A novel hybrid model based on Laguerre polynomial and multi-objective Runge–Kutta algorithm for wind power forecasting

Abstract

In recent years, the application of wind power forecasting in power systems has gained widespread recognition. However, most studies have only focused on improving forecasting accuracy and few have considered the stability of forecasting results. To solve these two issues simultaneously, this study proposed a novel hybrid wind power forecasting model. First, two types of Laguerre polynomials are used to construct the hybrid Laguerre neural network. Then, a multi-objective Runge–Kutta algorithm is proposed to optimize the weights of the neural network while enhancing the accuracy and stability of the forecasting. Finally, ensemble learning is introduced to further improve the forecasting capability of the model. To verify the effectiveness of the proposed hybrid forecasting model, a large number of comprehensive experiments are carried out using the wind power data of a wind farm in Xinjiang, China. The experimental results show that the proposed hybrid forecasting model had higher forecasting accuracy and better stability than other forecasting models.