基于MLE阈值规则的小波特征提取技术在气阀故障诊断中的应用

气阀故障是往复压缩机最常见的故障类型之一，占故障总数的60%以上，如果不及时发现并解决，往复压缩机的压缩效率将大大降低。针对目前往复压缩机气阀故障诊断中存在的问题，结合小波降噪技术，提出了采用基于最大似然估计（MLE：Maximum Likelihood Estimation）阈值规则对气阀早期故障弱冲击变化信号进行特征提取的方法，实现了气阀故障的早期预警。

Wavelet feature extracting technique based on maximum likelihood estimation threshold rule and its application in fault diagnosis of a gas valve

Faults of a gas valve often happen in reciprocating compressors,occupy more than 60%in total number of faults.If faults are not found and solved in time,the efficiency of a reciprocating compressor will be greatly reduced. Aiming at the current problems existing in reciprocating compressor valve fault diagnosis,combining with wavelet denoising, based on maximum likelihood estimation(MLE) threshold rule,a feature extracting technique for weak impact signals in early faults of a gas valve was proposed.The ear...