Fault diagnosis method for reciprocating compressors based on visibility graph

Reciprocating compressors are a common type of mechanical equipment widely used in many industries. Health condition detection and accurate fault diagnosis of reciprocating compressors can help maintain proper equipment operation. These detection and diagnostic methods can be achieved by monitoring parameters such as vibration, temperature, and pressure of the compressor. By analyzing and processing these parameters, problems that may lead to failures can be identified, but this method is susceptible to external interference and complex operation. In this paper, to solve the above problems, a fault diagnosis method based on visibility graph for reciprocating compressors is proposed. The method firstly represents the one-dimensional vibration signals by visibility graph, and then calculates the adjacent matrix eigenvalues and laplace matrix eigenvalues of each visibility graph to obtain each visibility graph index by the eigenvalues. Finally, the SVM classification algorithm is brought in to diagnose and classify the reciprocating compressor valve faults, and the classification results are obtained. This method can effectively extract fault features, avoid the influence of other factors on the fault diagnosis results, abandon the complex operation process, and improve the accuracy and reliability of fault diagnosis.