Elman neural network based short-term photovoltaic power forecasting using association rules and kernel principal component analysis

Photovoltaic power prediction for reducing the impact of the grid-connected photovoltaic power generation system on the power system is of great significance. Aiming at the power generation characteristics of the photovoltaic system, a method of Elman neural network based photovoltaic power forecasting using association rules and kernel principal component analysis (KPCA) is proposed in this paper. Gray relation analysis is a means of data mining and used for selecting several power days which are highly correlated with predicted days. In order to remove redundant information, the kernel principal component analysis (KPCA) is used to extract the feature of photovoltaic (PV) power time series. The Elman neural network is used for power prediction due to its dynamic recursive performance. In view of the fact that the prediction error of the Elman neural network prediction model at the peak of power fluctuation is large, the Markov method is proposed to revise and compensate the prediction value of the model to further improve the prediction accuracy. The model is validated by using real data from the National Renewable Energy Laboratory. The results show that the proposed method can effectively improve the prediction accuracy and enhance the generalization ability of the neural network model, which has a good feasibility.