A Hybrid Model Based on Grey Wolf Optimizer and Lagrangian Support Vector Regression for European Natural Gas Consumption Forecasting

Natural gas plays an important role in industry as a clean energy, with the intensification of Russia-Ukraine war, there is large-scale energy shortage Europe, and natural supply Europe has crisis due to cut-off Nord Stream No.1 pipeline. Therefore, it necessary accurately predict consumption gas. In order fulfill this requirement, paper uses Lagrangian Support Vector Regression model Sorensen kernel based on Nonlinear Auto-Regressive Grey Wolf Optimizer for 5-step forecasting monthly all European countries. Under three time lags, comparing results GWO-LSVR SVR, RF, LightGBM, XGBoost, MLP, those five models’ hyperparameters also optimized by GWO, found that smallest MAPE almost cases, numerical generated from 5.844% 11.622%, smaller step size, better effect. Moreover, compares difference GWO WOA, can obtained results. To sum up, proposed strong generalization performance robustness, reliable prediction model.