基于高精地图的物流配送车路径规划与跟踪控制

为了解决园区等场景下无人车多途经点配送问题，提出了一种基于矢量化高精地图的车道级全局路径规划、生成和跟踪控制方法。考虑配送车往返途经点顺序对行驶路径总长度的影响，基于高精地图采用A*算法计算各配送点间的最优路径，在此基础上，利用动态规划算法求解经过多个配送点的全局最优路径。应用贝塞尔曲线对规划的路径进行平滑，并根据道路曲率设定不同路径处的参考行驶速度，进而生成车道级的可用于跟踪的目标轨迹。利用车辆二自由度模型设计模型预测控制器进行轨迹跟踪，实现低速物流配送车的自主控制。在 CarSim/Prescan/Simulink联合仿真平台和实车平台上对提出的规划控制方法进行了试验。结果表明，相比传统的依据最近配送点策略确定的路径，所提出的方法搜索出的路径长度平均缩短了 6.15%。所设计的轨迹跟踪控制器能确保配送试验车与目标轨迹的横向偏差在 0.25 m 以内，航向角偏差在5°以内。

High-Definition Map-Based Route Planning and Tracking Control for Logistics Distribution Vehicles

To address the challenge of multi-waypoint delivery by unmanned vehicles in scenarios such as industrial parks, the paper proposes a lane-level global path planning, generation and tracking control method based on vectorized high-definition maps. Considering the influence of delivery point sequencing on the total path length, the A* algorithm is used based on high-definition maps to calculate the optimal path between each delivery point. And then, the dynamic programming algorithm is employed to obtain the globally optimal path that passes through multiple delivery points. The planned path is smoothed using Bezier curves,and the reference driving speed is set according to the road curvature at different points along the path,thereby creating a lane-level target trajectory suitable for tracking. Subsequently, a model predictive controller based on a two-degree-of-freedom vehicle model is designed for trajectory tracking to achieve autonomous control of low-speed logistics delivery vehicles. The proposed planning and control method is tested on a joint simulation platform of CarSim,Prescan and Simulink, as well as on a real vehicle platform.The results show, compared with the traditional path determined based on the nearest delivery point strategy,that the path length determined by the proposed method is reduced by an average of 6.15%. The developed trajectory tracking controller ensures that the lateral deviation of the experimental delivery vehicle from the target trajectory is maintained within 0.25 m and the yaw angle deviation is kept within 5°.