Time-Domain Predictable Trajectory Planning for Autonomous Driving Based on Internet of Vehicles

For the polynomial lane change method, the lane change trajectory is planned only at the initial time, and it cannot cope with the problem that other traffic participants enter the driving environment during the lane change process. This paper decomposes the polynomial lane change method into lateral displacement planning and longitudinal velocity planning. The Pontryagin minimum principle is used to solve the optimal lane change duration meeting the requirements of different driving conditions, and the polynomial method is used to plan the lateral displacement trajectory. In the longitudinal direction, the variable acceleration motion equation is used to describe the trajectory, so as to establish a prediction model, the real-time driving environment information is obtained through the internet of vehicles to realize the speed rolling optimization, the trajectory dynamic planning is carried out during the driving process, and the slack variable is introduced to solve the problem that the vehicle suddenly increases speed beyond the constraint range. Through Matlab/Simulink and Prescan co-simulation verification, the trajectory planned in this paper not only meets the requirements of comfort and lane change efficiency, but also has better avoidance capabilities for other traffic participants and is easy to follow in real vehicles.