Formation Control of Multiple Behavior-based robots

Abstract

Reactive control has the character of flexibility and timeliness. Aimed at the formation control of multiple robots, this paper presents a behavior-based method on it. In the process of moving to the goal, the robots take the leader-referenced technique. And the dynamic-dead-zone method is used to maintain the formation predetermined. The combination of potential-based avoidance and following-wall behavior ensure the robots are able to avoid the obstacles smoothly. To get rid of the following-wall behavior when it is not needed, a step of judging its necessity is added. Finally the experiment on the AmigoBot robots showed that the robots can return the formation after passing the obstacles