Abstract
Subspace identification techniques are reinterpreted via classical realization theory to formulate a wide class of subspace identification methods. Re-formulating subspace identification in terms of a low rank decomposition of a weighted Hankel matrix allows special cases such as impulse-based and step-based input signals, but also realization based on arbitrary input signals and correlation functions. Ideas are illustrated with a simulation example.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have