The Research On Path Following Control Based On Nonlinear Extended State Observer

The Unmanned Vehicle(UV)can be divided into four types based on the working environment,which includes Unmanned Aerial Vehicle(UAV),Unmanned Surface Vehicle(USV),Unmanned Underwater Vehicle(UUV),and Unmanned Ground Vehicle(UGV).With the development of science,technology,and the continuous expansion of human activities,the emergence of different types of unmanned vehicles can meet the needs of unmanned operations in various industries,which effectively prevents workers from safety-related accidents.Therefore,the development of unmanned vehicles has become the main trend of future traffic.The path following control is an important part of automatic driving technology and its robustness is related to whether the unmanned vehicle can perform well in practical applications.The active disturbance rejection control(ADRC)compensates the existing disturbances through the Extended State Observer(ESO),which shows good robustness.This paper starts with the convergence principle of the nonlinear ESO observer,then we develop a robust path following controller for the USV and the intelligent vehicle respectively based on the homogeneous nonlinear ESO observer(HNLESO).Overall,the main contributions include:(1)According to the Lyapunov stability theory and the related theorem of stochastic differential equations,the convergence of the NLESO observer for the mean square estimation of stochastic disturbances is analyzed,the convergence errors of the observer under assumptions are given.The theoretical analysis shows the feasibility of the NLESO observer in estimating stochastic disturbances.(2)Secondly,regarding a fully-actuated USV with stochastic lumped disturbances(stochastic external disturbances and model uncertainties)and the input saturation,a homogeneous nonlinear extended state observer(HNLESO)is designed to estimate stochastic lumped disturbances which will be compensated in controller design.The dynamic surface control(DSC)method is used to design the path following the controller considering input saturation.The subsequent contrast simulations illustrate the effectiveness of the proposed path-following controller.Compared with the traditional dynamic sliding mode surface(DSMC)method,the proposed controller can accurately follow the desired trajectory with smooth controls which invalidates the constraint.The controller is robust to stochastic lumped disturbances with satisfying effects.(3)Based on the two-degree freedom vehicle model,the robust path-following controller for the intelligent vehicle is developed.To improve the robustness of the control system,the corresponding HNLESO observer is designed to estimate the model uncertainties,and the nonlinear error feedback law is designed based on the global fast terminal sliding mode(GFTSM)method.Simulation results show that the proposed method in this paper can ensure that the vehicle has a good following effect under various working conditions and has strong robustness.