Research On Distributed Coordination Control Methods Of Multi-Robot Formation

With the development of control theory, bionics and artificial intelligence, distributed coordination control of multi-robot is getting more and more attention. The robots are often required to form and keep the desired formation in performing tasks, so the formation problem of multi-robot has always been one of the attractive research in scholars and experts.This paper focuses on the coordination of formation and tracking task in multi-robot. Firstly, the system architectures of single robot and multiple robots are introduced. Then the distributed control structure of multi-robot system is presented, and is adopted in the research of this paper. A robot model is presented, and each module of the robot is designed from the view of autonomy and openness, reflecting the robot intelligence and adaptability to environment. The location is the basic of the robot control. The wheeled robot motion model is described, and the dead reckoning is used to estimate its location. RFID (Radio Frequency Identification) technology is used as a location correction to help the robot have good location capability, and error correction model is given.A distributed coordination control method of multi-robot formation is the emphasis of the research. The communication mechanism and the method of formation of multi-robot are introduced, and the formation control in tracking problem is studied. Dynamic acquaintance network is used to maintain other robots’states in the local robot. A method of coordination behavior is proposed to keep formation, which is relied on the self assessment based on RFID to correct position and the relationship of acquaintance. Also, the principle of potential field-based is used to avoid collision and rules for the behavior of along the wall while in the process of movement. And the reactive collision avoidance based on sensors is designed for emergency. Behavior design is composed of two basic actions and packaged, which improves the reusability of behaviors.Finally, simulation and experiments are designed in order to verify the effectiveness of the proposed method in the research of the theory. Experiments are performed on the platform by using the real robots and virtual robots. The host computer communicates with the robots, and the formation and tracking is real-time displayed in the monitoring interface. By comparing and analyzing the experimental results, the method in this paper shows better stability of formation.