A new strategy to quantify uncertainties of wavelet-GRNN-PSO based solar PV power forecasts using bootstrap confidence intervals

• Published in: 2015 IEEE Power & Energy Society General Meeting pp. 1 - 5
• Main Authors: AlHakeem, Donna, Mandal, Paras, Haque, Ashraf Ul, Yona, Atsushi, Senjyu, Tomonobu, Tzu-Liang Tseng
• Format: Conference Proceeding
• Language: English; Japanese
• Published: IEEE 01.07.2015
• Subjects: Artificial neural networks; Bootstrap; Forecasting; generalized regression neural network; particle swarm optimization; Predictive models; Probabilistic logic; solar PV power forecasting; Uncertainty; wavelet transform; Weather forecasting
• ISSN: 1932-5517
• DOI: 10.1109/PESGM.2015.7286233

Quantification of uncertainties associated with solar photovoltaic (PV) power generation forecasts is essential for optimal management of solar PV farms and their successful integration into the grid. These uncertainties can be appropriately quantified and represented in the form of probabilistic rather than deterministic. This paper introduces bootstrap confidence intervals (CIs) to quantify uncertainty estimation of PV power forecasts obtained from the proposed deterministic hybrid intelligent model that uses an integrated framework of wavelet transform (WT) and a generalized regression neural network (GRNN), which is optimized by population-based stochastic particle swarm optimization (PSO) algorithm. This particular combination of deterministic hybrid intelligent model and bootstrap method for uncertainty estimation has not been applied in the area of solar PV forecasting. Test results demonstrate the high degree of efficiency of the proposed methods over the tested alternatives in multiple seasons including sunny days (SDs), cloudy days (CDs), and rainy days (RDs).