基于主成分分析法与聚类分析法的安全指标体系优化  

Safety Index System Optimization Based on Principal Components Analysis and Clustering Analysis Method

作　　者：马昕晖 智文书 陈景鹏 李忠光[3] 

机构地区：[1]装备学院航天装备系,北京101416 [2]装备学院研究生管理大队,北京101416 [3]军事交通学院装备保障系,天津300161

出　　处：《装备学院学报》2014年第6期58-62,共5页Journal of Equipment Academy

基　　金：部委级资助项目

摘　　要：航天发射试验安全评价中涉及众多风险指标,各指标的选取是否合理,将直接影响安全评估的结果。采用层次分析法对影响航天发射试验安全的设备操作人员评估指标体系进行权重确定,并对评估指标进行了重要性检验,删除相对不重要的指标。采用主成分分析法与聚类分析法结合的方式解决指标之间存在信息重叠且具有一定相关性的问题,实现了指标体系的优化。The safety evaluation of space launch test involves many risk indica selection of indicators is reasonable will directly affect the results of the safety tors. Whether the assessment. The weights of space launch hardware test operator safety evaluation index system was determined through analytic hierarchy process（AHP）, then the importance of indicators was tested and relatively unim- portant indicators was deleted. The overlap and some relevant problems between the index informa- tion were resolved by combining with principal components analysis and clustering analysis method, then the study optimized the index system.

关 键 词：航天发射试验 安全指标体系 主成分分析法 聚类分析法 

分 类 号：V55[航空宇航科学与技术—人机与环境工程]