Detection of power peaks in residential electricity consumption through data analysis

Abstract

The present work studies the consumption of electrical energy as it is essential in people's daily lives, electricity allows the development of work activities, industrial production, in the functionalities of household appliances and electronic devices; The objective was set: Model residential energy consumption through data analysis, identifying patterns; The data for analysis were obtained from a knowledge base with description of variables, the same that were acquired from devices made with Raspberry PI, and non-invasive current sensors. It has been considered important to use exploitative and descriptive research with data collection techniques: observation and experimental, they managed to determine predictive results. Using the Tableau software, it was possible to set the parameters to accurately understand the behavior of the graph of a dynamic discrete time system, whose output variable is determined by energy consumption, the parameters acquired from Tableau allowed a future estimate of energy residence of a specific residence, finding the model with the best consumption of adaptation to the prediction to the linear regression method, due to the behavior of its curve, it was also possible to quantify the determinants of residential energy consumption in households according to their socioeconomic characteristics, equipment and location In order to know the energy consumption curve, the representation of residential consumption patterns was carried out with the use of Tableau.