基于BL-Apriori的民航机务风险关联规则分析  

Analysis on risk association rules of civil aviation aircraft maintenance based on BL-Apriori

作　　者：刘伟伟 王华伟[1] 侯召国 LIU Weiwei;WANG Huawei;HOU Zhaoguo(School of Civil Aviation,Nanjing University of Aeronautics and Astronautics,Nanjing Jiangsu 211106,China)

机构地区：[1]南京航空航天大学民航学院,江苏南京211106

出　　处：《中国安全生产科学技术》2024年第4期27-33,共7页Journal of Safety Science and Technology

基　　金：国家自然科学基金项目(72271123)。

摘　　要：为了解决机务维修领域风险分析过程中文本数据挖掘不充分问题,提出1种基于二进制逻辑“与”运算改进的Apriori关联规则挖掘方法,采集到飞机维修企业2012—2021年机务维修领域不安全事件报告,最终成功挖掘出民航机务维修领域具有耦合性、关联性的风险要素,并对风险要素之间的关联规则进行分析。研究结果表明:改进后的算法运行时间从0.153 s降低至0.034 s,挖掘到机务风险中飞机检查不全面、人员遗忘/疏漏等为主要因素,与工程实际相符。研究结果可为机务安全管理提供决策支持。In order to solve the problem of insufficient text data mining in the risk analysis process of aircraft maintenance,an improved Apriori association rule mining method based on binary logic AND operation was proposed.By using the unsafe event reports of aircraft maintenance collected from aircraft maintenance enterprises from 2012 to 2021,the risk elements with coupling and correlation in the civil aviation maintenance field were successfully mined,and the association rules between risk elements were analyzed.The results show that the running time of the improved algorithm is reduced from 0.153 s to 0.034 s.The main factors identified in the mining of maintenance risks include incomplete aircraft inspections and personnel forgetting/negligence,which are consistent with the actual engineering situation.The research results can provide decision support for maintenance safety management.

关 键 词：民航机务维修 关联规则挖掘 APRIORI 数据挖掘 逻辑运算 

分 类 号：X949[环境科学与工程—安全科学]