Research On Trajectory Tracking Control Of Self-driving Vehicle

At present,the safety technology of vehicles is gradually developing towards active safety.Because the driver has more uncertain factors in the human-vehicle-environment system,it is considered to be the main cause of traffic accidents.Autonomous vehicles can significantly reduce or even avoid the adverse effects of the driver,thus greatly improving road traffic safety.The core technology of autonomous driving mainly includes three aspects:(1)environmental awareness.The vehicle acquires surrounding obstacles and environmental information through various sensors;(2)path planning and decision making.The vehicle plans an optimal driving route according to the obtained surrounding environment information;(3)trajectory tracking.The vehicle utilizes the steering and drive system to travel safely and steadily along the planned path,involving longitudinal speed following and lateral path control of the vehicle.A stable and accurate trajectory tracking control system is the underlying goal of self-driving cars.This paper mainly studies how to complete the trajectory tracking control of autonomous vehicles based on low-attachment roads.Aiming at the above problems,this paper firstly establishes the corresponding vehicle dynamics model and tire model,then designs the trajectory tracking control system based on the model predictive control theory,and adds the soft angle constraint of the lateral angle to the predictive control algorithm of the time-varying model.The stability of the vehicle is enhanced,and finally the trajectory tracking control of the vehicle is realized.The tire model is established according to the magic formula,and a dynamic model including longitudinal motion,lateral motion,yaw motion and front wheel rotation is established.The simulation test results show that the dynamic model and the tire based on the magic formula are obtained.The models have good accuracy and can meet the accuracy requirements of trajectory tracking control.Finally,the vehicle parameters are set in Carsim,and the vehicle simulation model is established.The effectiveness of the vehicle model is verified by comparing the simulation of Carsim model and Simulink model.The nonlinear model predictive control system is linearized,and the corresponding predictive model equations are proposed for the linear time-varying model prediction algorithm.At the same time,in order to realize the computer programming solution,the model optimization problem is transformed into the standard quadratic programming form.When establishing the front wheel active steering trajectory tracking controller,based on the optimization objective function and constraint conditions,the soft angle constraint of the side angle is introduced to maintain the stability of the vehicle at high speed,preventing the vehicle from losing the trajectory tracking ability due to the front axle slippage or The shaft slides sideways and ends.The Carsim/simulink joint simulation platform was established to complete the simulation of the double-shifting condition of the trajectory tracking control.The simulation comparison results show that the trajectory tracking controller for front wheel steering under low adhesion road has obvious superiority,and has certain adaptability to vehicle speed change and road surface adhesion conditions.