Short‐term forecasting of industrial electricity consumption in Brazil

Abstract

This paper presents short-term forecasting methods applied to electricity consumption in Brazil. The focus is on comparing the results obtained after using two distinct approaches: dynamic non-linear models and econometric models. The ®rst method, that we propose, is based on structural statistical models for multiple time series analysis and forecasting. It involves nonobservable components of locally linear trends for each individual series and a shared multiplicative seasonal component described by dynamic harmonics. The second method, adopted by the electricity power utilities in Brazil, consists of extrapolation of the past data and is based on statistical relations of simple or multiple regression type. To illustrate the proposed methodology, a numerical application is considered with real data. The data represents the monthly industrial electricity consumption in Brazil from the three main power utilities: Eletropaulo, Cemig and Light, situated at the major energy-consuming states, Sao Paulo, Rio de Janeiro and Minas Gerais, respectively, in the Brazilian Southeast region. The chosen time period, January 1990 to September 1994, corresponds to an economically unstable period just before the beginning of the Brazilian Privatization Program. Implementation of the algorithms considered in this work was made via the statistical software S-PLUS.

Neural networks for short-term load forecasting: a review and evaluation

“…They may be classified as time series (univariate) models, in which the load is modeled as a function of its past observed values, and causal models, in which the load is modeled as a function of some exogenous factors, specially weather and social variables. Some models of the first class suggested in recent papers are multiplicative autoregressive models [60], dynamic linear [27] or nonlinear [80] models, threshold autoregressive models [43], and methods based on Kalman filtering [46], [69], [81]. Some of the second class are Box and Jenkins transfer functions [34], [47], ARMAX models [91], [92], optimization techniques [94], nonparametric regression [11], structural models [36], and curve-fitting procedures [85].…”

Time series prediction using PSO-optimized neural network and hybrid feature selection algorithm for IEEE load data

“…Load forecasting as an active subject in the field of time series prediction is a vital and fundamental factor for successful operation of an energy system [43][44][45][46]. In order to operate the electrical system efficiently, the load should be correctly predicted even for individual households [47,48].…”

Electricity consumption probability density forecasting method based on LASSO-Quantile Regression Neural Network

“…where C is the constraint parameter corresponding to λ, which is negative correlation with punishment intensity. The optimal solution can be obtained by Eq 3and Eq (4). Because the penalty function is a first order function with absolute value, it cannot be solved by the calculus method.…”

“…Most commonly used methods for electricity consumption forecasting include regression models [3], time series models [4], the fuzzy theory [5], neural networks [6], Bayesian networks [7], hybrid method [8] and so on. Regression analysis and time series models are the most acclaimed modeling techniques in electricity consumption forecasting [9].…”