PREVISÃO DA DEMANDA RESIDENCIAL DE ELETRICIDADE NO BRASIL A PARTIR DA ANÁLISE DE COMPONENTES PRINCIPAIS
| Main Authors: | Bernalize do Rosário Vila Nova, Hiliene da Costa de Carvalho, Edson da Costa Bortoni |
|---|---|
| Format: | Article Journal |
| Bahasa: | por |
| Terbitan: |
, 2020
|
| Subjects: | |
| Online Access: |
https://zenodo.org/record/4330749 |
Daftar Isi:
- The projection of energy demand has been one of the most used tools in the area of energy planning, assisting in decision making related to energy policies. Therefore, this study had as main objective to apply the Principal Component Analysis (ACP) technique “Principal Component Analysis” (PCA), to evaluate the factors that most influence the residential demand for electricity and from the determined main components, perform the forecast for residential electricity demand for the year 2030. For this purpose, we used data from variables that influence residential energy demand on the website of the Brazilian Institute of Geography and Statistics (IBGE). The following variables were considered: Number of inhabitants, Income per capita, GDP per capita, Exchange Rate, Unemployment Rate, and Fertility Rate. Residential demand is also influenced by factors such as the price of electricity, the price of alternative energy, the rate of urbanization, the rate of access to electricity, however, these data are not available at year intervals that were considered for analysis. Principal component analysis (CP) was started by standardizing all variables in order to ensure that all of them have the same weight in the analysis, to have a zero mean and variance equal to one. After standardizing the variables, he performed the main component analysis to identify the variables with the greatest influence on residential energy consumption. Having defined the main components, it was possible to forecast demand for the year 2030 through multiple linear regression analysis using the least squares method. As a result, it was found that the proportion of contribution from the accumulated variation of the first three main components is 92.9%, that is, it reflects most of the information on demand for residential consumption. Only the main component one (CP1) already presents greater variability of the data (about 62.1%). Thus, in order to forecast residential electricity demand in Brazil, only the first two main components were considered, which are represented by income per capita, GDP per capita, exchange rate and number of inhabitants, these factors contribute with more explanatory variables. Another four main components were discarded, since they assume a very small proportion of the data variability (close to zero), therefore, it is estimated that they do not have as much influence. Thus, residential demand is a function of socioeconomic development and the Brazilian population growth. Using the least squares method, it was possible to forecast residential electricity consumption from 2019 until the year 2030. In this time interval there was a significant increase in residential consumption from 141258.7 MWh to 177076.5 MWh. This increase can be explained by population growth, socioeconomic development, the process of universal electric power, climatic conditions and greater availability of income for the acquisition of electrical equipment, especially air conditioning equipment. On the other hand, it was found that the growth in residential consumption has been decreasing in recent years and the forecast is that in the coming years this growth will average 2% per year.