The Collaborative Control Of Multi-manipulator System In Task Space

Since its inception,robots have been receiving widespread attention from society.The structure of the manipulator is generally more complicated,and the multi-link structure brings a high degree of flexibility to the manipulator,and also increases the difficulty of the manipulator control.In the early stage of the development of manipulators,only the control of a single manipulator is considered.However,with the increasing expectations of modern industries for manipulators,a single manipulator can no longer meet the requirements of more and more complex tasks,so the collaborative research of multiple manipulator systems has received extensive attention from domestic and foreign scholars.However,since the task description and completion of the manipulator are performed in the task space,the traditional joint space control method is difficult to adapt to different application scenarios,and the research on the manipulator control in the task space is gradually emerging.This thesis studies the cooperative control method of multi-manipulator in task space under undirected graph and directed graph.Under the undirected graph,a cooperative control algorithm based on the backstepping method is proposed for the symmetrical nature of the Laplacian matrix,and under the directed graph,an adaptive control algorithm including a sliding observer is proposed,so that The multi-manipulator system can achieve static consistency in the task space,that is,the position of the task space tends to be the same,and the task space speed tends to zero.And considering the uncertainty of the dynamic model and kinematics mo del of the manipulator and the presence of noise in the task space velocity measurement,a task space velocity observer was proposed,and a distributed algorithm without task space velocity measurement was designed.The stability theory proves the stability of the control algorithm including the task space velocity observer,and the effectiveness of the control algorithm is verified by simulation.In addition,for the manipulator with redundant degrees of freedom,the joint speed of the redundant manipulator's null space is designed to optimize the motion of each joint during the coordinated motion of the manipulator.For example,complete the sub-task goal of avoiding singularities and improve the manipulability of the manipulator without affecting the main tasks of the task space.The collaborative control algorithm and sub-task control algorithm simulation of the multi-manipulator system with redundant manipulator under the strongly connected directed graph verify the static consistency and manipulability of the multi-manipulator in the task space.In addition to the static consistency of the multi-manipulator system,this thesis also studies the dynamic problem of the trajectory tracking of the multi-manipulator system.The topological graph is a strongly connected directed graph,and the task space velocity cannot be measured.Considering the dynamics and kinematics uncertainty,a completely distributed control method is designed,which can still realize the synchronization of the multi-manipulator system and the overall trajectory tracking when only the leader knows the target trajectory.And the effectiveness of the control algorithm is verified by simulation.