Cost and performance comparison of holistic solution approaches for complex supply chains on a novel linked problem benchmark.

Ogunsemi, Akinola; McCall, John; Zavoianu, Ciprian; Christie, Lee A.

Abstract

Modern supply chains are complex structures of interacting units exchanging goods and services. Business decisions made by individual units in the supply chain have knock-on effects on decisions made by successor units in the chain. Linked Optimisation Problems are an abstraction of real-world supply chains and are defined as a directed network where each node is a formally defined optimisation problem, and each link indicates dependencies. The development of approaches to holistically solve linked optimisation problems is of high significance to decarbonisation as well as building robust industrial supply chains resilient to economic shock and climate change. This paper develops a novel linked problem benchmark (IWSP-VAP-MTSP) integrating Inventory Warehouse Selection Problem, Vehicle Assignment Problem and Multiple Traveling Salesmen Problem. The linked problem represents tactical and operational supply chain decision problems that arise in inventory location and routing. We consider three algorithmic approaches, Sequential, Nondominated Sorting Genetic Algorithm for Linked Problem (NSGALP) and Multi-Criteria Ranking Genetic Algorithm for Linked Problem (MCRGALP). We generated 960 randomised instances of IWSP-VAP-MTSP and statistically compared the performance of the proposed holistic approaches. Results show that MCRGALP outperforms the other two approaches based on the performance metrics used, however, at the expense of greater computational time.

Citation

OGUNSEMI, A., MCCALL, J., ZAVOIANU, C. and CHRISTIE, L.A. 2024. Cost and performance comparison of holistic solution approaches for complex supply chains on a novel linked problem benchmark. In Proceedings of the Genetic and evolutionary computation conference 2024 (GECCO'24), 14-18 July 2024, Melbourne, Australia. New York: Association for Computing Machinery (ACM) [online], pages 1327- 1335. Available from: https://doi.org/10.1145/3638529.3654163

• Presentation Conference Type: Conference Paper (published)
• Conference Name: Proceedings of the Genetic and Evolutionary Computation Conference
• Start Date: Jul 14, 2024
• End Date: Jul 18, 2024
• Acceptance Date: Mar 21, 2024
• Online Publication Date: Jul 14, 2024
• Publication Date: Dec 31, 2024
• Deposit Date: Oct 31, 2024
• Publicly Available Date: Oct 31, 2024
• Publisher: Association for Computing Machinery (ACM)
• Peer Reviewed: Peer Reviewed
• Pages: 1327-1335
• DOI: https://doi.org/10.1145/3638529.3654163
• Keywords: Linked optimisation; Genetic algorithm; Multi-criteria decision-making; Scheduling and planning
• Public URL: https://rgu-repository.worktribe.com/output/2408690