基于贝叶斯网络的铁路运营安全预警模型研究

Early Warning Model of Railway Operation Safety Based on Bayesian Network

准确掌握现场作业安全形势、科学预测安全趋势是确保铁路安全生产的重要基础。铁路现场日常安全检查过程中发现的安全问题比较客观地反映铁路运营的安全状况,在对铁路安全问题进行分类的基础上,应用贝叶斯网络理论构建铁路运营安全预警模型,通过数据验算,该模型可以在已知安全问题的条件下,对可能发生的事故类型及其概率进行预测;或者在已知发生某种类型事故的前提下,逆向推理可能导致该事故的安全问题,从而达到安全预警、辅助安全决策的作用。

Accurately grasping the safety situation of on-site operation and scientifically predicting the safety trend are important foundations for securing smooth railway transportation. The safety problems found during the daily safety inspection on railway sites can objectively reflect the safety status of railway operations. In this paper, based on the classification of safety problems, the Bayesian network theory was applied to build the early warning model of railway operation safety. By calculation with the actual data, the model can predict the types and probability of possible accidents in the condition of known safety problems. Or on the premise that a certain type of accident is known, the model can reversely infer the safety problems that may cause the accident, thus achieving the purposes of early safety warning and assisting safety decisions.