Dynamic Resource Allocation Using Multi-Agent Control for Manufacturing Systems
Mingjie Bi, Ilya Kovalenko, Dawn M. Tilbury, Kira Barton
Abstract
The COVID-19 pandemic brings highly dynamic effects to manufacturing environments, such as frequently shifting markets and unexpected disruptions. Such dynamic environments increase the demand for flexible and real-time manufacturing decision-making strategies. One essential problem is dynamic resource allocation to complete production tasks, especially when a resource disruption (e.g. machine breakdown) occurs. Multi-agent frameworks have been proposed to improve the flexibility and responsiveness of manufacturing systems in a distributed decision-making manner. This work introduces a clustering method based on resource agent (RA) capabilities and an RA coordination strategy that enables dynamic resource reallocation when the manufacturing system is subject to resource disruptions.