摘要
在ASIS系统上采集2017-2023年共计86917起不正常事件数据作为研究对象进行挖掘工作,并在此基础上构建不正常事件指标体系,通过对其进行风险评价来保障运行安全。在指标体系的基础上构建了云模型的评价模型,通过对影响因素的风险等级划分后计算云模型的期望、熵和超熵在不同风险等级下的值,并根据输出的云数字特征值将部分指标进行了标准云图的可视化,计算27个三级指标权重。然后,进行实例验证,以56个有详细案发信息的历史事件为样本,对各样本的风险状态进行推理,将样本计算风险等级与实际情况的拟合程度进行可视化分析,来验证所构建模型的有效性。最后,对各级评价指标进行深入分析,为航空运行安全管理工作提供指导意见。
A total of 86,917 abnormal events data from 2017 to 2023 are collected on ASIS system as the research object to carry out mining work,and on the basis of which the indicator system of abnormal events is constructed to ensure the operation safety by risk evaluation.Firstly,the evaluation model of cloud model is constructed on the basis of indicator system,and the values of expectation,entropy and super entropy of cloud model under different risk levels are calculated after the risk level division of influencing factors,and some indicators are visualized as standard cloud diagrams according to the output cloud digital eigenvalues,and 27 third-level indicator weights are calculated.Then,example validation is carried out to verify the validity of the constructed model by taking 56 historical events with detailed case information as samples,reasoning about the risk status of each sample,and visualizing and analyzing the degree of fit between the calculated risk level of the sample and the actual situation.Finally,the evaluation indexes at all levels are analyzed in depth to provide guidance for aviation operation safety management.
作者
杨昌其
姜美岑
林灵
贾志杰
吴戈
YANG Chang-qi;JIANG Mei-cen;LIN Ling;JIA Zhi-jie;WU Ge(Civil Aviation Flight University of China,Guanghan 618000,China;China Harbor Engineering Company Limited,Beijing 100000,China)
出处
《航空计算技术》
2025年第1期44-48,共5页
Aeronautical Computing Technique
基金
中国民用航空飞行学院横向科研项目资助(H2023-100)。
关键词
不正常事件
风险评价
云模型
权重
可视化
irregular events
risk evaluation
cloud model
weighting
visualization
