Research On Trajectory Tracking Algorithm Based On Improved MPC Under Ice And Snow Conditions

In recent years,the continuous development of society has led to an increasing demand for artificial intelligence technology.Driverless technology is a very important branch of artificial intelligence technology,and it has been widely studied and applied because it can simplify complex driving behavior and improve people’s travel efficiency.Path tracking technology is an important part of unmanned driving technology,and its main task is controlling the vehicle intelligently,so that it can accurately track the preplanned reference trajectory.The uncertain factors of road and traffic conditions in ice and snow environment bring great challenges to the path tracking control of autonomous vehicle.First,when the autonomous vehicle is driving on ice and snow roads,the road adhesion coefficient will change because the vehicle switches back and forth between non-ice and snow roads,and the path tracking of the autonomous vehicle may be offset when the road adhesion coefficient changes;Secondly,there may be dangerous areas on ice and snow roads that cannot be driven through but can only be bypassed,such as snow piles.Therefore,while tracking the path,it is necessary to monitor the dangerous area and avoid obstacles,so as to avoid traffic accidents.However,the existing path tracking algorithms can not meet the requirements of accurate path tracking and obstacle avoidance under ice and snow conditions.To solve these problems,a new trajectory tracking control system for collision avoidance is proposed in this paper.The system is mainly consituted of a path tracking controller,a local path replanner and a speed controller.The main task of the path tracking controller is to accurately track the preplanned reference path under ice and snow conditions;The task of the local path replanner is to output a local obstacle free reference path when the vehicle encounters a dangerous area in path tracking process;The task of the speed controller is to make the vehicle track and arrive at the designated place within the designated time,so as to realize the track tracking.The innovation points are as follows:(1)In order to reduce the deviation of autonomous vehicle when tracking on ice and snow roads,this paper proposes a new path tracking controller.The yaw angle disturbance is constructed in the model predictive control algorithm(MPC),and the optimal steering angle control input is obtained by fusing the proportional integral differential algorithm(PID)to control the vehicle to track to the reference track as soon as possible,which improves the path tracking ability of autonomous vehicle on ice and snow roads.(2)In order to meet the requirements of automatic driving vehicles to avoid dangerous areas while tracking paths on ice and snow roads,this paper proposes a local path replanner,which generates a new collision free path in real time through an improved artificial potential field method,and the new collision free path allows temporary deviation from the reference path to avoid dangerous areas.(3)In order to improve the stability of the artificial potential field method in searching target points and the local optimization ability in generating collision free paths,a new function of obstacle repulsion potential field is constructed in this paper.Its principle is to decompose the repulsion force generated by obstacles to vehicles into two directions:the direction of the connection between obstacles and vehicles and the direction of the target-point to the vehicle gravity,which improves the convergence accuracy of the artificial potential field method.(4)In order to meet the real-time tracking requirements of vehicles arriving at the designated destination within the specified time,this paper introduces a new speed controller into the MPC algorithm.Its principle is to calculate the reference speed of the reference point in real time according to the current vehicle position and the position of the reference point,so as to meet the real-time requirements of MPC algorithm path tracking.To check the performance of the algorithm,the proposed path tracking controller,model predictive control algorithm(MPC)and pure tracking algorithm(PP)are simulated and compared on straight path,lane changing path,uphill path and sinusoidal path in this paper.It is proved that the path tracking ability of the path tracking controller is stronger than other path tracking algorithms when the vehicle moves to the ice surface and offsets.Secondly,this paper artificially sets the dangerous area on the reference path,and proves the obstacle avoidance ability of the local path replanner from the perspective of the obstacle avoidance state of the autonomous vehicle and the replanned path.