Application of multi-objective simulation-optimization techniques to inventory management problems

Abstract

In this paper, we present how a solution framework developed for (a special case of) the multi-objective simulation-optimization problems can be applied to evaluate and optimally select the non-dominated set of inventory policies for two case study problems. Based on the concept of Pareto optimality, the solution framework mainly includes how to evaluate the quality of the selected Pareto set by two types of errors, and how to allocate the simulation replications according to some asymptotic allocation rules. Given a fixed set of inventory policies for both case study problems, the proposed solution method is applied to allocate the simulation replications. Results show that the solution framework is efficient and robust in terms of the total number of simulation replications needed to find the nondominated Pareto set of inventory policies.