High frequency circulating current
and its effect to control and driving circuit inside hybrid IPEM
ZENG Xiang-jun1, WANG Xiao-bao2, YANG Xu1, WANG Zhao-an1
(1. Electrical Engineering School, Xi’an Jiaotong University, Xi’an 710049, China;
2. Xi’an Power Electronic Research Institute, Xi’an 710049, China)
Abstract: Electromagnetic interference to the control and driver circuit caused by power circuit is a very important problem inside hybrid Integrated Power Electronic Module (IPEM). In this paper, simulation and experiments show that the high-frequency current oscillations caused by the stray inductance of DC link and parasitic capacitor of power device at the moment of switch-on or off are one of the main reason that leads to EMI inside IPEM. Research also shows that this current flows only between DC link and power devices like a circulating current, and does not flow through the load path. This high frequency circulating current brings disadvantageous effect to the control and driver circuit in the module.
Key words: integrated power electronic module; circulating current; EMI

Analysis on matrix converter operation in abnormal conditions
SUN Kai1, ZHOU Da-ning1, HUANG Li-pei1, SETO Makoto2, KIMATA Masahiro2
(1. Dept. of Electrical Engineering, Tsinghua University, Beijing 100084, China;
2. Mitsubishi Electric Corporation, Amagasaki 661-8661, Japan)
Abstract：Matrix converter is a direct three-phase AC to AC power conversion device. Due to the absence of dc link capacitors, the abnormal conditions at input side of matrix converter may influence the output performance directly. In this paper, the output characteristics of matrix converter have been theoretically analyzed in the conditions that the input power supply voltages are unbalanced and non-sinusoidal respectively. The unbalanced and non-sinusoidal power supply voltages introduce low order harmonic components into the output voltages of matrix converter and result in the distortion of the output current. The behavior of the input power supply short-time blackout and the rapid re-starting capability of matrix converter have been discussed in this paper. The experimental tests, in terms of unbalanced, non-sinusoidal power supply voltages and short-time blackout, have been carried out on a 2.7kVA prototype of matrix converter. Experimental results demonstrate the theoretical analysis.
Key words: matrix converter; unbalance; non-sinusoidal; short-time blackout

Current compensation method for soft-switching three-phase PWM inverter
by using alternate (positive & negative slope) sawtooth carrier
XU Chun-yu, CHEN Guo-cheng, SUN Cheng-bo, QU Ke-qing
(School of Electromechanical Engineering and Automation, Shanghai University, Shanghai 200072, China)

Abstract: In a soft-switching three-phase PWM inverter using alternate (positive &negative slope) sawtooth carrier, although soft-switching of the inverter can be realized, the output current waveforms for the inverter will be distorted. Based on flux linkage locus and PWM driven signal at zero-crossing of motor phase current, the reason of such distortion is analyzed, and a current compensation method was brought forward, namely, the PWM pulse width in a carrier period should be adjusted immediately after one phase current polarity changes, which can greatly improve output waveforms for the inverter. Experiment results proved that by using such current compensation method the output current waveforms for inverter are sinusoidal and have low harmonic components.
Key words: soft-switching; inverter; flux linkage locus; current compensation method

Research on fault diagnose technology of inverter
based on associative memory neural network
WANG Mu-lan, ZHANG Chong-wei, GU Sheng-gu
(Electrical Engineering Institute of Hefei University of Technology, Hefei 230009, China)
Abstract：An associative memory model of Artificial Neural Network (ANN) is introduced into the fault diagnose of inverter. Firstly, the diagnose algorithm is derived and the fault table is concluded. Secondly, the fault sample vectors are preprocessed by HADAMARD transform, and learned by the model. Finally, the samples and non-samples are parallel associatively recalled based on this network. Furthermore, the digital simulation results demonstrate that the novel strategy is reliable. Additionally, a hardware implementation thought is presented, which uses large-scaled Field Programmable Gate Array (FPGA) circuits.
Key words：artificial neural network；inverter；fault diagnose；field programmable gate array (FPGA)

Study of efficiency-optimized control based on accurate modeling for PMSM
SONG Ju-ming, SU Yan-min
(Xi’an Jiaotong University, Xi’an 710049, China)
Abstract: An accurate modeling of the permanent magnet synchronous motor (PMSM) is necessary for precise torque control and/or high efficiency drive control. This paper presented an accurate mathematical model of PMSM taking account of all kinds of losses, that is, the copper loss, iron loss, mechanical loss and stray loss. Based on the model, the efficiency-optimized control is presented, and the efficiency-optimized control equations are derived. The validity of the proposed mathematical model and the efficiency-optimized control are confirmed by the experiment.
Key words: mathematical model of PMSM; control; efficiency-optimized

Analysis of 3D eddy current field in cup-type axial magnetic field
vacuum interrupter
LIU Zhi-yuan, XIE Ke-song, WANG Zhong-yi, WANG Ji-mei
(Department of Electrical Engineering, Xi’an Jiaotong University, Xi’an 710049, China)
Abstract: A finite element analysis model that is the same as an electrode structure of a commercial cup-type axial magnetic field vacuum interrupter is built up. An arc that is treated as a cylinder conductor is included in the model. Eddy current field generated by an ac current passing the electrodes is analyzed. It is shown that maximum value of current density in contact at current peak becomes higher when the arc is considered and the eddy current influence in coils and the arc take effect. It is also found that the axial magnetic flux density is much higher on one side of a contact slot than on the other side at the current peak. B vector distribution on contact surface is also shown. It looks like a swirl. The axial component of B vector is higher in middle part of contact and it becomes lower near contact edge, where the radial component of B vector becomes higher. It is shown that axial magnetic flux density distribution on middle gap plane and contact surface changes to a tent shape at current zero when the arc is considered and eddy current influence in coils and the arc take effect, which also makes maximum axial magnetic flux density higher. It is pointed out that contact center is such a position that both phase shift time and axial magnetic flux density are high.
Key words: vacuum interrupter; axial magnetic field; vacuum arc

Design for non-fragile robust control
and selections of weighting functions of μ-synthesis
HENG Qing-hai
(Beijing Information Technology Institute, Beijing 100101, China)
Abstract：It is suggested that a class of design problem of the control systems can be formulated as a μ-synthesis by the concepts of the non-fragile robust performance. It can be used to solve the robustness problems of the perturbations of the controllers, which can’t be done by the conventional robust performance designs. It is pointed out that the system of μ-synthesis design can’t be of non-fragile robust performance when performance weighting function is selected incorrectly. The correct selections of the weighting functions are discussed and the μ-synthesis non-fragile robust PI controller is solved. The simulation results of a boiler superheated steam temperature system show that the system has better performance when the controlled members and the controllers are simultaneously perturbed.
Key words：selection of weighting function; controller perturbation; non-fragile robust control;
μ-synthesis; boiler superheated steam temperature

Accurate fault calculations for parallel power lines on same tower
LIN Jun, WEN Bu-ying
(Dept. of Electric Engineering, Fuzhou University, Fuzhou 350002, China)　　　　　　　　　　　　　　　　　　　　　　　　 
Abstract: This paper introduces accurate fault-calculation for parallel power lines on the same tower by eigenanalysis of complex matrix. This calculation method considers the effect of power line transposing and distributed capacitance. The method has high accuracy and can be used for calculating fault components of parallel power lines on the same tower when fault is on one line or between two lines or complex faults crossing lines. This method can be used for investigation the performance of relays and other devices in the parallel power lines and analysis of the accidents of parallel power lines.
Key words: parallel power lines; fault calculation; eigenanalysis

Sensitivity analysis of conducted EMI coupling paths
in switch mode power supply
HUANG Jiang-jiang, XUE Liang, HE Jun-ping, SUN Xu-dong, JIANG Jian-guo
(Dept. of Electrical Engineering, Tsinghua University, Beijing 100084, China)
Abstract: This paper first discusses how to build circuit model for the conducted EMI coupling paths in switch mode power supply. Based on the model, its sensitivity of a switch mode power supply had been analyzed by the OrCAD/PSpice software. Simulation and experimental results show that, building appropriate circuit and then analyzing the sensitivity by computer software, we can systemically analyze how each component takes effect on conducted EMI. This technique can bring effective guides to the designing and debugging of EMC of switch mode power supply.
Key words: switch mode power supply; conducted EMI; coupling paths; sensitivity analysis

Research on temperature compensation of optical fiber displacement sensor
based on RBF neural network
WEI Zhao-bi, LIU Ye, MA Zhi-ying, HU Guang-hui, SHI De-gang
(School of Electrical Engineering, Xi’an Jiaotong University, Xi’an 710049, China)
Abstract: In this paper a new software compensation method based on the theory of RBF neural network to solve the temperature effect on the optical fiber displacement sensor is presented. In this method, the outputs of the displacement sensor and the temperature sensor are processed using the neural network, and the temperature effect on optical fiber displacement sensor is basically overcome. A practically feasible way for actual design of optical fiber sensor with the temperature effect is proposed. After using the neural network, the experiment shows that the coefficient of temperature sensitivity of the optical fiber displacement sensor is reduced by two order of magnitude and both the accuracy of measurement and stability of the system are improved.
Key words: optical fiber displacement sensor；temperature sensitivity；compensation；RBF neural network

Comparison of several projects of high speed MAGLEV
ZHANG Rui-hua1,2, YAN Lu-guang1, XU Shan-gang1, WU Ying1,2
(1. Institute of Electrical Engineering, Chinese Academy of Sciences, Beijing 100080, China;
2. Graduate School of the Chinese Academy of Sciences, Beijing 100039, China)
Abstract: Since MAGLEV train has many advantages, many countries are busy in its study. The schemes of Germany, Japan are mature by far, and there are several typical schemes in American and some other countries. It is significant to study these typical schemes. We can use these schemes for reference in development of our MAGLEV. The article presents the three schemes of the high speed MAGLEV trains of Germany, Japan and America, and the detailed comparison of the three schemes is given. We can see that the scheme of Germany is the mature one.
Key words: MAGLEV; high speed; scheme comparison

Study on schemes of arcless on-load transformer for voltage regulation
ZHAO Gang, SHI Wei
（School of Electric Engineering, Xi’an Jiaotong University, Xi’an 710049, China）
Abstract: Three schemes of arcless on-load transformer for voltage regulation are studied in this paper. To the scheme of mechanical on-load tap changer combined with thyristors, its characteristic of arc suppression is analyzed in details, and a corrected scheme is then present. To the scheme of series transformer for voltage regulation controlled by thyristors, its performance of voltage regulation is illuminated. By physical tests, the effects of arc suppression and dynamic response characteristics of those schemes are figured out and compared with each other. Series transformer for voltage regulation controlled by thyristors is proved to be the best scheme in arcless on-load transformer for voltage regulation in this paper. 
Key words: power electronic; voltage control; on-load tap changer; arcless on-load transformer; thyristor

Pulsed power source in low field Nuclear Magnet Resonance system
ZHOU Ying, ZHANG Yi-ming, TANG Xiao-quan, JING Bo-hong
 (Institute of Electrical Engineering, Chinese Academy of Sciences, Beijing 100080, China)
Abstract:  Design of pulsed power source in low field nuclear magnet resonance (NMR) system with properties of practice, reliability, economy and simplicity is proposed in this paper. It is proved by theories and experimental results that switch power amplified complementary-output circuit satisfies the special requirements of pulsed power source in low field nuclear magnet resonance（NMR）system. A driving circuit for power MOSFET suited for less than 2MHz switch frequency is developed. The scheme has been used in low field nuclear magnet resonance system with advantage of high efficiency, wide frequency range, convenient tuning and easily searching for the resonance point.
Key words: low field NMR; pulsed power source; high speed power MOSFET driving

 
Current-source SVPWM rectifier based on DSP
YU Yong1, ZHANG Xing2, LIU Zheng-zhi1
(1. Institute of Plasma Physics, Chinese Academy of Sciences, Hefei 230031,China;
2. Hefei University of Technology, Hefei 230009, China)
Abstract: The three-valued logic modulation technique is usually used in conventional current-source PWM rectifier (CSR), and the SPWM control waveform could be got through the interphase decoupling, but the process is complex and the utilization ratio of direct current is low. So based on the conception of vector control, the second-level interrupt of DSP is taken full advantage. Consequently, the generation process of three-valued logic PWM and the realization of control scheme are simplified very much. The transform function of the system is also deduced.  Experiment results show the validity and the feasibility of the principle used in current SVPWM, technology of PWM waveform generating and the proposed.
Key words: PWM; current-source Rectifier; vector control

Modification of plasma reactor and matching with pulse power
NIE Yong, LI Wei, SHI Yao, TAN Tian-en
(Department of Environmental Engineering, Zhejiang University, Hangzhou 310027, China)
Abstract: The modified experiments of the decomposition of toluene by pulse corona discharge in wire-plate reactor were investigated. A pulse power based on thyratron switch, Blumlein pulse forming network and pulsed transformer was adopted. Different structures which influence the removal efficiency of toluene have been investigated, such as wire-to-wire spacing and the introduction of the baffle plate, the ceramic plate of dielectric barrier and the catalyst. The matching characteristics between the peak impedance of plasma reactor and the output impedance of pulse power were also analyzed. The results showed that the ratio of wire-to-wire spacing to wire-to-plate distance at 0.8 was optimum. The gas-flow distribution was “S” structure when the baffle plate was introduced in the reactor, and the removal efficiency was improved correspondingly. There were three types of introduction of the ceramic plate and the catalyst, including (a) no dielectric barrier, (b) ceramic plate barrier and (c) ceramic plate barrier with catalyst. The removal efficiency and energy yield of the three types of reactors was “a”<“b”<“c”. The peak impedance of the plasma reactor decreased when peak voltage increased, and at the same time, the ratio of the peak impedance of plasma reactor to the output impedance of pulse power was also approached to 1, the energy efficiency of pulse power improved and the maximum value was up to 92%. According to the current results, the optimization design for the reactor and the matching with pulse voltage source can be reckoned.
Key words: pulse power; plasma reactor; modification; matching; toluene

Development of large capacity active power filter
based on DFT detection method
LEI Wan-jun1, LI Hong-yu1, ZHUO Fang1, WU Long-hui1, WANG Zhao-an1, LIN Feng2, JI Yong2
(1. Xi’an Jiaotong University, Xi’an 710049, China;
2. Henan Electric Power Research Institute, Zhengzhou 450052, China)
Abstract: In this paper, a system of a 100kVA three-phase parallel active power filter with main circuit of multiple performances is introduced, which is based on DFT (Discrete Fourier transform) detection method. First, the basic operation principle of the parallel active filter is presented. Then, propose a digital real-time method to detect the harmonic and reactive element in the source current, which can offer the signal not only of the positive but also of the negative phase-sequence content of the fundamental and harmonic current. Third, accord to the control purpose, design the main circuit control method and the control circuit. At last, some experimental results are obtained. It shows that the control system meets the requires and the compensation property of the active power filter is very good.
Key words: active power filter；DFT；power quality

Preliminary application of BP neural network in predicting breakdown voltage
of large generator ground wall insulation
LIU Li-bing, GAO Nai-kui, MA Xiao-qin, XIE Heng-kun
(State Key Laboratory of Electrical Insulation for Power Equipment,
Xi’an Jiaotong University, Xi’an 710049, China)
Abstract: The breakdown voltage plays an important role in predicting remaining life of the large generator ground　 wall insulation. In this paper, we discussed a BP neural network that was used to predict the breakdown voltage of  the large generator ground wall insulation of rewind bars of the Jianbi Plant 7＃ generator. At first the neural network has been trained by the samples that include the parameters of dielectric loss factor tan , the parameters of partial discharge (PD) and breakdown voltage. We used the parameters of dielectric loss factor and partial discharge (PD) as the inputs of the BP neural network and breakdown voltage as the output. Then we tried to predict the breakdown voltage by the neural network that has been trained already. We found that it’s feasible and much accurate of neural network to predict breakdown voltage. This method can be applied to predict breakdown voltage of other generators ground wall insulation.
Key words: neural network; dielectric loss; partial discharge; breakdown voltage

Study on sterilization of n-TiO2 catalysis plasma discharge
XIA Tian1，YANG Xue-chang1，ZHOU Yuan-xiang1，KE Rui1，ZHAO Da-qing1，LI Ming-xian2，LI Ru-nan2
(1. Tsinghua University, Beijing 100084, China; 2.Beijing Huayi Institute of Purifying Science and Technology, Beijing 100084, China )
Abstract: A new approach on sterilization by nanometer-TiO2 (n-TiO2)catalysis plasma discharge is brought up in this paper. The basic knowledge and the apparatus of this approach are also introduced. Experiments on elimination of bacillus demonstrate the strong sterilization ability of this technology. It has been found that within 1 minute’s processing of n-TiO2 catalysis under plasma discharge, an amounts of 105 bacillus can be completely eliminated.
Key words: nanometer TiO2; plasma discharge; catalysis; sterilization; bacillus