摘要
为了提高高速公路应急管理能力,提升高速公路应急效率,从突发事件演化机理着手,基于贝叶斯网络构建了高速公路交通突发事件应急处置的态势评估模型,用于推理学习不同情况下的交通事故概率。评估模型首先确立了高速公路交通突发事件贝叶斯网络总体工作流程,将贝叶斯网络节点设置为高速公路突发事件的影响因素,贝叶斯网络结构根据各节点,即各影响因素之间的相互关系进行构建。其次,根据交通事故数据和专家知识对网络中每个节点进行条件概率赋值,再在条件概率的基础上根据现场初步事故信息运用联合树推理算法将贝叶斯网络转化为联合树,通过定义在联合树上的消息传递过程,计算后验概率;最后,在条件概率的基础上进行推理,建立评估模型实现对高速公路突发事件态势评估。实例分析随机抽取50次高速公路交通突发事件数据用于贝叶斯网络,通过软件Ge Nie2.2,以事故车辆类型,事故车辆数量,得到事故信息的时间,事发时天气及事发时段为证据信息,推理得出高速公路突发事件概率,预测结果表明基于贝叶斯网络的高速公路交通突发事件应急处置态势评估模型具有较高的准确性。
In order to improve the capacity and the efficiency of expressway emergency management,an emergency treatment situation assessment model of expressway traffic accidents is established based on Bayesian network for reasoning and leaning traffic accident probability under different conditions. First,the overall working process of the Bayesian network of expressway traffic is established in the model,the nodes of Bayesian network are set as the influencing factors of expressway,and the Bayesian network structure is constructed according to the relationship among the nodes（ various influencing factors）. Second,conditional probability is assigned to each node in the network according to traffic accident data and expert knowledge,then the Bayesian network is transformed into a joint tree using the joint tree inference algorithm according to the preliminary accident information based on the conditional probability,and the posterior probability is calculated through the message transfer process defined in the joint tree. Finally,the inference is conducted based on conditional probability,an evaluation model is established to evaluate the situation of expressway emergencies. Fifty expressway traffic emergency data are randomly selected for Bayesian network reasoning ina case study. By using the software Ge Nie2. 2,taking type of accident vehicles,number of accident vehicles,time of accident information,weather and the time period of the accident as the evidence information,the probability of the expressway emergency is derived by inference. The prediction result indicates that the emergency treatment situation assessment model of expressway traffic accidents based on Bayesian network has higher accuracy.
作者
赵朋
王建伟
孙茂棚
周雅欣
ZHAO Peng;WANG Jian-wei;SUN Mao-peng;ZHOU Ya-xin(School of Economics and Management,Chang＇an University,Xi＇an Shaanxi 710064,China)
出处
《公路交通科技》
CAS
CSCD
北大核心
2018年第9期115-121,共7页
Journal of Highway and Transportation Research and Development
基金
国家自然科学基金青年项目(41301130)
关键词
交通工程
突发事件态势评估
贝叶斯网络
高速公路
应急管理
traffic engineering
emergency situation assessment
Bayesian network
expressway
emergency management
