Short-term photovoltaic power prediction model based on VMD-DBO-KELM
Xiong Liu · Journal of Physics Conference Series · 2024
Abstract Aiming at the problem that the current photovoltaic power prediction model has poor generalization performance and cannot achieve accurate prediction under different environmental conditions, in this paper, a combinatorial prediction model of a kernel extreme learning machine is optimized by a dung beetle optimization algorithm after variational mode decomposition. A number of different time scales of photovoltaic power data are selected for the experiment. From the experimental results, compared with the traditional model, the proposed model has strong generalization performance and outstanding prediction results in all data sets, and has broad application prospects.