Sistema de detección y diagnóstico de fallas de un proceso térmico mediante inteligencia artificial

Abstract

In this document the development of a fault detection and diagnosis system of the thermal control process plant T5553-BD of the Amatrol brand is presented, using Artificial Intelligence. Among the equipment used is a low cost EDIBON data acquisition card, which is connected to the thermal control equipment in order to receive the signals from the temperature and flow sensors of the equipment and take them to a computer, in which LabView was used to export the data in Excel. Subsequently, using the MatLab software and programming language, an interface was designed to detect and diagnose faults using a Neural Network. The result obtained is a graphical user interface that informs the operator of the faults detected in one or more of the devices of the thermal control plant, through the data obtained during a laboratory practice, which are compared with the process test. in normal operating conditions.