基于链式变精度粗糙模糊集采场内因火灾分析  

Analysis of Internal-caused Fire in the Stopes Based on Chain Variable Precision Rough Fuzzy Set

作　　者：杨珊[1] 袁鸣珂 苏凯俊 徐梓桐 YANG Shan;YUAN Mingke;SU Kaijun;XU Zitong(School of Resources and Safety Engineering,Central South University,Changsha 410083,Hunan,China)

机构地区：[1]中南大学资源与安全工程学院,湖南长沙410083

出　　处：《黄金科学技术》2022年第1期93-104,共12页Gold Science and Technology

基　　金：国家自然科学基金青年基金项目“基于人工智能的矿山技术经济指标动态优化”(编号:51404305);国家自然科学基金项目“基于属性驱动的矿体动态建模及更新方法研究”(编号:51504286)联合资助。

摘　　要：针对地下矿山采场内因火灾影响因素多且复杂,涵盖自燃起火与火灾后果多个层面的情况,综合考虑矿石含硫量、含碳量和矿石温度等因素,建立了采场内因火灾多层面分析的指标体系,构建以影响指标为条件属性和以采场内因火灾自燃起火及火灾后果为决策属性的链式变精度粗糙集模型。对采场内因火灾的自燃起火可能性及火灾后果严重性进行了链式关联决策分析,得到了9条置信度为55%以上的采场内因火灾自燃起火决策规则及5条置信度为50%以上的采场内因火灾后果的决策规则。根据链式因果关系对两类规则进行归并整合,得到了采场内因火灾涵盖起因与后果的组合规律。结果表明:该方法能够科学全面地分析采场内因火灾等存在链式关联层面的不确定概率性问题,为矿山预判、应对采场内因火灾及保障安全生产提供了指导。The sulpfur content of ores in an underground pyrrhotite is nearly 50%.Internal-caused fires are prone to happen in such mines because of ore spontaneous combustion,and the internal-caused fire has become a major hidden danger to the sofety production of mine enterprises.In the underground mine stopes,there are many complex factors affecting the internal-caused fire.It covers multiple levels of the situation about cause and effects of the fire in the stopes.Considering 9 influencing factors such as sulfur content and temperature of ore,the impact indicators,probabilities of spontaneous combustion and its results of the internal-caused fire in the typical stopes of the mine were statistically analyzed.Then a multi-level analysis index system for fire in the stopes was constructed,and the knowledge system of rough fuzzy set was established.The condition attribute of the model is the impact indicators of internal-caused fire,and the decision attribute of the model are consequences severity and judgement result of spontaneous combustion of internal-caused fire in the stopes.For the above two kinds of knowledge systems,the sample object sets of the stope is divided into classes by equivalence partitioning according to condition attributes and decision attributes.The classification quality of the conditional attribute set of the spontaneous combustion knowledge system and fire consequences knowledge system is 70%and 67%respectively.Then,the conditional attribute set is reduced by using the theory of variable precision rough fuzzy set.After a chain correlation decision-making analysis which contained the possibility of spontaneous combustion and the severity of the internal-caused fire consequences in the stopes,9 decisionmaking rules for spontaneous fire with a confidence level above 55%were obtained,and 5 decision-making rules for fire consequences in the stopes were also obtained with a confidence level above 50%.According to the cause-effect chain,the two classes of rules were merged and integrated,and 30 combinatio

关 键 词：采场 内因火灾 粗糙模糊集 变精度 知识约简链式关联 矿山安全生产 

分 类 号：X43[环境科学与工程—灾害防治]