论文标题
改进的基于分解的启发式启发式卡车排
An improved decomposition-based heuristic for truck platooning
论文作者
论文摘要
储层计算是预测湍流的有力工具，其简单的架构具有处理大型系统的计算效率。然而，其实现通常需要完整的状态向量测量和系统非线性知识。我们使用非线性投影函数将系统测量扩展到高维空间，然后将其输入到储层中以获得预测。我们展示了这种储层计算网络在时空混沌系统上的应用，该系统模拟了湍流的若干特征。我们表明，使用径向基函数作为非线性投影器，即使只有部分观测并且不知道控制方程，也能稳健地捕捉复杂的系统非线性。最后，我们表明，当测量稀疏、不完整且带有噪声，甚至控制方程变得不准确时，我们的网络仍然可以产生相当准确的预测，从而为实际湍流系统的无模型预测铺平了道路。
Truck platooning is a promising transportation mode in which several trucks drive together and thus save fuel consumption by suffering less air resistance. In this paper, we consider a truck platooning system for which we jointly optimize the truck routes and schedules from the perspective of a central platform. We improve an existing decomposition-based heuristic by Luo and Larson (2022), which iteratively solves a routing and scheduling problem, with a cost modification step after each scheduling run. We propose different formulations for the routing and the scheduling problem and embed these into Luo and Larson's framework, and we examine ways to improve their iterative process. In addition, we propose another scheduling heuristic to deal with large instances. The computational results show that our procedure achieves better performance than the existing one under certain realistic settings.
