Study On Failure Trend Forecasting And Coping Strategies Of APU

Auxiliary Power Unit(APU) is an integral and important system in aircraft. However,its high failure rate affectes the normal operation of aircraft. In order to improve aircraft's normal rate and safety and to reduce airline's operating costs, research on failure trend forecasting and coping strategies of APU is carried out in this paper. The main research work is described as below:By analyzing research status of failure prediction both domestic and abroad, the difficulties and shortages of present failure prediction techniques are summarized. In view of the single traditional prediction method's defects that can not meet the demand of complex aircraft APU fault prediction, the combined method of chaotic time series prediction and fault diagnosis based on BP neural network is proposed to achieve fault prediction.The chaotic characteristics are verified for time series of numerous monitoring data that received from a airline's pratical operating data of APU. The predictive value of APU performance parameters is obtained by chaotic prediction. Compared with the results of exponential smoothing forecasting, the feasibility and accuracy of chaotic prediction is verified. The failure modes that consist of the results are diagnosed by using BP neural network, the outcome of which is the result of failure prediction. Given the accordance of computer simulation results and the expected ones, the validity of the proposed method of failure trend prediction is demonstrated. Chaotic time series prediction method has high accuracy and reliability, while BP network can produce a reasonable output through such data that not in the training set by its learning generalization ability. The combined method fully takes in the advantages of both method mentioned above in APU's failure prediction,which can provide airlines with the theory of APU's condition monitoring and maintenance decision-making in advance.At last, APU failure analysis is accomplished, based on which the coping strategies of fault prevention, conditionmonitoring, inspection and trouble shooting for APU failure trend are proposed.