Research On Hard Landing Diagnosis Of B737

The safety of aviation is the eternal issuse that attracts public concerns. Seven hazardous accidents happended in the domestic civil aviation transportation from 1996 to 2005, two of them resulted from hard landing. Hard landing becomes the major threat to the aviation safety, which attracts broad attention. In the base of investigation in the airways and analysis of related ragulations, the hard landing diagnosis method is insufficiently and not satisfyed by pilots and aircraft crew.Firstly, the causes of hard landing and related factors are analyzed in the view of mechanics, with the risk parameters being confirmed. Secondly, depending on the information of flight data recorded by QAR, signal processing method is used to decrease the noises. Then, the intelligent diagnosis model is established based on nerual network and fuzzy technology, and the model is trained and validated by the data from B737. the results show that the intelligent model can judge the hard landing effectively, feasibly and intelligently. The risk management is adopted, the pre-warning of hard landing is discussed by using grey system and exponential smoothing method, the results show that grey system is feasible for the hard landing pre-warning.Finally, the integrated system is developed in the platform of Delphi by using database technology and FOQA software. With its three subsystems : diagnosis system, pre-warning system and hard landing monitor database, the whole system could be provided to airways to decrease the risk of hard landing.