. The off-line calculated table for modulation was stored in the FPGA and d was determined by elapsed time from the last zero crossing. Electronic power supplies, which can also work as load, were used for both of dc- and ac sides. The power supplies controlled their voltage to be given values, that means the converter does not have responsibility to maintain voltages. For some real applications, dc voltage control must be done by power flow control of the converter. B. Waveforms Experimental waveforms with line cycle scale are shown in Fig. 10. Current set point, Iac ∗ , was set at nearly 5 A to achieve exactly 1-kW power conversion with φ = 0◦ and φ = 180◦ . Almost unity power factor and practically good sinusoidal


Fig. 9.

2927

Schematic diagram of the experimental setup including the controller.

Fig. 11. Experimental waveforms of the switch voltage and current with φ = 0 ◦ . (a) Controlled IGBT (used as IGBT). (b) IGBT which is kept OFF(used as a diode).

(a)

(b) Fig. 10. Experimental waveforms of ac currents in (a) φ = 0◦ , (b) φ = 180 ◦ (rectifier operation).

waveform of the current was observed. The experimental result confirms the proposed control and modulation technique. Experimental waveforms of switch voltage and current with φ = 0◦ are shown in Fig. 11. Fig. 11(a) and (b) were obtained from one switching device; however, the device acts as IGBT during a half cycle and as diode during the other half cycle. Waveforms with φ = 180◦ are shown in Fig. 12. ZCS of turnon and ZVS of turn-off were confirmed. Some amount of tail current was observed when the IGBT was turned OFF.

Fig. 12. Experimental waveforms of switch voltage and current with φ = 180 ◦ (rectifier operation). (a) Controlled IGBT (used as IGBT). (b) IGBT which is kept OFF (used as a diode).

2928

Fig. 13.


Oscillation due to a diode reverse recovery. Fig. 15.

Estimated semiconductor losses of the VSI and proposed converter.

result indicates that the design of the inductor is key design point of the proposed converter. IV. COMPARATIVE DISCUSSION WITH THE CONVENTIONAL CONVERTER Fig. 14.

Experimentally analyzed losses at 1-kW inverter operation.

In Fig. 11(a), small negative current flow, which is not shown in simulation, was observed in zero current period. This is caused by reverse recovery of free-wheeling diode (FWD) of the other side switch. Fig. 13 shows how the reverse recovery affects operation. The reverse recovery current of U flows into the inductor as shown as (2) since X is still turned OFF; on the other hand, the current flows into X in usual voltage source converters. After the reverse recovery current is shut down, the inductor current continues to flow and introduces current in X as (3). The current is also turned OFF and has reverse recovery current as (4). This behavior repeats some times, and observed as voltage share swinging between the upper and lower switches in the zero current period. The voltage share swinging due to the same reason was also observed but more clearly in rectifier operation as shown in Fig. 12(a) and (b). The phenomenon does not result in additional loss; however, can cause distortion in modulated current, and/or reduction of power conversion capability with same circuit parameters. Using MOSFETs is attractive for much higher switching frequency; however, poor reverse recovery characteristics of body diode can be another challenge. C. Loss Analysis To discuss advantage and disadvantage of the proposed converter, loss analysis in the inverter operation was performed and result is shown in Fig. 14. Total loss was measured by a digital power meter connected to ac side and dc side. Semiconductor conduction losses were calculated from measured I–V characteristics and device current waveforms. Switching losses could not be specified; therefore, it is included in “others.” Inductor losses were calculated from the measured flowing current and across the voltage. Almost all losses were specified and that means operation was correct and switching losses are relatively low. Inductor loss of Lac was relatively high. Increased copper loss in the winding and hysteresis loss in the core are thought to be reasons. The

A. Semiconductor Loss To make advantages and disadvantages of the proposed converter clear, this section discusses some aspects of the proposed converter in comparison to the conventional converter. Semiconductor conduction losses of the proposed converter and a voltage source inverter (VSI) were calculated based on the same condition with the experimental setup. For both converters, using the same semiconductor device, which was actually used for experiments, was assumed and conduction losses were calculated by simulated current waveforms and modeled on-state voltages of the device, which are available from the device datasheet. A pulse-width-modulation (PWM) with unipolar voltage switching was applied for the VSI. Fig. 15 shows the calculation results. Average flowing current in one device is same in both converters; however, resulting conduction loss in one device of the proposed converter is slightly large. The instantaneous device current in a switching period is almost constant and equal to the average in the VSI; on the other hand, it dynamically changes from zero to a peak in the proposed converter. This can be a drawback of employing DCM. The conduction loss of the switch D also introduces additional loss in the proposed converter. Switching loss depends on implementation issues; therefore, this paper avoids estimating based on the datasheet. The proposed converter has disadvantage in conduction loss; however, offers switching loss reduction; therefore, can be advantageous in higher switching frequency operation, which is attractive for the passive components design. Moreover the soft-switching feature has possibility to make devices whose design prioritizes on-state voltage be used [11]. B. Size of Magnetic Components Lac and the low-pass filter of the proposed converter naturally form a LCL filter, which can reduce the size of magnetic components also in conventional VSIs. A rough estimation of magnetic components was also performed. The filter designed for experiments achieves 17% of the voltage ripple in Cf and 8.2% of the current ripple in Lf in the proposed converter. For


TABLE IV MAGNETIC COMPONENTS CALCULATION

2929

circuit topologies is a complex task. This paper indicated that conduction loss is increased and the size of magnetic components can be decreased with the same switching frequency. However, total semiconductor loss depends on switching frequency; moreover, required inductance, size, and loss of magnetic components also depend on the switching frequency. It means that each circuit topology has different optimum switching frequency. Comprehensive comparison including losses and sizes of semiconductor, cooling device, and passive components is needed to conclude. REFERENCES

the VSI, single-L filter is assumed and the inductance was determined to achieve the same current ripple as that was achieved in the proposed converter. Switching frequency was 25 kHz for both converters and PWM with unipolar voltage switching was applied for the VSI. Table IV shows the calculation results. Peak current, Ip , and current in rms, Irm s , are calculated by simulation for Lac and Lf in the proposed converter and the grid connecting inductor, Lg , in the VSI. To discuss the size of magnetic components, Lac · Ip · Irm s , normalized by its value for Lg are also shown as a size index value. The size of a magnetic component can be thought to be proportional to the value, if current and flux densities and fill factors are same. The calculation results suggest that the proposed converter can be achieved with less than half of magnetic components compared to the VSI, even in the same switching frequency. However, it should be noted that the current in Lac is discontinuous; therefore, fill factor of the winding can be decreased by some reasons; for example, the use of Litz wire. V. CONCLUSION The proposed converter is a DCM operated grid connected converter with a soft-switching turn-off mechanism achieved by a small snubber capacitor, Cs , and an additional switch D. This paper proposed a modulation method and a design procedure of circuit parameters to achieve given specifications with assumption of negligibly small Cs . For much more high frequency operation, effects of Cs should be taken in account since relatively large Cs must be used to achieve sufficient soft-switching advantages, and it can be possible with more complex calculation and control implementation. The proposed converter offers: 1) easy control to achieve unity power factor and low harmonic distortion in ac current; 2) small magnetic components to connect with grid; 3) soft-switching operation; therefore, possibility to reduce the size of magnetic components further by use of higher switching frequency; 4) low dv/dt; therefore, low EMI and reduced filtering effort for EMI. However, it has some drawbacks; therefore, should be compared from some points of view. Comparison between different

[1] E. Román, R. Alonso, P. Ibañez, S. Elorduizapatarietxe, and D. Goitia, “Intelligent PV module for grid-connected PV systems,” IEEE Trans. Ind. Electron., vol. 53, no. 4, pp. 1066–1073, Jun. 2006. [2] I. Takahashi, “Power factor improvement of a diode rectifier circuit,” in Proc. IEEE Ind. Appl. Soc. Annu. Meet. 1990, vol. 2, pp. 1289–1294. [3] J. Lazar and S. Cuk, “Open loop control of a unity power factor, discontinuous conduction mode boost rectifier,” in Proc. 17th Int. Telecomun. Energy Conf. (INTELEC), 1995, pp. 671–677. [4] A. R. Prasad, P. D. Ziogas, and S. Manias, “An active power factor correction technique for three-phase diode rectifiers,” IEEE Trans. Power Electron., vol. 6, no. 1, pp. 83–92, Jan. 1991. [5] I. Patrao, E. Figueres, F. González-Esp´ın, and G. Garcerá, “Transformerless topologies for grid-connected single-phase photovoltaic inverters,” Renewable Sustainable Energy Rev., vol. 15, no. 7, pp. 3423–3431, 2011. [6] T. Takaku, T. Isobe, J. Narushima, H. Tsutsui, and R. Shimada, “Power factor correction using magnetic energy recovery current switches,” Electr. Eng. Jpn., vol. 160, no. 3, pp. 56–62, 2007. [7] J. A. Wiik, F. D. Wijaya and R. Shimada, “Characteristics of the magnetic energy recovery switch (MERS) as a series FACTS controller,” IEEE Trans. Power Del., vol. 24, no. 2, pp. 828–836, Apr. 2009. [8] J. A. Wiik, A. Kulka, T. Isobe, K. Usuki, M. Molinas, T. Takaku, T. Undeland, and R. Shimada, “Loss and rating considerations of a wind energy conversion system with reactive compensation by magnetic energy recovery switch (MERS),” Eur. Power Electron. J., vol. 18, no. 3, pp. 25– 30, 2008. [9] Y. Miyaji, T. Isobe, and R. Shimada, “A soft-switching active rectifier using a concept of magnetic energy recovery switch,” in Proc. Int. Power Electron. Conf.—ECCE Asia, 2010, pp. 2318–2323. [10] K. Taniguchi and Y. Nakaya, “Analysis and improvement of input current waveforms for discontinuous-mode boost converter with unity power factor,” in Proc. Power Convers. Conf.—Nagaoka 1997, vol. 1, pp. 399–404. [11] R. Shimada, J. A. Wiik, T. Isobe, T. Takaku, N. Iwamuro, Y. Uchida, M. Molinas, and T. M. Undeland, “A new AC current switch called MERS with low on-state voltage IGBTs (1.54 V) for renewable energy and power saving applications,” in Proc. 20th Int. Symp. Power Semicond. Devices Integr. Circuits, 2008, pp. 4–11.

Takanori Isobe (M’07) was born in Hamamatsu, Japan, in 1978. He received the M.S. degree in nuclear engineering and the Ph.D. degree in energy sciences from the Tokyo Institute of Technology, Tokyo, Japan, in 2005 and 2008, respectively. From 2008 to 2010 he was a Researcher with the Tokyo Institute of Technology, where from 2010 to 2012, he was an Assistant Professor at the Research Laboratory for Nuclear Reactors. He is currently a Researcher at the same university and with MERSTech, Tokyo, Japan. His research interests include static reactive power compensators and soft-switching power converters. Dr. Isobe is a Member of the Institute of Electrical Engineers of Japan and The Japan Institute of Power Electronics.

2930

Kyohei Kato was born in Hokkaido, Japan, in 1979. He received the Graduate degree from School of Information-Oriented Management, Sanno University, Kanagawa, Japan, in 2009. From 2001 to 2009, he was a Software Engineer and involved in design and development of mobile application software and data storage devices. In 2009, he joined MERSTech, Tokyo, Japan, where he is currently involved in development of power-electronics devices. Main scope of his work includes software development for embedded system in power-electronics devices, especially for renewable energy and smart grid.

Naoto Kojima was born in Nagoya, Japan, in 1971. He received the M.S. degree in electrical engineering and computer science from Nagoya Institute of Technology, Nagoya, Japan, in 1997. In 1997, he joined Nippon Telegraph and Telephone Corporation, Japan. From 1998 to 2001, he was with NTT Communications Corporation, Japan. From 2001 to 2008, he was with a software developing company for project management. In 2008, he joined MERSTech, Tokyo, Japan, where he was involved in intellectual property management and is currently the Manager of Engineering Department. Mr. Kojima is a Member of Information Processing Society of Japan and the Japanese Society for Artificial Intelligence.


Ryuichi Shimada (M’94) was born in Tochigi, Japan, in 1948. He received the B.Eng., M.Eng., and D.Eng degrees in electrical engineering from the Tokyo Institute of Technology, Tokyo, Japan, in 1970, 1972, and 1975, respectively. From 1975 to 1988, he was a Researcher on Nuclear Fusion Development and an Electrical Engineer at the Japan Atomic Energy Research Institute, Tokaimura, Ibaraki, Japan. He joined the development of the world largest Tokamak-type fusion experimental machine JT-60. He was a Group Leader of the power supplies development of JT-60. In 1988, he was a responsible Director of JT-60 operation and experiment. In 1983, he joined the Princeton Plasma Physics laboratory, Princeton University, Princeton, NJ, USA, to begin the large Tokamak Fusion Test Reactor. In 1988, he became an Associate Professor in the Department of Electrical and Electronics Engineering, Tokyo Institute of Technology. In 1990, he became a Professor and joined the Research Laboratory for Nuclear Reactors, Tokyo Institute of Technology, where he was also with the Department of Energy Science. He retired in 2013 and is currently a Professor Emeritus at the Tokyo Institute of Technology. Prof. Shimada received the 1985 Outstanding Achievement Award from the Institute of Electrical Engineers of Japan and the 1976 and 2000 Outstanding Paper Award from the Institute of Electrical Engineers of Japan. He received the 2003 Excellent Published Book Award from the Institute of Electrical Engineers of Japan.