论文标题
在混合流量中对自动驾驶汽车的最佳合并控制：最佳指数政策
Optimal Merging Control of an Autonomous Vehicle in Mixed Traffic: an Optimal Index Policy
论文作者
论文摘要
我们考虑了仅由人类驱动的车辆（HDV）组成的单个自动驾驶汽车（AV）的问题，目的是最大程度地减少参与合并过程中涉及的整个车辆的旅行时间和能源消耗。这是通过仅控制AV并确定最佳合并序列和与之相关的最佳AV轨迹来完成的。我们得出了最佳指数策略，该策略规定了AV在HDV组中的合并位置。我们还指定了最佳索引在所有HDV之前或所有HDV之后对应于AV合并的条件，在这种情况下，AV不需要与HDV的相互作用。包括仿真结果以验证最佳指数策略，并证明可以实现最佳合并的情况，而无需对人类驾驶行为有任何明确的假设。
We consider the problem of a single Autonomous Vehicle (AV) merging into traffic consisting only of Human Driven Vehicles (HDVs) with the goal of minimizing both the travel time and energy consumption of the entire group of vehicles involved in the merging process. This is done by controlling only the AV and determining both the optimal merging sequence and the optimal AV trajectory associated with it. We derive an optimal index policy which prescribes the merging position of the AV within the group of HDVs. We also specify conditions under which the optimal index corresponds to the AV merging before all HDVs or after all HDVs, in which case no interaction of the AV with the HDVs is required. Simulation results are included to validate the optimal index policy and demonstrate cases where optimal merging can be achieved without requiring any explicit assumptions regarding human driving behavior.
