Alsaedi, Yasir

Solar Price Forecasting Based on Various Univariate Time-Series Models: A Case Study from Queensland, Australia

Yasir Alsaedi1,2, Gurudeo Anand Tularam2, and Victor Wong3

1. Department of Mathematics, Umm Al-Qura University, Makkah, Saudi Arabia

2. Environmental Futures Research Institute , Griffith University, Brisbane

3. Department of Accounting, Finance and Economics, Griffith University, Nathan, Queensland

In the coming years, it is anticipated that the generation of power through solar means will play a key role in the electricity markets. Therefore, energy analysts and government organisations alike require guidelines to help them choose the most appropriate forecasting techniques for achieving accurate predictions of solar pricing trends.In this study, three types of univariate models are considered, namely the simple exponential smoothing (SES), Holt-Winters exponential smoothing (HWES) and autoregressive integrated moving average (ARIMA) models. In order to determine the most appropriate model, four different strategies are applied as selection criteria in order to quantify the accuracy of the model predictions, namely the mean squared error (MSE), root-mean-square error (RMSE), mean absolute error (MAE) and mean absolute percentage error (MAPE) strategies. The three models are compared by applying them to the time series of the average solar prices for Queensland, Australia. Data covering August 2012 to March 2018 are extracted from the database of Solar Choice, Australia’s solar power installation brokering and solar quote comparison service. The comparison indicates that the SES model performs better than the HWES model in terms of its predictive power, with a confidence interval of 95%. However, the ARIMA (1, 1, 1) model yields the best results, which leads us to conclude that this sophisticated and robust model outperforms the other simple, albeit flexible, models in relation to the solar market. This study aims to help policy makers and industry marketing strategists select the best forecasting method for the solar market.