Coordinated Control Of Backstepping Consensus For Networked Flexible Manipulator Systems

With the rapid advance of multi-agent system theory and its applications in the past two decades,the coordinated control of the networked Lagrange systems derived from the application background of cooperative multiple robots has attracted more and more interest and attention in many field of science and engineering.In particular the distributed consensus control of networked flexible manipulator systems described by under-actuated Lagrange dynamics has become one of the most active research topics in the field of dynamics and control.This is mainly because the flexible mechanical arm is viewed as the most common,important and versatile robot because of some potential advantages including more flexibility,high security and low energy consumption,and so it has a wide range of engineering applications in complex unstructured task environment such as lunar exploration and manned spaceflight,underwater surveys and deep space exploration,environmental monitoring and disaster rescue etc.On the other hand,networked flexible manipulator systems is regarded as a typical high-dimensional complex dynamic system with physical constraints and information interactions having multi-levels,multi-structure,and multi-scales,which also possess obvious under-actuated dynamic characteristics.As a result,the swarm coordinated control of a large-scale flexible manipulator systems under complex network environments is more challenging.Therefore,the study of distributed consensus control of networked flexible manipulator systems formulated by under-actuated Lagrange dynamics is of great scientific significance in theory,but also has very important practical value in engineering application.From the perspective of distributed control theory of multi-agent systems,by using the research mode and technical approaches in combination with convention analytical dynamics and modern network science,this theis is devoted to studying the issue of of swarm coordinated tracking control of networked flexible manipulator systems formulated by the under-actuated Lagrange dynamics.The main work of the thesis can be summarized as the following two aspects:1.Coordinated control of backstepping consensus for networked flexible manipulator systemsBy integrating backstepping method and sliding mode technique on nonlinear control systems,the primary objective is to develop the tracking control method for single flexible-joint manipulator for the study of networked flexible manipulator systems formulated by under-actuated Lagrange dynamics.The distributed coordinated algorithms of consensus tracking control for such networked robot systems are proposed under two cases of determinate system models and uncertain one respectively,and then the convergence analysis procedure of closed-loop control systems is presented based on the Lyapunov stability theory.A main feature of the proposed backstepping consensus algorithms is the introduction of the network topology describing the communication interaction among robots,and it also is fully distributed for controlling a large group of swarm robots.Finally,numerical simulations are performed to verify the effectiveness of the proposed control scheme.2.Consensus control of networked flexible manipulator systems based on RBF neural networkBy considering in practice the inevitable factors of system model uncertainty and external perturbation,etc,the proposed distributed consensus tracking algorithms is further developed based on radial basis function(RBF)neural networks.It is shown that the developed adaptive consensus tracking algorithm makes full use of the universal approximation performance of RBF,which can ensure to accurately achieve the desired consensus control objective.Meanwhile,it can make online adjustment and estimation of the model parameters,and also effectively inhibit and eliminate the chattering effect caused by the uncertainties and external disturbances.To this end,simulation examples are provided to show the correctness of the proposed intelligence algorithms.