Research On Robust Path Following Control Of Autonomous Vehicles

The path following control is one of the key technologies of autonomous vehicles,which aims to eliminate deviations between the vehicle and desired path,and enabling vehicles to accurately follow the desired trajectory.However,there are model uncertainties and unknown external disturbances in the autonomous vehicle system,as well as input saturation constraint in the actuator and measurement deviation in the speed sensor,all of which will affect the path following control performance.To solve the above problems,this paper investigates the robust path following control method for autonomous vehicles.The main results are summarized as follows.Firstly,a robust control strategy based on the mixed performance indices is proposed to attenuate the degradation of path following control performance caused by localization errors and tire slip angles.Based on vehicle kinematics,the path following model is established into a chained form with localization errors and tire slip angles,where the polytopic model is used to describe the uncertain vehicle model parameters.By using both the L1 and H∞ performance indices to represent the sensitivity of the system to external disturbances,the robust controller design problem under the constraints of mixed performance indices is transformed into a multi-objective optimization problem,and the gain of the controller is designed by solving the optimization problem.Simulation results show that the proposed control method can effectively address the degradation of control performance due to both localization errors and tire slip angles,and thus can guarantee the desired path following performance with strong robustness.Secondly,based on finite-frequency performance indices,this study presents a robust path-following control method for autonomous vehicles with modeling uncertainties,actuator input saturation,and disturbances in the finite frequency domain.An LPV model with actuator input saturation is built based on vehicle dynamics.The frequency domain characteristics of the disturbances are analyzed.By means of the GKYP lemma,the path following controller design problem is transformed into a convex optimization problem which satisfies some H∞performance indices in the frequency domain.Simulation results indicate that the proposed control method can effectively solve the problem of control performance degradation caused by input constraints and disturbances in finite-frequency domain,so as to ensure stronger anti-interference and good path following performance.Thirdly,considering the control performance degradation caused by the deviations between the actual vehicle velocity and its measurement,a previewbased robust path following control method is designed.By analyzing the effects of the inaccurate measurement of vehicle velocity on the preview distance,the preview distance and the vehicle velocity are selected as scheduling variables.Meanwhile,the uncertainties of the inaccurate scheduling variables on the weight coefficients of the polytopic model are analyzed.Sufficient conditions are developed to guarantee that the path following system is stable with H∞performance indices being satisfied.Simulation results demonstrate that the proposed control method can effectively solve the problem of relying on accurate measurements of vehicle velocity in the traditional preview-based control methods,so that the vehicle can follow the desired path stably and has a good preview control performance.Finally,an AET-based robust H∞ output feedback control method is designed with consideration of the limited communication bandwidth and DoS attacks.By analyzing the effects of event-triggered communication on the path following performance,an improved AET mechanism with nonmonotonic characteristics is designed to achieve a better tradeoff between the effective communication performance and accurate path following requirements.Meanwhile,a compensation strategy is presented to avoid the control performance degradation caused by DoS attacks.Simulation results demonstrate that,in the case of limited communication bandwidth and DoS attacks,the designed controller not only can achieve the desired path following performance,but also can significantly save communication resources.