CENTRALIZED VS. DECENTRALIZED PLANNING OF DYNAMIC FLOWS ON CAPACITATED TRANSPORT NETWORKS: A COMPARATIVE COMPUTATIONAL STUDY

Authors

  • Mojahid F. Saeed Osman American University of Sharjah, Department of Industrial Engineering, Sharjah, United Arab Emirates

DOI:

https://doi.org/10.23055/ijietap.2022.29.4.6835

Keywords:

industrial engineering, operations research, supply chain management, Network Flows

Abstract

This paper studies the problem of planning dynamic flows on capacitated transport networks. The main contribution of this study is to compare planning approaches that adopt similar routing and scheduling strategies with different levels of decentralization and investigate the impact of this difference on their computational efficiency. We compare the computational performance of two recent hybrid approaches, one centralized and the other decentralized, for routing and scheduling multi-commodity flows on capacitated transport networks using five performance and computational measures for twelve case problems. The numerical results of the performance and computational metrics for each case problem are recorded and analyzed. The main findings of this study are that solving each network flow case problem in the desterilized environment provides the same minimum flow times for all objects from their source nodes to predetermined terminal nodes with a significant decrease in the model size parameters and memory usage required by the solver, and increase in the number of sub-problems solved by the solver. All case problems reveal the same computational trade-offs, and that raises the credibility of the study findings.

Published

2022-08-15

How to Cite

Saeed Osman, M. F. (2022). CENTRALIZED VS. DECENTRALIZED PLANNING OF DYNAMIC FLOWS ON CAPACITATED TRANSPORT NETWORKS: A COMPARATIVE COMPUTATIONAL STUDY. International Journal of Industrial Engineering: Theory, Applications and Practice, 29(4). https://doi.org/10.23055/ijietap.2022.29.4.6835

Issue

Section

Operations Research