Research on Refining the Distributed Supply Chain Procurement Plans Based on CRL

Abstract

In this paper, we propose a method of coordinated reinforcement learning (CRL) to refine the distributed transportation plans generated by a group of agents using double auction market mechanism (DAMM) in the context of supply chain procurement problem. Through analyzing the drawbacks of DAMM due to its strict market rules, we develop a refinement method based on reinforcement learning, namely CRL. Numerical results demonstrate that CRL effectively refines the distributed transportation plans generated by the DAMM.