基于改进模糊聚类算法的城市快速路交通状态分类  

Traffic state classification of urban freeways based on improved fuzzy clustering algorithm

作　　者：雷宁 张光德[1] 陈玲娟[1] 

机构地区：[1]武汉科技大学汽车与交通工程学院,湖北武汉430058

出　　处：《广西大学学报（自然科学版）》2017年第5期1723-1729,共7页Journal of Guangxi University（Natural Science Edition）

基　　金：国家自然科学基金青年资金资助项目(51308425)

摘　　要：文中从城市快速路实时交通信息发布和缓解拥挤出发,对快速路交通状态进行有效分类,在选取分类指标时,在传统分类指标流量、速度、占有率的基础上引入了一个宏观分类指标——路网宽裕度。运用改进的模糊聚类方法对所选取指标进行聚类分析,以北京市五环路部分路网样本数据为例进行交通状态分类评估,同时与其他分类算法的分类结果及误差进行对比分析。结果表明,改进的模糊聚类算法在平均运行时间、目标函数值及分类误差上都低于其他算法的,由此验证了研究中所提算法的可靠性。The traffic state of urban freeways is effectively classified for the purpose of accurately releasing traffic information and relieving congestion. When selecting classification indicators, the road network ample degree is added on the base of traditional classification indicators such as traffic volume, speed and occupancy. The improved fuzzy clustering method is used to analyze the proposed indicators, and the data collected on the fifth ring road of Beijing is used as the sample to perform the traffic state classification evaluation. Then the classification results and its error arecompared with the those of other algorithms. Analysis results show that the average operation time, the objective function value and the classification error of the improved algorithm are all lower than those of other algorithms, which verified the reliability of the proposed algorithm.

关 键 词：交通信息发布 城市快速路 模糊聚类 交通状态分类 

分 类 号：U491[交通运输工程—交通运输规划与管理]