Research on Fault Diagnosis Method Based on Reciprocating Compressor

Abstract

Based on the research and analysis of related fault diagnosis technologies at home and abroad, a fault diagnosis method for reciprocating compressors based on support vector machine and gravity search algorithm is proposed. This method optimizes the kernel parameters of SVM. Combined with an example of a fault diagnosis model of a reciprocating compressor, the analysis results show that the fault diagnosis method of a reciprocating compressor using the GSA-SVM algorithm has higher recognition accuracy than the SVM algorithm.