摘要
针对目前Voronoi图路径规划算法实时性差、骨架提取过程耗时长、规划的路径曲折冗余的问题,提出一种基于自由区域的骨架关键点构建骨架图结构结合A^(*)算法的路径规划算法。对路径规划的栅格地图进行预处理,去除地图噪声和边缘毛刺,改进骨架提取算法,使其仅对地图中自由区域提取骨架,从而显著降低骨架提取时间。利用骨架关键点及其连接关系构建骨架图结构,骨架图结构能够显著简化环境信息,使路径规划算法在搜索过程中更加专注于关键路径,并将骨架图结构引入到A^(*)算法的评估函数中,有效解决Voronoi图路径曲折冗余问题。在家庭与多障碍物仿真环境中,与文献[6]的Voronoi改进算法对比实验表明:所提算法规划的路径长度平均减少13.23%,路径平滑度显著提升,消除了短距离规划中的冗余转折;规划时间平均降低34.46%,搜索效率明显改善。在实际机器人导航实验中,文中算法生成的路径在保持安全距离的同时更直接,验证了其在实际场景中的有效性与鲁棒性。文中算法所得的规划时间,规划路径长度较对比算法更少,显著提升了机器人路径规划的实时性和效率。算法在减少计算开销的同时生成更平滑、安全的路径,适用于复杂环境的实时导航需求,为机器人全局路径规划提供了高效可行的解决方案。
To address the challenges of poor real-time performance,high skeleton extraction costs,and redundant,tortuous paths in Voronoi-based path planning algorithms,a path planning algorithm was proposed that combined a free-space-based skeleton key point-constructed skeleton graph with the A^(*)algorithm.The grid map used for path planning was preprocessed to remove map noise and edge burrs,and the skeleton extraction algorithm was improved so that skeletons were extracted only from free regions of the map,thereby significantly reducing skeleton extraction time.Skeleton key points and their connectivity relationships were used to construct a skeleton graph structure,which significantly simplified environmental information,enabling the path planning algorithm to focus more on critical paths during the search process.The skeleton graph structure was incorporated into the evaluation function of the A^(*)algorithm,effectively resolving the problem of tortuous and redundant paths generated by Voronoi diagram-based methods.In simulations of home and multi-obstacle environments,comparative experiments with the improved Voronoi algorithm reported in Ref.[6]show that the proposed algorithm achieves an average reduction in path length of 13.23%,with path smoothness being significantly improved and redundant turns in short-distance planning being eliminated.The planning time is reduced by an average of 34.46%,and the search efficiency is markedly improved.In real robot navigation experiments,the paths generated by the proposed algorithm are kept safer while remaining more direct,which verifies its effectiveness and robustness in practical scenarios.Compared with the reference algorithms,shorter planning time and shorter path length are achieved,and the real-time performance and efficiency of robot path planning are significantly enhanced.While reducing computational overhead,smoother and safer paths are generated,making the algorithm suitable for real-time navigation in complex environments and providing an efficient and feasible solution for global robot path planning.
作者
王朝阳
蒋林
杨婉振
万乐
汤勃
朱建阳
WANG Zhaoyang;JIANG Lin;YANG Wanzhen;WAN Le;TANG Bo;ZHU Jianyang(Key Laboratory of Metallurgical Equipment and Control of Ministry of Education,Wuhan University of Science and Technology,Wuhan Hubei 430081,China;Institute of Robotics and Intelligent Systems,Wuhan University of Science and Technology,Wuhan Hubei 430081,China)
出处
《机床与液压》
北大核心
2026年第5期38-47,共10页
Machine Tool & Hydraulics
