New Short Term Load Forecasting models based on growth rate scaling and simple averaging

摘要

Load Forecasting is the link joining the strategic power system operation with the almighty mathematic algorithms in order to bring greater reliability, efficiency and economy in the system. This paper brings two novel forecasting engines, named SYGRSA and SYGRSAWP, for Short Term Load Forecasting (STLF), i.e., load forecasting with time leads ranging from one day to one week. STLF itself plays a crucial role in the control and scheduling operations of a power system. Modern techniques have been used to improve the accuracy of existing load prediction models using proper feature selection and consideration of necessary factors. The proposed models combine similar day approach, growth rate scaling and averaging techniques. SYGRSAWP is an optimised form of SYGRSA. Both the models have the potential to handle large historical data in short period of time. Moreover, they show remarkable forecasting accuracy. Further, a comparison between the two models has been analysed.