一种基于效益的多机器人避碰协调策略  

A benefit based collision avoidance strategy for multi-robot systems

作　　者：姚芝凤[1,2] 叶秀芬[2] 戴学丰[1] 朱玲[1] 孙明[1] 

机构地区：[1]齐齐哈尔大学计算机与控制工程学院,黑龙江齐齐哈尔161006 [2]哈尔滨工程大学自动化学院,黑龙江哈尔滨150001

出　　处：《计算机工程与科学》2014年第7期1324-1329,共6页Computer Engineering & Science

基　　金：国家自然科学青年基金资助项目(61100103);黑龙江省教育厅科技资助项目(12521601);齐齐哈尔大学青年教师科研启动支持计划资助项目(2012kM13)

摘　　要：多移动机器人系统在完成同时定位和地图构建SLAM任务时,机器人之间常常存在相互碰撞的问题,而这种碰撞的避免又不同于一般的避障,因为避障问题中的障碍物一般是不动的。为了解决机器人之间的避碰问题,提出了一种基于效益的多机器人避碰协调策略。该策略以提高多机器人系统探索效率为主,确定机器人通过交叉路口的顺序。同时考虑了动态协调避碰的情况,给出了确定机器人通过交叉路口顺序的算法。通过机器人在交叉路口实现避碰协调算法的仿真示例,对该方法的避碰协调过程进行了说明,并对仿真结果进行了分析,同时对仿真中机器人和目标位置的空间关系给出了合理的假设。The collision among robots often occurs when a multi-robot system performs the SLAM operation. However, the collision avoidance is different from the general obstacle avoidance, because the obstacles are generally not moving in the obstacle avoidance problem. In order to solve the collision avoidance problem among robots, a benefit based collision avoidance strategy for multi-robot systems is proposed. The strategy mainly focuses on improving the efficiency of multi-robot systems, and determines the order when the robots pass the intersection. Besides, the dynamic collision avoidance is considered and the algorithm of determining the order of robots＇ passing the intersection is given. Finally, a simulation example of robots＇ avoiding collision is given in order to illustrate the process of the strategy and the simulation results are analyzed. And the reasonable assumption is given for the spatial relationship between the robots and targets.

关 键 词：多机器人系统 避碰协调 代价 效用 效益 

分 类 号：TP302[自动化与计算机技术—计算机系统结构]