基于广义势场的多机器人避碰算法

针对目前多机器人避碰算法存在的计算复杂、缺乏普适性等的局限性,以传统人工势场法为基础,将协商和意愿引入人工势场范畴,提出了多机器人系统运动规划的广义势场算法.该算法以物理空间障碍为基础构造排斥势场,以意愿为基础构造吸引势场,并通过排斥势场与吸引势场的拓扑积构成人工势场.在人工势场中,通过对势场中微团运动趋势的求解就可得出所需的路径规划.实验证明,该算法不仅具有人工势场法系统模型构造简单、能够进行最优路径规划的优势,而且在系统运行过程中有效地避免了系统死锁的发生.

Multi-Robot Collision Avoidance Algorithm Based on Generalized Potential Field

The generalized potential algorithm for collision avoidance, base on traditional artificial potential algorithm, include consult and desire of the robots. The research proved that using the generalized potential algorithm can find the optimization path planning and avoid the system mobile collision. The constructing and using method of generalized potential algorithm are discussed in this paper.