AUTHOR=Hu Enze , He Jianjun , Shen Shuai 

TITLE=A dynamic integrated scheduling method based on hierarchical planning for heterogeneous AGV fleets in warehouses

JOURNAL=Frontiers in Neurorobotics

VOLUME=Volume 16 - 2022

YEAR=2023

URL=https://www.frontiersin.org/journals/neurorobotics/articles/10.3389/fnbot.2022.1053067

DOI=10.3389/fnbot.2022.1053067

ISSN=1662-5218

ABSTRACT=In modern industrial warehouses, heterogeneous and flexible fleets of automated guided vehicles (AGVs) are widely used to improve transport efficiency. However, as the AGV scale increases and their limit of battery capacity, the complexity of dynamic scheduling also increases dramatically. The problem is to assign tasks and determine detailed paths to AGVs to keep the multi-AGVs system running efficiently and sustainedly. In this context, a mixed-integer linear programming (MILP) model is formulated. A hierarchical planning method is used, which decomposes the integrated problem into two levels: the upper-level task assignment problem and the lower-level path planning problem. A hybrid discrete state transition algorithm (HDSTA) based on an elite solution set and tabu list method is proposed to solve dynamic scheduling problem with the aim to minimize the sum of costs of requests and tardiness costs of conflicts for the overall system. The efficacy of our method is investigated from computational experiments using real-world data.