A STUDY ON IMPROVING ACCURACY OF SHORT-TERM LOAD FORCASTING
Yang Shibai (Xi' an Jiaotong University)Li Yangxu (Guangdong Electric Power Test and Research Institute, Guangzhou)Yi Changqing (North West China Electric Power Administration, Xi' an)
This paper presents an algorithm based on the time series method for the short-term load forcasting, with the example of North - west China power system. Bad data are detected and corrected by the daily load curve and the data of the daily total energy requirement. An easy method is developed to utilize the fewer obtainable weather data. The program can be used on PC and the accuracy of the average daily forcasting reaches 2%-3%.