Pronóstico del Consumo de Energía Eléctrica Residencial para la Ciudad de Bogotá

Abstract

Electricity has become over the years an important component in the quality of life of human beings and is a determining factor for the development and economic growth of a society. Likewise, the growing demand for this service has led government agencies to increase the studies related to energy planning, in this way, predicting the consumption of electricity has become a necessary method for decision-making for entities, service providers as well as for public institutions. This research paper proposes three models of energy forecast, by means of regressions and analysis of explanatory variables. These models are based on econometric and statistical studies, thus presenting a useful tool in energy management. The research was developed based on the six socio-economic strata in which the City of Bogotá is organized, using historical data of surface temperature, gross domestic product, per capita income, subscribers, consumption and price of electric power, price of the service of natural gas from 2005 to 2016. The three models were developed in free software R studio, from mathematical language. The first and second model is a multiple linear regression, but the second model has an economic interpretation of coefficients, while the third model is developed in the form of double logarithm economic regression. The three models show a good behavior compared with the measured data and comply with the validations proposed in the methodology, analyzing each variable with respect to the socio-economic stratum, presenting as the most relevant result, the multiple linear regression model with economic interpretation. That model has a significant variable, the Gross Domestic Product, a variable that has been closely related to energy consumption. This model accurately represents the trend of energy consumption and complies with all the statistical and theoretical parameters to give an accurate energy prediction.