一种基于矢量控制的变流器故障诊断方法

基于空间矢量控制的变流器,提出一种以变流器输出三相电流信号为特征参数的变流器故障诊断模型。运用小波分析理论对电流信号进行降噪处理,通过FFT变换及对电流信号统计计算得到故障特征向量,将该特征向量作为神经网络的输入,通过对神经网络的训练,确定网络参数,对故障数据的分析,达到故障诊断的目的。实验数据表明,该故障诊断模型对采用空间矢量控制的变流系统的电力电子器件故障的检测与诊断是可行的。

An Approach to Converter Faults Diagnosis Based on Vector Control

Oriented to a converter under vector control, a model of its fault diagnosis with the three-phase current signals as characteristic parameters is established. The output current signals, from which noise is further reduced with wavelet analysis theory, are also used in the statistic calculation for characteristic fault vector with FFT transform. The characteristic fault vector acts as input to neutral network. Through the training of neutral network, the network parameters can be decided and the analysis of...