Neural Network Control Of Coordinated Manipulators

In recent years,problems about control of multiple robotic manipulators have aroused increasing interest of many scholars.Compared with one robotic manipulator system,the theoretical research about multiple robotic manipulators system is still relatively weak.The dynamics of coordinated manipulators are more complex than single manipulator system,especially about the problems of trajectory tracking,collision avoidance,load distribution of multiple manipulators.In our country,many institutions of higher learning and research institutions are actively studying about coordinated manipulators system and how to guarantee the stability of the system.This is a hot issue for many researchers who try to study about control of coordinated manipulators.In this paper,we aim to put forward a solution to tackle the problem about trajectory tracking of coordinated manipulators with uncertain parameters,and design a controller based on a barrier Lyapunov function to solve the output constraint,we also add Radial basis function neural network(RBFNN)into the controller to approximate nonlinear function.Based on the Lyapunov function and RBFNN,we design full state feedback controller and output feedback controller for coordinated manipulators system,and for system with output constraint,we also design a controller based on a barrier Lyapunov function.At last,the effectiveness of these methods is verified by MATLAB simulation platform.The dynamics of coordinated manipulator system is derived from dynamics of one rigid manipulator system,the system is consist of several rigid manipulators and an object.Manipulators grasp the object and make object move by following the predetermined trajectory.Compared with the dynamic of one manipulator system,multiple manipulator system is more complex and have more uncertain parameters which make it hard to have accurate dynamic model of coordinated manipulators.Therefore,we design two solutions which are full state feedback control method and output feedback control method for approximating the dynamic of system with RBFNN.,the trajectory tracking is achieved effectively.We also use Lyapunov function to insure the stability of system and prove that all of the parameters of system are semi globally uniformly bounded.At last,trajectory tracking is verified by MATLAB simulation.For coordinated manipulators system with output constraint,we analyze the influence caused by output constraint and propose a barrier Lyapunov function to ensure output constraint.We also use Lyapunov function to analyze the stability of system and prove that all of parameters of system are semi globally uniformly bounded.At last,we verify that this control method based on barrier Lyapunov function and RBFNN can make coordinated manipulator system track trajectory effectively.Based on coordinated rigid manipulator system model,we propose full state feedback control method and output feedback control method.We use RBFNN to approximate several unknown parts of dynamic of the system and regard these unknown parts as a whole.In this paper,we also analysis the stability of system and verify that all parameters of system are semi globally uniformly bounded.For coordinated system with output constraint,we put forward a barrier Lyapunov function to handle the influence cause by output constraint.Based on the existence theoretical research,this paper extends study about the coordinated manipulators system,propose control methods to handle problem of output constraint,verify the stability of system and achieve trajectory tracking effectively.