Maximum Power Load Forecasting Based On Combination Model

Power system is an important industry related to the nation's lifeline.Power load forecasting is an important part of dispatch,planning and operation in power system.It plays an essential guiding role in improving economic benefits,saving costs and arranging maintenance reasonably.The narrowly-defined load forecasting is the prerequisite to power generation planning.The generalized load forecasting has a great influence on the safety of the whole power system.Therefore,power load forecasting is regarded of great importance.In this paper we present a combined model called WPD-GSA-NMLP.According to WPD,the frequency characteristics of load data on different time scales can be revealed,and the distribution on different time scales is more clearly reflected.The NMLP model optimized by GSA is used to predict the frequency on different scales.Finally,different frequency prediction components are integrated into a final prediction result.On the basis of the experimental analysis of the maximum power load in some area of Shanxi province,different model prediction results and each model evaluation indicators(MAE?RMSE?MAPE and NMSE)are fully compared.With these trials accompished,WPD-GSA-NMLP,the combined model,is proved to be an optimal prediction model.Moreover,the prediction results and prediction accuracy are both improved,which demonstrates that this load forecasting method is effective.