Multi-scenario calibration method for double-Gaussian wake model based on DE-SQP cooperative optimization

Against the backdrop of global energy transition, the issues of power generation loss and equipment fatigue caused by wind farm wake effects are becoming increasingly prominent. This paper addresses the limited generalization capability of existing double-Gaussian wake models by proposing a Differential Evolution-Sequential Quadratic Programming (DE-SQP) cooperative optimization framework. This framework enhances multi-scenario prediction accuracy through the calibration of 15 empirical parameters within the model. Firstly, wake datasets under three atmospheric stability conditions – stable, neutral, and convective – are generated based on CFD simulations. Secondly, using the root mean square error (RMSE) of wind turbine power prediction in the wake region as the objective function, the DE algorithm is employed for global search of the parameter space, followed by local fine-tuning using the SQP algorithm. Finally, comparative validation demonstrates that the calibrated model reduces power prediction errors by 96.97%–99.21%, with the maximum error not exceeding 50.75 kW. This method significantly improves the robustness of the double-Gaussian model across the entire wake region, providing an efficient analytical tool for intelligent wind farm control.