Ensemble Techniques for Solar Energy Forecasts
A Chinese research group has sought to understand the relative performance of two weather prediction techniques based on ensemble modeling for solar energy forecasts. The scientists applied the two methods in combination with three classical post-processing methods.
PV Magazine International 6:25 pm on April 18, 2024
A research group compared the performance of analog ensemble (AnEn) and dynamical ensemble (DyEn) methods for solar energy forecasting using data from seven US locations. They found that AnEn had better raw calibration but introduced noises after post-processing, while DyEn showed better model consistency. The study, published in Solar Energy Advances, concluded that quantile regression emerged as the most suitable calibration method for both methods.
- Researchers compared AnEn and DyEn methods for solar energy forecasting
- AnEn had better raw calibration but introduced noises after post-processing
- DyEn showed better model consistency
- Quantile regression emerged as the most suitable calibration method for both methods
- Findings published in Solar Energy Advances
https://www.pv-magazine.com/2024/04/18/ensemble-techniques-for-solar-energy-forecasts/