| Efficient travel and production methods have always been the source of human technological progress.Autonomous driving technology has attracted much attention since its emergence because of its efficient and convenient application value and broad market prospect.In recent years,researches on autonomous driving technology have mainly focused on three aspects: environment perception,decision planning and motion control.Among them,the motion control part is the “hands and feet” of an autonomous vehicle.Every operation instruction to actuators on the vehicle is calculated and sent by the motion control module.A reliable motion control algorithm is the key to ensure whether the vehicle can run according to the expected route and ensure passenger comfort and safety.Taking different vehicle driving conditions as the starting point,this paper studies the motion control,and discusses the vehicle motion control algorithms on rough roads,unstable longitudinal longitudinal velocities and large curvature curves respectively,so as to ensure that the autonomous vehicle can complete the motion control task quickly,stably and accurately.First of all,in order to solve the problem that the autonomous driving vehicle cannot be accurately modeled due to disturbances on rough roads,a lateral control algorithm based on a nonlinear ESO and a nonlinear controller is proposed.The nonlinear ESO can effectively estimate the disturbance during driving,and compensate it in the controller to improve the robustness of the system.The nonlinear controller can make the vehicle quickly and accurately converge to the reference trajectory and improve the accuracy of vehicle lateral control.The Lyapunov method is used to prove the finite-time convergence property of the closed-loop system.The results of real-vehicle experiments verify the effectiveness of the proposed method.Secondly,considering the path tracking problem of the vehicle under unstable longitudinal velocity conditions,a vehicle lateral control scheme with finite-time convergence is proposed.A dual closed-loop control strategy is designed for the kinematics and dynamics models of autonomous vehicles.A trajectory tracking controller is designed for the outer loop system to track the reference trajectory of the vehicle.The inner loop system is designed with a nonlinear ESO and a yaw angular velocity controllers to ensure the tracking of the desired yaw angular velocity from the trajectory tracking controller.The finite-time convergence property of the double closed-loop system is proved by the Lyapunov method.The results of real-vehicle comparison experiments show the superiority of the method under the unstable longitudinal speed on the vehicle.Finally,for the path tracking problem on large curvature lanes such as turning lanes or roundabout lanes,a finite-time vehicle motion control algorithm with full state constraints is proposed.Considering the lateral and longitudinal subsystems of the motion control system on autonomous vehicles,a finite-time lateral controller and a finite-time longitudinal controller with full state constraints are designed respectively.At the same time,for the unmodeled dynamics in the lateral subsystem and the longitudinal subsystem,nonlinear observers are designed in the corresponding subsystems,respectively.The finite-time convergence property of the proposed control scheme is proved by the Lyapunov method.The results of real-vehicle comparison experiments verify the superiority of the proposed method in the process of tracking the roundabout lane. |
