A comparative study of lot sizing problem in MTO supply chain based on simulation optimization approach

Abstract

This paper deals with the lot sizing problem in Make To Order (MTO) supply chain solved by simulation optimization (SO) approach. A comprehensive case study which presented in detail involves a multi-stage, multi-product, multi-location, multi-resource with setup, capacity constraints and stochastic demand. The case study objective is to determine a fixed optimal lot size for each manufactured product type that will ensure Order Mean Flow Time (OMFT) target value for each finished product type. For a comparative purpose, three SO methods are used: (1) A metamodel method based on Response Surface Methodology (RSM), (2) A metamodel method based on Factorial Design, and (3) Global Search method based simultaneously on Tabu Search, Neural Networks, and Scatter Search. Methods (1) and (2) have been treated in the literature with a manual based approach, while the method (3) using the OptQuest Software will be more addressed in this paper,. The results of the comparative study show that the first method which is based on RSM is very effective while the third one is very practical and gives a good satisfactory solution.