摘要
研究了一种新颖的动态复杂不确定环境下的机器人路径规划方法和动态避障码蚁预测算法.该方法模拟蚂蚁的觅食行为,由多组蚂蚁采用最近邻居搜索策略和趋近导向函数相互协作完成全局最优路径的搜索.在此基础上用虚拟蚂蚁完成与动态障碍物碰撞的预测,并用蚁群算法进行避障局部规划.理论和仿真实验结果均表明,即使在障碍物非常复杂的地理环境,用文中算法也能迅速规划出优化路径,且能安全避碰.
Based on Ant Colony Optimization （ACO）, this paper presents a novel algorithm underlying the robot path planning and dynamic obstacle avoidance in a complex and unfamiliar environment. By mimicking the food hunting behavior of ant colony, this algorithm can search for the global optimal path by adopting the nearest-neighbor search strategy combining an approximating direction function used by multiple ant groups. The virtual ants, based on this algorithm, are able to predict their potential collision with the moving obstacles. The subsequent local plans for avoiding such collisions are then scheduled under the ACO. The analytical and computer experiment results demonstrate that this novel algorithm can plan an optimal path rapidly in a cluttered environment. The successful obstacle avoidance is achieved, and the model is robust and performs reliably.
出处
《计算机学报》
EI
CSCD
北大核心
2005年第11期1898-1906,共9页
Chinese Journal of Computers
基金
江苏省教育厅自然科学基金(2001SXXTSJB111)资助
关键词
机器人
路径规划
移动障碍物
蚂蚁预测
蚁群优化算法
robot
path planning
moving obstacles
ants predictive algorithm
ant colony optimization algorithm
