一种改进蚁群算法在捡球机器人多目标路径规划中的应用  

An Improved Ant Colony Algorithm and Application in Picking Robot Multi-objective Path Planning

作　　者：曹洁[1,2] 耿振节[1] 

机构地区：[1]兰州理工大学电气工程与信息工程学院,兰州730050 [2]甘肃省制造业信息化工程技术研究中心,兰州730050

出　　处：《小型微型计算机系统》2015年第10期2384-2389,共6页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(61263031)资助;国家科技支撑计划课题项目(2012BAF12B19)资助

摘　　要：在静态环境下,针对蚁群算法在进行多目标路径规划时存在搜索时间长以及容易陷入局部最优的特点,介绍一种可对捡球机器人进行多目标路径规划的改进蚁群算法.首先采用栅格法将捡球机器人的网球场环境划分建模;然后自适应调整算法中与信息素相关的系数以及种群规模,提高算法搜索速度;并引入交叉操作,改善算法停滞问题,增强算法的逃逸能力;最后实现改进蚁群算法在捡球机器人多目标路径规划中的应用.仿真实验结果证明,所提算法可以加快搜索速度,找到全局最优路径.In the static environment, to solve searching for long time and easy to fall into local optimum in ant colony algorithm, an improved ant colony algorithm was proposed for multi-objective path planning of picking robot. First adopts grid method modeling the working tennis court environment of picking robot;Then adaptive adjustment the pheromone related coefficients and population size, improve the search speed of the algorithm, and introduce crossover operation ,improve algorithm stagnation problem, enhance algorithm escape ability;Finally achieve improved ant colony algorithm in the application of a multi-objective path planning for picking robot. Simulation results show that the algorithm can speed up the search speed,find the global optimal path.

关 键 词：改进蚁群算法 捡球机器人 多目标路径规划 自适应调整 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]