Diseño e implementación de una red neuronal artificial (RNA) para detección y diagnóstico de fallas de un proceso de control de presión

Abstract

This project has the purpose of designing and implementing an artificial neural network (RNA) that allows the detection of faults in a pressure control process. The neural network is trained from data on the operation of the process under normal operating conditions, from a remote monitoring system. The detected faults generate an alarm in the monitoring system, in order to prevent damage to the systems and help the personnel in charge to diagnose and make the appropriate decisions regarding the process. Currently in the Technological Headquarters of the Francisco José de Caldas District University, the Test Bench "T5555-AAU Pressure Process Control Learning System" does not have this system, so it will be used as a prototype for the implementation of the RNA also in order to apply the fault detection system because it is applicable to the industry in general in order to train students in pressure control systems.