Implementación de mantenimiento predictivo para una bomba centrífuga utilizando machine learning

Abstract

This document presents an application of machine learning as predictive maintenance in a simulated centrifugal pump, it is a strategy of how to use this branch of artificial intelligence as a maintenance technique. By means of discrete simulations data are extracted from dynamic variables such as torque, pressure and flow from a centrifugal pump in a common water supply system, data is taken from the pump at two speed ranges in the pump rotor