Fault Diagnosis Based on Improved Elman Neural Network for a Hydraulic Servo System

Due to the nonlinear, time-varying, ripple coupling property existed in the hydraulic servo system, and slow convergence speed and the instability of BP network, a two-stage improved Elman neural network model is developed to realize failure detection. The first-stage Elman network is adopted as a failure observer to realize the failure detection. The trained Elman observer, working concurrently with the actual system, accepts the input voltage signal to the servo valve and the measurements of the ram displacements, then rebuilds the system states. The output of the system is accurately estimated. By comparing the estimated output with the actual measurements, residual signal is generated and then analyzed to report the occurrence of faults. The second-stage Elman neural network can locate fault occurred through the residual and net parameters of the first-stage Elman observer. Improved Elman neural network adds internal self-connections signal of the context nodes, so fasten convergence speed and can better identify the nonlinear dynamic system. The experimental results indicate that the improved Elman neural network model is effective in detecting the failure of the hydraulic servo system