基于Frenet坐标系的智能车辆多场景轨迹规划

Multi-scene Trajectory Planning of Intelligent Vehicle Based on Frenet Coordinate System

轨迹规划是智能车辆行为决策模块的重要组成部分,也是车路协同技术的重要一环。Frenet坐标系相对于传统的笛卡尔坐标系,在车辆规划算法中具有明显优势。因此提出了一种基于Frenet坐标系的智能车辆多场景轨迹规划模型。首先在Frenet坐标系的基础上,根据车辆动力学模型解耦并构建横、纵向两个维度的轨迹规划集模型,然后根据各场景需求分别建立高速、低速、跟驰等多种场景的轨迹优化函数。最后通过仿真验证,证明该模型具有可实现性,并能高效准确地解决多种场景下的轨迹规划问题。

Trajectory planning is an important part of behavior decision-making module in intelligent vehicle,and it is also an important part of cooperative vehicle infrastructure system.Compared with the traditional Cartesian coordinate system,Frenet coordinate system has obvious advantages in vehicle planning algorithm.A multi-scene trajectory planning model for intelligent vehicle based on Frenet coordinate system is proposed.Firstly,based on Frenet coordinate system and vehicle dynamics model,trajectory planning models in horizontal and vertical dimensions are built.Then the trajectory optimization functions in high-speed scene,low-speed scene,car following and other cases are established.Finally,simulation is performed and the results show that the model is reliable and accurate to realize trajectory planning in various scenes effectively.