Research on short term power system load forecasting method based on improved VMD and LSTM

Aiming at the characteristics of nonlinear and non-stationary of power system load data, and the low accuracy of shortterm power system load forecasting, a combined short-term power system load forecasting model based on improved VMD and LSTM methods is proposed. Firstly, the historical load data is preprocessed. Secondly, for the problem that the artificial selection of core parameters in VMD has a poor decomposition effect, the improved whale optimization algorithm is used to search for the VMD parameter combination to obtain the optimal modal component, so as to reduce the complexity of the load data sample set. Then, an IWOA-VMD-LSTM prediction model is constructed to predict the modal components after decomposition, as well as festival characteristics and meteorological characteristics. The predicted results of different components are superimposed to obtain the final load forecasting value. Finally, through simulation experiments, it is verified that the forecasting effect of the IWOA-VMD-LSTM model is better than other forecasting models, and it shows good performance in short-term power system load forecasting.