The lot sizing problem with multiple production plants and stochastic demand: formulations and solution methods

• Grant number: 25/09621-5
• Support Opportunities: Scholarships in Brazil - Doctorate
• Start date: October 01, 2025
• End date: June 30, 2029
• Field of knowledge: Engineering - Production Engineering - Operational Research
• Principal Investigator: Reinaldo Morabito Neto
• Grantee: Alex Paranahyba de Abreu
• Host Institution: Centro de Ciências Exatas e de Tecnologia (CCET). Universidade Federal de São Carlos (UFSCAR). São Carlos , SP, Brazil
• Associated research grant: 22/05803-3 - Cutting, packing, lot-sizing, scheduling, routing and location problems and their integration in industrial and logistics settings, AP.TEM

Abstract

This project addresses the multi-plant capacitated lot sizing problem (MPCLSP) under demand uncertainty using a multi-stage stochastic programming approach. The MPCLSP involves determining production quantities and potential inter-plant product transfers to minimize setup, production, transfer, and inventory holding costs, while respecting capacity constraints. Although most of the existing literature treats the problem deterministically, the inherent uncertainty in production processes underscores the importance of optimization under uncertainty. Given the multi-period planning horizon, multi-stage stochastic programming enables more realistic modeling, leading to higher-quality solutions for decision-makers. However, developing effective formulations and solution methods that can handle large-scale instances remains challenging, requiring the development of sophisticated approaches that exploit specific structural characteristics of the problem. The objective of this research is to develop formulations and both exact and heuristic methods for the MPCLSP under uncertain demand using multi-stage stochastic programming, leveraging effective modeling and computational techniques in this context, particularly decomposition techniques, branch-and-cut algorithms, dynamic programming, integer L-shaped method, among others. Beyond theoretical contributions, the project includes collaboration with one of the world's largest orange juice producers, incorporating real-world data and requirements, as well as the application of the developed methods to a real problem, promoting advances in computational tools for decision-making support.