International Journal of Industrial Engineering & Production Research
Iran University of Science & Technology
Tue, May 26, 2026
[Archive]
1- PhD candidate in Industrial Engineering
2- Associate Professor of Industrial and Systems Engineering, Tarbiat Modares University ,m_kargari@modares.ac.ir
3- Assistant Professor of Industrial and Systems Engineering, Tarbiat Modares University
2- Associate Professor of Industrial and Systems Engineering, Tarbiat Modares University ,
3- Assistant Professor of Industrial and Systems Engineering, Tarbiat Modares University
Abstract: (29 Views)
Determining optimal product mix under uncertain demand and capacity is a critical challenge in the lubricant industry. This study proposes four MILP models: deterministic, robust scenario-based, downside-risk two-stage stochastic, and CVaR two-stage stochastic. All models incorporate real-world constraints including multi-period, multi-product settings, dual-warehouse inventory, backlog/lost-sale shortages, and mandatory production of unprofitable products. Using real data from a major Iranian lubricant refinery, the models improve profit from the company's actual 300 million monetary units (MU) to 535 (deterministic) and 504 million MU (CVaR). Among stochastic models, the robust approach provides the highest worst-case profit and most stable performance. This is the first systematic comparison of these three stochastic approaches in the lubricant industry under identical constraints, demonstrating the value of uncertainty-aware production planning.
Keywords: Product mix, inventory and shortage system, uncertainty, robust model, two-stage stochastic model, lubricant industry
Type of Study: Research |
Subject:
Production Planning & Control
Received: 2025/10/14 | Accepted: 2026/05/17
Received: 2025/10/14 | Accepted: 2026/05/17
Send email to the article author
| Rights and permissions | |
 | This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
