Artificial neural networks applied to heat exchangers problems: a review

Autores

  • Thomas S. Pereira
  • Thiago A. Alves
  • Hugo V. Siqueira
  • Yara de Souza Tadano

Palavras-chave:

Artificial Neural Networks, Heat Exchanger, Comprehensive Review

Resumo

Artificial Neural Networks (ANN) are computational algorithms inspired by the nervous system of
animals. They are powerful tools to predict patterns and behaviors for problems in different fields of study. Heat
exchangers are devices created to improve heat transfer that are hard to model by conventional methods but are
highly important in many applications. ANN have been used to model heat exchange problems, thus, helping to
predict and analyze patterns and behaviors not easily predicted by traditional methods. This review discusses the
application of ANN to heat exchanger problems, evaluating the improvement in the field over the last decades. To
achieve this goal, the number of publications was first analyzed, and the studies were divided into groups according
to the research goals. It was also analyzed the number of publications each year. It was considered the keywords:
"artificial neural network" and "heat exchanger" in the Science Direct platform. One hundred nine papers were
found, and around 50% were published in the last five years. 68% of the articles focused on the evaluation of the
ANN rather than utilizing it to optimize heat exchangers, showing that the method is still in development even if
it has become more important in the last decade.

Downloads

Publicado

2024-05-30

Edição

Seção

Artigos