小波包能谱熵与神经网络在断路器故障诊断中的应用

提出了一种以振动信号小波包能谱熵为特征量的断路器故障神经网络诊断方法。利用小波包分解原理将高压断路器振动信号分解到不同频段中,计算各频段的能谱熵值,以此构造小波包能谱熵向量作为神经网络的输入向量,并利用遗传算法对网络的连接权值进行了优化。引入置信度的概念,对改进神经网络输出的故障模式识别结果进行评价。通过试验分析结果表明了该方法的有效性,改进后的神经网络具有新故障模式的识别功能。

Applying wavelet packet energy entropy and neural networks to diagnose circuit breaker faults

We presented a new fault diagnosis method based on wavelet packet energy entropy and an improved neural network.A wavelet packet was used to decompress the vibration signal into different frequency bands using the theory of wavelet packet decomposition and reconstruction.Wavelet packet energy entropy was then extracted to construct characteristic vectors of signals and is used as an input of the neural network,which was optimized by genetic algorithm.Finally,the degree of confidence concept was introduced to evaluate the test results.This method was proven to be effective by the pattern recognition results of the circuit breaker fault.Furthermore,the improved neural network can recognize new fault patterns.