Study on Short-Term Load Forecasting Considering Meteorological Similar Days
A short-term load forecasting method based on meteorological similar days and error correction is proposed in this paper. First, SPSS software is used to carry on the regression analysis of meteorological factors, select most significant meteorological factors in each season, and determine the weight of each factor as the basis for selecting the weather similar days. Then the historical forecast error data sample set is set up. For a certain forecast date, the error data samples from the similar days are extracted to establish a set, and the probability density distribution model is established. Finally, the error fluctuation of the forecast point is analyzed to get the compensated value of forecast error. The sampling error closest to the error compensation value is selected as the fitted error values and added to the predicted value to improve the forecast accuracy.
KeywordsMeteorological similar days Probability density distribution Error correction Volatility analysis
- 1.Kang, C., Xia, Q., Zhang, B.: Review of power system load forecasting and its development. Autom. Electr. Power Syst. 28(17), 1–11 (2004)Google Scholar
- 2.Tai, N., Hou, Z.: New short-term load forecasting principle with the wavelet transform fuzzy neural network for the power systems. Proc. CSEE 24(1), 24–29 (2004)Google Scholar
- 3.Liu, W., Yang, K., Da, L., et al.: Day-ahead electricity price forecasting with error calibration by hidden markov model. Autom. Electr. Power Syst. 33(10), 34–37 (2009)Google Scholar
- 4.Li, X., Chen, Z.: Correctly using SPSS software for principal components analysis. Stat. Res. 27(8), 105–108 (2010)Google Scholar