Fault diagnosis of hydraulic actuator based on improved convolutional neural network

This paper proposes a fault diagnosis approach for hydraulic actuator based on short-time Fourier transform and convolutional neural network. The common failure modes of hydraulic actuator include external leakage, internal leakage and crawling, while it is difficult to measure and diagnose above failures with traditional fault diagnosis method. This paper focuses on the signal variance of pressure of rodless chamber of actuator, extract the effective fault features with Short-Time Fourier Transform (STFT) and use convolutional neural network to carry out the fault diagnosis of the leakage and crawling of actuator with time-frequency image. Simulation results show that the proposed method has good accuracy in distinguishing classic failures under different operating conditions.