Research Of Tracking Control On Multi-robot Based On Multi-agent

With the rapid development of society and the rapid progress of science and technology of robots, the multi-robot systems are increasingly concerned about by people. The coordination control of multi-robot is an important issue in the field of robotics research, and in addition, multi-robot system based on agent theory is a hotspot of current research. In this thesis, tracking control problems of multi-robot system are researched based on multi-agent system theory.Firstly, the background and current status of the research are presented, the important theoretical and practical significance of the subject are presented, and the basic theories of multi-agent tracking control method are described. Then nonholonomic constraint model of mobile robot is analyzed. The motion model of robot is established from the two aspects of kinematics and dynamics. With reference to the theory of multi-agent system and the model of robot, the architecture of multi-robot is designed.Secondly, the tracking control problems of single robot based on agent theory are studied. According to the robot kinematics model, the control law of point stabilization is designed by constructing a Lyapunov function. According to the kinematics and dynamic model of the robot, the control law of trajectory tracking is designed by constructing Lyapunov functions. The control laws of point stabilization and trajectory tracking problem are validated respectively by matlab simulation.Thirdly, the tracking control problems of multi-robot based on multi-agent theory are studied in a simple environment, and the problems are divided into three types including dynamic tracking, flock tracking and formation tracking. The three kinds of tracking problems are analyzed respectively by using the robot kinematics model. The control law of dynamic tracking is designed by using consistency theory. The control law of flock tracking is designed by using the combination of consistency theory and potential energy function. The control law of formation tracking is designed by using leader-follower method and the trajectory tracking algorithm of single robot. In a simple environment, the control laws of the three kinds of tracking problems are validated respectively by matlab simulation.Finally, the tracking control problems of multi-robot based on multi-agent theory are studied in a complex environment. In the environment with obstacles, the obstacle avoidance problems of multi-agent are studied in the process of tracking control, and the problems are discussed through two ways that are autonomous obstacle avoidance and collaborative obstacle avoidance. The law of autonomous obstacle avoidance, which can be used in the three types of tracking problems including dynamic tracking, flock tracking and formation tracking, is designed according to potential energy function method. The law of collaborative obstacle avoidance is designed according to environment aware computing and conditional rules, and it is suitable for formation control.