基于SPA-Markov的飞行安全态势评估与预测研究

为提高航空公司飞行安全风险控制和预测水平，利用QAR数据作为支撑，结合集 对分析和马尔科夫理论建立飞行安全态势评估模型。该模型以QAR超限事件为评估指标 ，采用集对分析中的联系度来描述安全风险等级；运用马尔科夫理论确定安全状态的转 移概率矩阵，以预测飞行安全动态变化趋势。最后以某公司A320机队6个月发生频率最 高的7类超限事件对模型进行验证。结果表明:该公司2月、6月和预测月份的安全状态 为一般风险，其他月份为低风险；总体安全状态也为一般风险，并有增加趋势。

Research on assessment and prediction of flight safety situation based on SPA-Markov

In order to improve the risk control and prediction level of flight safety for airlines, taking the QAR data as support, an assessment model of flight safety situation was established by combining set pair analysis (SPA) with Markov theory. In this model, the QAR transfinite events were taken as the assessment indexes, then the connection degree in SPA was adopted to describe the safety risk level, and the Markov theory was applied to determine the transition probability matrix of safety state, so as to predict the dynamic variation trend of flight safety. Finally, the model was verified based on the data of 7 kinds of transfinite events which occurred most frequently in A320 fleet of a certain airline within 6 months. The results showed that the safety status of this airline in February, June and the predicted month are general risk, and those in the other months are low risk. The overall safety status is also general risk, with an increasing trend.