Advanced
A Load Compensator Based on One-Cycle Control with Plug-In Repetitive Control
A Load Compensator Based on One-Cycle Control with Plug-In Repetitive Control
Journal of Power Electronics. 2015. Mar, 15(2): 443-454
Copyright © 2015, The Korean Institute Of Power Electronics
  • Received : July 16, 2014
  • Accepted : December 02, 2014
  • Published : March 20, 2015
Download
PDF
e-PUB
PubReader
PPT
Export by style
Share
Article
Author
Metrics
Cited by
TagCloud
About the Authors
Jian Hu
College of Electrical Engineering, Zhejiang University, Hangzhou, China
Zhaohui Sun
College of Electrical Engineering, Zhejiang University, Hangzhou, China
Hao Ma
College of Electrical Engineering, Zhejiang University, Hangzhou, China
Guozhu Chen
College of Electrical Engineering, Zhejiang University, Hangzhou, China
gzchen@zju.edu.cn

Abstract
This study proposes a novel one-cycle control scheme with a plug-in repetitive controller for load compensator. The novelty of this scheme lies in the combination of high dynamics and the simplicity of a one-cycle controller and good steady-state harmonic suppression ability of the repetitive controller. In addition, the proposed scheme can reduce the effect of the harmonics in phase voltage for the existence of the repetitive controller. Finally, experimental results on a three-phase, four-wire, three-level load compensator are reported to validate the effectiveness of the proposed control scheme.
Keywords
I. INTRODUCTION
Development of the power system has resulted in a huge growth in nonlinear loads, such as adjustable speed drives, electric arc welders, and switching power supplies, which usually generate large amounts of harmonic currents in both industrial and domestic fields. Consequently, line current harmonics is one of the greatest challenges in power systems because it causes voltage distortion and increases power loss [1] , [2] .
An easy solution to this problem is to employ a passive filter. However, such filters are often heavy, large in size, and can be easily affected when the impedance of the grid changes [3] . In contrast, a load compensator (LC) is a better solution because of its smaller size and weight compared with the passive filter, as well as its insensitiveness to the grid parameters [1] - [3] .
The goal of LC is to generate a current with the same magnitude but the opposite phase to the harmonic that the nonlinear loads generated. A typical controller for LCs consists of two parts: a harmonic extractor that extracts the harmonic components of load currents to generate the reference for the current controller, and a double-loop controller that forces the output current of the LC to follow the reference and controls the DC-link voltage [4] , [5] . Therefore, both current controller and harmonic extractor can affect LC performance. The harmonic extractor has already been widely employed by filters [6] - [8] ; schemes based on the instantaneous reactive power theory [9] or Discrete Fourier Transformation (DFT) [10] . However, all these schemes have problems in balancing the steady-state and transient performance of its output. Thus, the harmonic extractor degrades both the accuracy and the transient behavior of LCs.
Alternatively, an indirect method is introduced in [11] - [13] , where the line current is sensed and sinusoidally regulated as the same phase of the grid voltage. Without the harmonic extractor, this scheme is free from tracking errors and delays that the harmonic extractor causes and can be implemented with only one set of current sensors. However, an additional set of high-performance voltage sensors for the grid voltage is needed because the reference of the current is generated directly from the grid [12] or from the d-q reference frame that needs the information from the grid [13] .
One-cycle control (OCC) based LCs [14] - [16] do not explicitly require information on the grid and load current; thus, these LCs can be implemented with only line current sensors and DC-link voltage sensors. Moreover, semiconductor devices operate at a constant switching frequency, which simplifies the design of the output filter. However, the control bandwidth of OCC based LC is limited because its bandwidth is equivalent to a proportional controller that forces the current to follow the waveform of the grid voltage. A detailed model of OCC based LC will be discussed in section II. Therefore, the OCC based LC cannot compensate high order harmonics, and can be easily affected by distorted grid voltage.
The design of the current controller becomes a challenging task because LCs need to generate a high-order harmonics current. The bandwidth of the current loop has to be sufficiently wide. In the literature, various current controllers have been developed, such as proportional-integral (PI) control [17] , [18] , hysteresis control [19] , [20] , proportional-resonant (PR) control [21] , [22] , and vector PI (VPI) control [23] , [24] . The bandwidth of PI control is limited, which is making it unsuitable for LCs. Although hysteresis control is simple and robust, it also has flaws—the varying switching frequency poses huge difficulty difficulties in the design of the output filter. Moreover, hysteresis control also has problems in balancing the accuracy of the current output and switching frequency. The PR controller can provide a high bandwidth for the current controller with a constant switching frequency. However, undesired frequency peaks usually appear in the closed-loop frequency response because of the delay in the digital implementation of the controller [23] . In addition, the complexity and computation burden also increase significantly when tracking high-order harmonics because each resonant controller can only regulate one harmonic component. In [23] , a VPI controller based on a fundamental reference frame is introduced. This new scheme can greatly reduce the number of resonant controllers and eliminate the peaks from delay. Unfortunately, this scheme can only work under a three-phase three-wire system with a balanced load, and the performance of the controller under L-C-L filter is has not yet been reported.
Repetitive controller based on internal-mode principals is introduced in [25] . The repetitive controller can regulate all harmonics with one controller because the harmonics of line current are periodic signals, and the delay caused by the digital implementation can be easily compensated. Applying repetitive controllers to LCs under a typical structure with two parts have been attempted many times [26] , [27] . However, as discussed before, LC performance is still degraded by the harmonic extractor.
In this paper, a modified one-cycle control scheme with plug-in repetitive controller is proposed. By combining the two controllers, the proposed scheme is capable of greatly increasing the harmonic rejection ability of the LC. In addition, the repetitive controller can reduce the effect of harmonics from the phase voltage. The desired features of one-cycle controller is retained because the main structure of the system is still OCC based—no phase lock loop (PLL) and harmonic extractor are needed for the LC, and the switching device works at a constant frequency.
The remaining parts of this paper are organized as follows: The configuration and design of the proposed controller is discussed in Section II. A comparison between the performance of the proposed controller and that of the existing OCC controller are given in Section III. The description of the experimental setup and the result of experiments are given and summarized in Section IV. Finally, the main conclusions are highlighted in Section V.
II. PROPOSED REPETITIVE CONTROLLER
- A. System Configuration
A typical circuit of neutral point clamp (NPC) based on a three-level three-phase four-wire load compensator with an input L-C-L or L filter is shown in Fig. 1 . Repetitive control is an excellent way to eliminate all the harmonics from the system because all the harmonics in the AC system are periodic. The control system configuration is depicted in Fig. 2 . A band-pass filter is used to extract the fundamental components of the line current iref , and the error signal for the repetitive controller is generated by subtracting the line current with its fundamental components. The repetitive control is separated from the OCC current loop because the repetitive controller is used as a “plug-in” structure and only deals with the error left by the one-cycle control so that fast dynamic performance can be obtained.
PPT Slide
Lager Image
Set up of the system.
PPT Slide
Lager Image
Control system configuration.
- B. Model of Current Loop of Existing OCC Scheme
To design the repetitive controller of the proposed OCC based LC, the basic working principle of the existing OCC based LC is first explained and the model of its current loop is established.
A three-phase four-wire system can be regarded as three independent single-phase systems. Therefore, one such system is used as an example for analysis, as shown in Fig. 4 .
One-cycle control is based on a double-loop structure, which is similar to other control schemes. The voltage loop regulates two DC voltages with two PI controllers, and the output of these voltages is defined as V m 1 and V m 2 . The two sawtooth waveforms with peaks of V m 1 and V m 2 are generated using an integrator with a reset and an adder ( Fig. 3 ). Similar to the two-level case present in [28] , the key control equation of the existing one-cycle controller is given in
PPT Slide
Lager Image
where d 1 and d 2 represent the duty ratios of high output and low output of an inverter leg, which represents the line current. A detailed implementation of one-cycle control scheme for a three-phase, four-wire three-level LC is given in Appendix A.
PPT Slide
Lager Image
Control block diagram of existing OCC scheme.
PPT Slide
Lager Image
Model of one phase in a three-phase four-wire load compensator
Output voltage vo of the inverter is expressed using an average model as follows:
PPT Slide
Lager Image
where U dc 1 and U dc 2 represent the voltages of the upper and lower capacitors respectively.
It can be assume that the outputs of the two voltage loops are constants when analyzing the current loop because the bandwidth of the voltage loop is low compared with the current loop. Thus, as shown in Appendix A, the output of the voltage loop is given in
PPT Slide
Lager Image
The emulated resistance R e represents the effective per-phase load resistance seen by the AC source.
Combining Eqs. (1), (2), and (3) yields
PPT Slide
Lager Image
Eq. (4) shows that the magnification of one-cycle control is R e . Thus, the control diagram of the current loop is depicted in Fig. 5 (a), where G 0 represents the transfer function of the filter.
PPT Slide
Lager Image
Control diagram of the existing one-cycle control.
If the grid voltage v s has no distortion, then the active power balance of the whole system yields
PPT Slide
Lager Image
where isf represents the real part of the fundamental component of i s . Magnification of the one-cycle control is also R e , and thus the control diagram can be transformed into Fig. 5 (b) through replacing v s with i sf . Therefore, the current loop of the existing OCC is a proportional controller that lacks the ability to track high-order harmonics.
Fig. 5 (b) shows that v s works as a reference of the current system loop, and the nonlinear load current il is regarded as the disturbance. Thus, if the grid voltage v s is distorted, then the performance will be greatly degraded because the control system cannot suppress the harmonic component of phase voltage v h , which is injected from the reference as shown in Fig. 5 (c). Consequently, the controller has no ability to suppress the error from the distorted phase voltage.
- C. Design of Plug-in Repetitive Controller
The current loop of Fig. 2 can be redrawn in the z-domain with additional details, as shown in Fig. 6 (a). The phase difference between the fundamental of the line current and phase voltage can be neglected because the one-cycle controller exhibits good performance at the fundamental frequency of the system. Thus, the output of band-pass filter iref and the real part of the fundamental component isf can be combined. The control system can be simplified through combining these two components, as shown in Fig. 6 (b). The load current is treated as a disturbance. This structure is a classical plug-in repetitive control. The plant of the repetitive controller P ( z ) can be considered as a close-loop transfer function of one-cycle control as [29] shows:
PPT Slide
Lager Image
PPT Slide
Lager Image
Control diagram of the proposed control.
The parameters of LC is is listed in TABLE I and the design of the repetitive controller is based on them.
PARAMETERS OF LOAD COMPENSATOR
PPT Slide
Lager Image
PARAMETERS OF LOAD COMPENSATOR
As the core of the repetitive controller, the modified internal model 1/(1- Q ( z ) z -N ), which integrates the error on cycle bases, is designed first. The sampling frequency of the load compensator is 10 kHz, and the fundamental frequency of the system is 50 Hz. One repetitive control cycle contains N = 200 error sampling. The Q ( z ) filter, which can increase the robustness of the repetitive control, is set as a low-pass filter or a constant near 1. In this case, Q ( z ) is set as 0.9.
Repetitive control can achieve high gain under the fundamental and harmonic frequencies, but such high gain on high frequency may affect the system stability. Therefore, a filter F ( z ) is added to eliminate the error signal from the bandwidth of the current loop and thus maintain the stability of the controller. Due to the existence of the delay stage z-N which makes the control system causal, an advance stage zk is added together with F ( z ) to the diagram to compensate for the phase lag of P ( z ) and F ( z ).
Various R e should be considered because the response of the one-cycle control is determined using R e , which is based on the nonlinear load. Fig. 7 shows that advance stage z 2 can compensate the phase lag of P ( z ) within the bandwidth of the current loop under various R e . To evaluate the stability of the proposed control scheme, the transfer function from error signal i err to reference signal i sf can be expressed as
PPT Slide
Lager Image
PPT Slide
Lager Image
Phase-frequency characteristic of P(z) and z-2.
The characteristic equation of the system is
PPT Slide
Lager Image
To achieve system stability, all the roots of the above must be inside the unity circle that is centered at the origin of the z-plane. A sufficient condition for system stability is presented in [30] :
PPT Slide
Lager Image
Define H ( ejωt )= Q ( ejωt )- F ( ejωt ) ejωt P ( ejωt ). Eq. (9) means that the end of vector H ( ejωt ) should never exceed the unity circle. The Nyquist plot of H ( ejωt ) under various R e are given in Fig. 8 . The entire locus remains well within the unity circle, as seen in Fig. 8 .
PPT Slide
Lager Image
Nyquist plot of H(ejωt) under different Re.
- D. DC-Link Voltage Control Loop
The goal of the outer DC-link voltage loop is to keep the DC-link voltage of the LC constant through a simple PI regulator, whose output is the amplitude of the sawtooth waveforms V m 1 , V m 2 as follows:
PPT Slide
Lager Image
where Kpdc and Kidc are the proportional and integrator gains of the PI controller, Udc * is the reference of the DC-link voltages, and U dc 1 and U dc 2 are the measured DC-link voltages. The model of the voltage loop given in Appendix B can be expressed as
PPT Slide
Lager Image
A and ω 0 can be expressed as
PPT Slide
Lager Image
where Vs is the root mean square (RMS) value of the phase voltage, and Udc and Vm are the quiescent operation points of he voltage loop.
Based on the instructions in [31] , the proportional gain of the controller Kpdc is derived as
PPT Slide
Lager Image
to transform the bandwidth voltage loop into fc . The corner frequency of the PI controller is set as fc /10 to achieve sufficient phase margin for the controller. Thus, the integrator gain of the controller is derived as
PPT Slide
Lager Image
- E. Frequency Adjustment
Harmonic rejection ability of the repetitive controller can be degraded because the grid frequency varies in a narrow range. Thus, a frequency adjustment block should be added to adjust the sampling time and make the error samplings per period N a constant when the grid frequency changes.
The filtered current iref is compared with 0 to reshape the waveform into a square wave. A timer detects the high-level length of the square, which is equal to half of the grid period. The output of the timer is then divided with N /2, which is the right number of samplings per half period. After the division, the result is set as the new sampling time. Frequency adjustment block is shown in Fig. 9 .
PPT Slide
Lager Image
Frequency adjustment block.
III. PERFORMANCE OF PROPOSED CONTROL
- A. Harmonic Rejection Ability
The transfer function of the repetitive controller in Fig. 6 (b) is
PPT Slide
Lager Image
The transfer function of existing OCC in Fig. 5 (b) is
PPT Slide
Lager Image
The open loop gain of | G ( z )| and | Grp ( z ) G ( z )| are compared in Fig. 10 , which indicates that the proposed controller provides higher gains at harmonic frequencies than the existing one-cycle controller. Therefore, the proposed controller has better harmonic suppressing ability than the existing one-cycle controller.
PPT Slide
Lager Image
Open loop gain of |Grp(z)G(z)| (blue) and |G(z)|(red).
- B. Sensitivity to Phase Voltage Distortion
Section II shows that the existing one-cycle controller lacks the ability to suppress distortion from the phase voltage. Investigating the effect of voltage harmonics on the proposed control is necessary. The input of the repetitive controller is from the band-pass filter, and thus contains no harmonics from the grid. To understand the suppression ability of the repetitive controller, the phase voltages are separated into two parts: fundamental component vsf and harmonic component vsh . The former, which contains no harmonics, is combined with the output of the band-pass filter. The diagram of the control blocks is shown in Fig. 11 .
PPT Slide
Lager Image
Control diagram of repetitive control under distorted line voltage.
The transfer function M ( z ) from harmonic voltage vsh to error signal ierr can be expressed as
PPT Slide
Lager Image
The transfer function of M ( z ) in the existing one-cycle control in Fig. 5 c can be expressed as
PPT Slide
Lager Image
The amplitude-frequency characteristics of M(z) is drawn in Fig. 12 ( R e =1). M ( z ) has a smaller amplitude with the repetitive controller under harmonic frequencies. This characteristic indicates that most of the distortion from the harmonic component in phase voltage can be suppressed using repetitive control.
PPT Slide
Lager Image
Magnitude-frequency between ierr and vsh.
PPT Slide
Lager Image
Laboratory prototype.
IV. EXPERIMENTAL RESULTS
- A. Experimental Setup
A scaled-down laboratory prototype of LC was built to test the performance of the proposed control. This prototype was tested on a three-phase system that had a phase voltage of 80 V. The reference voltage of both DC links is set as 160 V. To test the performance of the proposed control under a distorted voltage and its ability to suppress the effect of grid frequency variations, a Chroma programmable AC source 61512 is used as power source for the whole system. More detailed parameters used for this prototype can be found in TABLE I .
The nonlinear load of the system is a three-phase diode rectifier with two resistors on the DC-side. The neutral point of the rectifier is connected to the system, as shown in Fig. 14 .
PPT Slide
Lager Image
Nonlinear Load.
The controller is programmed with an existing one-cycle control Fig. 3 and a proposed control as Fig. 2 with the same switching frequency. The performance of both control schemes are compared under two scenarios: ideal phase voltage and distorted phase voltage.
- B. Time and Spectrum Results under Ideal Phase Voltage
A set of experiments was conducted under ideal phase voltage to compare the performance of the two control schemes. Under this condition, the results are shown in Fig. 15 with voltage and currents of phase A. The existing one-cycle control cannot handle the step change of the line current because of the limited bandwidth in the current loop. Therefore, spikes appear in the waveform of line current, as shown in Fig. 15 (a). The proposed control scheme has a high gain under harmonics through employing a plug-in repetitive controller. Thus, the current spikes are eliminated using the proposed control scheme, as shown in Fig. 15 (b).
PPT Slide
Lager Image
Comparison of existing one-cycle control and proposed control.
Fig. 16 compares the current spectra obtained from the two control schemes with the International Electrotechnical Commission (IEC) standard. The 11th- and 13th-order of harmonics in the existing OCC scheme exceeds the limitation of the IEC standard, whereas the current spectra of the proposed control meet the requirement of the IEC standard, as shown in Fig. 16 . The line current total harmonic distortion (THD) is significantly reduced from 14.8% to 3.6% using the proposed control scheme. The proposed scheme presents better harmonic-elimination capability over the existing control and can reduce the amplitudes of low-order harmonics, which the existing control cannot do.
PPT Slide
Lager Image
Measured current spectra at ideal phase voltage.
- C. Time and Spectrum Result under Distorted Phase Voltage
In practical circumstances, especially in low-voltage applications, phase voltage is normally distorted. Thus, a set of experiments was performed to compare the performance of the two controls under distorted phase voltage.
The experiments are from a Chroma 61612 programmable three-phase AC source. A 5th order of harmonics whose amplitude is 20% of the fundamental is then injected into the base voltage.
Fig. 17 demonstrates the results of the experiments. The existing one-cycle control almost loses the ability of harmonic suppression. In addition to the spikes, a large amount of 5th-order harmonics also appears in the current waveform. The proposed control can cancel out the effect of phase voltage harmonics using a repetitive controller that leads to good performance. Little spikes and 5th-order harmonics appear in the current waveform. Fig. 18 compares the spectra of the line current of the two controls with the IEC standard. For the existing control scheme, the amplitude of the 5th-order harmonics of the existing control is 29% and greatly exceeds the IEC standard, which is mostly from the injected voltage harmonics that the control scheme does not suppress. Under the proposed control scheme, the amplitude of the 5th-order harmonics is reduced to 3.5%, which meets the IEC standard. The THD of the line current is reduced from 36% to 7.2% through replacing the existing one-cycle control with the proposed control scheme. The proposed control shows better rejection ability of phase distortion than the existing OCC and can work effectively even with a highly distorted phase voltage.
PPT Slide
Lager Image
Comparison of existing one-cycle control and proposed control under distorted phase voltage.
PPT Slide
Lager Image
Measured current spectra at distorted phase voltage.
- D. Frequency Variation
A set of experiments was conducted with different grid frequencies to test the harmonic rejection ability of the proposed control under various grid frequencies. The results of the experiments can be found in Fig. 19 .
PPT Slide
Lager Image
Test of the proposed controller under various grid frequencies.
The results show that the proposed control scheme retains harmonic rejection ability when grid frequency varies. The THD of the line current is less than 3.5%, which suggests that the proposed control has the ability to adjust the variation of the grid frequency.
- E. Dynamic Performance
The dynamic performance of the proposed controller is also evaluated under step load changes. Fig. 20 shows the transient response of line current is . A step change of the nonlinear load from 50% to 100% is performed. In the first cycle after the transient of load current, the system behaves similarly to the existing one-cycle control, and the amplitude quickly changes because the repetitive controller needs one cycle for output change. Moreover, the repetitive control does not affect the fast dynamic of the on-cycle control. Finally, the repetitive controller begins to change its output, and the line current soon becomes sinusoidal.
PPT Slide
Lager Image
Dynamic performance of the system.
V. CONCLUSION
A novel one-cycle control scheme for a three-phase, four-wire, three-level load compensator based on a plug-in repetitive controller was proposed. The voltage loop of this scheme is similar to the existing one-cycle control, in which a plug-in repetitive controller is added to the current loop. The detailed design of parameters of the repetitive controller is presented in this study through analyzing the loop gain and phase shift of the one-cycle control. A series of experiments was conducted to evaluate the performance of the proposed controller. A comparison between the performance of the proposed controller and that of the existing OCC confirmed that the former offers a superior harmonic suppression capability and lower sensitivity to the phase voltage distortion than the existing one-cycle controller, which suggests that the proposed scheme may be a good candidate under such condition.
Acknowledgements
This work is supported by the National Natural Science Foundation of China (NSFC) (#51177147 and #51177149).
BIO
Jian Hu was born in Nanjing, China, in 1990. He received his B.S. in Electrical Engineering degree from Southeast University, Nanjing, China. He is currently working toward his Ph.D. degree in the College of Electrical Engineering, Zhejiang University, Hangzhou, China. His research interests include power electronics equipment, renewable energy, power quality and parallel operation of the inverters.
Zhaohui Sun was born in Zhejiang, China. He received his B.S. degree in Electrical Engineering from Zhejiang University, Zhejiang, China, in 2013, and is currently working toward his Ph.D. degree at the same university. His current interests include the application of advanced control methodologies in power electronic converters and the application of the DC micro-grid.
Hao Ma was born in Hangzhou, China, in 1969. He received his B.Sc., M.Sc., and Ph.D. degrees in Electrical Engineering from Zhejiang University, Hangzhou, China, in 1991, 1994, and 1997, respectively. He is currently a Professor at the College of Electrical Engineering, Zhejiang University. From September 2007 to September 2008, he was a Delta Visiting Scholar at North Carolina State University, Raleigh, USA. His research interests include advanced control in power electronics, fault diagnosis of power electronic circuits and systems, and application of power electronics.
Guozhu Chen (M’03) was born in Hubei, China, in 1967. He received his B.S. degree in Electrical Engineering from Hangzhou Commerce University, Hangzhou, China, in 1988, and the M.S. and Ph.D. degrees in Electrical Engineering from Zhejiang University, Hangzhou, China, in 1992 and 2001 respectively. Since 1992, he has been with the faculty of the College of Electrical and Engineering at Zhejiang University, China, where he was an Associate Professor from 2000 to 2005, and a Professor since 2005. From January 2001 to April 2004, he was a Visiting Scholar in the University of California, Irvine, USA. His current research interests include high-power electronics applications and their digital control; active power quality control such as APF, UPQC, SVC, dSTATCOM and dFACTS; grid connection of renewable energy/distributed power generation; and power electronic system integration. He holds more than 20 Chinese patents, and has contributed to more than 160 academic papers.
References
Green T. C. , Marks J. H. “Control techniques for active power filters” in Proc. IEE Electric Power Applications 2005 Vol. 152, No. 2 369 - 381
Akagi H. 2002 “Trends in active power line conditioners,” IEEE Trans. Power Electron. 9 (3) 263 - 268    DOI : 10.1109/63.311258
Watanabe E. H. , Aredes M. , Akagi H. 2007 Instantaneous Power Theory and Applications to Power Conditioning John Wiley & Sons chap. 2
Akagi H. , Kanazawa Y. , Nabae A. 1984 “Instantaneous reactive power compensators comprising switching devices without energy storage components,” IEEE Trans. Ind. Appl. IA-20 (3) 625 - 630    DOI : 10.1109/TIA.1984.4504460
Choi W.-H. , Lam C.-S. , Wong M.-C. , Han Y.-D. 2013 “Analysis of DC-link voltage controls in three-phase four-wire hybrid active power filters,” IEEE Trans. Power Electron. 28 (5) 2180 - 2191    DOI : 10.1109/TPEL.2012.2214059
Rahmani S. , Mendalek N. , Al-Haddad K. 2010 “Experimental design of a nonlinear control technique for three-phase shunt active power filter,” IEEE Trans. Ind. Electron. 57 (10) 3364 - 3375    DOI : 10.1109/TIE.2009.2038945
Cirrincione M. , Pucci M. , Vitale G. , Miraoui A. 2009 “Current harmonic compensation by a single-phase shunt active power filter controlled by adaptive neural filtering,” IEEE Trans. Ind. Electron. 56 (8) 3128 - 3143    DOI : 10.1109/TIE.2009.2022070
Kanieski J. M. , Cardoso R. , Pinheiro H. , Grundling H. A. 2013 “Kalman filter-based control system for power quality conditioning devices,” IEEE Trans. Ind. Electron. 60 (11) 5214 - 5227    DOI : 10.1109/TIE.2012.2226412
Herrera R. S. , Salmeron P. , Hyosung K. 2008 “Instantaneous reactive power theory applied to active power filter compensation: Different approaches, assessment, and experimental results,” IEEE Trans. Ind. Electron. 55 (1) 184 - 196    DOI : 10.1109/TIE.2007.905959
Chandrasekaran S. , Ragavan K. “Sliding DFT assisted instantaneous symmetrical components method for estimating reference current to active power filter,” in Proc. MWSCAS 2012 1168 - 1171
Corasaniti V. F. , Barbieri M. B. , Arnera P. L. , Valla M. I. 2009 “Hybrid active filter for reactive and harmonics compensation in a distribution network,” IEEE Trans. Ind. Electron. 56 (3) 670 - 677    DOI : 10.1109/TIE.2008.2007997
Singh B. N. , Singh B. , Chandra A. , Rastgoufard P. , Al-Haddad K. 2007 “An improved control algorithm for active filters,” IEEE Trans. Power Del. 22 (2) 1009 - 1020    DOI : 10.1109/TPWRD.2006.886790
Singh M. , Khadkikar V. , Chandra A. 2011 “Grid synchronization with harmonics and reactive power compensation capability of a permanent magnet synchronous generator-based variable speed wind energy conversion system,” IET Power Electron. 4 (1) 122 - 130    DOI : 10.1049/iet-pel.2009.0132
Sreeraj E. S. , Prejith E. K. , Chatterjee K. , Bandyopadhyay S. 2014 “An active harmonic filter based on one-cycle control,” IEEE Trans. Ind. Electron. 61 (8) 3799 - 3809    DOI : 10.1109/TIE.2013.2286558
Taotao J. , Smedley K. M. 2006 “Operation of one-cycle controlled threephase active power filter with unbalanced source and load,” IEEE Trans. Power Electron. 21 (5) 1403 - 1412    DOI : 10.1109/TPEL.2006.880264
Chongming Q. , Smedley K. M. 2003 “Unified constant-frequency integration control of three-phase standard bridge boost rectifiers with power factor correction,” IEEE Trans. Ind. Electron. 50 (1) 100 - 107    DOI : 10.1109/TIE.2002.804980
Zhong C. , Yingpeng L. , Miao C. 2012 “Control and performance of a cascaded shunt active power filter for aircraft electric power system,” IEEE Trans. Ind. Electron. 59 (9) 3614 - 3623    DOI : 10.1109/TIE.2011.2166231
He Y. J. , Liu J. J. , Wang Z. A. , Zou Y. P. 2014 “A pi control algorithm with zero static misadjustment for tracking the harmonic current of three level APFs,” Journal of Power Electronics 14 (1) 175 - 182    DOI : 10.6113/JPE.2014.14.1.175
Lam C.S. , Wong M.C. , Han Y.D. 2012 “Hysteresis current control of hybrid active power filters,” IET Power Electron. 5 (7) 1175 - 1187    DOI : 10.1049/iet-pel.2011.0300
Suresh Y. , Panda A.K. , Suresh M. 2012 “Real-time implementation of adaptive fuzzy hysteresis-band current control technique for shunt active power filter,” IET Power Electron. 5 (7) 1188 - 1195    DOI : 10.1049/iet-pel.2011.0371
Yi H. , Zhuo F. , Li Y. , Zhang Y. J. , Zhan W. D. 2013 “Comparison analysis of resonant controllers for current regulation of selective active power filter with mixed current reference,” Journal of Power Electronics 13 (5) 861 - 876    DOI : 10.6113/JPE.2013.13.5.861
Lascu C. , Asiminoaei L. , Boldea I. , Blaabjerg F. 2009 “Frequency response analysis of current controllers for selective harmonic compensation in active power filters,” IEEE Trans. Ind. Electron. 56 (2) 337 - 347    DOI : 10.1109/TIE.2008.2006953
Quoc-Nam T. , Hong-Hee L. 2013 “An advanced current control strategy for three-phase shunt active power filters,” IEEE Trans. Ind. Electron. 60 (12) 5400 - 5410    DOI : 10.1109/TIE.2012.2229677
Lascu C. , Asiminoaei L. , Boldea I. , Blaabjerg F. 2007 “High performance current controller for selective harmonic compensation in active power filters,” IEEE Trans. Power Electron. 22 (5) 1826 - 1835    DOI : 10.1109/TPEL.2007.904060
Sato H. , Sueno T. , Toyama T. , Mikawa M. , Toda T. , Matsumoto S. “High accuracy magnet power supply for proton synchrotron by repetitive control,” in Proc. PESC 1991 812 - 816
Garcia-Cerrada A. , Pinzon-Ardila O. , Feliu-Batlle V. , Roncero-Sanchez P. , Garcia-Gonzalez P. 2007 “Application of a repetitive controller for a three-phase active power filter,” IEEE Trans. Power Electron. 22 (1) 237 - 246    DOI : 10.1109/TPEL.2006.886609
Mattavelli P. , Marafao F. P. 2004 “Repetitive-based control for selective harmonic compensation in active power filters,” IEEE Trans. Ind. Electron. 51 (5) 1018 - 1024    DOI : 10.1109/TIE.2004.834961
Smedley K. M. , Zhou L. , Qiao C. 2001 “Unified constant-frequency integration control of active power filters steady-state and dynamics,” IEEE Trans. Power Electron. 16 (3) 428 - 436    DOI : 10.1109/63.923776
He L. , Zhang K. , Xiong J. , Fan S. 2014 “A repetitive control scheme for harmonic suppression of circulating current in modular multilevel converters,” IEEE Trans. Ind. Electron. 30 (1) 471 - 481
Franklin G. F. , Powell J. D. , Emami-Naeini A. 2010 Feedback Control of Dynamic Systems 6th ed. chap. 5
Franklin G. F. , Powell J. D. , Emami-Naeini A. 2010 Feedback Control of Dynamic Systems 6th ed. Addison-Wesley chap. 6
Chongming Q. , Smedley K. M. 2002 “Three-phase bipolar mode active power filters,” IEEE Trans. Ind. Appl. 38 (1) 149 - 158    DOI : 10.1109/28.980369