Intelligent estimation of wind farm performance with direct and indirect ‘point’ forecasting approaches integrating several NWP models

Ghali Yakoub, Sathyajith Mathew, J. B. Leal

Energy · 02 November 2022

In one sentence. Genetic-programming AutoML wind power forecasts combining three NWP sources cut errors by 8–22% at a Norwegian wind farm using direct and indirect turbine-level schemes.

Abstract

Reliable wind power forecasting is essential for trading wind energy in the electricity market and efficiently integrating wind power into the power grids. We propose short and medium-term wind power forecasting systems targeted to the Nordic electricity market, which integrates inputs from three numerical weather prediction (NWP) sources. A point forecasting scheme is adopted, which forecasts the power at individual turbine. Both direct and indirect forecasting approaches are considered and compared. An automated machine-learning pipeline, built and optimized using genetic programming, is implemented for developing the proposed forecasting models. The point forecasts, using different approaches, are then combined to a single forecast using a weighting method based on recent forecast errors. These are then aggregated for the wind farm power estimates. The proposed forecasting schemes are implemented with data from a Norwegian wind farm. We found that in both the direct and indirect forecasting approaches, the forecasting errors could be reduced between 8% to 22%, while inputs from several NWP sources are used together. The wind downscaling model, which is used in the indirect forecasting approach, could significantly contribute to the model accuracy. The performance of both the direct and indirect forecasting schemes are comparable for the studied wind farm.

Labels

Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.