基于出租车GPS数据的交通热区识别方法  

A novel method for traffic hotspots recognition based on taxi GPS data

作　　者：郑运鹏 赵刚[1] 刘健[1] 

机构地区：[1]北京信息科技大学信息管理学院,北京100192

出　　处：《北京信息科技大学学报（自然科学版）》2016年第1期43-47,共5页Journal of Beijing Information Science and Technology University

基　　金：国家自然科学基金资助项目(61272513)

摘　　要：为了使交通热区识别结果清晰、准确,降低交通热区划分成本,利用城市出租车GPS(global positioning system)数据,提出了一种基于网格的K-Means交通热区识别方法。通过上下车位置提取、滑动窗口检测、区域外数据过滤等数据预处理手段,结合网格划分和聚类的思想,对目标城市进行交通热区识别,并将结果输出到地图上。利用南京市出租车GPS数据对南京市交通热区进行识别,识别的交通热区边界清晰、结果准确,对分析城市交通流、交通趋势具有重要的参考价值。In order to make the recognition results clear and accurate, and reduce the recognition cost as well, a novel grid-based K-Means clustering method is proposed for the recognition of traffic hotspots with Taxi GPS data. GPS data are selected from the city taxies and preprocessed through detecting sliding window and filtering data outside the region. By combining grid division and clustering, the Grid-based K-Means clustering method is used to recognize the traffic hotspots in the target city, and output the results on the map. The experiment based on GPS data of Nanjing city taxies automatically recognizes the traffic hotspots in Nanjing, and verifies the validity of this partition method. The experimental results show that this novel clustering method is effective and gives a good reference value for the analysis of city traffic flow and trends.

关 键 词：交通热区 聚类 网格划分 GPS K-MEANS 交通趋势 

分 类 号：TP309.2[自动化与计算机技术—计算机系统结构]