Interval Prediction of Photovoltaic Output Based on WOA-LSTM-LSSVM Combined Model

Existing photovoltaic power prediction methods often only predict the definite value of future photovoltaic output, and the prediction model is single and has limited accuracy. In view of this, this paper proposes an interval prediction method for photovoltaic output based on a combined model. The proposed combined model consists of the whale optimization algorithm (WOA), the long-short term memory (LSTM) neural network, and the least square support vector machine (LSSVM). WOA is used to optimize some relevant parameters in the LSTM and LSSVM. The LSTM is adopted to predict the high-frequency signals of original photovoltaic output sequence. The LSSVM is applied to predict the medium-frequency and low-frequency signals. Finally, the effectiveness of the proposed method is verified through actual calculation example analysis. The results show that the prediction accuracy of the proposed method is higher than that of the existing methods.