The aim of this study are to compare periodic and stochastic models generated by using FFT fre- quencies with periodic and stochastic models generated by using astronomic frequencies. Fourier series are used to simulated periodic model, and autoregressive method is employed to simulate stochastic model. In this study, 253 dominant FFT frequencies and 9 astronomic frequencies are used as constituent frequencies to simulate periodic models. 512 hours of data are used for the model. Results of this study present that the periodic and stochastic models using FFT frequencies give more accurate results than periodic and stochastic models using the astronomic frequencies.
Copyrights © 2014