摘要
传统的路径规划算法很大程度上是依赖于改进的加权最短路径算法,在大规模路网中效率较低而且没有考虑实际交通中的各种因素,得到的是理想情况下的最优路径。针对这种情况,根据出租车的轨迹数据提出一种路径规划方法,主要包括三个部分:首先,利用出租车数据挖掘司机在路径选择上的经验,提取经验轨迹形成经验轨迹集;然后,根据出租车在各经验路段各时段的速度和频次利用贝叶斯分类器对路网进行分层,构建分层路网;最后,使用分层路径规划算法实现层次路径规划。以北京市GPS数据为研究对象,将该方法与经典路径规划算法的结果进行比较。结果表明,该方法得到的路径可以综合考虑各种因素,得到实际行驶中的较快路径。
Traditional path planning algorithms are largely dependent on the improved weighted shortest path algorithm, their efficiency in large-scale road networks is rather low and the various factors of practical transportation do not taken into account as well, what they get are the optimal paths under ideal conditions. In light of this, we introduce in this paper a path planning method based on the tracks data of taxis, the method mainly includes the following three parts. First, it mines cab drivers＇ experience on path selection using taxi data and extracts the experience tracks to form empirical tracks set. Then, it stratifies the road network with Bayesian classifier according to the speed and frequency of taxis in each period on various experience road sections and builds the stratified road network. Finally, it implements the hierarchical path planning with stratified path planning algorithm. Taking the GPS data of Beijing as the research object, we compare the results of the method and the classical path planning algorithm. Results show that the paths derived by this method can consider various factors comprehensively and get a faster path in practical drives.
出处
《计算机应用与软件》
CSCD
2016年第1期68-72,共5页
Computer Applications and Software
基金
辽宁省高等学校优秀人才支持计划项目(LJQ2012011)
辽宁省自然科学基金项目(20102175
201102200)
关键词
智能交通
路径规划
轨迹数据
贝叶斯分类器
分层路径规划算法
Intelligent transportation Path planning Tracks data Bayesian classifier Stratified path planning algorithm
