Many studies have been carried out using vehicle trajectory to analyze traffic conditions, for instance, identifying traffic congestion. However, there is a lack of a systematic study on the appropriate number of prob...Many studies have been carried out using vehicle trajectory to analyze traffic conditions, for instance, identifying traffic congestion. However, there is a lack of a systematic study on the appropriate number of probe vehicles and their sampling interval in order to identify traffic congestion accurately. Moreover, most of related studies ignore the streaming feature of trajectory data. This paper first represents a novel method of identifying traffic congestion considering the stream feature of vehicle trajectories. Instead of processing the whole data stream, a series of snapshots are extracted. Congested road segments can be identified by analyzing the clusters＇ evolution among a series of adjacent snapshots. We then calculated a series of parameters and their corresponding congestion identification accuracy. The results have implications for related probe vehicle deployment and traffic analysis; for example, when 5% of probe vehicles are available, 85% identification accuracy can be reached if the sampling time interval is 10 s.展开更多
交通事件的检测与确认是交通事件管理中的首要问题。基于线圈和视频数据的检测方法由于成本高,检测效果不明显,在实际应用中受到限制。提出了一种基于离群点挖掘的交通事件检测算法。该算法通过使用浮动车(floating car data,FCD)技术...交通事件的检测与确认是交通事件管理中的首要问题。基于线圈和视频数据的检测方法由于成本高,检测效果不明显,在实际应用中受到限制。提出了一种基于离群点挖掘的交通事件检测算法。该算法通过使用浮动车(floating car data,FCD)技术得到路况信息,并提取交通事件特征,建立特征向量。算法简单、高效、易于部署。实验结果表明,同模式识别方法相比,该算法具有较高的准确度,能有效区分常规拥堵与交通事件。展开更多
文摘Many studies have been carried out using vehicle trajectory to analyze traffic conditions, for instance, identifying traffic congestion. However, there is a lack of a systematic study on the appropriate number of probe vehicles and their sampling interval in order to identify traffic congestion accurately. Moreover, most of related studies ignore the streaming feature of trajectory data. This paper first represents a novel method of identifying traffic congestion considering the stream feature of vehicle trajectories. Instead of processing the whole data stream, a series of snapshots are extracted. Congested road segments can be identified by analyzing the clusters＇ evolution among a series of adjacent snapshots. We then calculated a series of parameters and their corresponding congestion identification accuracy. The results have implications for related probe vehicle deployment and traffic analysis; for example, when 5% of probe vehicles are available, 85% identification accuracy can be reached if the sampling time interval is 10 s.
文摘交通事件的检测与确认是交通事件管理中的首要问题。基于线圈和视频数据的检测方法由于成本高,检测效果不明显,在实际应用中受到限制。提出了一种基于离群点挖掘的交通事件检测算法。该算法通过使用浮动车(floating car data,FCD)技术得到路况信息,并提取交通事件特征,建立特征向量。算法简单、高效、易于部署。实验结果表明,同模式识别方法相比,该算法具有较高的准确度,能有效区分常规拥堵与交通事件。
