大数据环境下公路网突发事件预警与应急决策研究

With the increase of traffic infrastructure, the road network is becoming more complex and intelligent. But due to the natural factors (agglomerate fog, smog etc.) and man-made factors (unsafe driving behaviors etc.), there are a lot of traffic incidents every year that have caused heavy casualties and property losses. How to further improve highway emergencies early warning and handling capacity has been an important problem faced by the intelligent transport.The development of the Internet of things, cloud computing and big data technology provides a new technical means to solve this problem. Based on the big road network data, this project closely deals with the requirements of emergency management, builds decision models for data driven emergencies, studies mainly secondary traffic incidents warning data and fast matching rules with situation data and real-time data, quickly generates emergency plans and efficient distribution channel strategy for secondary traffic incidents. The research has important theoretical significance and practical value.

随着国家对交通基础设施投入的持续增加，交通路网越来越复杂，越来越智能。但是由于团雾、雾霾等自然因素以及不安全行驶行为等人为因素的影响，每年都有很多交通事件发生，带来重大的人员伤亡或财产损失。如何进一步提升公路网突发事件预警和应急处理能力一直是智能交通面临的一个重要问题。物联网、云计算和大数据技术的发展为解决此问题提供了新的技术手段。本项目紧密结合公路网突发事件应急管理需求，基于公路网智能系统采集的大数据，构建情景数据、交通事件预警数据与实时数据关联规则与快速匹配模型，快速生成突发事件应急预案和次生交通事件高效分流疏导策略。本项目研究具有重要的理论意义与实践价值。

从交通事件的预测与控制需求出发，以复杂网络及其相关特性为视角，研究交通事件发生机理及其引起的路网全局的态势演变。主要从两个方面入手：一是基于小世界网络特征和质网络度分布指数，研究交通网络重要节点量化模型，提出多节点的控制策略。 二是，将事件演化追溯并归结于在约束条件下的网络路径选择问题，并给出了基于路径选择机制的构建交通事件演化网的具体方法，进而提出一个交通事件安全评判与控制策略。