基于K均值聚类算法的交通状态判别方法研究

Study on traffic state identification method based on K-means clustering algorithm

在交通供需矛盾持续深化的情况下,研究交通常发性拥挤是现代交通管理的重点课题.通过分析国内外交通状态判别方法与交通流特性研究,运用K均值聚类算法,提出在离线状态下有效判别交通状态的定量方法.以青岛市环湾快速路交通流数据对该方法的实效性进行验证分析,结果表明判别方法能够快速处理大量交通流数据,判别交通流运行状态,识别率较高,方便,高效,研究成果可以为协同优化交通控制系统和交通流诱导系统提供方法依据.

In the context of the continuously deepening contradiction between supply and demand of transportation,it has been a key issue of modern traffic management to study the frequent-type of congestion.By analyzing the methods of traffic state identification and traffic flow characteristics at home and abroad,a quantitative method that can effectively identify traffic state off-line is proposed by using K-means clustering algorithm.The validity of the method is verified by the traffic flow data of Huanwan Expressway in Qingdao.The results show that this method can quickly process a large amount of traffic flow data,and identify the state of traffic flow with high recognition rate,convenience and efficiency.The research results can provide methodological basis for collaborative optimization of traffic control system and traffic flow induction system.