| | |
Summary: A PSO-Subspace Algorithm for
Identification of Hammerstein Models
S. Z. Rizvi
H. N. Al-Duwaish
King Fahd University of Petroleum & Minerals, Dhahran, Saudi
Arabia (e-mail: srizvi@kfupm.edu.sa )
King Fahd University of Petroleum & Minerals, Dhahran, Saudi
Arabia (e-mail: hduwaish@kfupm.edu.sa )
Abstract: A new method for hammerstein model identification is proposed in this work by
modeling the linear dynamic part with state-space model and the nonlinear part with radial
basis function neural network (RBFNN). A unique PSO-Subspace algorithm is proposed which
makes use of particle swarm optimization (PSO) for estimation of RBFNN parameters and
numerical algorithm for subspace state-space system identification (N4SID) for estimation of
parameters of the linear part. Simulation examples are included to illustrate the performance
of proposed algorithm.
Keywords: Particle Swarm Optimization, Subspace Identification, Static Nonlinearity,
Dynamic Linearity, Radial Basis Function Neural Network.
1. INTRODUCTION
|
