Adaptive Cooperative Control Of Multi-manipulators System Based On Multi-agent And Its Application

With the development of science and technology, intelligent is the theme in the modern era of science and technology, and on behalf of the intelligent robot introduction of science and technology. Robot in the continuous development of our demand for intelligent growing at the same time, scientists thought of using multiple robots to solve this kind of demand, so now the study of multi-robot system become a hot issue of control field. As more intelligent multi-agent system, on the other hand, the concept of abstracting, has been studies by many scholars both at home and abroad. To solve the problem of coordinated control of multiple robot arm system in this paper, we use multi-agent research methods to study the coordination control problem of multi-manipulators can simplify the complexity of the problem, to be able to design a feasible system control protocol. Therefore, this article will first of all a kind of multi-agent system to carry out the study, and applies this method to multi-manipulators more deeper study in the system. Paper's main work is as follows:First, this paper has studied in the directed graph a class of non-affine coordinate control problem of multi-agent systems, providing that the objective can only get partial information in each agent, and the leader of the state system is time-varying. But the system presented a non-affine form of the whole system. This paper made use of the mean value theorem and the implicit function theorem split controller, introduce RBF neural network approximation of nonlinear part, designed distributed adaptive coordinated controller, used Lyapunov stability theory proof system in the limited time to achieve consistent stability and maintain stability, which ensure each agent's information or leader of the information on the state to track or state.Second, as to multi-manipulators system, more complex item as multi-manipulators systems are non-affine systems. We used the non-affine algorithm design ideas, used mean value theorem of non-affine systems to separate controller and design a distributed adaptive control. We also used Lyapunov stability theory to prove all slave manipulator system can track the active manipulators system in a limited time.This paper proposes that distributed adaptive coordinated controller can achieve this kind of tracking problem. Two simulation results will show the effectiveness of the algorithm.