Hard Landing Risk Assessment And Forecast Research Based On QAR Data

Hard landing is the main unsafe event in approach and landing phase. Because of the high frequency and wide scope, it has become a serious threat to flight safety. Currently, the study of hard landing focused on determination and maintenance, while, with the increasing in civil aviation safety management concept, how to make risk identification, assessment, control, and prediction for hard landing event in the perspective of risk has become a problem to be solved.The vertical load overrun events recorded by QAR in an airline from January 2010 to December 2013 with 300,000 takeoffs and landings have been treated as the data base to evaluate and forecast hard landing events. Firstly, according to the characteristics of hard landing risk identification and common identification method, SHELL model was selected to analysis the vertical load overrun event recorded by QAR, then 34 risk factors have been identified out from the angle of pilot flying, pilot flying with related personnel, pilot flying with hardware, pilot flying with software, pilot flying with environment. Secondly, Entropy TOPSIS method was used to improve risk matrix to make risk assessment, then the risk factors have been divided in 10 levels. Within it, undue stress strength was treated as the most serious risk. Entropy TOPSIS removed the heterogeneity of different indicators, reduced risk knot, which will help the efficient allocation of resources. Thirdly, in the basis of risk assessment, countermeasures were proposed for the risk factors which closeness less than 0.8. Fourthly, with the randomness of vertical load overrun events, autoregression model was selected to forecast, after the verification, the model fit the trend of events and the accuracy is about 90%. Finally, a closed-loop management system was designed based on forecast date, which support airlines to publish the early warning information of hard landing in time.