Research on Fault Diagnosis Technology of Nonintrusive Current Detection Electromagnetic valve

Abstract

This article has carried out the research on the fault diagnosis technology of electromagnetic valve based on the current detection of the drive end, collected the drive end current signal of the electromagnetic valve under different conditions, and explored the fault pattern recognition method of the electromagnetic valve. Firstly, the research team set four typical conditions of electromagnetic valve normal operation, spring broken, spool stuck, and spool stuck slightly; secondly, the research team collect the current signal in the process of electromagnetic valve energization and pull; thirdly, the research team use wavelet packet decomposition method to decompose and reconstruct the current signal; fourthly, the energy value of each frequency band is extracted as the characteristic value of the electromagnetic valve under different conditions; fifthly, the characteristic vector of the electromagnetic valve state based on the frequency band energy is constructed; finally, the research team carry out the method based on the support vector machine to identify the electromagnetic valve fault mode. Aiming at the problem of low recognition accuracy and slow recognition speed in the traditional neural network recognition process, the electromagnetic valve fault mode recognition method based on the support vector machine is adopted. The experimental results show: (1)the method based on “entropy-fault” can realize the diagnosis of typical electromagnetic valve faults; (2)the fault diagnosis method of electromagnetic directional valve based on multi-class support vector machine can improve the recognition accuracy by 30.43% compared with the traditional neural network recognition method.