Port Wind Speed Prediction Model Based on VMD and GRU-MLP

This study proposes a combined prediction model integrating variational mode decomposition (VMD), gated recurrent unit (GRU), and multi-layer perceptron (MLP) to address the low accuracy and poor fitting performance of existing short-term port wind speed prediction models. The model’s performance is evaluated using root mean square error (RMSE) and mean absolute percentage error (MAPE). Experiments were conducted with real-time wind speed data collected from wind speed sensors on large cranes to predict wind speed for the next 10 minutes. Comparative analysis results demonstrate that the VMD-GRU-MLP model achieves higher predictive accuracy for short-term forecasts.