论文信息 - On the stability of dynamic models obtained by least-squares identification

On the stability of dynamic models obtained by least-squares identification

The stability properties of the least-squares (LS) models fitted to stochastic dynamic systems described by difference equations are investigated. It is shown through a counterexample that the LS method can, for such systems, give unstable models. Sufficient conditions for the stability of the LS models in question are also included.

Petre Stoica | Torsten Söderström | T. Söderström | P. Stoica

[1] T. Söderström,et al. The Steiglitz-McBride identification algorithm revisited--Convergence analysis and accuracy aspects , 1981 .

[2] J. Makhoul,et al. Linear prediction: A tutorial review , 1975, Proceedings of the IEEE.

[3] J.H. McClellan,et al. A simple proof of stability for all-pole linear prediction models , 1979, Proceedings of the IEEE.