The Study Of APU Intelligent Fault Diagnosis Based On Sensitive Parts Detection

Auxiliary power unit(abbreviated APU) is an important equipment on the aircraft, which is second only to the aero-engine. The health of APU directly affects the airline's operating costs and social benefits. In the repair process of APU, in order to solve the excessive maintenance, less maintenance, low level of automation and intelligent fault diagnosis and other issues, this paper carried out research of APU fault intelligent diagnosis method based on APU's monitor sensor.This paper firstly discussed the wavelet algorithm's relevant characteristics and demonstrated the feasibility of wavelet filtering algorithm to deal with the APU's monitor sensor's data. The improved "Min-Max" algorithm was creatively used to judge whether the sensor was malfunction. The essence of fault intelligent diagnosis is through analysis the variation trend of APU's performance parameters. Therefore, on the basis of sensor fault diagnosis and combined with neural network technology. The paper build fault diagnosis model based on improved BP neural network and improved RBF neural network.The result of simulation and actual experiment indicated that, the intelligent fault diagnosis model achieved expected target. At the same time, the results also indicated that the fault diagnosis model which based on RBF network have a higher learning efficiency and higher diagnostic accuracy. This paper's result have an important meaning in improve the level of automation maintenance and airline's economic benefits.