Electromagnetic valve fault diagnosis based on multi-source signal feature fusion

This paper proposes a new electromagnetic valve fault diagnosis method, which can effectively expand the number of diagnosable fault types for electromagnetic valves and achieve high fault diagnosis accuracy. First, dynamic, sealing, and flow data are collected according to the different working states of the solenoid valve. Then, feature extraction and fusion of these three types of data are performed. In the process of extracting dynamic characteristic data, a comparative analysis is conducted between the method of data feature extraction based on physical nodes and the method of variational mode decomposition -energy moment data feature extraction. During data fusion, a comparative analysis is performed between feature-level fusion and decision-level fusion methods. To evaluate the effectiveness of different methods of data feature extraction and data fusion, deep learning is performed by using an long short-term memory network model, and the learning results are analyzed.