Solar Radiation Forecasting using Multi-Reservoir Computing and Performance Analysis

Accurate and reliable predictions are crucial for optimizing the efficiency of solar energy systems. Reservoir computing, particularly in the forecasting of complex time series, is distinguished by its low computational costs and high accuracy. Multi-reservoir architectures offer a more flexible and robust framework for data modeling. In this study, the multi-reservoir computing (M-RC) method is employed for solar radiation forecasting, and its performance is analyzed using normalized solar radiation data for Sivas, obtained from the National Solar Radiation Database (NSRDB. The method achieved the lowest error rate with a mean square error (MSE) of 0.0471. The results indicate that M-RC is an effective approach, particularly in complex and dynamic systems, enabling more accurate compared to classical methods.