A New Hybrid Multi-criteria ABC Inventory Classification Model Based on Differential Evolution and Topsis

Abstract

During the last decades, many companies have taken seriously the task of managing the inventory efficiently because of the surplus of stock and the need to make more profits for their financial and logistical well-being. For this purpose, the ABC classification is one of the most frequently analysis used in production and inventory management domains, in order to classify a set of items in three predefined classes A, B and C, where each class follows a specific management and control policies. In this paper, we present a new hybrid approach for the ABC multi-criteria inventory classification (MCIC) problem using the evolutionary algorithm namely the Differential Evolution (DE) with the multi-criteria decision making method (MCDM), called Topsis. This hybrid approach is modeled by using DE, the parameters of which (criteria weights) are optimized and tuned by using a Topsis method. To evaluate objectively the performance of our proposed model, an estimation function based on the inventory cost and the fill rate service level is used, and also represents the objective function of our approach DE-Topsis, which consists of minimizing the inventory cost. The aim of our proposed approach is to exploit the robustness and usefulness of both DE and Topsis methods, to reduce the inventory cost, to provide acceptable performance and to comply with the constraints of the ABC MCIC problem. A comparative study is conducted to compare our proposed hybrid approach with other ABC classification models of the literature by using a widely used data set. We have established that the proposed model enables more accurate classification of inventory items and better inventory management cost effectiveness for the ABC multi-criteria inventory classification problem.