Safe and Efficient Trajectory Planning Considering Longitudinal and Lateral Coupled Limits

Autonomous vehicles need to coordinate longitudinal and lateral motion to achieve safe and efficient driving when facing complex and dangerous scenarios. This work explores the longitudinal and lateral coupled constraints of vehicle motion from the frictional limit of each tire and proposes a general longitudinal and lateral coupled trajectory planning framework, especially improving the extreme performances of autonomous vehicles. Although the multiple tire force constraints are complex, through the analysis of the relationship between vehicle motion and tire forces, the problem could be simplified. According to the vehicle dynamics model and the tire friction elliptical cones, the discrete data of the coupled envelope of the feasible longitudinal, lateral, and yaw angular acceleration of the vehicle were obtained. An analytical model for the coupled limits of the longitudinal, lateral, and yaw motion, ACORN, was proposed and adopted to fit the discrete coupled envelope. Based on the Frenet frame, the longitudinal and lateral information of the trajectory was modeled in the form of multi-shooting and was solved simultaneously. Combined with a trajectory tracking controller, the proposed planning module was verified by simulations and experiments with a real vehicle platform, which was effective for both safer driving in extreme maneuvers and faster racing on the course.