电动汽车逆变器故障诊断

Fault Diagnosis of Electrical Vehicle Inverter

针对电动汽车Z源逆变器,根据开路故障下两相之间的输出线电压特性,以故障线电压作为故障信息。提出了一种基于BP神经网络的故障诊断方法,通过频谱分析获得的线电压直流分量、基波幅值、基波相位及2次谐波相位作为故障特征向量。利用神经网络的自学习和非线性映射能力实现Z源逆变器的故障诊断。实验结果表明,该方法具有很好的故障识别能力,能快速准确定位故障源。

According to the open-circuit fault characteristics of output line voltage between the two phases, taking fault line voltage as the fault information, a fault diagnosis method is proposed based on BP neural network, spectrum anal- ysis is utilized to extract line voltage DC component, the amplitude, fundamental wave phase and the second harmonic phase as the fault characteristics.Using neural network self-learning and nonlinear mapping ability to realize the fault diagnosis of Z source inverter.Test results show that the method has good fault identification ability, can quickly and accurately locate the source of fault.