基于多目标的动态环境智能车路径规划算法

针对传统算法无法适用多目标及动态环境的智能车路径规划问题,文中基于改进A*算法与势场蚁群算法进行了面向多目标的动态环境智能车路径规划算法研究。根据多目标的特征,采用改进A*算法识别完整的周边环境,并进行全局路径规划。对于实验场景中出现的局部变化或障碍物移动,将人工势场算法与蚁群算法相结合,获得了改进势场蚁群算法,以实现在原有全局路径规划基础上的局部修改。通过优化仿真得到了文中所提算法的最优参数值,并与蚁群算法进行对照测试。结果显示,所提算法相比对照组路径长度缩短了2.7%,具有良好的综合性能。

Radial Based Probabilistic Roadmap Motion Planning Method in Sparse Environment

A new dynamic path planning method in high dimensional workspace, radial based probabilistic roadmap motion (RBPRM) planning method, is presented. Different from general probabilistic roadmap motion planning methods, it uses straight lines as long as possible to construct a path graph, so the final path obtained from the graph is relatively shorter and straighter. Experimental results show the efficiency of the algorithm in finding shorter paths in sparse environment.