自动驾驶车辆主动换道轨迹规划方法研究

Research on Active Lane Change Trajectory Planning of Autonomous Vehicle

为了解决复杂换道环境下自动驾驶车辆的实时轨迹规划问题,将结构化道路空间中的车辆的定位信息转化到曲线坐标系下,通过基于轨迹采样与成本优化相结合的轨迹规划方法生成最优换道轨迹,以满足换道过程中安全性及平稳性要求。为克服该方法被动换道存在的局限性的缺陷,在成本函数中设计了主动换道因子,用于车辆的主动换道轨迹规划,最后将该方法设计的局部轨迹规划系统部署在实车测试平台下进行试验,试验结果表明该方法皆能完成车辆的主、被动换道任务,且能保证采样轨迹的完整,提高自动驾驶车辆局部轨迹规划系统的灵活性。

In order to solve the problem regarding complex lane changing environment of real-time trajectoryplanning for Autonomous vehicles, the vehicle positioning information in structured road space has been transformed intocurvilinear coordinate system. By means of a trajectory planning method combined with trajectory sampling and costoptimization, the optimal trajectory planning method of lane changing trajectory is generated to meet the safety and comfortrequirements in the process of lane changing. To overcome the defects of the limitation of passive lane changing method,active lane changing factor is designed in the cost function, which aims to apply in vehicle active lane changing trajectoryplanning. Then the deployment of the approach is tested in real vehicle test platform. The test results show that the methodcan not only complete the vehicles’ active and passive lane changing task, but also guarantee the completeness of samplingtrajectory so that it can improve the flexibility of local trajectory planning system for autonomous driving vehicle.