Uma abordagem em máquinas de estados para o teste de controladores para processos avícolas

Abstract

This work presents the development of a state machine model for testing controllers aimed at poultry processes. Brazil, as one of the largest global producers of chicken meat, uses control systems to optimize the environment in poultry houses and improve the process of animal protein production. However, many commercial controllers have limitations, being based on reactive control without refined control zones, which can cause discomfort to the birds and affect production results. Furthermore, the development process of these controllers is typically manually programmed, making the development process less efficient and more prone to errors. The proposal of this work aims at the reengineering of models previously developed in the literature, subsequently applying an automatic synthesis approach to a new controller. The study involves three main stages: the reproduction of an existing poultry control model, the development of a new controller using automatic synthesis techniques, and practical implementation and comparison with commercial controllers. Validation will be performed through a dashboard that captures logs, allowing the comparison between the commercial controller and one obtained through formal modeling.