Fault diagnosis of marine air compressor based on VMD multi-scale permutation entropy and SVM
Marine machinery vibration signals have nonlinear and non-stationary problems,and fault features are difficult to extract.Through variational mode decomposition multiscale permutation entropy and support vector machine fusion fault diagnosis method to study the vibration signal.Taking the air compressor as an example,first,6 kinds of air compressor working conditions are simulated,the thermal parameters of each working condition are tested,the degree of change of the thermal parameters of each working condition is analyzed,and the collected vibration signals are analyzed in the frequency domain.Then the vibration signal is decomposed by VMD to obtain a series of natural modal components,and the cross-correlation coefficient with the original signal is calculated to screen out the sensitive natural modal components.Finally,the multi-scale permutation entropy of the sensitive natural mode components is calculated,which is used as a feature vector and input into the SVM for fault identification.The experimental results show that the air compressor fault identification method based on the fusion of VMD multi-scale arrangement entropy and SVM can effectively identify the fault type,and the overall accuracy rate can be maintained at 98.6667%.The comparison of this method with other methods proves the effectiveness of this method.
marine reciprocating air compressorvariational mode decompositionmultiscale permutation entropyfault diagnosis
NETL
NSTL
万方数据
