Search

This paper investigates the gate voltage (V gs ) oscillation of paralleled SiC MOSFETs during fast switching transient. Based on the study, the paper proposes to add the coupled inductance and/or the additional path on Kelvin source of paralleled MOSFETs to suppress the amplitude of gate voltage oscillation. The paper considers 1-in-1 and 2-in-1 power modules, and devices with on-chip current sense, which are used in many HEV/EV (Hybrid Electric Vehicle/Electric Vehicle) traction inverters. Since the proposed methods only require the modifications on the low power signal circuit instead of the high power circuit, they are easy to be implemented with small volume and weight, low power consumption, and low cost, which are required by automotive application.

Direct bypass of DC-DC boost converter in traction inverter has many advantages including increased efficiency and pass-through capability. This paper explains different practical aspects of direct bypass of the boost stage of traction inverter using diode. Bypass diode enters into the linear operating region (DC Q-point) when the battery discharges through the bypass diode to drive the electric motors. Therefore, thermal stress of the DC-link capacitor is shared between the input and DC-link capacitors through the bypass diode. During battery charging operation, the inverters introduce voltage ripple in the DC-link terminal which results in unwanted energy circulation through the bypass diode. Both of these phenomena have been investigated in details. In addition, this paper proposes and validates a control strategy to eliminate the energy circulation during battery charging mode using minimum boosting at a reduced frequency to optimize the power loss of the boost converter.

The IGBTs are dominantly used in traction inverters for automotive applications. Because the Si-based device technology is being pushed to its theoretical performance limit in such applications during recent years, the gate driver design is playing a more prominent role to further improve the traction inverter loss performance. The conventional gate driver design in traction inverter application needs to consider worst case scenarios which adversely limit the semiconductor devices' switching speed in its most frequent operation regions. Specifically, when selecting the gate resistors, the IGBT peak surge voltage induced by di/dt and stray inductance must be limited below the device rated voltage rating under any conditions. The worst cases considered include both highest dc bus voltage and maximum load current. However, the traction inverter operates mainly in low current regions and at bus voltage much lower than the worst case voltage. This paper proposed a gate driver circuit that can mitigate the compromise between low switching loss and high device stresses. Compared with conventional gate driver circuits, only a low voltage PNP BJT, two Zener diodes and a resistor are added in the proposed circuit. The implementation is easy and the potential cost increase is low. The proposed gate driver can separately control the different phases of turn-off transient to achieve short turn-off delay, fast dv/dt and low surge voltage across IGBT. Because the proposed method does not require any feedback circuits, logic and control circuits which could limit the circuit bandwidth, this method can be used for fast switching applications.

With the increasing demand for environment-friendly and high fuel economy vehicles, most of the automotive manufacturers are working on the development of their electric and hybrid electric vehicles (EVs/HEVs). The traction inverter plays a significant role in EV/HEV system and converts energy between the battery and the electric motors. One of the key challenges during the traction inverter development is how to evaluate its performance in the lab. This paper proposes a method to evaluate the performance of traction inverter in EVs/HEVs. Comparing with the traditional lab test methods, the proposed method uses the real driving cycles of the vehicle and illustrates the traction inverter performance in real driving conditions. The experimental results show the benefits of the proposed method.

The application of cross-links at the cable joints and the arrester-groundings at the cable terminals undoubtedly reduces the ground potential and the sheath current in a cable system, however, such configuration builds an open circuit in the grounding conductor and stops the partial discharges (PD) pulses to travel. Thus very few pulse energy goes into the grounding wire, which is an industrial accepted checkpoint for PD defects through conventional sensors such as high frequency current transformers. To solve the problem, this paper introduced an innovative broadband PD sensor by using a capacitive jumper to bridge the PD propagation path. In this way, the problem of impedance mismatch at PD frequency is avoided, while an `open circuit' is maintained at power frequency. The sensor is tested in the laboratory with an oscilloscope for its sensitivity and detection bandwidth, while the latter reaches about 200 kHz ~ 87MHz. Such sensor was tested in a real 110 kV cable and shown the convenience and effectiveness in PD detection.

Physical-layer network coding (PNC) has obtained much attention due to its ability to improve throughput in relay-aided communications. In this paper, we introduce orthogonal frequency division multiplexing-continuous phase modulation (OFDM-CPM) signals into a two-way relay (TWR) channel adopting the PNC protocol to form a new scheme. With PNC, the TWR communication is performed during two phases. In the multiple access (MAC) phase, the two users send OFDM-CPM signals concurrently to the relay, and the relay broadcasts the XORed version of the two users' data in the broadcast phase. We focus on the MAC phase and present a decoding scheme for the relay receiver. The error performance of the new scheme is provided by simulation results, and it is shown that the proposed scheme outperforms that of OFDM-PSK for PNC over frequency-selective fading channels.

In this work, an innovative and feasible solution is proposed for a Variable Voltage Converter (VVC) with direct bypass for traction drive inverters in HEVs. With the proposed method, the VVC capability is significantly increased and power loss is dramatically reduced. The advantages associated with this method include 1) Mitigation of VVC oscillation issues occurring during fast power flow transients, since the VVC output minimum voltage is directly clamped to the HV battery voltage anytime the output side voltage is lower than that of the HV battery voltage. 2) A significant increase in VVC pass-through operation capability; as a result, both vehicle EV drive maximum speed and duration are increased, 3) significantly reduced VVC power loss and noticeably increased vehicle fuel economy. This solution has been proved to be very useful and feasible especially for Plug-in HEV applications.

When Petri net modeling complicated system may lead to state blast of model easily. In this paper, we combine the advantage of Agent and Petri net and propose an Agent oriented Petri net (AOPN) model to describe complicated system by an intuitionistic and natural way. At last, we apply AOPN to fighting action modeling of missile.

Computer Generated Forces (CGF) lacks an effective collaboration method. As a result of this, Based on the traditional contract net protocol, a integrated contract net protocol suitable for CGF collaboration is proposed in which strategies such as interaction trust, witness trust and threshold value are integrated. In order to ensure integrity, correctness and reliability of integrated contract net protocol, formalized method colored net is used to model integrated contract net protocol and liveness, accessibility and bound ness of protocol are analyzed and validated. Finally, integrated contract net protocol is applied to collaboration simulation of missile attack and defense CGF system. Simulation results show that integrated contract net protocol effectively reduces communication cost of collaboration and increases completion task rate of system.

Focusing on the systematicness and validity of the evaluation system of the patient satisfaction, large number of factors impacting the patient satisfaction were considered comprehensively, the analytical method combining the Attribute Reduction in Rough Set with AHP was adopted. First, the evaluation factors were analyzed to establish the evaluation hierarchical model of the patient satisfaction, the indexes were reduced by Attribute Reduction based on the information quantity; then, the significance of attribute was used to determine the index weight; lastly, the total sequence weight of sub-criteria was calculated by AHP method and the comprehensive evaluation value was obtained. This model was used to evaluate and analyze the five hospitals in one area and the validity of this method was tested. This method reduces the amount of calculation, improves the computational efficiency and provides a method of quantitative analysis.

Educating and training outstanding engineers involves the government, the institution of higher learning and enterprises. The chief problems of implementing “A Plan for Educating and Training Outstanding Engineers (PETOE)” are how to build an guarantee evaluation index system and determine the guarantee index weight and interaction among indexes. In this paper, the method of Analytic Network Process (ANP) were adopted, eleven indexes, such as the policy, capital and teachers involved in implementing guarantee were analyzed, the guarantee Model and supermatrix of implementing PETOE Based on ANP were built, the SD software was used to find out the convergence state of the associated supermatrix, the stable limit supermatrix, the index weights were calculated and their total sequences of affecting indexes was sorted. The results show that the guarantee model of implementing PETOE based on ANP is reasonable.

In this work, an innovative and feasible evaluation method is proposed to fully verify the Variable Voltage Converter (VVC) functionality for vehicle electrification applications. With the proposed method, the VVC can be fully tested for all four operation modes which can occur during vehicle real driving conditions. The proposed method provides a solution which allows hardware self-evaluation with HEV drive cycles during development stages in the laboratory. The advantages associated with this method includes 1) the capability of testing the VVC up to full power under all operating modes, 2) minimal external power demand and energy savings; 3) enable hardware to be tested with real drive conditions and significantly reduces costs and resources which are required by vehicle road tests.

In the latest video coding standard H.264/AVC, coding efficiency is improved by introducing the intra-prediction. At the same time, it brings extremely high computational complexity and makes it a bottleneck for a real-time implementation of the H.264/AVC encoder. A fast and efficient intra-prediction mode decision algorithm is proposed in this paper. Firstly, it makes a choice between the INTRA_4×4 and INTRA_16×16 based on the complexity of the image. And then it adopts the high correlation of the prediction mode between current 4×4 block and adjacent blocks, strong correlation between the sum of absolute transform difference (SATD) and rate distortion (RD) performance to skip some unlikely prediction modes efficiently. So it reduces the intra-prediction complexity greatly. Experimental results show that the proposed algorithm can save the encoding time up to 65.5% averagely without noticeable loss of coding efficiency.

In applications where batteries work together with power electronic circuits, the current ripple generated by the power electronics will be shared by both the battery and passive components in the circuit. The amount of ripple absorbed by the battery depends on its impedance at the switching frequency of power electronics. This paper presents an impedance based high frequency battery model derived from test results of a NiMH battery using a novel battery impedance tester. The possible reasons for the battery impedance characteristics in high frequency region, including skin effect and proximity effect, are also discussed. This battery model can be directly used in current ripple analysis, passive components design and control strategy optimization of power electronic circuits. The effect of the passive component values on the battery current ripple is analyzed using the ac equivalent circuit of the test setup.

In this work, the IGBT turn-on and turn-off waveforms are normalized in order to analyze the closed-loop gate drive switching losses. The relationships of energy losses and controlled switching speed, voltage overshoot, current overshoot are derived. Theoretical analysis results show good agreement with experimental results. The proposed analysis methodology and results can provide guidelines for the gate drive design in real applications.

To overcome the drawbacks of the conventional gate drive, in this paper a closed-loop gate control method is proposed. In this novel method, the switching speed, specifically di/dt, is measured from the voltage across the parasitic inductance of the IGBT module and a closed-loop control is employed to adaptively adjust the gate drive voltage to control the switching speed according to a preset control reference. As a result, both the voltage overshoot and current overshoot can be effectively controlled. This new gate drive method enables programmable control of switching speed and allows full use of the capability of the power devices. The oscillations during IGBT switching are also reduced and associated EMI problems can be mitigated. The relationships between the controlled voltage overshoot, current overshoot and associated energy losses are derived and can provide guidelines for practical design. Also, the proposed gate drive fully utilizes gate drive signal information of amplitude and duty cycle, and provides a new capability for the system modulator to effectively control both the power electronic circuit's steady-state and transient behaviors simultaneously.

An active fault protection method is proposed to prevent the IGBTs from being damaged should they be under over-current and short circuit fault conditions. A closed-loop gate drive method is used to effectively control the IGBT turn-off switching speed independent of the fault type or resultant over-current level. Compared with conventional solutions, the proposed method can limit fault current level, minimize energy loss with fast shut down transient, and control voltage overshoot. Also, no additional hardware circuit or soft shut down process is need. Even if an over-current fault condition can not be effectively detected, the automatically controlled IGBT turn-off switching speed can avoid a dangerous voltage overshoot across the power device. This new feature can significantly improve the robustness of the power system and cannot be achieved with the conventional approaches.

A full-bridge dc-dc converter employing diode rectifier in the output experiences severe voltage overshoot and oscillation problem across the diode rectifier caused by the interaction between the junction capacitance of the rectifier diode and the leakage inductance of transformer. The pronounced reverse-recovery current of high power diodes enforces these issues significantly by increasing power loss and voltage overshoot. Conventional energy recovery clamping circuits suffer from high voltage overshoot if the input voltage of converter is wide. In this work, a novel energy recovery clamp circuit is proposed to overcome this problem. Performance of the proposed circuit is verified both theoretically and experimentally with a 70 kW dc-dc converter.

A dead-time elimination method is presented in this paper for PWM controlled inverters. In comparison to using expensive current sensors, this method precisely determines the load current direction by detecting the operating conditions of active devices and their anti-parallel diodes. A low-cost diode-conducting detection circuit is developed to measure the operating status of the anti-parallel diode. In comparison with complicated compensators, this method features with simple logic and flexible implementation. This method significantly reduces the output distortion and regains the RMS value. The principle of proposed dead-time elimination method is described. Simulation and experimental results are given to demonstrate the validity and features of the dead-time elimination method.

A 260 kVA inverter has been developed with a passive soft-switching snubber. The main circuit configuration and prototyping are described. Optimization of the passive soft-switching snubber parameters has been made to control the voltage overshoot across the switches, minimize snubber and inverter losses, and effectively limit dv/dt and di/dt during activation. An efficiency higher than 98.5% has been achieved as well as other prospected features. Experimental results are given to demonstrate the validity and features of the inverter.