基于小波变换的发动机噪声诊断

A New motor Noise Diagnosis Approach Based on Wavelet Transform

提出了一种发动机噪声故障诊断的新方法。该方法在分析故障产生机理和噪声信号特征的基础上 ,构造一种新型的小波—指数衰减型小波函数 ,应用它对发动机噪声信号进行处理 ,提取表征故障的间歇性撞击声信号特征 ;为了获得可靠的诊断结果 ,给出发动机正常噪声和故障噪声的复合假设 ,构造相应统计量进行假设检验判决 ,检测故障。试验验证表明 ,该方法能有效地诊断摩托车发动机存在的活塞击缸故障。

This paper discusses a new approach of detecting faults in motorcycle motors.The approach uses the nose signals generated by motors to achieve fault diagnosis.Based on the analysis of fault mechanism and signal characteristic,a new type of wavelet--Exponential Damping Wavelet is constructed to recognize the intermittent impacting sounds caused by mechanical faults from the mass of motor noises.In order to get reliable diagnosis results,a composite hypothesis of the normal nose signal and faulty noise signal is constructed,and by testin that hypothesis,faults can be detected accurately.