QAR数据在飞行安全评价中的应用

Study on flight safety evaluation based on QAR data

为提高航班飞行安全水平,帮助航空公司制定有针对性的飞行技术提升措施,采用QAR数据建立飞行安全评价模型。模型以QAR超限事件为评价指标,通过确定白化权函数,并利用离差最大化原理得到各指标权重;再应用灰色聚类方法评价一定时期内的飞行安全水平;最后以某航空公司B737-800机队2014年度QAR超限频率较高的5类实例事件对评价模型进行验证。结果表明:该公司6月、9月、12月运行情况很好,而1月、7月运行情况很差,这一结果与实际情况基本吻合,证明模型适用于飞行安全评价。

The aim of this study is to establish a safety evaluation model for evaluating the flight safety level objectively and precisely based on the flight QAR data. The model built by the authors chose QAR exeeedanee events as evaluation indexes. By using white function and maximizing deviations method, the weight of each index could be determined. Then, the grey clustering method was applied to get the evaluation results. Finally, data on five kinds of most frequently occurred QAR exeeedance events of a certain airline＇s B737-800 fleet during 2014 was used to verify the model. The results show that the flight performances in June, September and December are the best while the ones in January and July are the worst. The evaluation result conforms basically with the practical situation. Therefore, the model can be used to objectively and precisely evaluate the flight safety level.