Hydraulic System Fault Diagnosis Based On Emd And Elman Network Technology Research

Construction machinery is widely used in various industries projects. The improvement of the degree of automation increases the difficulty of maintenance of equipment at the same time, complex maintenance processes and maintenance of technology to make fault diagnosis technology develop rapidly. As information processing technology and remote communication technologies become more sophisticated, remote fault diagnosis system can complete the function of real-time monitoring, predictive, diagnostic and troubleshooting, to ensure the normal operation of the machinery and resolve the complex issues more timely which traditional fault diagnosis method does not do. Intelligent fault diagnosis technology has a profound theoretical significance and broad application space.This paper mainly studies the remote fault diagnosis system of intelligent fault diagnosis technology, and analyses the main intelligent diagnosis technologies widespread used at present. The neural networks fault diagnosis model is used in diagnosing the type of fault of construction machinery hydraulic system. Pressure, temperature, flow, and vibration signals are selected as fault diagnosis parameters. According to the characteristics of each parameter pre-treatment is different, the dominant signal of pressure, temperature and flow are normalized, and the noise of vibration signal is reduce by wavelet analysis, and then the fault feature vector is extracted by empirical mode decomposition method based on GM(1,1) and mirrorzing extension. The four kinds of signal after processing as the neural network inputs are used to diagnose the type of fault.In order to improve the computing ability and precision of the diagnostic model, the Elman neural network architecture is selected. The inertia weight and learning factors of the particle swarm optimization is improved in this paper, and then the improved particle swarm optimization is used to optimizing the weights and threshold matrix of the Elman neural network which convergence rate and diagnostic accuracy is better. Final the interactive interface is designed to provide the intelligent diagnosis model with expositive features and equipment information storage and query functions.Experimental results show that, intelligent diagnosis model based on empirical mode decomposition and the Elman neural network can effectively analyze the failure of the hydraulic system. Compared to the BP network model the method in this paper has been improved the accuracy and effectiveness, reduced diagnostic errors, improved diagnostic accuracy, and possess practical significance.