基于贝叶斯网络的堰塞湖风险评估方法研究

Study on risk assessment method of dammed lake based on Bayesian network

堰塞湖风险问题一直是国内外学者研究的热点,合理的风险评估方法有利于科学的应急响应。提出了一套基于贝叶斯网络的堰塞湖风险快速定量评估方法,包括堰塞湖危险性评价模块、溃决洪水演进模拟模块和溃决损失评估模块。该模型根据堰塞湖溃决灾害链因果关系提出风险逻辑拓扑结构,通过统计数据、经验模型和逻辑推断结合蒙特卡洛方法对各模块网络节点进行量化,以此得到堰塞湖溃决可能性和各受灾地区的生命损失、经济损失率,从而对堰塞湖风险进行综合评估。并根据国家关于事故损失界定的条例和堰塞湖寿命分布概率结合贝叶斯模型计算堰塞湖风险界定范围,提出了堰塞湖风险等级划分表,最后对白格堰塞湖进行风险评估,评估结果显示白格堰塞湖综合风险值为0.7339,属于极高风险,应进行Ⅰ级红色警报。研究方法可对突发性堰塞湖整体风险进行快速定量评估,为快速应急响应决策的制定提供参考。

The risk of dammed lake has always been the research hotspot of scholars at home and abroad.A reasonable risk assessment method is conducive to scientific emergency response.In this paper,a rapid quantitative assessment method of dammed lake risk based on Bayesian network is proposed,including dammed lake risk assessment module,outburst flood routing simulation module,and outburst loss assessment module.The model puts forward the risk logic topology according to the causality of the dammed lake outburst disaster chain and quantifies the network nodes of each module through statistical data,empirical model,and logical inference combined with the Monte Carlo method,to obtain the possibility of dammed lake outburst and the life loss and economic loss rate of each affected area,to comprehensively evaluate the dammed lake risk.According to the national regulations on the definition of accidental loss and the probability of the life distribution of the dammed lake,combined with the Bayesian model,the risk definition range of the dammed lake is calculated,and the risk classification table of the dammed lake is proposed.Finally,the risk assessment of Baige dammed lake is carried out.The assessment results show that the comprehensive risk value of Baige dammed lake is 0.7339,which belongs to the highest risk,and class I red alarm should be carried out.This method can quickly and quantitatively evaluate the overall risk of a sudden dammed lake and provide reference for decision-making of rapid emergency response.