Research on Fault Feature Extraction Method for Gas Valve Signal Based on SSA-VMD
In order to solve the problem of signal overdecomposition and underdecomposition caused by the artificial preset decomposition layer number K and penalty factor α of VMD algorithm,a method for extracting fault features of diaphragm compressor based on SSA-VMD is proposed.This method optimizes the VMD with the minimum envelope entropy of each component as a fitness function.By comparing the SSA-VMD method with the original VMD method,the multi-scale dispersion entropy(MDE)values of each component of the two decomposition methods are calculated separately.The appropriate entropy values are selected as feature vectors,and the random forest algorithm is used to classify and identify the faults,so as to diagnose the various faults of the diaphragm compressor air valve.The results show that the average accuracy of fault identification using the original VMD method is 90.63％,while the average accuracy of fault identification using the SSA-VMD method is as high as 99.32％.This result indicates that using SSA-VMD method to decompose signals significantly improves the accuracy of fault diagnosis.
万方数据
维普
