Long-term Wind Power Generation Prediction of Guangdong Province Based on Machine Learning Algorithm

Wind power generation(WPG) plays a crucial role in China's "double carbon" strategic goal, boasting broad development prospects in the country. Guangdong Province, in particular, is endowed with rich WPG resources and great potential. Thus, forecasting long - term WPG in Guangdong is of great significance. This paper commences by introducing diverse machine - learning algorithms in the context of the "two - carbon" target and long - term wind power prediction. Subsequently, historical data on WPG in Guangdong Province from 2009 to 2023, along with its potential influencing factors, are collected. After screening these influencing factors, appropriate independent variables related to energy factors are identified as the feature input for machine - learning algorithms.. According to the fitting effect of different algorithms, those suitable for long-term wind power generation in Guangdong Province are selected, including RR, Enet, Sufit, Stepwise and SVM (linear kernel function). Finally, based on the historical data of energy factors, the paper models and predicts the WPG in the future of Guangdong Province under the five algorithms. The predicted values of WPG in Guangdong for 2030, 2035, 2050, and 2060 are obtained by taking the average of the results from these five algorithms.