三端口双向直流变换器断路故障综合诊断方法

三端口双向直流变换器是现代微电子技术中的新型核心装置，其故障的快速综合诊断可促进电子技术的发展。通过建立SHEV电气系统的三端口双向直流变换器的仿真拓扑模型，考虑功率器件的位置条件，以直流侧母线输出电流作为特征量，细致划分每种故障类别，并分析不同故障下输出电流的特性，通过FFT变换分解信号获取各个频率段，以20k Hz(k=1，2，…，20)频率段直流侧电流信号能量作为故障诊断的特点量，结合遗传算法BP神经网络，综合实现故障类别的区分与诊断。实验结果表明，此方法在诊断与区分交直流变换器的断路故障时，诊断最小误差仅为0.033，诊断结果精；且可在一个开关周期内实现开关管断路故障诊断，具有较高准确率及效率。

Comprehensive Diagnosis Method for Open-circuit Fault of Three-port Bidirectional DC Converter

Three-port bidirectional dc converter is a new core device in modern microelectronics technology. Through the establishment of three ports of electrical system of SHEV bidirectional dc converter topology model of the simulation, considering the conditions of the location of the devices, dc bus output current as the characteristic, the detailed for each type of fault classification, and analysis of the characteristics of the output current under different fault, by FFT transformation decomposition signal for each frequency to 20 k Hz (k = 1, 2,... As the characteristic quantity of fault diagnosis, the energy of the dc side current signal in the frequency segment is combined with the genetic algorithm BP neural network to realize fault classification and diagnosis. The experimental results show that the minimum error of this method is 0.033, and the diagnosis result is accurate. In addition, the fault diagnosis of switch tube breaking can be realized within one switching cycle, with high accuracy and efficiency.