The article presents a reinforcement learning approach to solve the Stochastic Vehicle Routing Problem with Time Windows (SVRP). The goal is to optimize goods delivery by reducing travel costs. The authors develop a novel SVRP formulation that takes into account uncertain travel costs and demands, as well as specific customer time windows. An attention-based neural network trained through reinforcement learning is used to minimize routing costs. The model outperforms traditional methods and proves robust in diverse environments, making it a potential benchmark for future SVRP studies and industry applications.
Publication date: 16 Feb 2024
Project Page: Not provided
Paper: https://arxiv.org/pdf/2402.09765
