Research On Local Trajectory Planning And Path Following Control Of Intelligent Vehicle

Trajectory planning and tracking algorithm is an important part of vehicle intelligence.This paper takes intelligent vehicle as the research object,and studies the motion planning and control problems in the course of vehicle driving.In order to meet the real-time,comfort and driving constraints of vehicle to motion planning algorithm in the planning process,a local trajectory planning algorithm of intelligent vehicle is designed according to the given reference trajectory.Aiming at the problem that the traditional model predictive controller can not guarantee the tracking accuracy of vehicle on curved road,a lateral control strategy of path tracking based on state feedback is proposed.Based on the optimal control method,a local trajectory planning algorithm with discrete sampling optimization is designed.According to the vehicle road relationship,the Frenet coordinate system is established to reduce the solving dimension of the trajectory planning problem.The road discrete points are sampled in the parameter space,and the trajectory set is screened by vehicle kinematics and dynamics constraints and obstacle information,and the trajectory set is constructed according to different driving scenes.The optimal trajectory is selected by the trajectory evaluation function,and the speed optimization algorithm is designed to determine the safe driving speed according to the actual road and obstacles,which is used as the reference speed for vehicle trajectory planning.Based on the vehicle dynamics model,a path tracking model predictive controller considering the constraints of tire slip envelope is established.The time-domain parameters of the controller are selected according to the vehicle speed.The vehicle tracking error model is established with the vehicle centroid position as the control point,and the state feedback regulator is established with the yaw angle deviation of the vehicle current position.The attitude of unmanned vehicle is corrected by LQR optimal control method.Finally,MATLAB/Simulink and Carsim software are used to verify the proposed motion planning and control algorithm.Through the simulation analysis and verification,the quintic polynomial trajectory planning method based on discrete sampling can plan a reasonable trajectory,and the generated trajectory is safe,stable and meets the comfort requirements;at the same time,the speed and acceleration curves are continuous through the speed optimization method.The simulation results of path tracking controller on typical double lane changing road shows that the lateral deviation of vehicle path tracking is reduced by more than 16%and the yaw angle deviation is reduced by more than 33%at low and medium speed.The designed controller can effectively improve the accuracy of vehicle path tracking and ensure the adaptability and driving stability of vehicles on variable curvature curves.