Solar Power Forecasting Based on Numerical Weather Prediction and Physical Model Chain for Day-ahead Power System Dispatching

Accurate solar power forecasting has a decisive effect on the formulation of day-ahead power system dispatch strategies. At present, there is every confidence that paring numerical weather prediction with a physical model chain is the state-of-the-art solar forecasting method suitable for grid integration. Leveraging this two-stage solar power forecasting framework, the total cost benefits of day-ahead power system dispatch have been quantified based on a modified IEEE 24-bus system. The numerical experiment results reveal that the two-stage forecasting method saves 2.1 million dollars (1.33%) per year, as compared to the benchmark.