Research On Distributed Coordination And Tracking Control Of Multi Arm System Based On Relative Output Feedback

With the continuous innovation and development of science and technology,it is more and more common to use intelligence instead of manual in high repetition work and dangerous environment.In the face of increasingly complex modern industrial needs,the demand for intelligent control is becoming higher and higher,and a single intelligent body is no longer enough to complete some complex tasks.Therefore,researchers combine multiple agents to achieve industrial requirements by controlling multi-agent collaboration.Because in the actual control system,the state variables of the system can not be measured completely or the measurement is expensive,but the relative output feedback control structure is simple,easy to realize physically,low cost and high reliability.Therefore,relative output feedback is usually used to solve the problem of multi-agent coordinated tracking.As a practical application case of multi-agent network system,multi-link manipulator system will be of guiding significance to the research of distributed coordinated tracking control of multi-agent network system.This paper focuses on the research of distributed coordinated tracking control of multi-link manipulator system based on relative output feedback.The main contents of the paper are:Firstly,this paper summarizes the research significance and relevant background,introduces the research status of distributed coordination control of multi-arm system and multi-agent system,and expounds the main research contents of this paper.Secondly,the dynamic analysis of multi-single link mechanical arm system and the mathematical model are established.At the same time,the algebraic graph theory and related matrix theory,adaptive control and system stability theory are briefly summarized.Thirdly,this paper mainly studies the distributed coordinated tracking control problem of multi-link manipulator system based on relative output feedback.According to the multi-link manipulator system with leader,the state information in the system is not measurable.Only using the output information of the system,a kind of tracking control protocol based on relative state output feedback is designed.By means of Lyapunov equation,two algorithms are proposed to determine the output feedback gain matrix,one is iterative algorithm,the other is to construct a non-iterative algorithm of Lyapunov matrix with special structure after proper coordinate transformation of the system.The sufficient conditions for the system to achieve consistency tracking and the controller design method based on the two algorithms are given.Finally,the simulation experiment of multi-link manipulator system is carried out to verify the effectiveness of the control protocol and the two methods,and the two methods are compared in simulation and algorithm.Fourthly,this paper studies the fully distributed adaptive coordinated tracking control problem.For the multi-single link arm system with leaders,a new fully distributed adaptive control protocol is proposed without global information of communication network topology,using matrix theory and algebra graph theory to convert the original system into error system and analyze the stability by Lyapunov method.The simulation experiment shows that the distributed adaptive coordination tracking controller can achieve the consistent tracking goal of multi-single link arm system.Finally,this paper summarizes the research work of the full text,points out the advantages and disadvantages of the theoretical results,and plans and prospects for the future work direction.