Using enhanced crow search algorithm optimization-extreme learning machine model to forecast short-term wind power

• The enhanced crow search algorithm outperforms the state-of-the-art variants. proposed optimizes parameters of extreme learning machine. wind power forecast model comparison models. Accurate prediction reduces operating cost system. strong volatility and randomness impact grid reduce voltage quality when is connected to in large scale. sector takes abandonment measures ensure stability. optimization-extreme machine (ENCSA-ELM) accurately short-term improve utilization efficiency clean energy. (1) (ENCSA) applied forecast. convergence performance test revealed that local development global exploration capabilities ENCSA were enhanced, result outperformed other well-known nature inspired algorithms CSA variants; (2) output input forecasting models determined by analysis factors samples autumn, winter spring forecasted ENCSA-ELM model; (3) results analyzed multiple evaluation indexes. simulation experiments error interval indexes methods, traditional ELM optimized algorithms. RMSE value MAPE controlled below 20% 4%. maintains stability increases