Deep Reinforcement Learning for Stochastic Dynamic Vehicle Routing Problem

Taukekhan Mustakhov, Yernar Akhmetbek, Aigerim Bogyrbayeva

2023

DOI 10.1109/icecco58239.2023.10147154

Data confidence 70 medium

How this score is calculated

It is built from checkable properties of the record itself: whether the article has a DOI, how many independent sources confirmed its data, whether the required fields are filled in, and how recent the latest confirmation is. Disagreement between sources and a missing identifier lower it. A value of 80 or above means the record was published without further checking; 60 to 79 means it is published at reduced confidence, with a manual check advisable. A record verified by an editor by hand is never pushed below 90 by a recalculation. This scores our data about the article, not the article itself.

Abstract

In this study, we propose solving the Stochastic Dynamic Vehicle Routing Problem with Deep Reinforcement Learning. We propose a fully attention-based model consisting of a dynamic encoder and decoder to learn the routing policy of a single vehicle, where the customer orders appear over time. To train the model, we utilize a reinforcement learning setting, in which a reward signal guides an agent to select nodes to be visited. The computational analyses demonstrate the suggested strategy's superiority over the chosen benchmark algorithms.