Fault Diagnosis of Hydraulic Syetem Based on Neural Network

Abstract

Because neural network has the advantages of fast parallel processing, associative memory, self-organizing and self-learning, it is widely applied in the fault diagnosis of hydraulic system. Present in this paper is a fault diagnosis approch to a typical failure in hydraulic system which is leakage of hydraulic cylinder.The fault diagnosis approch is based on monitoring preesure singal,time domain feature and neural network. According to the method, the time domain feature is extracted from the pressure singal and costitutes the eigenvectors at first, then these eigenvectors are input into neural network to identify faults. The experimental results show that three modes of no leakage, slighter leakage and severe leakage are correctly identified and it can be used in the fault diagnosia of hydraulic syetem.