基于DBSCAN算法的城市交通拥堵区域发现
2015-04-29刘畅李治军姜守旭
刘畅 李治军 姜守旭
摘 要:现阶段我国车辆的数目不断增加,这必然会导致现有的交通条件等不再能够满足现在的交通状况。因时空数据蕴含丰富的信息,本文将通过分析车辆的时空轨迹数据,即GPS轨迹数据,并利用DBSCAN算法挖掘出城市中交通最拥堵的城市区域,并将得到的结果映射到城市路网上,使得相关部门能够清楚的了解问题,并制定一系列相应措施,如调整道路规划等,保证该区域的问题能够迅速得到解决。本文的实验结果清晰地标明了城市中具体的拥堵区域,本文下一步将根据城市拥堵区域的发现结果预测区域的下一次拥堵的时间。
关键词 : DBSCAN算法;时空轨迹数据;城市拥堵区域
中图法分类号: TP39 文献标识码: A
Discovery of Heavy Traffic Areas based on DBSCAN Algorithm
LIU Chang, LI Zhijun, JIANG Shouxu
(School of Computer Science and Technology, Harbin Institute of Technology, Harbin 150001, China)
Abstract: At present, the number of vehicles is increasing in our country. This will inevitably cause that the existing road and so on would no longer able to meet the current traffic conditions. Because the spatio-temporal data contain rich information, this paper will analyze the GPS data of the vehicle. This paper will mining the heaviest traffic congestion areas in a city by using the DBSCAN algorithm, and the mining result will be mapped to the city road network so that the relevant departments can clearly understand where the problems are. The departments also can develop a series of measures to solve those problems, such as adjusting the road planning and so on. This will ensure that the problems can be resolved quickly, and the experiment results of this paper clearly indicate the specific congestion areas in a city. The next work is the time prediction of congestion areas.
Key words: DBSCAN Algorithm; Spatio-temporal Data; Heavy Traffic Area
0 引 言
我国现阶段经济迅速发展,私家车、公交车、出租车等的数量正与日激增,这一态势必将导致现有的部分地区交通道路已经难以满足不断增长的运营与运力需求,所以如何做好城市交通道路规划即已归作城市建设中需要重点解决的关键问题之一[1],一个科学、合理的交通道路规划则是建造一座现代智能城市的基础必备技术手段。
时空轨迹数据[2]蕴含非常丰富的地理信息、语义信息,连续的轨迹数据还能反映出其移动规律以及相应变化。通过相同种类大量不同对象的轨迹数据,规划团队还可以迅速了解某一种类对象的特点特性。而当规划者使用城市中的车辆轨迹数据时,这些数据的动态分布则能实时有效地反映城市的交通变化[3],这就为城市交通道路规划提供了极其重要的信息,并使得城市的交通道路能够进一步做出相应的调整。……
