Research On The Formation Of Multi-robots Based On The Hybrid Algorithm

Mobile robot is a comprehensive system that integrates a variety of functions, such as theenvironment perception, dynamic decision-making, behavior control and execution. With thedevelopment of the society, the single robot has been unable to meet with the needs of humanproduction and life. Multi-robots system can well implement the tasks that the single is unableto conduct. The multi-robots system has the characteristics of distribution, robustness andfault tolerance. Multi-robots coordination technology has become the focus of the robotresearch. The AmigoBot produced by ActivMedia Robotics Corporation was taken as theresearch platform in this study.A single robot path planning was deeply studied, on this basisdifferent control methods of multi-robots system formation were also studied.The formationof three robots in different environment was realized.Firstly, the research status and significance of multi-robots system were introduced. Theproblems of Multi-Robot system were also described in detailSecondly, the control methods of multi-robots formation were discussed and apreliminary comparation between various control methods was made. The standards of theselecting of reference point and formation of robots team were given. Meanwhile theexperiment platform was also introduced.Thirdly, based on the traditional artificial potential field (APF), an advanced artificialpotential field that had the ability to weaken the jitter which existed in traditional artificialpotential field was given.Finally, different control methods are implemented. Both APF and feed-back controlmethods were used in the process that the leader moved to the goal. Leader-follower controlmethod and dynamic APF method were used to keep the formation during the robots moving.In order to avoid the obstacle during the robots moving, wall-follow and APF were used basedon the environment. In addition to that, the behavioral inhibition method was adopted as theselection mechanism of robots formation’s behavior. When using dynamic APF, the followersmight be left by the leader in the process of obstacle avoidance, because the largest speed offollower was used. Thus the feed-back between leader and followers was given. If thedistance between leader and followers was too large, the leader’s speed would be changed towait the followers to keep the formation steady.