基于安全域与风险场的网联双车协同换道轨迹规划

Cooperative Lane Change Trajectory Planning for Connected Dual Vehicles Based on Safety Domains and Risk Fields

摘要:

为解决智能网联汽车在高速路换道场景下，因两车预设意图时空重叠而引发的决策冲突与轨迹规划耦合问题，提出一种考虑3次筛选的协同轨迹规划方法。首先，基于五次多项式生成两车轨迹组合解空间；其次，利用双曲正切函数建立椭圆安全边界模型以完成轨迹组合初筛；然后，融合风险场理论与横纵向冲击度构建总风险函数以进行二次轨迹筛选；最后，结合车辆动力学约束与多目标换道成本函数，对轨迹组合整体进行评价，筛选出最优轨迹组合。双车道仿真结果表明，在正常与紧急换道场景下，所提算法较传统风险评估方法换道成本均有所降低，能够有效协调安全性与舒适性，为网联车协同决策算法的开发提供参考。

Abstract:

To address the decision conflict and trajectory planning coupling issues arising from the spatio-temporal overlap of predefined intentions between vehicles during lane-changing scenarios on highways for intelligent connected vehicles （ICVs）， a collaborative trajectory planning method incorporating a three-stage screening process was proposed. Firstly， a quintic polynomial was employed to generate the combined trajectory solution space for both vehicles.Secondly， a hyperbolic tangent function was utilized to establish an elliptical safety boundary model for the initial trajectory screening. Thirdly， integrating the risk field theory with lateral and longitudinal impact intensity， a total risk function was constructed for the secondary theory screening. Finally， by combining vehicle dynamics constraints with a multi-objective lane-change cost function， overall trajectory combinations were evaluated to select the optimal combination. Two-vehicle lane-change simulation results demonstrate that under both normal and emergency lane-change scenarios， the proposed algorithm reduces lane-change costs compared to traditional risk assessment methods and effectively balances safety and comfort， which provides a reference for developing cooperative decision-making algorithms for connected vehicles.