Hammerstein System Identification with Stochastic Approximation

W. Greblicki


System identification, Hammerstein system, recursive identification, nonparametric identification, stochastic approximation


This article presents recursive algorithm to recover the nonlinear characteristic of the memoryless part of the Hammerstein system. The a priori information about the system is nonparametric; no functional form of the nonlinearity is known. The algorithm is derived from the idea of stochastic approximation. Its convergence to the characteristic is shown, and the convergence rate is given.

Important Links:

Go Back