基于混合因果逻辑的尾矿坝事故知识图谱构建与应用  

Construction and Application of Tailings Dam Accident Knowledge Graph Based on Hybrid Causal Logic

作　　者：郭梨[1] 高元[1] 吴昊[1] 杨震[1,2] GUO Li;GAO Yuan;WU Hao;YANG Zhen(School of Resource Engineering,Xi′an University of Architectural Science and Technology,Xi′an 710055,China;Xi′an key Laboratory of Perceptual Computing and Decision-making in Intelligent Industry,Xi′an 710055,China)

机构地区：[1]西安建筑科技大学资源工程学院,陕西西安710055 [2]西安市智慧工业感知计算与决策重点实验室,陕西西安710055

出　　处：《金属矿山》2025年第1期233-242,共10页Metal Mine

基　　金：国家自然科学基金面上项目(编号:51974223);陕西省杰出青年基金项目(编号:2020JC-44);陕西省自然科学基金面上项目(编号:2022JM-274);陕西省重点科技创新团队项目(编号:2023-CX-TD-12)。

摘　　要：针对尾矿坝事故风险分析的复杂性和不确定性,提出了一种基于混合因果逻辑的尾矿坝事故知识图谱构建与应用方法。该方法首先设计了尾矿坝事故风险分析的混合因果逻辑模型框架,针对尾矿坝自身风险,识别确定性因果逻辑关系;针对人为组织失误,识别非确定性的因果关系。在此模型中,事件序列图位于最顶层,用于风险逻辑演化和计算事故发生概率;中间层为故障树,探究关键事件发生的原因;贝叶斯网络位于最底层,分析具有变化性且相互关联的事件或因子的影响,评估人为和组织失效的概率。然后根据所得到的节点及其之间的逻辑关系,采用Python+Neo4j方法转化为基于混合因果逻辑的尾矿坝事故知识图谱。以降雨引发的尾矿坝事故为例,分析了尾矿坝事故的主要原因和影响因素,以及它们之间的因果关系,利用混合因果逻辑模型对尾矿坝事故风险进行了定量和定性的推理和分析,并构建了相应的知识图谱。研究结果表明:该方法能够综合考虑尾矿坝事故风险的复杂性和不确定性,从多个角度以图形化方式描述事故的演化机理,为尾矿坝风险管理提供了一种有效工具。Aiming at the complexity and uncertainty of tailings dam accident risk analysis,a method of construction and application of tailings dam accident knowledge map based on mixed causal logic is proposed.Firstly,the mixed causal logic model framework is designed for tailings dam accident risk analysis,and the deterministic causal logic relationship is identified according to the tailings dam′s own risk.Identify nondeterministic causal relationships for human organizational failures.In this model,event sequence diagram is at the top level,which is used for risk logic evolution and calculating accident probability.The middle layer is the fault tree to explore the causes of critical events.Bayesian networks are at the bottom,analyzing the effects of variable and interrelated events or factors and assessing the probability of human and organizational failure.Then,according to the obtained nodes and their logical relations,the method of Python+Neo4j is used to transform the knowledge map of tailings dam accident based on mixed causal logic.Taking tailings dam accidents caused by rainfall as an example,this paper analyzes the main causes and influencing factors of tailings dam accidents,as well as the causal relationship between them.The mixed causal logic model is used to conduct quantitative and qualitative reasoning and analysis of tailings dam accident risks,and the corresponding knowledge map is constructed.The results show that this method can comprehensively consider the complexity and uncertainty of tailings dam accident risk,and graphically describe the evolutionary mechanism of the accident from multiple perspectives,which provides an effective tool for tailings dam risk management.

关 键 词：混合因果逻辑 知识图谱 尾矿坝事故 风险评估 

分 类 号：TD77[矿业工程—矿井通风与安全]