Event Margin Model For Flight Operation Quality Assurance

The development of FOQA has played a positive role in improving the civil aviation safety and reducing the accident rate. Nowadays, the method of FOQA is limited to the flying event statistics and decision, and the comprehensive quantitative analysis of the multiple parameters in the process of flight is insufficient. The event margin model for FOQA is studied.Firstly, chose the description of the event margin. Through the comprehensive analysis of the requirement of flight safety, quantitative description of events degree is given in flight operation quality assurance. According to flight quality monitor rules, combined with the feature of kurtosis and skewness value, with the statistical analysis of QAR data, the optimal method for computing event margin is achieved.Secondly, establish the correlative event set. Through the statistical analysis of events at various stages in the process of flight, with the theory of Bayesian, calculating the conditional probability of different events, the correlative event set is established and applied to the declining stage. The results show that the method is available in the analysis of flight quality.Finally, build the event margin model based on BP neural network, and carry on the instance validation. According to the flight quality monitoring standard, the margin model is established with the theory of BP neural network. Fight data is used to validate the event of vertical landing load big,the results show that the event margin model can analysis the event accurately. Apply the model to the whole process of flight, the quantitative monitoring of flight quality can be realized.