基于贝叶斯网络的煤矿火区瓦斯爆炸事故情景推演

Scenario deduction of gas explosion accidents in coal mine fire areas based on Bayesian network

为防止煤矿火区瓦斯爆炸事故造成严重的人员伤亡和财产损失,利用知识元、情景构建及推演理论,研究煤矿火区瓦斯爆炸事故的演变过程。以2014年7月某矿瓦斯爆炸事故为例,从情景状态、应急目标、处置措施、孕灾环境四个维度出发,构建基于动态贝叶斯网络预测模型的煤矿火区瓦斯爆炸事故情景网络,计算事故情景发生概率,分析其演变规律。结果表明,煤矿火区瓦斯爆炸事故推演结果与实际发展情景基本吻合,验证了该预测模型的合理性。研究成果可为火区瓦斯爆炸事故风险的防范、应急预案的完善、应急决策等提供技术支持。

In order to prevent serious casualties and property losses caused by gas explosion accidents in coal mine fire areas,the evolution process of gas explosion accidents in coal mine fire areas is studied by using knowledge elements,scenario construction,and deduction theory.Taking the gas explosion accident in a certain mine in July 2014 as an example,a coal mine fire area gas explosion accident scenario network based on dynamic Bayesian network prediction model is constructed from four dimensions:scenario state,emergency objectives,disposal measures,and disaster environment.The probability of accident scenario occurrence is calculated,and its evolution law is analyzed.The results indicate that the inferred results of gas explosion accidents in coal mine fire areas are basically consistent with the actual development scenario,verifying the rationality of the prediction model.The research results can provide technical support for the prevention of gas explosion accident risk in fire areas,the improvement and improvement of emergency plans,and emergency decision-making.