This article has been accepted for publication in a future issue of this journal, but has not been

fully edited. Content may change prior to final publication.

New Method to Achieve AC Harmonic Elimination and Energy Storage Integration for 12-Pulse Diode Rectifiers
Sanzhong Bai, Student Member, IEEE, and Srdjan Lukic, Member, IEEE

AbstractThe 12-pulse rectifier is often used to supply high power industrial loads. Its ability to effectively and cheaply mitigate the harmonics on the AC-side has ensured its dominance in the industry even as active front ends become cheaper and more reliable. Despite its many benefits, the 12-pulse rectifier is not able to reduce the AC-side harmonics to a level acceptable by the pertinent IEEE standards without additional filtering. In this paper we outline a novel method to profile the rectifier output current to be triangular which results in low AC-side harmonics. The novelty in the proposed approach is that we exploit the DCside filter design to minimize the VA rating of the current source used to profile the DC-side rectifier current. Additional benefits of the proposed method include lower VA rating of the DC filter, simple integration of DC energy storage, and effective AC harmonic control even when the initial rectifier current is discontinuous. Index TermsActive Power Filter, Storage, 12-pulse diode rectifier. harmonics, Energy

I. INTRODUCTION VAILABILITY of reliable and inexpensive semiconductor devices has led to numerous power electronics intensive industrial loads requiring DC power supply for operation. When a number of DC powered loads are in proximity, it becomes viable for them to share a common DC bus. Many such systems benefit from local DC storage to (1) reduce the power demand from the grid, (2) provide backup power, and (3) store locally-generated renewable power rather than feeding it back to the grid. Local DC distribution has been considered for data centers [1], DC-level plug-in vehicle charging stations [2], more electric ships and aircrafts [3]. The 12-pulse rectifier is a common, simple, and cost effective method to provide a stable DC supply, while at the same time minimizing the grid side harmonics. The twelve pulse rectifier is a good tradeoff between harmonic reduction and subsystem complexity and is commonly used in the industry. However, the 12-pulse rectifier alone does not sufficiently reduce the AC harmonics to a level prescribed by relevant standards, such as IEEE 519 [4].
Copyright 2009 IEEE. Personal use of this material is permitted. However, permission to use this material for any other purposes must be obtained from the IEEE by sending a request to pubs-permissions@ieee.org

A standard way to eliminate AC-side harmonics is to use active, passive or hybrid power filters [5-7]. Though these methods are well understood and widely used, passive filters are bulky, while active filters require complex control and specialized power electronics. An alternative method controls the current draw from the DC-side so as to minimize the AC harmonics. Researchers have shown that for the 12-pulse rectifier, the AC-side harmonics are minimized by shaping the two sixpulse rectifier output currents to be triangular and out of phase. The resulting total harmonic distortion (THD) is as low as 1% [8, 9]. In [8-13], researchers exploit this property; while effective at reducing the harmonics, the circuitry used to shape the current is placed in path of the load current, resulting in substantial VA rating. Furthermore the proposed approaches in [8-13] are only effective for continuous rectifier currents, and are not capable of DC storage integration. In this paper we propose a new method of profiling the DCside rectifier current by using current sources placed in parallel with each six-pulse rectifier bridge to inject current and shape the rectifier output current. The method is simple to implement, and makes use of standard series connected IGBT or MOSFET modules. The property of the resulting current sources is that they are able to deliver power to the load, allowing for the integration of an energy storage system on the DC side. The VA rating required to profile the current is substantially reduced by a proper choice of the DC-side LC filter parameters. The remainder of the paper is organized as follows: in Section II we review the proposed methods for harmonics reduction by shaping the rectifier output current; in Section III we establish the proposed harmonics reduction topology and control strategy. We present simulation and experimental results in Section IV, and close with conclusions in section V. II. REVIEW OF EXISTING METHODS FOR AC HARMONIC ELIMINATION BY PROFILING THE DC RECTIFIER CURRENT FOR THE 12-PULSE RECTIFIER A. Principle of Harmonic Elimination In analyzing the operation of 12-pulse rectifier (Fig. 1), some methods are dedicatedly designed to shape the output current of 12-pulse rectifier (irec1, irec2) so as to minimize AC-side

Copyright (c) 2011 IEEE. Personal use is permitted. For any other purposes, permission must be obtained from the IEEE by emailing pubs-permissions@ieee.org.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication.

harmonics (ia, ib, ic) formed by the operation of the rectifier. The exact shape of the rectifier output current that results in complete elimination of AC input current harmonics has been derived in [8, 9]. Authors in [8-13] show that the triangular waveform, shown in Fig. 2, presents a good approximation of the ideal rectifier output current that completely eliminates the AC current harmonics. Assuming that the rectifier current is profiled to be triangular, its Fourier series expansion gives: 1 (1) n irec1 =I L + 4 I L 2 2 cos(6nt ) n n =1 (1) 1 (1) n irec 2 =I L + 4 I L 2 2 cos 6 nt + 6 n n =1 where is the line frequency and IL is the load current delivered to the DC load, and irec1 and irec2 are defined in Fig. 1. Based on the operation of the six-pulse rectifier, the Fourier series expansion of rectifier output voltage is: = vrec1_ dc + vrec1_ ac vrec1
(1) n cos(6nt ) n =1 36n 2 1 = vrec 2 _ dc + vrec 2 _ ac vrec 2 = +

B. Existing Methods for Current Profiling Numerous topologies have been proposed to shape the rectifier output current in order to eliminate the AC harmonics.

Fig. 1. The 12-pulse diode rectifier

3 V p

6 Vp

(2)

where Vp is the peak value of the AC input voltage and vrec1 and vrec2 are defined in Fig. 1 as the rectified voltages. Assuming that the profiled current is triangular as expressed in (1), it is quite simple to calculate the harmonics in the ACside line current. Namely, by realizing that the AC-side phasea current ia can be expressed as: (3) ia = ic1 ) 3 + ia 2 (ia1 where the currents are defined in Fig. 1. We can express the ac-side current ia as the sum of the Fourier expansions of the rectified currents irec1 and irec2 i S 30 irec1 S 150 (4) ia rec1 + irec 2 S 0 3 where S is defined as: 2 3 1 1 S sin(t ) sin(5t ) sin(7t ) = 5 7 (5) 1 1 + sin(11t ) + sin(13t ) 11 13 which is the switching function shown in Fig 2. By numerically evaluating (4), the AC-side line current THD is shown to be close to 1% [8, 9]. Analyzing (1) and (2), both vrec and irec have a strong AC component at six times the source frequency. The fundamental AC components of the voltage and the current are in phase for each six-pulse rectifier, while these AC components are out of phase when comparing the upper and lower rectifiers, as evident from Fig. 2 and (1) and (2). The fact that the fundamental AC components of the voltage and the current are in phase for each six pulse rectifier is exploited in the LC filter design as discussed in later sections.

3 V p

6 Vp

(1) n cos 6 nt + 2 6 n =1 36n 1

Fig. 2. Construction of the input current

They generally rely on directly shaping the rectifier current [12, 13] or on inserting a voltage source between the capacitor bank and the rectifier [8-11]. In [12] the authors propose to directly shape the rectifier output current by using active switches in the current path. The approach outlined in [13] uses a dedicatedly-sized reactor to shape the output current; however, the approach is only effective for a fixed load value. Other approaches insert a voltage source into the rectifier output loop [8-11]. The voltage source generates a rectangular pulse voltage with controlled amplitude and 6 times the line frequency, resulting in triangular rectifier output current. Researchers present a rough implementation of this idea in [10] by using the tap-changer, while [8] uses a 5-level fixedamplitude voltage source. The control accuracy of these setups

Copyright (c) 2011 IEEE. Personal use is permitted. For any other purposes, permission must be obtained from the IEEE by emailing pubs-permissions@ieee.org.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication.

is limited. In [9, 11] authors shape the rectifier output current by using PWM controlled active converter to control the inserted voltage source. The problem with the approaches in [8-11] is that they require continuous rectifier output current to function. In addition, the rectifier output voltage needs to be ripple-free to have expected results, requiring a large filter capacitor. Moreover, the current through the 360 Hz transformer (for 60 Hz grid), has an amplitude of 2IL resulting in a large conduction loss. The noise generated by the low frequency transformer poses another problem. III. PROPOSED METHOD FOR HARMONIC REDUCTION AND ENERGY STORAGE INTEGRATION A. Proposed Approach A novel approach proposed in this paper inserts three current sources into the circuit as shown in Fig. 3. The two current sources (is1 and is2) in parallel with each rectifier are used to shape the rectifier currents irec1 and irec2 in order to eliminate the AC-side harmonics. The third current source is3 is used to inject active power from an energy storage system. In addition to providing the ability to inject real power on the DC side, we will show that with the correct choice of the LC filter parameters, the VA rating of the current source used for filtering can be substantially reduced compared, for example, to the direct current profiling case. A direct comparison with other proposed solutions is more difficult, since the VA ratings of the voltage source will be a function of the design of the entire system.

The parallel connected current sources can be used to interface energy storage system to inject active power; Additional 360 Hz transformer presented in approaches outlined in [8-11] is eliminated; The proposed method works even if the initial rectifier current is discontinuous;

Fig. 4. Configuration of the proposed AC/DC rectifier with current sources formed by two buck-and-boost converters

Since the current sources only need to compensate the current difference, the VA rating of the current sources will be comparable to (or less than) those in [8-11].

Fig. 3. Proposed method with parallel connected current sources

The hardware implementation of the three current sources shown in Fig. 3 is given in Fig. 4. The current sources are implemented using two cascaded buck-and-boost converters which provide two independent currents, is1 and is2. The third current, is3, results from a combination of is1 and is2 (is3= is1 is2). The DC bus of the cascaded buck-and-boost converter can be supplied by a capacitor bank, much like in active filter designs; otherwise, the DC bus can be used to interface energy storage. In comparison with the existing methods discussed in Section II, the proposed method has the following advantages:

B. System analysis and parameter optimization We propose to shape the rectifier current by inserting a current source in parallel with each six-pulse rectifier. To improve the efficiency of the proposed approach, we minimize the VA rating of these current sources through appropriate choice of filtering components. The equivalent circuit of the 12-pulse rectifier is shown in Fig. 5. Considering the DC and fundamental component, each six-pulse rectifier can be represented by a DC and AC source in series (vrec_ac and vrec_dc). Looking at the equivalent circuit of the system, the two DC voltage sources are connected in series, and will produce a DC current in the load which is also the DC component of irec . The DC load current can be calculated as: 6 Vp IL = (6) RL To attenuate the effect of vrec_ac on the DC load, an LC filter is inserted between the rectifier and the load. The inductor is formed by the leakage inductance of the transformer Lleak, and additional inductor Lf on the DC side. In our investigations, rather than using the LC filter to attenuate the effect of vrec_ac on the load, we use the LC filter to shape the rectifier output current, irec, to be as close as possible to the desired shape irec_ref , as illustrated in Fig. 6. As described by (1) and (2) the fundamental component of vrec_ac is dominant, and it is in phase with the fundamental component of the triangular current reference irec_ref, given in (1). The

Copyright (c) 2011 IEEE. Personal use is permitted. For any other purposes, permission must be obtained from the IEEE by emailing pubs-permissions@ieee.org.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication.

THD of irec_ref can be calculated based on (1), as 12.12%, which suggests that the fundamental component of irec_ref is also dominant. To ensure that vrec1 and irec1 remain in phase, the LC filter should be purely resistive (i.e. at resonance) at the frequency of interest, namely 6. Using the superposition principle, and assuming resonance in the two LC filters at 6 results in a current flow only in the respective LC filter. This is the result of the infinite impedance presented by the parallel resonance in the other branch, as depicted graphically in Fig. 7 1 for vrec1_ac; the analysis is similar for vrec2_ac. Therefore, the AC component of irec will only be determined by vrec_ac and the equivalent series resistance (ESR) of the LC filter (which is not shown in the figure). By only considering the fundamental component of vrec_ac and irec, and assuming irec is continuous, the peak value of the AC component of irec can be calculated: Vp 6 I rec _ ac ,= (7) peak 35 ESRLC where ESRLC is total ESR of the LC filter. The optimal value of ESRLC will depend on the load current, and the ESR can only be optimized for one value of load. To derive the optimal ratio between ESRLC and RL we define the parameter M as: I rec _ ac , peak m= (8) IL Solving numerically for a value of M that gives the minimum difference between the fundamental component of irec_ac and the triangular reference irec_ref, we get: m = 0.77 (9) The rectifier output current and triangular reference are shown in Fig. 6 for m=0.77. It is apparent that the two waveforms are quite close. Combining (6), (7), and (8) results in: ESRLC 1 = (10) RL 35m Based on (9) and (10) the optimized ESR can be chosen for a certain load resistance. C. Injected Current Flow Next, using the superposition principle, we analyze the current flow from the inserted current source. Based on the equivalent circuit given in Fig. 8, three paths are possible. Ideally, the current should flow primarily through path I. Assuming the worst-case where the initial rectifier output current is purely DC, the current source will need to inject: 1 (1) n (11) 4 I L 2 2 cos(6nt ) is = n n =1 Therefore, the fundamental of the current source will be at 360Hz. It is apparent that if Lf is larger than Lleak the higher order harmonics will take the desired path. At 360Hz, when Lf and Cf are in resonance at 360 Hz, only paths I and II are possible, as shown earlier. To ensure the current takes the desired path, it is preferable to minimize the transformer
1 In reality, the impedance of the LC circuit is determined by the ESR of the LC, and the resistance of the load. In practice the resulting current through the branch is negligible.

leakage inductance and ESR (which is not shown in Fig. 8) so that most of the current takes Path I. D. Control Strategy This section describes the proposed control strategy for the three current sources shown in Fig. 3. Our approach is to control the currents is1 and is2 through the two inductors in Fig. 4. The control diagram of the system is given in Fig. 10. Similarly to active power filter control a simple proportional controller is used to act on the error between the references is1_ref and is2_ref and the inductor currents is1 and is2.

Fig. 5. Equivalent circuit of the 12-pulse rectifier

Fig. 6. Rectifier output voltage and current

Fig. 7. Decoupling of the currents produced by the fundamental of vrec1_ac

Fig. 8 Equivalent circuit for the upper current source using the superposition principle

Copyright (c) 2011 IEEE. Personal use is permitted. For any other purposes, permission must be obtained from the IEEE by emailing pubs-permissions@ieee.org.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication.

The current references for the two inductors is1_ref and is2_ref are made up of three components. The first component ensures the tracking of the triangular reference. The desired triangular references irec1_ref and irec2_ref, are generated by multiplying the averaged rectifier output current (Irec1 and Irec2) and the triangular reference which is synchronized with the line voltage. The dynamics of the averaging can be fairly slow, since the result of a mismatch between the estimated and actual value of IL simply results in real power injection/absorption by current source into the load during the transient. The second component balances the voltages of the energy storage elements (batteries or capacitors) that supply the current source. Since the voltage sources (CDC1 and CDC2 in Fig. 4) represent DC signals with fairly large time constants, a proportional-integral control is used to ensure zero steady-state error. The third component is used to inject active power into the DC bus from the energy source. Overall, the control is much simpler than that of the AC-side active power filters [14, 15]. IV. SIMULATION AND EXPERIMENTAL RESULTS Two systems are presented in this paper: we analyze a large industrial 1MVA system and validate the proposed concept on a smaller experimental setup. The main limitation of the experimental setup is the high leakage inductance and the high ESR of the transformer. A. Simulation Results We model the 1MVA rectifier using Sim-Power-Systems Toolbox in Simulink. The specifications of the setup are given in Table I. The LC filter ESR is optimized according to (10) to give the minimum current source VA rating at RL = 0.5 (with RL = 0.37 corresponding to full load resistance). To assess the required current source ratings, we subtract the actual rectifier current from the ideal triangular waveform. Simulation results summarized in Fig. 11 show that the RMS current of current source normalized by the load current reaches a minimum at RL = 0.54 which is close to the designed value of RL = 0.5. The small difference comes from the omission of higher order harmonics in the original analysis, and neglected cross-coupling between LC filters due to finite ESR of the LC resonant circuit. At the point when the load resistance is 0.7, there is a mode change from continuous to discontinuous rectifier current. In order to account for the non-idealities of the control loop and the practical implementation of the current source, we simulated a switching model of the current source. The results are given in Fig. 12. The VA rating remains low, especially at optimized point. The system can be optimized at different load conditions depending on application requirements. Note that in all cases, the THD is below 5%. It is also worth noting here that the rectifier current becomes discontinuous at 0.56 (p.u.), yet the filter is still able to control the harmonics all the way down to 0.2 (p.u.). This is one of the benefits of the proposed approach compared to other approaches that shape the 12pulse rectifier output current. Additionally, the LC filter VA rating is kept quite low.

A. Experiment results A small scale experiment setup (Fig. 13) is built to verify the functionality of the proposed system. The specifications of this setup are given in Table II. The 3-phase transformer consists of three single phase transformers (VPT48-10400) and the secondary windings are rewound to get desired voltage level. The 12-pulse diode rectifier consists of two 3-phase diode bridges (SC50VB80). Two inductors are connected in series to get Lf. The filter inductors were hand-wound around a ferrite E80 core. The MOSFETs used for the buck-and-boost converters are IXFH40N50Q2. We used the TDS5034B oscilloscope from Tektronix and associated measurement and analysis software to capture the experimental waveforms and perform power quality analysis. The experiment setup has some limitations. First, based on the analysis in Section III, the setup requires a high current injection from the current sources due to the relatively high leakage inductance and ESR of the transformers.

Fig. 9. Equivalent circuit of the 12-pulse rectifier with the current sources included

Fig.10. Control diagram TABLE I PARAMETERS OF THE AC/DC RECTIFIER FOR SIMULATIONS PARAMETER Output power Transformer primary voltage (line to line RMS) Transformer secondary voltage (line to line RMS) Transformer leakage inductance (equivalent) Value 1 MW 4160 V 240 V (two in series) 5 H

Copyright (c) 2011 IEEE. Personal use is permitted. For any other purposes, permission must be obtained from the IEEE by emailing pubs-permissions@ieee.org.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication.

DC-side inductor Total ESR of Lf and Cf Rectifier output capacitor (Cf1,Cf2) Buck-and-boost switching frequency Buck-and-boost DC link voltage Buck-and-boost output filtering inductor

44.3H 18.6m 3.6 mF 20 kHz 1200 V 100 H

Fig. 11. Verification of optimized ESR

optimized point (where the VA rating is minimized) is when load current is 2.6 A (with load resistance as 25). While according to (7) and (8) and Table II, the calculated optimal point is when load resistance is 22. The VA rating is relatively high because of the high leakage inductance and ESR of the transformer. According to the current flow analysis described in the previous section, we can see that for the filter parameters from Table II, only 57.1% and 56.1% of the compensation current goes through loop I. Therefore, almost half of the compensation current goes towards reducing the LC filter current rather than towards shaping the rectifier current into the triangular reference. The active power injection function of the current source is3 is also verified in experiment. A DC source is connected to the DC link of the two buck-and-boost converters to provide active power. Fig. 19 shows the transition to active power injection. When is3 injects active power, is1 has a positive offset and is2 has a negative offset. The input line current decreases because of this active power injection. In addition, the lower current from the rectifier results in a smaller compensation current from the filter.
TABLE II PARAMETERS OF THE AC/DC RECTIFIER FOR EXPERIMENT PARAMETER VALUE Output power 270 W Transformer primary voltage (L-L RMS) 60 V Transformer secondary voltage (L-L RMS) 25 V (two in series) Transformer leakage inductance 0.3 mH ESR of upper/lower transformer (equivalent) (0.28/0.31) DC-side inductor (Lf1,Lf2) 2.6 mH Rectifier output capacitor (Cf1,Cf2) 66 F ESR of Lf1/ Cf1/ Lf2/ Cf2 (0.52/0.015/0.48/0.015) Buck-and-boost switching frequency 20 kHz Buck-and-boost DC link voltage 130 V Buck-and-boost output filtering inductor 2 mH

(a) THD of line current

(b) Combined VA rating of the current sources Fig.12. Simulation results with switching model of proposed system

Fig. 14 shows the experimentally obtained waveforms with and without the compensation current when the load current IL is 4.2A. Fig. 15 shows the compensation currents is1 and is2 at the same load. The improvement of the line current harmonic level is apparent: the THD drops from 4.60% to 2.41%. The magnitude of each harmonic is plotted in Fig. 16. It is interesting to note that even the initial harmonics are fairly low and much lower than for the 12-pulse rectifier with LC filter design for DC ripple minimization. This is due to the choice of the LC filter that amplifies the 360Hz harmonic in the current. In Fig. 14, results show that the rectifier current is successfully shaped into the triangular form. Further comparisons are given in Fig. 17, with different load conditions. Fig. 17 shows that the proposed method is effective at reducing the harmonics in the input line current even as the rectifier output current becoming discontinuous. The VA rating of the current sources at different load conditions are given in Fig. 18. It can be seen that the

Fig. 13. Experimental Setup

Copyright (c) 2011 IEEE. Personal use is permitted. For any other purposes, permission must be obtained from the IEEE by emailing pubs-permissions@ieee.org.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication.

Ch1: line voltage (phase A), 50V/Div Ch2: line current (phase A), 2A/Div Ch3: irec1, 5A/Div Ch4: irec2, 5A/Div (a) Without rectifier current profiling

Fig.17. Comparison of THD with and without current profiling

Fig.18. VA rating of the current sources is1+is2

Ch1: line voltage (phase A), 50V/Div Ch2: line current (phase A), 2A/Div Ch3:irec1, 5A/Div Ch4: irec2, 5A/Div (b) With rectifier current profiling Fig.14.Comparison of line current and rectifier output current

Ch2: line current (phase A), 2A/Div Ch3: is1, 2A/Div Ch4: is2, 2A/Div Fig. 19. Active power injection

V. CONCLUSIONS
Ch1: line voltage (phase A), 50V/Div Ch2: line current (phase A), 2A/Div Ch3:is1, 2A/Div Ch4: is2, 2A/Div Fig.15. Compensation current when the load current IL is 4.2A

Fig. 16. Harmonics distribution with and without the current profiling when load current IL is 4.2A

In this paper we presented a new method for active profiling of the 12-pulse rectifier output current. The chosen topology acts as a current source in parallel with the input rectifier, and is therefore not in the path of the load current. In addition, the current source is constructed from a voltage fed buck-andboost converter, allowing for seamless integration of dc storage into the system. The VA ratings of the current profiling circuitry is further minimized by exploiting the resonance of the DC-side LC filter. In comparison with similar approaches, the proposed method has apparent advantages of eliminating the use of low frequency transformers for shaping the rectifier output current, providing energy storage interface to the system, and functioning regardless of the continuality of the initial rectifier output current. Simulation of a high power industrial AC/DC

Copyright (c) 2011 IEEE. Personal use is permitted. For any other purposes, permission must be obtained from the IEEE by emailing pubs-permissions@ieee.org.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication.

rectifier verified the parameter optimization of the system, while experimental results show significant reduction of harmonics in the line current. The DC energy storage integration functionality is also verified experimentally. VI. REFERENCES
[1] V. Sithimolada and P. W. Sauer, "Facility-level DC vs. typical ac distribution for data centers: A comparative reliability study," in Proc. TENCON 2010, 2010, pp. 2102-2107. B. Sanzhong, Y. Du, and S. Lukic, "Optimum design of an EV/PHEV charging station with DC bus and storage system," in Proc. IEEE Energy Conversion Congress and Exposition, 2010, pp. 1178-1184. Z. He, F. Mollet, C. Saudemont, and B. Robyns, "Experimental Validation of Energy Storage System Management Strategies for a Local DC Distribution System of More Electric Aircraft," IEEE Trans. on Ind. Electronic, vol. 57, pp. 3905-3916, 2010. "IEEE Recommended Practices and Requirements for Harmonic Control in Electrical Power Systems," IEEE Std 519-1992, p. 0_1, 1993. L. An, T. Ci, S. Zhi Kang, Z. Wei, R. Fei, and Z. Ke, "A Novel Three-Phase Hybrid Active Power Filter With a Series Resonance Circuit Tuned at the Fundamental Frequency," IEEE Trans. Ind. Electron., vol. 56, pp. 2431-2440, 2009. A. Bhattacharya, C. Chakraborty, and S. Bhattacharya, "Parallel Connected Shunt Hybrid Active Power Filters Operating at Different Switching Frequencies for Improved Performance," IEEE Trans. Ind. Electron., vol. PP, pp. 1-1, 2011. S. Ostroznik, P. Bajec, and P. Zajec, "A Study of a Hybrid Filter," IEEE Trans. Ind. Electron., vol. 57, pp. 935-942, 2010. Y. Liu, N. Watson, and J. Arrillaga, "A new concept for the control of the harmonic content of voltage source converters," in Proc. IEEE Conf. Power Electronics and Drive Systems 2003, pp. 793-798 Vol. 1. S. Choi, P. N. Enjeti, H. H. Lee, and I. J. Pitel, "A new active interphase reactor for 12-pulse rectifiers provides clean power utility interface," IEEE Trans. Ind. Appl., vol. 32, pp. 1304-1311, 1996. S. Miyairi, S. Iida, K. Nakata, and S. Masukawa, "New method for reducing harmonics involved in input and output of rectifier with interphase transformer," IEEE Trans. Ind. Appl., pp. 790-797, 1986. S. Fukuda, M. Ohta, and Y. Iwaji, "An auxiliary-supply-assisted harmonic reduction scheme for 12-pulse diode rectifiers," IEEE Trans. Power Electronics, vol. 23, pp. 1270-1277, 2008. M. E. Villablanca, J. I. Nadal, and M. A. Bravo, "A 12-pulse AC DC rectifier with high-quality input/output waveforms," IEEE Trans. Power Electronics, vol. 22, pp. 1875-1881, 2007. T. Tanaka, N. Koshio, H. Akagi, and A. Nabae, "A novel method of reducing the supply current harmonics of a 12-pulse thyristor rectifier with an interphase reactor," in Proc. IEEE Industry Applications Conference, 1996, pp. 1256-1262 vol. 2. C. Po-Tai, S. Bhattacharya, and D. M. Divan, "Application of dominant harmonic active filter system with 12 pulse nonlinear loads," IEEE Trans. Power Delivery, vol. 14, pp. 642-647, 1999. J. H. Sung, S. Park, and K. Nam, "New hybrid parallel active filter configuration minimising active filter size," Proc. IEE Electric Power Applications Conf., vol. 147, pp. 93-98, 2000. Sanzhong Bai (S09) received the B.S. and M.S. degree in Electrical Engineering from Sichuan University, Chendu, China, in 2003 and 2006, respectively. Between 2006 and 2008 he was a Research Assistant in the Electric and Technology Institute of Sichuan University. In 2008 he joined the Future Renewable Electric Energy Delivery and Management (FREEDM) systems center at North Carolina State Universiy. Currently he is a Research Assistant and working towards a Ph.D degree. His primary areas of interest are Electric Vehicle charging station, resonant power

conversion and power quality issues. Srdjan M. Lukic (S02, M07) received the M.S. and Ph.D. degrees in electrical engineering from the Illinois Institute of Technology. From 2002 to 2004, he was with Firefly Energy Inc., where he was responsible for optimizing certain aspects of carbon/graphite foam-based lead acid batteries for novel automotive applications. Currently he is an assistant professor in the department of electrical and computer engineering at North Carolina State University, Raleigh. He serves as the distributed energy storage devices subthrust leader at the Future Renewable Electric Energy Delivery and Management (FREEDM) Systems Engineering Research Center. Dr. Lukic served as a Guest Editor of the IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS; and serves as Associate Editor of the IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS. His research interests include design, and control of power electronic converters and electromagnetic energy conversion with application to wireless power transfer, energy storage systems and electric automotive systems.

[2]

[3]

[4]

[5]

[6]

[7] [8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

Copyright (c) 2011 IEEE. Personal use is permitted. For any other purposes, permission must be obtained from the IEEE by emailing pubs-permissions@ieee.org.