一种新的自动驾驶轨迹规划方法

摘要

提出了一种轨迹规划算法,该算法可以为自动驾驶生成舒适安全的轨迹,障碍物通过摄像头或车间通讯的协同感知消息((CAM))和分散化环境通知消息(DENM)方式感知.该算法主要分为3个部分,首先由轨迹生成模块生成舒适的轨迹簇,其次由于传感器可能存在的误差,对其他交通车的位置和速度进行估计,最后由轨迹选择模块选择最适合的轨迹.为了帮助无人车做出决策,建立了碰撞概率模型.在MATLAB环境下的仿真证明了该算法安全可靠.%A local trajectory-planning algorithm that generates safety and comfort trajectories is described. The obstacles are detected online by receiving cooperative awareness messages (CAMs) and decentralized environmental notification messages (DENMs). Our approach has three main steps. The first step uses a trajectory generation module to generate comfort trajectories. The second step predicts the position and velocity of obstacle vehicles because of transmission delay. The final step chooses a best trajectory to follow by operating the trajectory selection module. A collision probability model is then proposed to help autonomous vehicle make decision. The trajectory-planning algorithm described in this paper was tested on MATLAB, and the simulation results show they are secure and reliable.