Research On Backstepping Adaptive Trajectory Tracking Control Of Nonholonomic Mobile Robot

Mobile robot with three wheels due to suffer the nonholonomic motion constraints in the motion process,the mobile robot system is a typical nonlinear system,so it is referred to as the "nonholonomic mobile robot".Restricted by nonholonomic constraint,general smooth state feedback cannot be applied to this system,and the uncertainty of disturbance and model parameters is existed in mobile robot system under unknown environment.Therefore,aimed at containing the uncertainty of nonholonomic mobile robot motion control research is of great significance.This paper aimed at containing uncertainty trajectory tracking motion control of nonholonomic mobile robot to do the following research work:Under the condition of the model parameters are known,firstly,after considering the complexity of the calculation in the design process of the control law,combining Backstepping design method and dynamic surface sliding mode control with first-order low-pass filter to design trajectory tracking control law with known model parameters and the upper bound of disturbance.Then,in view of the above control law,the disturbance boundary value selection is difficult,an adaptive Backstepping dynamic surface sliding mode control method is proposed to realize the trajectory tracking control with the unknown upper bound of disturbance.Finally,aiming at the existence of torque chattering problem in the above two control methods,a dynamic surface sliding mode Backstepping adaptive fuzzy control method with weak chattering based on disturbance compensation is proposed,The fuzzy system is used to estimate the unknown disturbance in order to reduce the torque chattering.Under the condition of the model parameters are unknown,an adaptive trajectory tracking control law is designed by using the RBF neural network and the Backstepping method.Further considering the RBF neural network to bring the problem of the large amount of calculation,an adaptive Backstepping control method based on neural network minimum parameter method is proposed,online adjustment of single parameter is used to replace weights of RBF neural network in order to achieve the purpose of reduce amount of calculationFinally,the stability analysis and simulation verification of the designed control law are carried out,by using Lyapunov function can be concluded that the designed control law can guarantee the closed-loop system is uniformly stable.Through the analysis of the system simulation curve can get the following conclusion:When the model parameters are known,control system at the same time of complete the trajectory tracking control can overcome the unknown disturbance,and the adaptive fuzzy system can effectively reduce the torque chattering;The control law based on RBF neural network is able to overcome the influence of model parameters and unknown disturbances in the case of unknown model parameters,and the control law designed by the method of minimum parameter of neural network can effectively reduce the amount of calculation.