基于小波分析的制冷压缩机气阀故障诊断方法的研究

基于小波多尺度分解、重构的方法，对制冷压缩机气阀故障进行了诊断。把从压缩机缸盖上测得的振动信号经小波变换后，分解成高频和低频两个频带，阀片的自振频率成分出现在低频带。由于受工频及其谐波成分的影响，气阀的自振频率成分很微弱，难以辨识，不宜作为气阀故障的特征量。为此构造一个特征量，即高频带平均功率与总平均功率之比，利用此特征量来诊断压缩机气阀故障。大量的实验表明，该方法是有效的。

Application of the Wavelet Transform to Fault Diagnosis in Compressor Valves

This paper presents a wavelet transform-based method for valve failure diagnosis of a compressorused in a trailer refrigeration unit. The vibration signalsmeasured on the cylinder heads of the compressor are de-composed into low frequency parts and high frequencyparts by wavelet transform. The components of naturalfrequency of the valves are contained in low frequencyparts, but they are buried in the synchronous componentsand their harmonics. Hence, the weak components of in-herent frequency of the valves are not suitable to detectthe valve failures. To this end, a vibration signaturefunction is constituted, which is defined as the ratio ofthe average power spectrum of the high frequency parts tothat of all frequency region. This signature function issensitive to the valve failures, therefore it is an importantindicator of presence of valve failures. An intensive ex-periment with a real machine has confirmed the effective-ness of the above procedure.