Research On Fault Diagnosis Of Retraction And Extension Of Aircraft Flap And Slat Based On QAR Data

The flat and slat system of aircraft is a subsystem of the aircraft manipulator system,which is an important system for aircraft to increase lift and drag.Retraction and extention fault of flat and slat is a common fault of aircraft manipulator system,and aircraft QAR(Quick Access Recorder)data record various parameters about the operation state of aircraft,which is an important basis for aircraft fault maintenance and prediction.The present studying takes the flat and slat as the research object,taking the aircraft QAR data as the foundation,according to the actual engineering research data,using the artificial neural network algorithm and the mathematical statistics principle,carring out the research on retraction and extension fault condition monitoring and fault diagnosis method of the flat and slat.The research can provide new methods and ideas for the design and development of domestic civil aviation aircraft health management system and the application of aviation engineering.In terms of artificial neural network algorithm,flap asymmetry fault is a typical fault of flat and slat system.Because Elman neural network is sensitive and dynamic to historical data and the network itself has a strong ability to process dynamic information,the present studying uses Elman neural network to establish a prediction model of left-right angle difference of trailing edge flap.Improved beetle antennae search is used to optimize initial weights and threshold parameter of the Elman neural network model.The input parameters for the model are determined by the study on operating principle and dynamic impact of flat and slat.In the end,Elman neural network model of the trailing edge flap about angle difference is established by improved beetle antennae search algorithm when the parameters of the neural network and beetle number are determined.The network model after training is validated by test data,and conclusion is obtained that the IBAS-Elman prediction model about trailing flaps angle difference has the hightest prediction accuracy,the best generalization,and the strongest ability of learning by the Elman,BAS-Elman,IBAS-Elman evaluation index of three kinds of prediction model.In terms of mathematical statistics,for further research on flat and slat fault,firstly,the trailing edge flaps asymmetric risk model is established based on the Monte Carlo method that improves the deficiency of the data,and the risk values are quantified into four grades.The reliability and the practicability of the model are verified by collecting the fault cases,and conclusion is obtained that the left-right angle difference of trailing edge flap of different plane are different at the same risk value.The model can provide a new method and idea for trailing edge flaps asymmetric fault.Secondly,the time of retraction and extention of the flat and slat is studied based on the leading edge flat and slat fault case.Combining with the retraction and extention operating principle of the flat and slat,the follow-up response rate is put studied.The follow-up response rate is divided into several dimensions by two states of air and ground.In order to assist airlines in unplanned maintenance,a retraction and extention monitoring platform of aircraft flat and slat is developed based on ASP.NET.