Abstract: Vehicle lane change is a common traffic scene, this paper obtains the trajectory of autonomous vehicles based on five polynomials, and uses a fuzzy controller to dynamically adjust the weight coefficient of the evaluation function, so as to select the optimal trajectory according to the actual conditions. The simulation results show that compare with the traditional fixed weight coefficient trajectory planning method, under the working conditions with a large collision risk, the trajectory planning based on fuzzy change can be reasonably adjuste to adjust the weight vertical safety, planning a reasonable change trajectory with higher efficiency and smaller collision risks to adapt to the current conditions.

Key words: Autonomous vehicle; Trajectory planning; Fuzzy control; Weighting coefficien

摘要： 车辆换道为一种常见交通场景，基于五次多项式可得到自动驾驶车辆换道轨迹簇，文 章采用模糊控制器动态调整评价函数的权重系数，从而根据实际工况选取最优换道轨迹。仿 真结果表明，与传统固定权重系数的换道轨迹规划方法相比，在碰撞风险较大的工况下，基 于模糊变权重的换道轨迹规划可合理调整权重值，优先考虑换道效率，提高纵向安全性，规 划出适应当前工况的换道效率更高、碰撞风险更小的合理换道轨迹。

关键词: 自动驾驶车辆；轨迹规划；模糊控制；权重系数