Emprego de inteligência artificial na identificação de parâmetros eletromagnéticos de motores elétricos

Abstract

This work investigates the development of an artificial intelligence application to identify electromagnetic parameters in electric motors. Using the FEMM software, a dataset of magnetic flux curves by current for various ferromagnetic materials was created, serving as the basis for training neural networks. These networks were able to determine the equation of these curves from a minimal number of points, regardless of the material type, implying possible improvements in the efficiency and performance of electric motors.