Dynamic Grouping of Wind Power Clusters Based on Ultra-Short- Term Wind Power Forecasting with NWP

Research on wind power prediction control can help to solve the problem of insufficient absorptive ability of power grid, and have great significance in the power system safety and stable operation. Therefore, the paper puts forward a method of multiple time scale dynamic wind cluster partition based on BP neural network. The method takes into account numerical weather prediction to forecast the wind ultra-short-term power, according to the trend of the wind power prediction power of wind farm output power dynamic category, in order to classify wind farm output power, and establish wind farm output power cluster priorities. The validity and rationality of the proposed method are verified by simulating and analyzing cases with domestic wind power clusters. The results show that the numerical weather prediction can effectively improve the accuracy of wind power prediction, and the absolute mean error and root mean square error are improved by 5% and 4% respectively. And the dynamic clustering of wind farms lays the foundation for the regulation of wind power.