摘要
针对传统势场蚁群算法在机器人路径规划中表现出的3个关键不足——转折点精度不足、收敛特性与解质量不匹配、局部最优规避效能低下,本研究提出了一种基于边界势场矢量合成的改进算法。该算法的核心创新体现在:构建势场-蚁群协同搜索架构,采用角度优选概率模型提升路径连续性;设计动态收敛调节因子,实现勘探与开发的智能平衡;建立融合转向特征的信息素更新机制。多尺度栅格环境测试表明,改进算法使迭代代数减少37.4%,路径转折量下降54.7%,全局最优概率提升至91.3%,同时显著优化了计算时效性和路径质量,有效提升了移动机器人的导航性能,搜索时间明显缩短,路径转折频率次数明显降低,机器人路径规划效率明显提升。
To address three critical limitations of conventional potential field ant colony algorithms in robotic path planning—inadequate turning point precision,mismatched convergence characteristics and solution quality,and insufficient local optimum avoidance capability—this study proposed an enhanced algorithm based on boundary potential field vector synthesis.The core innovations of this algorithm included:1)constructing a1 potential field-ant colony collaborative search framework that employs an angle-optimized probability model to improve path continuity;2)designing a dynamic convergence regulation factor to intelligently balance exploration and exploitation;and 3)establishing a pheromone update mechanism incorporating turning features.Multi-scale grid environment tests demonstrated that the improved algorithm reduces iteration generations by 37.4%,decreased path turns by 54.7%,and elevated global optimum discovery probability to 91.3%,while significantly enhancing computational efficiency and path quality.These advancements substantially improved mobile robot navigation performance,evidenced by markedly reduced search time,significantly lower path turning frequency,and notably enhanced path planning efficiency.
作者
刘学文
任兴贵
陈金鑫
刘雪婷
LIU Xuewen;REN Xingui;CHENG Jinxin;LIU xueting(Guangzhou Huaxia Vocational College,School of Vehicles and Automation,Guangzhou 510935,China)
出处
《现代农业装备》
2025年第4期19-24,85,共7页
Modern Agricultural Equipment
基金
广东省青年人才创新项目(2019GKQNCX092)。
关键词
蚁群算法
轨迹优化
农用机器人
势场引导
模型迭代
ant colony optimization
trajectory optimization
agricultural robot
potential field guidance
model iteration
