Load forecasting is the basis for planning of the grid.The accuracy of the forecast is related to the safe,reliable and economic operation of the grid.In order to solve the problem of low prediction accuracy due to the unstable original data,this paper proposes an ARIMA prediction model based on the ensemble empirical mode decomposition (EEMD),which performs the noise processing on the monthly load of a certain area,and then performs empirical mode decomposition to make the components smoothed.Each component is predicted by ARIMA model.Finally,the component prediction results are added to obtain the final predicted value.The example shows that the prediction accuracy of the regional monthly load of the EEMD-ARIMA model is higher than that of the ARIMA model.