A combined forecasting strategy for the improvement of operational efficiency in wind farm

Providing precise and stable forecasts for wind speed is a priority for promoting the efficiency of the performance and the economic effectiveness of wind power generation systems. Based on this actual demand, various strategies have been proposed to solve complicated nonlinear forecasting problems like wind speed forecasting. However, the previous models ignore the decisive role of data preprocessing and the limitations of a single model, which is the main reason leading to the inaccurate prediction. Hence, an updated hybrid forecasting system built on a data preprocessing strategy is proposed to effectively predict the wind speed sequence, which includes a data preprocessing module, a multi-objective optimization module, and a forecast module. Data preprocessing contributes to better seizing the traits of the data, a multi-objective optimization algorithm is recommended to optimize the precision and stability of predictions simultaneously, and the different models are combined into a new model for accurate prediction. Moreover, a 10-min wind speed sequence is utilized to affirm the strengths of the hybrid forecasting system, and the empirical studies also affirm that the model proposed has higher exactness and stability over other models.