出租车GPS大数据的道路行车可视分析

发布时间：2018-02-24 07:43

本文关键词： 大数据可视分析云计算平台离散编码连续编码箭头图出租车GPS大数据　出处：《计算机辅助设计与图形学学报》2014年12期 　论文类型：期刊论文

【摘要】：针对出租车GPS数据因其数据量庞大和时空信息特征复杂而带来的分析难题,提出一种基于出租车GPS大数据的道路行车可视分析方法.该方法用OpenStreetMap得到开阔道路的地图,采用离散和连续型2种编码方式对道路上的车流量、行车方向和速度等情况进行分析;离散编码采用箭头图表示,并用速度区间聚类算法优化颜色布局;连续编码采用栈图表示,并用特征点提取算法加速图表绘制.最后以杭州市出租车GPS数据为样例,将数据分布式存储在云计算平台上,采用MapReduce加快数据查询和处理,应用文中的2种可视编码方式进行可视分析,结果表明,该方法能准确地反映杭州市道路交通状况.
[Abstract]:In view of the difficult problem of analyzing taxi GPS data due to its huge data volume and complex space-time information characteristics, this paper presents a visual analysis method based on taxi GPS big data, which uses OpenStreetMap to get the map of open road. Discrete and continuous coding methods are used to analyze the traffic flow, driving direction and speed on the road, the discrete coding is expressed by arrow graph, and the color layout is optimized by speed interval clustering algorithm. The continuous coding is expressed by stack diagram and accelerated by feature point extraction algorithm. Finally, taking Hangzhou taxi GPS data as an example, the data is distributed stored on cloud computing platform, and MapReduce is used to speed up data query and processing. The results show that this method can accurately reflect the road traffic situation in Hangzhou.
【作者单位】：浙江工业大学计算机科学与技术学院;
【基金】：国家自然科学基金(61379017,61202205) 浙江省自然科学基金(LY12A01027) 浙江省钱江人才计划项目(2013R10054) 浙江省公益技术应用研究项目(2014C33077)
【分类号】：P228.4;U495

【参考文献】