Multivariate autoregressive model applied to conditioned spectral analysis of complex systems

Risto Suoranta, Seppo Rantala

Research output: Chapter in Book/Report/Conference proceedingConference article in proceedingsScientificpeer-review

1 Citation (Scopus)

Abstract

A method of performing out conditioned spectral analysis without the drawbacks of traditional Fourier-transform-based methods is introduced. The method is based on the estimation of the multivariate autoregressive (MAR) model. Because the MAR model is a black-box model and can describe systems with feedback loops, it is a suitable tool for the analysis of complex systems. Two different approaches for the conditioned spectral matrix in the context of the MAR model are presented. They are the reduced conditioned spectral matrix and the noise conditioned spectral matrix. These spectral quantities offer possibilities in the analysis of systems where no exact prior knowledge about internal structures is available
Original languageEnglish
Title of host publicationI 1990 IEEE International Conference on Systems Engineering
PublisherIEEE Institute of Electrical and Electronic Engineers
Pages190-193
ISBN (Print)0-7803-0173-0
DOIs
Publication statusPublished - 1990
MoE publication typeA4 Article in a conference publication

    Fingerprint

Cite this

Suoranta, R., & Rantala, S. (1990). Multivariate autoregressive model applied to conditioned spectral analysis of complex systems. In I 1990 IEEE International Conference on Systems Engineering (pp. 190-193). IEEE Institute of Electrical and Electronic Engineers . https://doi.org/10.1109/ICSYSE.1990.203130