基于改进A*算法的智能车路径规划研究
2020-02-25沈刚邵金菊谭德荣牛亚明马晓田
沈刚 邵金菊 谭德荣 牛亚明 马晓田
摘 要:针对智能车使用A*路径规划算法存在转折点和冗余点的问题,提出一种考虑智能车静态特性的改进A*路径规划算法。在已知静态环境信息的栅格地图上,考虑到智能车自身存在实际宽度,对障碍物进行膨胀扩展;其次根据路径上前后节点相对方向的改变提取必要的转折点,并依次连结前后转折点,若转折点连线不经过障碍物,删除连结转折点之间冗余的转折点;重复上述操作,直至所有冗余点被删除,保留关键转折点。仿真结果表明,该方法可以实现车辆安全无碰撞地到达目标终点。关键词:路径规划;A*算法;障碍物膨胀;轨迹平滑中图分类号:V323.19 文献标识码:B 文章编号:1671-7988(2020)02-28-03
Abstract: In order to solve the problem of turning points and redundant points in using A* path planning algorithm for intelligent vehicles, an improved A* path planning algorithm considering the static characteristic of intelligent vehicles is proposed. On the grid map with known static environmental information, considering the actual width of the intelligent vehicle itself, the obstacles are expanded. Secondly, the necessary turning points are extracted according to the change of the relative direction of the front and rear nodes on the path, and the front and rear turning points are connected in turn. If the turning points are connected without passing through the obstacles, the redundant turning points between the turning points are deleted. Repeat the above operations until all redundant points are deleted and key turning points are reserved. The simulation results show that the method can ensure the vehicle to reach the destination safely without collision.Keywords: Path Planning; A* Algorithm; Obstacles Expansion; Track SmoothingCLC NO.: V323.19 Document Code: B Article ID: 1671-7988(2020)02-28-03
前言
近年来,智能车越来越成为国内外学者和各大汽车厂商研究的热点和重点。导航是智能车非常重要的功能模块,路径规划技术作为导航的基础和关键[1,2],直接影响着智能车行进中的安全避障性和快速通过性。早期的路径规划技术主要针对移动机器人进行主要研究,最常见的路径规划算法主要分为两类,一类是基于节点搜索的以A*算法和Dijkstra算法为代表的传统路径规划算法[3,4],一类是基于遗传式的以蚁群算法和粒子群算法为代表的智能路径规划算法[5,6]。在全局静态路径规划领域,传统的A*算法具有计算量小、搜索精度高等优点,将移动机器人作为质点完成路径规划,却无法满足智能车在实际导航中的综合需求,存在忽略智能车自身静态特性与动态特性的问题。……
