首页|基于避障寻优改进蚁群算法的机器人路径规划

基于避障寻优改进蚁群算法的机器人路径规划

扫码查看
针对蚁群算法在处理路径规划过程中存在收敛速度慢,规划路径冗余等问题,提出了一种基于避障信息和快速寻优策略的改进蚁群算法.为了改善蚁群的首次搜索效率和精度,引入切比雪夫距离改进距离启发函数,在转移概率中增加目标点对机器人的引导作用;采用自适应转移概率调整路径规划过程中节点的选择方式,并根据节点周围的障碍物分布设置初始信息素,使得蚂蚁首次生成有效路径的比率从60%提高至92%;同时删除生成路径的垃圾信息,提高最优路径节点的信息素浓度,平衡了蚁群的局部和全局搜索能力,加快了最优路径的速度;通过平滑生成的路径,减少机器人转弯次数,缩短了路径距离.选择SSA、ACO、IACO、I-ACO等算法在3种栅格环境上进行性能测试.结果表明,改进的蚁群算法路径寻优上优于其他算法.
Robot path planning based on obstacle avoidance optimization and improved ant colony algorithm
Aimed at the problems of slow convergence speed and redundant planning paths in the processing of path planning by ant colony algorithm,an improved ant colony algorithm based on obstacle avoidance information and fast optimization search strategy was proposed.In order to improve the first search efficiency and accuracy of the ant colony,the Chebyshev distance was in-troduced to improve the distance heuristic function,and the guidance of the target point to the ro-bot was enhanced in the transfer probability.The adaptive transfer probability was used to adjust the selection method of nodes during path planning and the setting of initial pheromones based on the distribution of obstacles around the nodes,and the percentage that the ants generate effective paths for the first time increased from 60%to 92%.The garbage information of the generated paths was removed,increasing the pheromone concentration of the optimal path nodes,balancing the local and global searching ability of the ant colony,and speeding up the optimal path.By smoothing the generated paths,the number of robot turns was reduced and the path distance was shortened.The algorithms SSA,ACO,IACO,and I-ACO were selected for performance testing on three grid environments.The results show that the improved ACO algorithm outperforms the other algorithms on path optimization.

robot path planningobstacle avoidance optimizationant colony optimization algo-rithmgrid mappath smoothing

贺兴时、陈慧园

展开 >

西安工程大学 理学院,陕西 西安 710048

机器人路径规划 避障寻优 蚁群优化算法 栅格地图 路径平滑

陕西省自然科学基础研究计划

2023-JC-YB-064

2024

西安工程大学学报
西安工程大学

西安工程大学学报

CSTPCD
影响因子:0.473
ISSN:1674-649X
年,卷(期):2024.38(3)
  • 13