摘要
基于政府A部门提供的502起建设工程事故调查报告,综合应用比较分析、数据挖掘、网络层次分析法(ANP)和超级决策(SD)决策软件,建立建筑安全管理绩效评价框架,包括6个一级指标和24个二级指标;通过分析各因素之间的关系,修正各个因素的权重,得出事故致因的传递路径。根据和其他偏定性研究的比较还可以发现,基于数据挖掘的ANP和SD的综合运用,更利于对真实情况进行研究模拟,在应对较大建设工程事故风险时更具有针对性和前瞻性。
Based on the data of 502 construction accident investigation reports from Chinese government A department,and by comprehensively applying comparative analysis,data mining,Analytic Network Process (ANP) and the Super Decisions (SD) decision-making software,a construction safety management performance evaluation framework is established,including 6 first-level indicators and 24 second-level indicators. Through analyzing the relationship among the indicators,the weight of each indicator is revised and the transmission path of accident causation is obtained. According to the comparison with other partial qualitative research,it can be found that data mining and application of ANP and SD are more favorable to the simulation of actual situations and are of more direction and foresight in the control of major construction accident risk.
出处
《中国安全科学学报》
CAS
CSCD
北大核心
2010年第4期110-115,共6页
China Safety Science Journal
基金
国家科技支撑计划子课题资助(2006BAJ01B04-03)