摘要
针对快速行进树算法(fast marching tree,FMT^(*))在移动机器人路径规划中存在冗余探索多、路径规划时间长、路径拐点多等问题,提出了一种基于引力势能修正的快速行进树算法(gravitational potential energy FMT^(*),GPE-FMT^(*))。该算法借鉴人工势场法的思想,通过引入有限的引力势能场限制路径探索范围,从而有效减少冗余搜索;同时,采用基于引力势能值的路径树修正策略,在路径树扩展过程中对其生长方向进行引导,使路径树的生长方向更靠近终点,从而减少路径中的拐点,缩短规划时间。仿真实验结果表明,在相同计算资源下,GPE-FMT^(*)显著减少了路径规划时间,降低了路径拐点数和算法迭代次数,提升了路径质量。
To address the issues of redundant exploration,long planning time,and numerous turning points in mobile robot path planning using the FMT^(*)algorithm,this paper proposed a GPE-FMT^(*)algorithm.This algorithm drew from the concept of artificial potential fields by introducing a finite gravitational potential energy field to constrain the exploration range,effectively reducing redundant searches.Additionally,it employed a path tree correction strategy based on gravitational potential energy values to guide the growth direction of the path tree during its expansion,moving it closer to the target and thus reducing tur-ning points and shortening planning time.Simulation results demonstrate that,under equivalent computational resources,GPE-FMT^(*)significantly decreases path planning time,reduces the number of turning points,and lowers the number of algorithm iterations,ultimately enhancing path quality.
作者
徐正宇
贾小林
顾娅军
袁雷
Xu Zhengyu;Jia Xiaolin;Gu Yajun;Yuan Lei(School of Computer Science&Technology,Southwest University of Science&Technology,Mianyang Sichuan 621010,China;Mobile Internet of Things&Radio Frequency Identification Technology Key Laboratory of Mianyang,Mianyang Sichuan 621010,China)
出处
《计算机应用研究》
北大核心
2025年第8期2416-2420,共5页
Application Research of Computers
基金
国家自然科学基金面上项目(61471306)
四川省自然科学基金资助项目(2022NSFSC0548)
绵阳市移动物联射频识别技术重点实验室专项资助项目(MYZD23032302)。
关键词
FMT^(*)算法
人工势场法
引力势能修正
路径规划
FMT^(*) algorithm
artificial potential field method
gravitational potential energy correction
path planning
