Forecasting for smart energy: An accurate and efficient negative binomial additive model

Abstract

Smart energy requires accurate and efficient short-term electric load forecasting to enable efficient energy management and active real-time power control. Forecasting accuracy is influenced by the characteristics of electrical load particularly overdispersion, nonlinearity, autocorrelation and seasonal patterns. Although several fundamental forecasting methods have been proposed, accurate and efficient forecasting methods that can consider all electric load characteristics are still needed. Therefore, we propose a novel model for short-term electric load forecasting. The model adopts the negative binomial additive models (NBAM) for handling overdispersion and capturing the nonlinearity of electric load. To address the seasonality, the daily load pattern is classified into high, moderate, and low seasons, and the autocorrelation of load is modeled separately in each season. We also consider the efficiency of forecasting since the NBAM captures the behavior of predictors by smooth functions that are estimated via a scoring algorithm which has low computational demand. The proposed NBAM is applied to real-world data set from Jericho city, and its accuracy and efficiency outperform those of the other models used in this context.