Meta-heuristic Approaches for Multi-objective Simulation-based Optimization in Supply Chain Inventory Management - Archive ouverte HAL Access content directly
Book Sections Year : 2010

Meta-heuristic Approaches for Multi-objective Simulation-based Optimization in Supply Chain Inventory Management

(1) , (2) , (2)
1
2

Abstract

A supply chain (SC) is a complex network of facilities with dissimilar and conflicting objectives. Discrete-event simulation is often used to model and capture the dynamic interactions in SCs and to provide performance indicators. However, a simulator by itself is not an optimizer. Optimization algorithms can be coupled with a simulation module in order to find the most suitable SC policies. Nevertheless, because the simulation of an SC can take a considerable amount of time, the optimization tool must be well chosen. This chapter considers the hybridization of evolutionary algorithms, well known for their multi-objective capabilities, with an SC simulation module in order to determine the inventory policy (order-point or order-level) of a single product SC, taking into account two conflicting objectives: maximizing customer service level and minimizing total inventory cost. Four algorithms (SPEA-II, SPEA-IIb, MOPSO and NSGA-II) are evaluated on five different SC configurations to determine which algorithm gives the best results and makes the best use of the simulator. The results indicate that SPEA-2 favours a rapid convergence and that modifying its crossover or its archive truncation rule (variant SPEA-IIb) may improve the results even further.

Dates and versions

hal-02497360 , version 1 (03-03-2020)

Identifiers

Cite

David Sanchez, Lionel Amodeo, Christian Prins. Meta-heuristic Approaches for Multi-objective Simulation-based Optimization in Supply Chain Inventory Management. Artificial Intelligence Techniques for Networked Manufacturing Enterprises Management, pp.249-269, 2010, Springer Series in Advanced Manufacturing, 978-1-84996-119-6. ⟨10.1007/978-1-84996-119-6_9⟩. ⟨hal-02497360⟩

Collections

CNRS UTT LOSI
10 View
0 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More