Design Of Fuzzy Controller For Multi-Degree-of-Freedom Manipulators

With the wide application of multi-degree-of-freedom manipulators in manufacturing,non-manufacturing and extreme environments,higher requirements are put forward for the trajectory tracking speed,accuracy and stability of multi-degree-offreedom manipulators.However,in practical engineering applications,multi-degreeof-freedom manipulator is a highly coupled multi-input and multi-output complex system with strong non-linearity,and there are uncertainties such as external timevarying disturbances,unmodeled dynamics,parameter perturbations,load timevarying and internal friction.It is difficult to establish an accurate dynamic model,which makes it difficult to achieve high-precision trajectory tracking control of multidegree-of-freedom manipulators.Therefore,two kinds of high precision,fast and stable trajectory tracking controllers for multi-degree-of-freedom manipulators are designed in this paper.The main research contents are as follows:The establishment of the mathematical model of a manipulator is the basis of the design of the trajectory tracking controller of the manipulator.Therefore,the dynamic model and kinematics model of the multi-degree-of-freedom manipulator are analyzed respectively.Firstly,given the joint variables of the multi-degree-of-freedom manipulator,the position and attitude of the end-effector of the multi-degree-offreedom manipulator relative to the basic coordinate system,that is,the positive kinematics model of the multi-degree-of-freedom manipulator,is established by using the D-H method.Secondly,the inverse kinematics model of the multi-degree-offreedom manipulator is analyzed.Finally,the dynamics of the multi-degree-offreedom manipulator is modeled by Langrange method,and the dynamics model and characteristics of the multi-degree-of-freedom manipulator are analyzed.Aiming at the uncertainty of mathematical model of multi-degree-of-freedom manipulator,an adaptive control method of the manipulator based on RBF neural network is proposed.Firstly,the dynamic model of the multi-degree-of-freedom manipulator is identified by using RBF neural network through off-line training and on-line learning.Then,the adaptive control algorithm of RBF neural network is designed to approximate the friction and gravity items in the dynamic model of the manipulator to obtain the compensation control quantity.Robust control is designed for time-varying disturbance and RBF neural network approximation error to overcome the influence of many uncertainties.At the same time,the stability of the designed control system is analyzed by constructing Lyapunov function.Finally,simulation results show that the proposed composite control method has higher trajectory tracking accuracy,speed and stronger interference ability when the mathematical model of the manipulator is uncertain.However,when the initial time,friction and disturbance jump,the trajectory tracking error of the multi-degree-of-freedom manipulator is slightly larger and the convergence speed of the tracking error is slower by using RBF neural network adaptive robust controller.On the basis of RBF neural network adaptive control,the fuzzy logic control is introduced,and a multi-degree-of-freedom manipulator adaptive robust controller based on fuzzy RBF neural network is designed.Firstly,the fuzzy RBF neural network adaptive control algorithm is used to approximate and compensate the non-linear function including friction and gravity term in the dynamics model of the manipulator.Then,the robust controller is designed to compensate the uncertainties such as fuzzy RBF neural network approximation error,load time variation,parameter perturbation and external time-varying disturbance.The Lyapunov function is used to analyze the stability of the proposed adaptive robust control system of a manipulator based on fuzzy RBF neural network.Finally,the simulation results verify the effectiveness of the proposed control system.