Order Sequencing Problem in a Robotic Mobile Fulfillment System

Abstract

With rapid development of Business-to-Customer (B2C) e-commerce, enormous goods assortment and fluctuated demand are presented in customer orders. Traditional manual warehouses have the demerits of low picking efficiency and high human cost, which misfit B2C e-commerce. To resolve the difficulties, a Robotic Mobile Fulfillment System (RMFS) is introduced and implemented. This paper studies the order sequencing problem in an RMFS with the situation that a rack can be reused among multiple picking stations in one rack schedule. In order to solve the problem, a Q-learning-based differential evolution algorithm is proposed. Compared with the existent algorithms, numerical experiments show that the proposed algorithm makes an evident improvement on order picking efficiency.