Research On Flight Risk Forecast Method Based On QAR Data Analysis

In order to realize flight risk forecast based on QAR dada, this thesis developed the research of applying statistical process control into the flight risk forecast based on QAR data.Currently, there are certain amount of deficiencies in Flight Operational Quality Assurance,one of them is sending alarm only when exceedance event happens and this might waste a lot of information which is within limit, besides, lacking the management of QAR data noise,ignoring alarm or sending false alarm could also cause the unexpected results. In order to fill up these deficiencies, this thesis came up with the solution of applying statistical process control into the flight risk forecast based on QAR data, and introduced the rationality as well.Practices have proved that most QAR data exceedance events happens for a reason, and there are certain “signs” before its happening. This thesis took the QAR data of B737-800 as an example, combined rare event theory, developed a series of criterions which could decide whether the flight is stable or not. These criterions are the structure of flight risk forecast model based on QAR data. The new criterions are more intensive so it would not cause any conflicts with the current Flight Operational Quality Assurance. Finally this thesis demonstrated flight risk forecast with actual QAR data, the result indicated that this new method could make scientific risk forecast based on QAR data, it could realize risk management on both individuals and group after making minor modifications based on specific background.After the realization of real-time QAR data transmission, the flight risk forecast system based on real-time QAR data could be designed based on the method in this thesis. It would make more contribution to aviation safety management.