Forecasting China's wind energy generation using a novel all-information seasonal grey model

In the clean energy sector, wind power forecasting plays a crucial role in energy planning and system scheduling. A novel grey forecasting framework (AMSGM(1,1)) is developed in this research, incorporating dynamic seasonal adjustments and fractional-order operators to analyze China's quarterly wind power generation. This model transforms one-dimensional time series into a two-dimensional data structure. It introduces fractional-order generation operators and all-information dynamic seasonal control units. Model parameters are calibrated through a swarm intelligence-based optimization method. Through these approaches, the model can effectively capture the seasonal fluctuations and complex trends of the data. The model continuously updates its all-information control unit based on the identification of seasonal patterns, enabling the integration of seasonal characteristics and trends while eliminating the positive and negative impacts of random disturbances. Empirical analysis shows that the AMSGM(1,1) model has significant advantages in prediction accuracy compared with other competing models. During the performance testing phase, the model demonstrated significantly better performance and smaller errors compared to other models. The proposed model was applied to project China's wind power generation from 2025 to 2030, revealing a nonlinear growth trend with seasonal fluctuations. The quarterly wind power generation data exhibit pronounced fluctuations, with an overall trend of non-linear growth. Based on the prediction results, suggestions are put forward for the development of wind power in China from multiple aspects. This study provides critical references for China's energy planning and carbon emission policy-making. Meanwhile, it also advances the application and development of grey prediction models in the electricity field.