Communications in Information and Systems

Volume 6 (2006)

Number 4

Recursive system identification by stochastic approximation

Pages: 253 – 272

DOI: https://dx.doi.org/10.4310/CIS.2006.v6.n4.a1

Author

Han-Fu Chen

Abstract

The convergence theorems for the stochastic approximation (SA) algorithm with expanding truncations are first presented, which the system identification methods discussed in the paper are essentially based on. Then, the recursive identification algorithms are respectively defined for the multivariate errors-in-variables systems, Hammerstein systems, and Wiener systems. All es- timates given in the paper are strongly consistent.

Keywords

System identification, Hammerstein system, Wiener system, errors-in-variables (EIV), stochastic approximation, recursive algorithm, strong consistency

Published 1 January 2006