摘要
对航班运行风险因素进行评估,筛选风险因素,分析航班运行风险传播动态,以减少由航班安全问题而带来的损失。研究运用随机森林算法对各风险因素节点进行排序筛选,结合中国民用航空局2015年发布的咨询通告《航空承运人运行控制风险管控系统实施指南》(AC-FS121-2015-125)实现对风险因素的降维,共统计了华北区域225天航班运行风险数据的97个节点。再根据民航不安全事件记录,分析事件因果关系,构建有向加权航班运行风险网络。结果显示,当不对风险因素进行控制时,风险传播网络峰值为13.69,当按总度值排序对风险因素进行控制时,风险传播网络峰值降为7.18,下降率49.3%。
Assessing flight operation risk factors,screening risk factors,and analyzing flight operation risk propagation dynamics are important to reduce losses caused by flight safety issues. The research uses the random forest algorithm to sort and screen each risk factor node,and combines the advisory circular "Air Carrier Operation Control Risk Management and Control System Implementation Guide"(AC-FS121-2015-125)issued by the Civil Aviation Administration of China in 2015 to realize the risk factors. A total of 97 nodes of 225-day flight operation risk data in North China are counted. Then,according to the records of civil aviation unsafe events,the causal relationship of the events is analyzed,and a directional weighted flight operation risk network is constructed. The results show that when the risk factors are not controlled,the peak value of the risk propagation network is 13.69. When the risk factors are controlled according to the total degree value,the peak value of the risk propagation network is reduced to 7.18,and the decline rate is 49.3%.
作者
赵宁宁
林典毅
ZHAO Ningning;LIN Dianyi(College of Air Traffic Management,Civil Aviation Universily of China,Tianjin,300300,China)
出处
《现代交通与冶金材料》
CAS
2022年第5期12-19,共8页
Modern Transportation and Metallurgical Materials
基金
国家重点研发项目(2020YFB1600101)。
关键词
航班运行风险
随机森林算法
有向加权网络
传播动力学
flight operation risk
random forest algorithm
directed weighted network
propagation dynamics
