Analysis and Prediction of Electrical Quantities for Energy Management in Public Buildings
Power quality. Forecasting of electrical variables. Neural networks. Energy management system.
This work aims to develop an approach for the analysis and prediction of electrical variables in the facilities of the Legislative Assembly of the State of Rio Grande do Norte (ALRN), through the use of digital power meters and data science techniques. For this purpose, two WEG MMW04 meters were installed, positioned in the Main Low Voltage Switchboard (QGBT) and in the essential loads panel connected to the generator busbar, enabling continuous acquisition of parameters such as voltage, current, power, power factor, and harmonic distortions. Communication is established via the TCP/IP protocol, allowing systematic data collection and export in digital format. The methodology includes data preprocessing, exploratory analysis of power quality, disturbance identification, and subsequent application of predictive models of an exploratory nature, encompassing both statistical and machine learning methods. The development of a computational prototype will enable the integration of data collection, analysis, and prediction, along with automated report generation and alert issuance to support decision-making. The results are expected to provide a basis for energy management and predictive maintenance strategies at ALRN, contributing to greater reliability, operational efficiency, and sustainability in the use of electrical energy.