论文部分内容阅读
针对交通事件自动检测多以高速公路、城市快速路为对象以及使用数据源单一的现状,本文提出一种基于多源数据融合的城市道路交通事件检测方法。在对信号控制下交通事件引起的交通流变化进行分析的基础上,利用杭州市城区浮动车、SCATS、Citilog、OD系统提供的实时交通数据,基于CUSUM算法构建差分流量和速度交通事件检测模型。该模型可以有效抑制交通信号对于交通流的周期性影响,协同视频、行程时间构成两方式三指标的事件检测体系。实验表明,模型在高峰时段和平峰时段均能快速准确检测交通事件。
In view of the fact that the traffic accidents are detected automatically in the expressway and the urban expressway and the single data source is used, this paper presents a method of urban road traffic incident detection based on multi-source data fusion. Based on the analysis of the change of traffic flow caused by traffic accidents under signal control, the real-time traffic data provided by the floating car, SCATS, Citilog, OD system in Hangzhou urban area are used to construct the differential traffic and speed traffic incident detection model based on CUSUM algorithm. The model can effectively restrain the periodic impact of traffic signal on traffic flow, and the collaborative video and travel time constitute the two-way three-index event detection system. Experiments show that the model can quickly and accurately detect traffic accidents during peak hours and peak hours.