模块化群体机器人构型分析与自组装控制

Configuration Analysis and Self-assembly Control for Modular Swarm Robots

源于社会性昆虫的自组织原理,自组装为群体机器人的协作提供了有效的行为模式。提出一种分布式自组装控制方法,并基于自主开发的新型自组装模块化机器人Sambot进行试验研究。采用构型连接状态表的方法对多个Sambot组成的集合体机器人的运动构型进行描述与分析,建立运动构型库。提出一种由种子机器人、对接机器人与连接机器人组成的分布式自组装系统,其中种子与连接机器人决定自组装规则,并控制构型增长;对接机器人没有任何目标构型与全局坐标信息,采用基于行为的控制器,实现与种子及连接机器人的的自主对接。用5个Sambot机器人成功演示了典型蛇型与4足构型的自组装试验,验证了控制算法的有效性。

Inspired by the swarm behaviors of the social insects,the self-assembly provides an effective behavioral paradigm for collaboration of the swarm robots.Based on the Sambot platform,a distributed self-assembly control algorithm for a given robotic structures is presented.A configuration connection state table is used to represent the configuration of the robotic structures formed by multiple Sambots.A distributed self-assembly control model is proposed.There are three types of Sambots,i.e.SEED,connecting and docking Sambots.The SEED and connecting Sambots are used to decide the self-assembly rules and control the growth of the configuration.Docking Sambots,without any target configuration and global coordinate information,adopt behavior-based controller to realize autonomous docking with SEED and connecting Sambots.Five Sambots are used to successfully demonstrate the self-assembly experiments of snakelike and quadruped configurations,which verifies the effectiveness of the control algorithm.