A hybrid fuzzy-stochastic multi-criteria ABC inventory classification using possibilistic chance-constrained...

Area of Science:

• Operations Research
• Supply Chain Management
• Decision Science

Background:

• Inventory classification is crucial for effective inventory policy development.
• Traditional methods often struggle with multi-criteria and uncertain environments.
• Decision-making often involves heterogeneous data types, including stochastic and fuzzy information.

Purpose of the Study:

• To propose a mathematical modeling-based approach for inventory classification that addresses heterogeneous decision-making.
• To incorporate stochastic demand and fuzzy cost information into the inventory classification process.
• To develop a hybrid algorithm for solving the proposed inventory classification model.

Main Methods:

• A hybrid algorithm combining chance-constrained programming (for stochastic demand) and possibilistic programming (for fuzzy costs) was developed.
• Mathematical modeling was used to represent the inventory classification problem with uncertainty.
• The approach was applied to a case study involving 51 inventory items.

Main Results:

• The hybrid algorithm successfully classified inventory items into three importance levels: Class A (extremely important), Class B (moderately important), and Class C (relatively unimportant).
• The classification resulted in 22% of items in Class A, 39% in Class B, and 39% in Class C.
• The method also determined minimum inventory levels to manage demand stochasticity.

Conclusions:

• The proposed mathematical modeling and hybrid algorithm provide an effective solution for inventory classification in uncertain environments with heterogeneous data.
• This approach enhances inventory control policy by accurately assigning items to importance classes.
• The study demonstrates the practical applicability and effectiveness of the method through a real-world case study.