Engineering Mechanics

Proceedings Vol. 12 (2006)

ENGINEERING MECHANICS 2006

Copyright © 2006 Institute of Theoretical and Applied Mechanics, Academy of Sciences of the Czech Republic, Prague

 

J. Tůma
pages 394 - +9p., full text

The paper deals with methods for frequency spectrum estimation by autoregressive modeling. Estimate of the autoregressive model parameters is the first step in this way of spectral analysis. The method used to do it are based on solving the system of normal equations, resulting from the least square method, or on solving the Yule-Walker equation and on Burg’s method. The least square method uses the Gram-Schmidt ortogonalisation and its modification. The YuleWalker equations are solved using Cholesky’s factorization and Levinson’s iterative method. The AR model parameters determine the transfer function of a linear system with white noise as an input signal. The frequency transfer function and the model error variance determine the frequency spectrum.

Text and facts may be copied and used freely, but credit should be given to these Proceedings.

All papers were reviewed by members of the scientific committee.