动态障碍物环境下移动机器人路径规划  

Robot Path-Planning in a Moving-Obstacle Environment

作　　者：王于[1] 林良明[1] 颜国正[1] 

机构地区：[1]上海交通大学电子信息学院,上海200030

出　　处：《上海交通大学学报》2002年第10期1430-1434,共5页Journal of Shanghai Jiaotong University

摘　　要：以机器人基本避障策略与启发式动态规划法相结合的方法 ,研究了在障碍物轨迹已知的动态环境中机器人的路径规划问题 .根据机器人与单个障碍物可能的碰撞情况 ,分别提出几种避障策略 ,策略内容为子目标和相应的速度 .根据这些策略 ,在机器人当前视窗内 ,将形成一个策略树 .采用启发式动态规划算法 ,依据时间和路径费用最小原则 ,找出一个最优策略集 ,即最优子目标集 .In a dynamic environment, a robot's collision free algorithm combining robot's basic collision avoidance strategies with the heuristic dynamic programming was presented. Several collision situations between the robot and single obstacle were analyzed, and for each situation, several basic collision free strategies were studied. The content of each strategy includes one sub goal and robot's corresponding velocity. According to these strategies, in the robot's current view, a tree of sub goals is created, and a heuristic dynamic programming is adopted to find a set of optimal strategies, or a set of optimal sub goals.

关 键 词：动态障碍物 移动机器人 路径规划 启发式动态规划 避障策略 时间费用最小原则 路径费用最小原则 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]