摘要
针对蚁群算法在静态环境下全局路径规划存在迭代稳定次数过多、求解目标单一等问题,提出一种多因素A^(*)蚁群算法。通过栅格法搭建机器人工作地图,并规定行进方式以保证路线的安全性;利用改进A^(*)算法得到的路径设定地图的初始信息素,降低蚁群算法前期搜索盲目性;引入路径的转弯次数、颠簸程度作为蚁群选择道路的考量因素,避免路径规划以距离作为唯一目的,以满足机器人的实际工作需求;引入启发函数动态调节因子,让启发函数的作用随着迭代次数增加而减弱,避免蚁群算法陷入局部最优。仿真实验表明,改进算法迭代次数减少,且路径综合性能指标优于对比文献算法和传统算法。
In order to solve the problem that the ant colony algorithm has too many iterations and a single solution objective,this paper proposes improved A^(*)multi-factor ant colony algorithm.In the grid map,the path obtained by the improved A^(*)algorithm is used to set the initial pheromone of the map which reduce the blindness of the previous search of the ant colony algorithm.In order to meet the actual working needs of the robot and avoid taking distance as the sole purpose of path planning,the turning times and bumpiness of the path are introduced as the factors considered by ant colony to select the road.In order to avoid ant colony algorithm falling into local optimization,the dynamic adjustment factor of heuristic function is introduced to weaken the role of heuristic function with the increase of iteration times.Simulation experiments show that the iteration times of the improved algorithm is reduced,and the path comprehensive performance indicators are better than comparative document algorithms and traditional algorithms.
作者
徐劲力
柳佳
司马立萱
XU Jin-li;LIU Jia;SIMA Li-xuan(School of Mechanical and Electronic Engineering,Wuhan University of Technology,Wuhan 430070,China)
出处
《组合机床与自动化加工技术》
北大核心
2022年第8期21-25,共5页
Modular Machine Tool & Automatic Manufacturing Technique
基金
基于5G网络和云平台的新能源汽车智能制造关键技术集成与应用(桂科AA21077016)。
