基于场景驱动与分层规划的自主代客泊车系统研究

定义了自主代客泊车系统设计与测试验证的功能场景，并基于场景定义建立了环境模型。自主代客泊车系统的功能实现主要包含路径规划层与行为层。基于V2I通信系统服务与Reeds-Shepp Car曲线生成了自主代客泊车系统的全局参考路径，采用三次样条拟合曲线进行了路径平滑，利用车辆非线性约束模型、预瞄控制理论与动态面控制技术设计了用于轨迹跟踪的横向控制器。最后，在PreScan-MATLAB联合仿真环境下实现对自主代客泊车系统的功能集成与有效性验证。

Research on AVPS Based on Scenario Driven and Hierarchical Planning

Functional scenario of design and testing of the AVPS was initially defined, and the environmental models were set up based on the defined scenario. The functions of the AVPS mainly included the path planning layers and the behavior layers. Global reference paths of the AVPS were generated based on the V2I communication service and the Reeds-Shepp Car curves. The cubic spline fitting curves were used for path smoothing. The nonlinear constraint vehicle model, the preview control theory and dynamic surface control were used for design lateral controllers to achieve trajectory tracking. Finally, the function integration and validity verification of the AVPS were verified in the PreScan-MATLAB co-simulation environments.