A Research On Reciprocating Compressor Fault Diagnosis Method Based On LCD And MSE

Reciprocating compressor is an essential power equipment to compress and transport gas and used widely in petroleum and chemical industry. With the increasing emphasis on production safety in present society, it is necessary to implement fault diagnosis technique for reciprocating compressor. Owing to the complex structure and numerous excitation sources, it is necessary to conduct further investigation of the fault diagnosis technique for reciprocating compressor. In view of the reciprocating compressor vibration signals with nonlinear,non-stationary characteristics, this article proposes a method based on the local characteristic scale decomposition(LCD) and multiscale entropy(MSE) for fault diagnosis of reciprocating compressor.First of all, based on the basic structure, the reciprocating compressor working principle,and a brief overview of the reciprocating compressor fault diagnosis technology, the common faults of cylinder and transmission parts was taken as the study objects, and their fault mechanism are summarized. The experimental research and data collection of 2D12 type reciprocating compressor was conducted.Secondly, according to the non-stationary characteristics of vibration signals of reciprocating compressors, an adaptive time-frequency analysis method-LCD was introduced.For LCD method adopts linear transformation, the phenomenon of the waveform distortion in the baseline signal occurs, and an improved method was put forward by using the Hermite interpolation instead of the linear transformation. Through the analysis of the simulation signal and measured signal, the proposed method was verified.Thirdly, according to the nonlinear characteristics of vibration signals of reciprocating compressors, the method based on LCD and MSE is proposed for reciprocating compressor fault diagnosis. The ISC components of LCD decomposition results which contain the main information was chosen by using the correlation coefficient, then the MSE was used to describe characteristics of ISC components. The eigenvalues of state feature matrix was extracted by singular value decomposition as the final eigenvectors.Finally, the eigenvectors of faults was extracted by the proposed method. Taken SVM as pattern classifier, the eigenvectors of faults were diagnosed, and superiority of this method is verified by comparing the recognition results of other methods.