The research paper focuses on the use of artificial neural networks (ANN) to model the acoustic behavior of induction motors. The study aims to minimize the discomfort caused by noise emissions from these motors. Two different types of ANNs are proposed to evaluate the acoustic quality of induction motors. The simpler models are more interpretable but less accurate, while the more complex models provide higher accuracy but hide the cause-effect relationship. The study concluded that product unit neural networks achieved the best results for both mean squared error (MSE) and standard error of prediction (SEP).
Publication date: 31 Jan 2024
Project Page: https://doi.org/10.1016/j.apacoust.2020.107332
Paper: https://arxiv.org/pdf/2401.15377
