摘要
针对水面无人艇在动态环境下的路径规划难以满足全局最优和实时避障需求的问题,提出了一种改进D^(*)算法和改进动态窗口法相融合的算法,即D^(*)DWA。首先,对环境地图进行栅格化建模,利用层次聚类法根据障碍物的坐标位置对地图进行区域划分;然后,建立区域障碍物复杂度量化指标向量对D^(*)算法中的代价函数进行优化,获取全局最优路径的基本信息;最后,根据全局最优路径中关键节点信息设计动态窗口法的评价函数,快速规划出全局最优光滑路径。实验将所提出的D^(*)DWA与其他路径规划算法进行了仿真对比。实验结果表明,该算法提高了路径规划的效率,增加了路径的平滑度。
To solve the problem that path planning for unmanned surface vehicles in dynamic environments struggles to meet the requirements of global optimality and real-time obstacle avoidance,an algorithm that integrates the improved D^(*)algorithm and the improved dynamic window approach is proposed,namely D^(*)DWA.Firstly,the environmental map is modeled using a grid-based approach,and a hierarchical clustering method is employed to partition the map into regions based on the coordinates of obstacles.Then,a quantitative index vector for regional obstacle complexity is established to optimize the cost function in the D^(*)algorithm,thereby obtaining basic information about the globally optimal path.Finally,based on the key node information in the globally optimal path,an evaluation function for the dynamic window approach is designed to quickly plan a globally optimal and smooth path.The proposed D^(*)DWA is compared with other path planning algorithms through simulation experiment.The experimental results demonstrate that this algorithm improves the efficiency of path planning and enhances path smoothness.
作者
段求辉
DUAN Qiuhui(The 10th Research Institute of China Electronic Technology Group Corporation,Chengdu 610036,China)
出处
《控制工程》
北大核心
2026年第1期129-134,共6页
Control Engineering of China
关键词
水面无人艇
路径规划
层次聚类法
改进D^(*)算法
动态窗口法
Unmanned surface vehicle
path planning
hierarchical clustering method
improved D^(*)algorithm
dynamic window approach
