Estimation Method of Wind Power Credible Capacity With Large‐Scale Wind Power System in Extreme Cold Surge Weather

This is the authors' abstract. We don't add key points for this paper.

The violent fluctuation of wind power under extreme cold surge weather brings great challenges to power grid dispatching. Aiming at the limited scheduling reference caused by the large error of wind power prediction in extreme weather, we proposed a dynamic wind power credible capacity assessment method based on an equivalent reliability Monte Carlo approach (ERMC). First, we proposed the concept and assessment method of dynamic wind power credible capacity, which is derived from the traditional credible capacity. Second, we proposed a wind power prediction method based on XGBoost‐Transformer, considering the wind turbines protection logic under extreme weather. Furthermore, the ERMC is proposed to dynamically assess wind power credible capacity based on the prediction results. Finally, experimental verification is conducted on a reduced 30‐bus system of a power grid in North China. The results show that: (1) Compared with other methods, XGBoost‐Transformer can reduce the mean absolute error by 197 MW under an extreme cold wave. (2) The ERMC achieves a full‐period mean proximity gap of 12.73 MW and a conservative hit rate of 100%, both outperforming the benchmark methods. The research can provide theoretical support for the operation and dispatching of wind power grid‐connected systems under extreme weather.