Photovoltaic Array Power Prediction Model Based on EEMD and PSO-KELM

In order to improve the accuracy of short-term photovoltaic (PV) power prediction, the PV array power prediction model based on ensemble empirical mode decomposition (EEMD) and kernel extreme learning machine (KELM) optimized by particle swarm optimization algorithm (PSO-KELM) is proposed in this paper. The EEMD algorithm is used to decompose the power sequence of the PV array into a series of intrinsic mode functions (IMFs) with different features, the IMFs with similar features are divided into new components by sample entropy (SE), and establishing the KELM power prediction model corresponding to the new component. The PSO is used to optimize the output weight of KELM, seeking the optimal solution of upper and lower limits, and improving the performance of KELM. The actual data is used to compare the proposed prediction model with the traditional back-propagation (BP) and extreme learning machine (ELM) prediction models. The results validate that the proposed model can predict power of PV array with good accuracy and calculation speed.