More and more studies have aimed to optimize the ground vehicle (GV) and unmanned aerial vehicle (UAV) system in which GVs function as mobile satellites and UAVs are dispatched from GVs for last-mile deliveries. From the two-echelon scheme perspective, GV routes originating at the depot are on one echelon, and UAV routes originating at satellites are on the other echelon. A change in a GV route may affect some UAV routes, which indicates the satellite synchronization. In the past decade, the optimization of vehicle routes in two-echelon networks has attracted increasing attention in the operations research community. We classify routing problems of two-echelon networks based on the modeling mechanism connecting the two echelons. Different formulations for describing connection mechanisms of the two-echelon scheme, especially constraints on capacitated satellites, satellite synchronization, vehicle coupling/decoupling at satellites, etc., are briefly introduced. There are several modeling challenges of optimizing delivery routes for a fleet of GV–UAV combinations that include new connection mechanisms between the two echelons. Some important variants, especially those involving mobile satellite synchronization, and GV–UAV flexibly coupling/decoupling, require new mathematical formulations and algorithms catering to more general and practical situations.

越来越多的研究旨在优化地面车辆 (GV) 和无人机 (UAV) 系统，其中 GV 充当移动卫星，UAV 从 GV 调度用于最后一英里交付。从两梯队方案来看，始发站的GV航线在一个梯队，始发于卫星的无人机航线在另一梯队。GV 路线的变化可能会影响一些无人机路线，这表明卫星同步。在过去的十年中，两梯队网络中车辆路线的优化越来越受到运筹学界的关注。我们根据连接两个梯队的建模机制对两梯队网络的路由问题进行分类。描述两梯队方案的连接机制的不同公式，特别简要介绍了对有容量卫星的限制、卫星同步、卫星上的车辆耦合/解耦等。为一组 GV-UAV 组合优化交付路线存在几个建模挑战，其中包括两个梯队之间的新连接机制。一些重要的变体，尤其是那些涉及移动卫星同步和 GV-UAV 灵活耦合/解耦的变体，需要新的数学公式和算法来满足更普遍和实际的情况。