Forecasting the industrial solar energy consumption using a novel seasonal GM(1,1) model with dynamic seasonal adjustment factors

Abstract

Due to influences of natural and social factors, the data of solar energy consumption generally show the characteristic of seasonal fluctuations. In order to forecast data with seasonal fluctuation, this study proposed a novel seasonal grey model with dynamic seasonal adjustment factors, the factors reflect the proportion of the actual value deviating from the average trend value due to seasonal effects. The proposed model, grey model, seasonal grey model and a seasonal autoregressive integrated moving average model were built based on monthly data of the industrial solar energy consumption in the United States from 2005 to 2017. The results demonstrate that the grey model only show a growth trend of the original data. The seasonal autoregressive integrated moving average model and seasonal grey model can identify and predict the seasonal characteristics of series, but fail to accurately forecast their fluctuation effects. Moreover, the proposed model is able to effectively identify dynamic change process of seasonal adjustment factors and significantly improves prediction accuracy. The predicted results showed that the year-on-year growth rate of monthly consumption of the industrial solar energy in the United States maintained at 40–50% in 2018–2019, and the monthly consumption still had obvious periodic seasonal fluctuation characteristics.