Integrated lot sizing and blending problems under demand uncertainty

• Grant number: 23/09205-6
• Support Opportunities: Scholarships abroad - Research Internship - Doctorate (Direct)
• Start date: January 01, 2024
• End date: December 31, 2024
• Field of knowledge: Engineering - Production Engineering - Operational Research
• Principal Investigator: Silvio Alexandre de Araujo
• Grantee: Maurício Rocha Gonçalves
• Supervisor: Raf Jans
• Host Institution: Instituto de Biociências, Letras e Ciências Exatas (IBILCE). Universidade Estadual Paulista (UNESP). Campus de São José do Rio Preto. São José do Rio Preto , SP, Brazil
• Institution abroad: École des Hautes Études Commerciales (HEC Montréal), Canada
• Associated to the scholarship: 23/02210-4 - Integrated lot sizing and blending problems under demand uncertainty, BP.DD

Abstract

This research project focuses on industrial production planning problems that involve blending components to assemble end products, where the proportion of components used in the final composition can vary. The objective is to yield plans that minimize overall costs of blending components to meet external demand of end products. The plans also need to consider capacity time constraints and satify quality requirements of end products. To address this problem, we propose mathematical programming approaches that integrate lot sizing decisions for component purchasing and end product production over a finite and discrete planning horizon. We tackle the problem under demand uncertainty through a novel two-stage stochastic formulation, drawing upon strategies for stochastic lot sizing problems. To solve the resulting stochastic formulation, we employ the Sample Average Approximation (SAA) scheme and utilize Benders decomposition techniques with commercial solvers to optimize the mixed-integer Linear programming problems resulting from the SAA application. Computational experiments have already been conducted, and as part of this project, we plan to further enhance our proposed solution approaches, conduct additional tests, and explore alternative stochastic formulations. (AU)