Agradecimentos: Os autores gostariam de agradecer aos integrantes do Laboratório de Inteligência Computacional (LInC) e a Universidade Federal de Alfenas (UNIFAL-MG) pelo apoio e disponibilização da infraestrutura para o desenvolvimento da pesquisa, e a Faculdade de Engenharia Mecânica (FEM) da Universidade Estadual de Campinas (Unicamp) por disponibilizarem os dados Ver menos

The Main goal of Two Phase Flow Pattern Identification is to pinpoint the correct pattern in terms of flow in a given section of a pipe. The identification of these patterns has applications in several areas, such as chemical and oil industries. When technologies of identification in real systems are intended to be used, several aspects must be considered, such as data size, memory, processing power. The objective of this paper is to investigate the possibility of embedding a methodology to identify flow patterns in devices that have limited processing power and memory capacity. In this sense, several statistical operations were made aiming to reduce the information needed to identify the patterns. Tests were conducted simulating reductions in data size, resulting in no loss quality at the identification process. From tests performed, it was discovery an efficient identification model and computationally easy to be embedded in a small devices Ver menos