Research On Multi-Mode Flocking Motion System For Multi-Robot Under Complex Environment

In nature, fish, insects, birds and other creatures will spontaneously form a regular and well-organized group to move. Flocking, therefore, is a group as a whole that is aggregated in certain formation, performing a collective movement. Flocking system is a complex system, which has the characteristics of noncentralized control, self-organization emergent and partially interactive properties. Currently in the research of mobile robot system, the multi-robot flocking system is a very important research direction. With the multi-mode design for the system, the property of self-adaptability to complex environments could be enhanced. In this dissertation, the multi-robot multi-mode flocking motion control system will be presented. The main works are as follows:1. The theoretical significance of multi-robot flocking motion control system is introduced. The present study status and methods of multi-robot flocking motion control system are summarized.2. A sequential control algorithm of multi-robot flocking motion is introduced, then, the flocking motion control systems with or without a leader are discussed respectively. On the basis, the concept of "virtual force" is introduced. Furthermore, the transformation method from Agent to mobile robot motion control is analyzed according to kinetics rule, which makes the algorithm of multi-robot flocking motion control can be applied to a real mobile robot. In addition, an experimental platform for the simulation of multi-robot flocking motion, MuRoS, is introduced.3. A multi-robot multi-mode flocking motion control system is designed. Aiming at tackling the "local minimum" problem which exists in the algorithm of robot’s obstacle avoidance of complex environment, the rigid’behavior of collision avoidance and the behavior of collision avoidance when it moves along the wall are combined and a multi-mode switching mechanism is designed, which will make multi-robot flocking system have the capability of real-time response and can be adaptive to the environment. And the simulation results verify the efficiency and effectiveness of the algorithm in dynamic and irregular environment.4. Aiming at tackling the trajectory tracking problem of the nonholonomic mobile robot, a simple controller is designed combining Backstepping method and Lyapunov method, which can make the system achieve global stability. The experimental results show that the controller possesses simple structure, ease of implementation, and excellent real-time performance.5. The motion control algorithm of nonholonomic mobile robot and the multi-mode obstacle avoidance algorithm for multi-robot flocking system in complex environment are integrated. Multi-mode motion control algorithm of Leader robot based on Backstepping is presented, which can make the Leader robot get to destination more quickly and accurately in complex environment. The simulation results verify the effectiveness and real-time performance of the algorithm.