关键词: 机械装备, 电液伺服阀, 故障诊断, BP神经网络, GA+LM算法

Abstract: The BP neural network had some shortcomings, such as the learning speed was very slow, easy to fall into local minima, and was sensitive to initial parameters. To improve the precision and speed of fault diagnosis of electro hydraulic servo valves, a new combination optimization method was proposed. GA was used to optimize the initial weights and thresholds of the neural network to improve the training speed and to reduce the BP neural network's sensitivity to initial parameters. And LM algorithm was used to train accurately and search for the global optimal solution in the local solution space. In the premise of preserving the mapping ability of the BP neural network, the method may improve the learning speed and accuracy of the networks, and thus the efficiency and accuracy of fault diagnosis of electro hydraulic servo valves may be greatly improved. At last, the MOOG D761-2716A servo valve's diagnosis results were given, which explained the practicality and efficiency of this method.

Key words: machinery equipment, electrohydraulic servo valve, fault diagnosis, BP neural network, genetic algorithm(GA)+Levenberg-Marquardt(LM) algorithm