Research on short-term power load forecasting based on ARIMA-LSTM model

As an important part of power development planning, power load forecasting is an important prerequisite for power generation, transmission and reasonable distribution of electric energy, and improving the accuracy of load forecasting is an important means to guarantee the stable operation of the power grid and improve the economic benefits of the power system. Aiming at short-term power load forecasting, this paper used the Autoregressive Integrated Moving Average (ARIMA) model to forecast the power load time series data, and combined the climate data, the main influencing factor of power load, to optimize the ARIMA residual by Long Short Term Memory (LSTM) model. A combined forecasting model based on ARIMA-LSTM was proposed by fully combining the advantages of both models. Taking the power load in Europe as the prediction object, the results showed that the root mean square error (RMSE) of the combined model was only 0.7835 MW, and the mean absolute percentage error (MAPE) was only 0.06 %. Compared with the LSTM model and ARIMA model, RMSE had been reduced by 91.18 % and 98.48 %, respectively. The combined model has improved the prediction accuracy and had specific practical application value.