基于HMM算法的矿山水灾害链复杂演化网络模型构建及应用

Construction and application of complex evolutionary network model of mine water disaster chain based on HMM algorithm

矿山水灾害可视为由各种相互联系的因素构成的复杂系统,通过研究矿山水灾害链中节点事件对演化过程的影响,可为矿山断链减灾提供依据。采用灾害链理论结合复杂演化网络模型,提出了灾害节点演化性子模型、灾害节点影响度子模型、灾害节点重要度子模型、灾害综合中心性模型4个子模型,利用HMM算法对子模型进行优化,构建了基于HMM算法的矿山水灾害链演化网络模型;参照前人的案例分析结果,对比分析矿山水灾害链演化网络模型的可靠性,结果表明:该模型能更有效地分析出各节点事件对整体灾害链演化的影响程度,推算出最有可能发生的矿山水灾害演化链,从而为矿山灾害链的风险分析以及断链减灾提出更合理的建议。

Mine water disaster can be regarded as a complex system composed of various interrelated factors.Studying the influence of node events on the evolution process of mine water disaster chain can provide the basis for breaking and shortening mine disaster chain.Based on disaster chain theory and complex evolutionary network model,four sub-models subject to disaster node evolution,disaster node impact degree and disaster node importance degree and disaster comprehensive centrality were proposed.The sub-models were optimized by HMM algorithm and the evolutionary network model of mine water disaster chain based on HMM algorithm was constructed.With reference to the results from previous case studies,the reliability of the mine water disaster chain evolution network model was compared and analyzed.The results show that the model can be applied in effective analysis of the influence level of each node event on the evolution of the whole disaster chain and to determine the evolution chain that mine water disaster is most likely to take place so as to make reasonable suggestions on risk analysis and for breaking or shortening the mine water disaster chain.