基于模糊粗糙集的化工园区多米诺事故救援时机分析  

Research on the Rescue Opportunity for Chemical Industry Park Domino Accident Based on Fuzzy Rough Set

作　　者：段在鹏[1] 钱新明[1] 多英全[2] 侯云娟[1] 刘储朝 蒋锦辉[1] 

机构地区：[1]北京理工大学爆炸科学与技术国家重点实验室,北京100081 [2]中国安全生产科学研究院,北京100012

出　　处：《工业安全与环保》2015年第6期39-42,共4页Industrial Safety and Environmental Protection

基　　金："十二五"国家科技支撑计划项目(2012BAK13B01)

摘　　要：本文旨在通过粗糙集约简手段研究化工园区多米诺事故救援时机问题。首先将初始事故等级、间距系数、环境等级、临近企业脆弱性、初始事故发生时间等5个因素作为多米诺事故发生条件,应用Kmeans聚类方法实现模糊粗糙集数据离散化,之后应用属性约简及属性值约简方法分析案例信息,最后提取规则并从以上5个因素角度确定火灾多米诺事故的救援最佳时机,结果表明救援规则简洁明了,可执行性较高。The thesis intends to use the rough set reduction method to study the rescue opportunity for Domino accident in the chemical industry park.Firstly,level of primary accident,distance,environmental level,vulnerability of adjacent enterprise,time of occurrence for the primary accident are the conditions for Dominion accidents,then historical data together with the five factors are used to build fuzzy rough set decision tables; K-means clustering is used for continuous data discretization.The ‘distance＇and ‘time of occurrence for the primary accident＇are both continuous-value attributes,reference to other attributes categories,so we divide the both two continuous-value attributes into three levels with K-means clustering method firstly.After that,the attribute reduction and attributive value reduction are used to analyze and draw rules for the occurrence of Domino accidents in the chemical industrial park.Finally,the optimum rescue opportunity is confirmed for Domino accidents.The rescue tasks are urgent and complicated in the rescue scene.So the straightforward rescue rules are easier to be carried out.

关 键 词：化工园区 多米诺事故 模糊粗糙集 聚类分析 属性约简 属性值约简 规则提取 

分 类 号：TQ086[化学工程]