足球机器人动态角色分配策略仿真研究

Research on Dynamic Role Assignment Strategy in Robot Soccer

研究足球机器人角色分配问题,角色分配是足球机器人决策协调的关键,在比赛中要兼顾最优性与实时性进行角色分配是足球机器人技术的难点。针对足球机器人比赛的特点,提出一种基于市场机制的动态角色分配算法。为了使角色分配接近最优,综合考虑影响角色分配的距离、角度以及其它多种不确定性因素来设计竞标函数;为了保证实时性,系统根据赛场态势,采用分级竞标的策略动态调整机器人角色。实验表明,市场动态算法在最小时间内提供接近最优化的角色分配方案,兼顾了足球机器人角色分配的实时性与最优性。

The role assignment strategy for the robot soccer game is important. Considering the demands of the dy- namics and the optimality, we presented an improved market-based algorithm for dynamic role assignment. The im- proved bidding functions were designed, which fully considered the related factors. According to the current situation, the improved algorithm adjusts the roles in the system and set priorities, and then assigns the roles in two steps on the basis of the priority. That guarantees the stability, the rationality and the real time of the algorithm. The experimental results show that the improved role assignment algorithm is reliable and rational, and can increase the intelligence of the robots. It improves the overall ability of the multi-robot system.