小波包特征熵在级联式变频器故障诊断中的应用

为精确诊断级联式变频器功率器件开路故障,提出了一种基于小波包特征熵的故障信号提取方法.对采集到的级联式变频器相电压信号进行三层小波包分解,提取特征熵构造电压信号的特征熵向量,并以此作为故障诊断样本,利用概率神经网络进行故障诊断.仿真结果表明,基于小波包特征熵的信号提取方法在级联式变频器故障诊断的应用中具有较高的有效性与可行性.

Fault Diagnosis for Cascaded Converter with Characteristic Entropy of Wavelet Packet

In order to accurately diagnose the open circuit fault of power device of cascaded converter,a fault signal extraction method based on wavelet packet characteristic entropy is presented.The three-layer wavelet packet is adopted to decompose the voltage signals of cascaded converter and extract characteristic entropy vector of voltage signals.The extracted characteristic entropy vector is used as the fault diagnosis sample and then the probabilistic neural network is used to diagnose the faults.The simulation results show that the method of signal extraction based on wavelet packet characteristic entropy is more effective and feasible in the application of cascaded inverter fault diagnosis.