基于贝叶斯网络的电力安全突发事件情景推演方法研究

Research on Electric Power Emergency Scenario Deduction Method Based on Bayesian Network

电力安全突发事件难预测、演化复杂、危害大,而传统推演方式灵活性、经济性差,容错率低。以美加大停电为例,采用情景分析法分析其演化过程,由典型大停电案例总结出电力安全突发事件的一般演化过程;围绕情景状态、处理方法和处理目标3个要素分析了美加大停电事件演化的多条路径,构建了事件发展的情景网络,计算网络节点间的条件概率。基于Netica仿真软件搭建了事件的贝叶斯网络并实现了多路径推演。推演结果表明,不同的处理方法能改变事件演化路径和情景概率,该模型能较好地模拟真实事件。

Electric power safety emergencies are difficult to predict and evolve in a complex manner and harmful. Taking the outages in America and Canada as an example,this paper analyzes the evolution by scenario analysis method,and summarizes the general evolution of power safety emergencies through typical outages;it also analyzes the paths of outages evolution in America and Canada based on scenario state,treatment method and treatment target,and constructs the situation network of event development and calculates the conditional probability among network nodes. Based on Netica simulation software,the event Bayesian network is built and multi-path deduction is realized. The result shows that different treatment methods can change the event evolution path and scenario probability,and the model can simulate the real event.