Study on Predicting Energy Storage and Load of Photovoltaic Plant in Xusheng
In view of obvious features such as fluctuation and intermittence of PV plants'load and energy storage,a PV generation prediction model based on long-and short-term memory neural network and a PV plant load prediction model based on BP neural network were proposed in this study.In PV generation prediction model,Pearson correlation coefficient was used to analyze influencing factors of PV power generation,and factors with higher correlation were input into the long-and short-term memory neural network model.Then the data were classified into four seasonal types by K-means clustering analysis,and the processed dataset was put into the prediction model for training.In PV plant load prediction model,the processed historical load data were input into time sequence-based BP neural network for iterative training,while the weights and thresholds of the model were adaptively optimized.Simulation results verified that the proposed prediction model has considerably high prediction accuracy,and the PV plant load prediction model has higher accuracy and stability compared with PSO-BP and GA-BP neural networks.
NETL
NSTL
万方数据
