A study of the beverage production process and solution approaches based on mixed integer linear optimization techniques

• Grant number: 23/08977-5
• Support Opportunities: Regular Research Grants
• Start date: November 01, 2023
• End date: October 31, 2025
• Field of knowledge: Engineering - Production Engineering - Operational Research
• Principal Investigator: Deisemara Ferreira
• Grantee: Deisemara Ferreira
• Host Institution: Centro de Ciências e Tecnologias para a Sustentabilidade (CCTS). Universidade Federal de São Carlos (UFSCAR). Sorocaba , SP, Brazil

Abstract

In view of the recent computational advances and solution methods, mathematical models and solution methods for integrated lotsizing and scheduling problems have become increasingly sophisticated (Wörbelauer et al., 2018). Despite this, beverage production planning problems still require specific models and solution methods, given their peculiarities, such as, multiple stages, synchronization between stages, and sequence-dependent changeover times/costs (Bo et al., 2021).The scientific literature suggests that solution methods based on mathematical models (MIP and matheuristics) is a promising line of research, since using parts of the models in heuristic methods makes it easier to control the more complicated constraints of the problem, such as synchrony between stages.In this research project, we intend to study efficient solution methods for solving the non-alcoholic beverage production problem. The focus will be on developing methods that combine heuristics and mixed integer linear optimization models (MIP and Matheuristics and local search heuristics such as Variable Neighborhood Search (VNS)) and their application in the solution of large instances. In addition, because of the importance and particularities of production processes in the beverage sector and the lack of specific tools to help the decision-maker when choosing a more appropriate optimization method, in this research project, it is intended to establish a review on this class of problems. The objective is to establish the differences and similarities that make up a beverage production planning problem in such a way that it is possible to classify and facilitate the choice of the most suitable model/method for certain or similar beverage production processes. None of the existing literature reviews discussed and analyzed the mathematical programming models from the perspective of their applicability to real-life beverage production processes. (AU)