Recursive Identification of Hammerstein Structure

A novel recursive algorithm for identification of Hammerstein structures is developed. The linear and nonlinear parameters are separated and estimated recursively in a parallel manner, but each updating algorithm employs the estimation produced by the other at the previous time instant. Hence, it is termed the Alternately Recursive Least Square (ARLS) algorithm. When compared with Recursive Least Squares (RLS) algorithm applyed to the over-parametric representations of the Hammerstein structure, ARLS demonstrated superior performance over extensive numerical simulations.

ISBN:978-1-887706-53-7

Publication year:2011

Keywords:System identification, Hammerstein system, Muscle models, Alternating least squares, Functional electrical stimulation