Integrating multi-agent reinforcement learning and 3D A* search for facility layout problem considering connector-assembly

Qiaoyu Zhang () and Yan Lin
Additional contact information
Qiaoyu Zhang: Dalian University of Technology
Yan Lin: Dalian University of Technology

Journal of Intelligent Manufacturing, 2024, vol. 35, issue 7, No 21, 3393-3418

Abstract: Abstract The paper proposes an innovation method of integrating a multi-agent reinforcement learning and 3D A* search, , to tackle the facility layout problem with the goal of minimizing connector-assembly cost, including wires and pipes, in a finite three-dimensional space subjected to complex constraints. The mathematical model is established through the bounding boxes and grids, and an improved 3D A* algorithm is used to evaluate the route quality of connectors. After defining facility states in the ranking way, the position and orientation of each facility are iteratively solved by the multi-agent reinforcement learning, Wolf Policy Hill-Climbing (Wolf-PHC), with three actions of horizontal movement, rotation and swap. Then a pure water manufacturing system has been taken as a case to carry out a series of combination comparative experiments with other approaches (Dijkstra; Ant colony optimization; Genetic algorithm) used in the relevant literatures and the parameter values of these approaches have been sampled by Latin Hypercube. The results show that the stability, efficiency and solving quality of are better.

Keywords: Facility layout problem; Multi-agent reinforcement learning; A* search; Latin hypercube; Assembly design (search for similar items in EconPapers)
Date: 2024
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://link.springer.com/10.1007/s10845-023-02209-x Abstract (text/html)
Access to the full text of the articles in this series is restricted.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:joinma:v:35:y:2024:i:7:d:10.1007_s10845-023-02209-x

Ordering information: This journal article can be ordered from
http://www.springer.com/journal/10845

DOI: 10.1007/s10845-023-02209-x

Access Statistics for this article

Journal of Intelligent Manufacturing is currently edited by Andrew Kusiak

More articles in Journal of Intelligent Manufacturing from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().