Short-term photovoltaic power generation prediction based on VMD-MSI-GWO-KELM

To improve the accuracy of photovoltaic power prediction, a combined prediction model integrating multiple algorithms is proposed. First, key factors are selected, and the Gaussian Mixture Model (GMM) is used to cluster and generate historical days with high correlation to the forecast day. Then, a photovoltaic power prediction model based on Variational Mode Decomposition (VMD) and an improved Grey Wolf Optimizer Kernel Extreme Learning Machine (MSI-GWO-KELM) is constructed. Finally, actual data from a photovoltaic power station in Australia is used as a case study. The results show that the VMD-MSI-GWO-KELM model is suitable for different weather conditions and offers better prediction accuracy than other models.