摘要
为解决矿山日常安全隐患排查数据不规范、非结构化、生命周期短和利用程度低等问题,运用大数据分析技术研究与开发矿山安全隐患辨识与预警系统。根据矿山安全隐患大数据分析场景,在大数据平台和云计算平台基础上规划系统整体架构、功能体系与大数据分析逻辑。通过矿山安全隐患语料库、停用词库的构建,规范矿山常见隐患主题词,借助词库对安全隐患进行标准化分词,运用“六何分析方法”构建以时间、地点、类型、程度、致因、责任主体为要素的多维数据集。针对矿山安全隐患大数据特征,运用主题挖掘、语义网络分析、关联规则、分类与预测模型完成危险源和安全风险主题辨识、安全隐患知识图谱构建、安全隐患致因规律挖掘以及安全风险预警,借助商务智能分析工具完成可视化系统的搭建。系统应用于国内大型矿山企业,实现了安全隐患智能识别、诊断、预测与预警,大幅提高安全管理人员对安全隐患认识、管理和决策能力。
In order to solve the problems of nonstandard,unstructured,short life cycle and low utilization of daily mine safety hazard investigation data,the big data analysis technology is used to research and develop mine safety hazard identification and early warning system.According to the big data analysis scenario of mine safety hazards,the overall architecture,functional system and big data analysis logic of the system are planned on the basis of big data platform and cloud computing platform.Through the construction of mine safety hazard corpus and stop use thesaurus,the paper standardizes the subject words of mine common hazards,uses the thesaurus to conduct standardized word segmentation for safety hazards,and uses the"5W1H analysis method"to build a multidimensional data set with time,place,type,degree,cause and responsibility subject as the elements.According to the characteristics of big data of mine safety hazards,theme mining,semantic network analysis,association rules,classification and prediction model are used to complete the theme identification of hazard sources and safety risks,the construction of safety hazard knowledge map,the mining of safety hazard causes and safety risk warning,and the construction of visualization system is completed with the help of business intelligence analysis tools.The system has been applied to large domestic mining enterprises to realize the intelligent identification,diagnosis,prediction and early warning of potential safety hazards,and greatly improve the safety management personnel′s awareness,management and decision-making ability of potential safety hazards.
作者
李国清
李学玉
侯杰
强兴邦
王浩
国祯翔
赵威
LI Guoqing;LI Xueyu;HOU Jie;QIANG Xingbang;WANG Hao;GUO Zhenxiang;ZHAO Wei(School of Civil and Resource Engineering,University of Science and Technology Beijing,Beijing 100083,China;Sanshandao Gold Mine,Shandong Gold Group Mining(Laizhou)Co.,Ltd.,Laizhou 261442,China)
出处
《金属矿山》
CAS
北大核心
2022年第6期129-137,共9页
Metal Mine
基金
国家自然科学基金项目(编号:52074022)。
关键词
矿山安全
隐患辨识
事故预警
大数据分析
系统研发
mine safety
hidden danger identification
accident warning
big data analysis
system development