A Short-Term Forecasting Method Based on Time Series Filtering with Savitzky-Golay for Power Load Curve

The accurate prediction of the load curve is not only vital to establishment generation scheduling, unit start-stop, and maintenance plan, but also has an important influence on ensuring smooth operation side consumption side, reducing cost power improving economic benefits. Based this, a short-term forecasting method based Savitzky-Golay time series filtering constructed. Firstly, relevant factors that affect are used as input components model, such holidays, weather, etc. Secondly, filter sequence weaken adverse effects local fluctuations forecast. Finally, combining with training testing process Deep Neural Network (DNN) samples were constructed by using after filtering, weather other data, so realize short-time in next seven days. effect model verified daily data influencing factor collected from certain plant. simulation results show accuracy cooperation DNN can reach 95%.