A New Short-Term Load Forecasting in Power Systems
2014 ◽
Vol 494-495
◽
pp. 1631-1635
Keyword(s):
For realizing highly accuracy load forecasting, a new method is proposed. Power load time series belongs to chaotic series. Firstly, for obtaining three parameters in chaotic theory, namely time delay, embedding dimension and the number of the nearest neighbors, self-correlation function method and G-P algorithm are used to reconstruct the phase space of chaotic time series. Secondly, ant colony optimization approach is introduced to more accurately acquire forecasting reference points, considering distance and relativity of phase points evolution in this paper. Finally, GM (1, 1) Model is applied to forecast daily load data. The actual forecasting results prove that the new approach has better forecasting accuracy and convergence.
2013 ◽
Vol 816-817
◽
pp. 766-769
Keyword(s):
Keyword(s):
2006 ◽
pp. 984-993
◽
Keyword(s):