Research On Local Trajectory Planning And Trajectory Tracking Control Of Intelligent Vehicle

With the rapid increase of the total number of cars,traffic accidents occur frequently,and people pay more and more attention to the safety problems in the process of car driving.Driver’s human error is one of the main causes of traffic accidents,and the development of autonomous driving technology can reduce traffic injuries caused by human factors.Local obstacle avoidance trajectory planning and trajectory tracking control,as the key components of autonomous driving technology,have a decisive impact on the safety and comfort of intelligent vehicle driverless driving.Therefore,based on the idea of numerical optimization and spline interpolation,a local obstacle avoidance trajectory planner is studied in this paper.Based on model predictive control and active disturbance rejection control theory,a trajectory tracking controller is studied.The main research contents are as follows:The first,through the magic formula tire model,the longitudinal and lateral characteristics of tire are analyzed,and the linear relationship between tire lateral force and side Angle is obtained.The linear vehicle dynamics monorail model in geodetic coordinate system is built based on this relation.Based on this model,the linear trajectory tracking error model in Frenet coordinate system is rewritten for the design of lateral path tracking controller.The second,based on the transformation relationship between geodetic coordinates and Frenet coordinates,the global path is optimized and the reference line is generated.The local obstacle avoidance trajectory planning is decoupled into path planning and velocity planning.For urban structured scenes,a local obstacle avoidance path planner was designed based on expectation maximization algorithm and spline interpolation.Considering the curvature discontinuity of high-order polynomial curve,the path optimization algorithm was improved on subsection jerk optimization algorithm.Considering the influence of point and ellipse vehicle shape models in traditional algorithms on opening up convex space,the mathematical model of vehicle volume is improved to a corner model.The discrete S-T diagram is established,the sampling points are connected based on spline interpolation method,and the multi-objective cost function considering safety and comfort is designed.The dynamic programming theory is used to solve the speed decision problem and open up the convex space.Then the improved optimization algorithm is used to optimize the planning speed twice.The feasibility of the proposed local trajectory planning algorithm is verified by simulation experiments.The third,a prediction model was established based on the trajectory tracking error model in Frenet coordinate system,and the prediction model was improved to an incremental model considering the jump of the control quantity in each cycle.Considering the characteristics of urban driving conditions,combined with vehicle dynamics and other constraints,the objective cost function was established,the QP solver was used to solve the quadratic cost function,and the MPC based lateral controller was obtained.Aiming at the deficiency of PID control method,the longitudinal controller based on cascade ADRC is established,and the calibration table of throttle and brake is made.The Matlab/Simulink-Carsim co-simulation platform was built to verify the performance of the controller.The fourth,an integrated simulation platform of local trajectory planning and tracking control based on Prescan-Matlab/Simulink-Carsim was built to carry out multicondition simulation tests on urban structured roads,and the test results were analyzed to verify the performance of the planning and control algorithms studied in this paper.