Application of Improved Support Vector Machine Based on Shuffled Frog Leaping Algorithm in Wind-Photovoltaic-Battery Power Forecasting

Abstract

Formulating reasonable and accurate wind-photovoltaic-battery generation system power forecasting strategy can improve the security and stability of new energy access to the grid. An improved support vector machine model based on shuffled frog leaping algorithm is proposed to forecast wind power and photovoltaic power in wind-photovoltaic-battery generation system. Based on the historical data of normal operation as input, using the shuffled frog leaping algorithm (SFLA) to optimize the parameters which influences the regression performance of support vector machine and establish the model, then training the model and forecasting the generating power. Finally, the simulation proves that SFLA has better optimization ability, the model has higher accuracy which can effectively forecast wind and photovoltaic power in wind-photovoltaic-battery generation system.