塔式起重机事故案例关联规则挖掘与分析

Mining and analysis of association rules in tower crane accident cases

为充分利用塔式起重机事故案例信息,深入挖掘事故特征,提出一种改进的Apriori算法挖掘模式,快速有效挖掘塔式起重机事故关联规则。首先,收集200份具有详细事故调查报告的塔式起重机事故案例,分析并提取事故调查报告中事故的属性数据,按照事故属性的概念层次结构编码;然后,基于经典的Apriori算法,提出一种适用于多维多层关联规则挖掘的模式,挖掘塔式起重机事故属性与致因间多维多层的关联规则;最后,根据挖掘结果,分析并总结塔式起重机事故特征。结果表明:关联规则能有效利用塔式起重机事故调查报告信息,用定量的方式挖掘事故特征;塔吊事故属性间以及事故致因间有较强关联关系。

In order to make full use of tower crane accidents information and mine deep into tower crane accident characteristics,an improved Apriori algorithm mining pattern was proposed,association characteristics of tower crane accident can be quickly and effectively mined.Firstly,200 detailed accident investigation reports of tower crane were collected,accident attribute data in accident investigation report were analyzed and extracted,and coded according to the hierarchical structure.Secondly,a pattern suitable for multi-dimensional association characteristic was proposed based on classic Apriori algorithmt,association characteristics between tower crane accident attributes and causes were mined.Finally,tower crane accident association characteristics were analyzed and summarized according to mining results.The results showed that association rules algorithm can effectively utilize investigation of accident reports and mine characteristic of tower crane accidents in a quantitative way.Between tower crane accident attributes and accident causes have a strong association relation.