Comparação da aplicação de meta-heurísticas para o problema de otimização de mistura de grãos de café

Abstract

In this study, the application of optimization techniques to solve the challenge of coffee bean blending is explored, inspired by the complexity faced by the coffee market in evaluating various combinations. This problem is crucial for the coffee industry, as time optimization and quality standardization can reduce costs and increase profits. Additionally, this research also implements the possibility of an application capable of achieving specific sensory profiles in the resulting blends. To address this issue, we use the meta-heuristics Genetic Algorithm and Particle Swarm Optimization, comparing their performances with sample data provided by the company Open Science. After calibrating the parameters of the heuristics, we evaluate their performance in terms of execution time and solution quality. The results revealed that the Genetic Algorithm stood out, offering superior results, albeit with longer processing time compared to Particle Swarm Optimization.