引用本文: 辛煜, 梁华为, 杜明博, 梅涛, 王智灵, 江如海. 一种可搜索无限个邻域的改进A*算法[J]. 机器人, 2014, 36(5): 627-633.
XIN Yu, LIANG Huawei, DU Mingbo, MEI Tao, WANG Zhiling, JIANG Ruhai. An Improved A* Algorithm for Searching Infinite Neighbourhoods[J]. ROBOT, 2014, 36(5): 627-633.
 Citation: XIN Yu, LIANG Huawei, DU Mingbo, MEI Tao, WANG Zhiling, JIANG Ruhai. An Improved A* Algorithm for Searching Infinite Neighbourhoods[J]. ROBOT, 2014, 36(5): 627-633.

## An Improved A* Algorithm for Searching Infinite Neighbourhoods

• 摘要: 传统A*算法在栅格地图上进行路径规划时，求解得到的路径长度不是最短并且转折点较多．针对这些不足，提出了一种改进A*算法，将传统 A*算法的可搜索邻域个数从离散的8个拓展为无限个，可以沿任意方向进行搜索．这样不仅求解出来的路径长度更短，并且大大降低了其转折点的个数．该算法被应用于自主研发的“智能先锋”号系列无人驾驶车辆上，实车试验以及它们在“中国智能车未来挑战赛”中的优异表现证明该方法能够在栅格地图中求解出一条更优的可行驶路径，可以显著提升无人驾驶车辆行驶的效率和平稳性．

Abstract: The paths obtained by conventional A* algorithm in grid map are usually not the optimal solutions, which have many turning nodes. To solve these problems, an improved A* algorithm is presented. In this method, the number of searchable neighbourhood is extended to infinite (basic A* algorithm has 8 discrete searchable neighbourhoods), and the search direction can be arbitrary. As a result, the path produced by the improved A* is much shorter and the number of turning nodes is greatly reduced. The proposed method is implemented on our self-developed unmanned vehicle, Intelligent Pioneer. The vehicle experiments and its good performance in the Intelligent Vehicle Future Challenge of China show that the resulting path of our method is more optimal, which observably improves the driving efficiency and stationarity of the unmanned vehicles.

/

• 分享
• 用微信扫码二维码

分享至好友和朋友圈