A Vehicle Local Trajectory Planning Method in Structured Road Scenes

Abstract

A vehicle local trajectory planning method in a structured road scene proposed by the present invention firstly obtains the motion state vector of the vehicle under the reference coordinate system at the current moment by establishing a reference coordinate system based on the lane reference line; then generates The longitudinal sampling trajectory set and the horizontal sampling trajectory set under the reference coordinate system of different speed intervals, the curve forms of all sampling trajectories in the longitudinal sampling trajectory set are the same, and the curve forms of all sampling trajectories in the horizontal sampling trajectory set are the same; then the horizontal and vertical The sampling trajectory is synthesized; then the synthesized sampling trajectory is converted back to the Cartesian coordinate system, and the sampling trajectory that does not satisfy the trajectory curvature and collision detection is eliminated; finally, the cost of each sampling trajectory is calculated, and the trajectory with the smallest cost is the original Invent the resulting optimal trajectory. The invention can improve sampling efficiency by constructing sampling tracks at high speed and low speed, and has strong adaptability to complex constraint conditions.

Description

Technical field[0001]The invention belongs to the technical field of trajectory planning in intelligent vehicle decision-making, and particularly relates to a vehicle local trajectory planning method in a structured road scene, which is used for online rapid calculation and generation of a vehicle starting point as a starting point in a complex dynamic traffic scene. Collision trajectory.Background technique[0002]According to the different application scenarios, the local trajectory planning of autonomous vehicles can be divided into local trajectory planning in unstructured scenarios and local trajectory planning in structured scenarios. Structured scenes such as highways and urban arterial roads are characterized by known lane shapes, strong environmental dynamics, high vehicle speeds, and short response time left to the automatic driving system. At the same time, in this scenario, the result of the local trajectory planning of the vehicle is a spatiotemporal curve, so the solutio...

AI Technical Summary

Problems solved by technology

[0005] In summary, although the existing local trajectory planning methods based on sampling trajectories have been widely used in structured scenarios, they still have certain drawbacks.

Images