Monitoreo de la red eléctrica del Laboratorio de Máquinas de la Facultad de Informática y Electrónica aplicando inteligencia artificial.

Abstract

The present investigation is an artificial intelligence application to monitor a three-phase electric system of electric-machine laboratory at the Faculty of Computer Science and Electronics of the School of Industrial Networking and Control Engineering. With this work the user was able to determine the instantaneous demand and a future forecasting prediction. The inductive method was used to gather information from each of the three-phase system transmission lines, the programming and the design, whereas the experimental method was used to develop the test programs and error correction of the laboratory demand forecasting running. This monitoring system has a Siemens Sentron PAC3100 device-data stored in Excel and obtained in real time from each of the transmission lines can be read. Besides, a LabVIEW program was done previously to store data. Modbus RS 485 communication program was set up to have access data. Of the data gotten from the three-phase system, the type of artificial intelligence to be applied in the MATLAB program was determined. In this work, a neuronal network was applied and the type and the way of neuronal network training were defined. In the program the input and output variables were determined. In the neuronal network training results in Matlab, the probable forecasting of the energetic consumption demand the laboratory may have during classes, practices and exams, was determined using graphics display. We recommend to observe precautions determining the input variables for the neuronal network so that data can be entered data correctly in the programs to be used.