考虑多气象因素的EMD-PCA-LSTM组合模型在光伏功率短期预测中的应用OA

To address the challenges of strong non-stationarity,stochastic fluctuation,and feature redundancy in photovoltaic(PV)power forecasting under multiple meteorological influences,a hybrid prediction model based on empirical mode decomposition(EMD),principal component analysis(PCA),and long short-term memory(LSTM)is proposed.First,EMD is employed to decompose meteorological variables into multi-scale components,enabling effective extraction of temporal features at different scales.Then,PCA is applied to reduce the dimensionality of high-dimensional features while retaining the dominant information and mitigating redundancy.Finally,the processed features,together with historical PV power data,are fed into an LSTM network to perform short-term forecasting.Case studies based on real PV operation data demonstrate that the proposed model outperforms conventional LSTM and EMD-LSTM models in terms of mean absolute error(MAE),root mean square error(RMSE),and coefficient of determination(R²).The results verify the effectiveness and robustness of the proposed method under complex weather conditions.