有无人驾驶后视效应和前车加速度信息的双速度差模型

Two-Velocity Difference Model with Rear View Effect of Unmanned Driving and Acceleration Information of the Preceding Vehicles

摘要: 本文在无人驾驶(Unmanned Driving, UD)环境下研究车辆跟驰模型。在双速度差模型(Two-Velocity Difference, TVD)的基础上，考虑无人驾驶汽车后视效应和前车加速度信息对车辆跟驰行为的影响，提出一种改进的跟驰模型(BL-ATVD)。利用线性稳定性方法得到了BL-ATVD模型的稳定性条件，并分析了稳定区域的扩大情况和占比率；利用约化摄动方法导出了BL-ATVD模型分别在稳定区域和不稳定区域内的Burgers方程、mKdV方程，给出相应的孤立波解、扭结–反扭结波解。数值模拟结果表明，BL-ATVD模型的交通流稳定性好于全速度差(Full Velocity Difference, FVD)模型，且稳定性随后视效应权重的减小或前车加速度敏感系数的增大而增强。因此，BL-ATVD模型适合用于描述无人驾驶车辆的跟驰特性。

Abstract: In this paper, the car-following model in the unmanned driving (UD) environment is studied. Based on the two-velocity difference model (TVD), an improved car-following model (BL-ATVD) is proposed by considering the influence of the rear view effect of the driverless cars and the acceleration in-formation of the preceding ones on the car-following behavior. The stability conditions of the BL-ATVD model are obtained by using the linear stability method, and the expansion and propor-tion of the stable region are analyzed. Burgers equation, mKdV equation, and the corresponding solitary wave solutions, kink-anti-kink wave solutions are given by using the reduced perturbation method. Numerical simulation results show that the BL-ATVD model has better traffic flow stability than the full speed difference model (FVD), and the stability is strengthened by the reduction of the apparent effect weight or the acceleration sensitivity coefficient of the preceding vehicles. Therefore, the BL-ATVD model is suitable for describing the car-following characteristics of unmanned vehi-cles.