基于复杂网络的化工事故致因分析

Analysis on causation of chemical accidents based on complex network

为明确导致化工事故发生的各风险因素间的关联性以预防事故的发生,将收集到的2015-2019年间143起事故调查报告进行数据挖掘与清洗,从人的因素、物的因素、环境因素、管理因素4类主风险因素出发逐步细化确定了26类子风险因素.采用Gephi软件构建化工事故致因网络模型,通过对模型的网络规模与密度、加权度、边权、聚类系数、PageRank值分析,研究模型的整体特性及各节点的联系,确定关键节点并对未来化工管理的重点与方向提出建议.

In order to clarify the correlation between various risk factors leading to chemical accidents,and prevent chemical accidents from happening,data mining and cleaning are done upon the collected investigation reports for 143 accidents happened between 2015 and 2019. From four main risk factors such as human factors,material factors,environmental factors,and management factors,26 types of sub-risk factors are identified.Gephi software is employed to establish a network model for causation of chemical accidents.Through analyzing the network scale and density,weighting degree,edge weight,clustering coefficient,and PageRank value,the overall characteristics of the model and the connections between nodes are studied,key nodes are identified,and the suggestions about future chemical management priorities and directions are provided.