A Short-Term Wind Power Prediction Method Based on Hyperparameter Optimization Using Snow Ablation Optimizer

With the widespread adoption of renewable energy, especially wind energy, wind power forecasting has become increasingly important in power system scheduling and optimization. Accurate wind power forecasting is crucial for ensuring the safe and stable operation of the system. However, forecasting errors are common, and data quality directly affects the accuracy of the predictions. Data cleaning plays a vital role in improving data quality, and this process is crucial in forecasting. To address this, this paper proposes a time-series denoising and feature selection method based on Variational Mode Decomposition (VMD) and Random Forest (RF), which effectively extracts features that reflect the relationship between meteorological factors and wind power output. Additionally, artificial neural networks are increasingly used in wind power forecasting due to their strong nonlinear modeling capabilities, making them one of the key technologies for improving forecast accuracy. However, the performance of artificial neural networks heavily depends on the selection of hyperparameters. Proper adjustment and optimization of hyperparameters are essential for achieving accurate predictions. In this paper, we propose a shortterm wind power combined forecasting method based on the Glacier Volcano Snowmelt Algorithm (GVSAO), and construct a combined forecasting model based on CNN-BiGRU-MSA, which achieves automatic hyperparameter optimization. GVSAO is an improved optimization algorithm based on the Snowmelt Algorithm (SAO), which can more effectively adjust the hyperparameters of the forecasting model, enhancing its accuracy. Finally, we validate the effectiveness of the proposed method in improving forecasting accuracy using data from a wind farm in Henan, demonstrating its practical application potential in engineering.