Photovoltaic power prediction based on improved moth-flame optimization algorithm and extreme learning machine
Jing Zhang, Jing Gao · 2022 7th International Conference on Intelligent Computing and Signal Processing (ICSP) · 2022
Accurate photovoltaic power forecasting is a direct method to deal with the uncertainty of photovoltaic power generation. Therefore, this paper proposed a photovoltaic power prediction model. Firstly, in order to improve the optimization performance of the moth-flame optimization (MFO) algorithm, the differential mutation operator was introduced into the MFO algorithm, and the improved MFO (IMFO) algorithm was proposed. Secondly, the IMFO algorithm was applied to determine the optimal parameters of extreme learning machine (ELM), and the photovoltaic power prediction model based on IMFO-ELM was established. Finally, the photovoltaic dataset provided by Desert Knowledge Australia Solar Center was applied to test the prediction performance of the IMFO-ELM model. The results showed that the proposed model accurately predicts the photovoltaic power prediction and control the mean absolute percentage error below 5%.