基于浮动车数据的交通事件参数特性研究

Traffic Parameter Features in Traffic Incidents Based on Probe-car Data

利用微观交通仿真软件Vissim搭建交通事件环境,对在不同堵塞程度和不同交通流量的交通事件情况下,浮动车采集的浮动车速、行程时间、浮动车位置等移动检测交通参数变化特性进行了分析比较,总结出交通事件情况下的浮动车移动检测各交通参数变化规律,并以此为基础对模式识别法、统计预测法、时间序列和智能算法等4种基于浮动车数据的交通事件检测算法进行了探讨,具有覆盖面广、成本低等特点。

Existing traffic incident detection methods are mostly based on analyzing the traffic parameter characteristics including volume,velocity and occupation,and on using fixed traffic detector,which are limited by the costs and position.In order to obtain the solutions of this problem,this paper analyzes and compares the probe-car traffic parameter features including velocity,travel time and probe car position,in various traffic incidents based on micro traffic simulation software Vissim,and gets the regulations of probe car traffic parameters in incidents.On the basis of the conclusion above,four traffic incident detection algorithms are discussed including pattern distinguishing algorithm,statistics and forecast algorithm,time sequence algorithm and nerve network algorithm,which have the benefits on position and costs.The algorithm in this paper is potentially applicable in practice.