Abstract: A fault diagnosis method for turboengine based on wavelet and fuzzy network is presented. In this method, wavelet is used to extract fault characteristics and neural network is used to diagnose the faults; the inputs and outputs of the neural network are all memberships concretely, the inputs are the memberships of the crest factor, pulse factor, margin factor, bias factor, kurtosis factor and maximum spectrum on each characteristic domain, and the outputs are memberships of each fault. Applications on a turboengine show that the effect of this method is significant.

Key words: Wavelet transform;Artificial neural network;Fuzzy algorithm;Fault diagnosis;Turbojet engine

摘要： 提出了一种基于小波和模糊神经网络的涡喷发动机故障诊断方法。即利用小波变换获取特征域 ,取特征域上的峰值因子、脉冲因子、裕度因子、偏态因子、峭度因子及频谱最大值作为神经网络的输入 ,并对神经网络的输入、输出进行模糊化处理 ,以神经网络进行诊断。将该方法成功地应用于某型涡喷发动机的故障诊断 ,结果表明 ,该方法诊断效果明显。

关键词: 小波变换;人工神经元网络;模糊算法;故障诊断;涡轮喷气发动机