Hammerstein model identification of continuous stirred tank reactor based on least squares support vector machines

Authors:
Zhang Jianzhong;Wang Qingchao
Affiliations:
School of Energy Science and Engineering, Harbin Institute of Technology, Harbin, China;School of Energy Science and Engineering, Harbin Institute of Technology, Harbin, China
Venue:
CCDC'09 Proceedings of the 21st annual international conference on Chinese control and decision conference
Year:
2009

Citing 5
Cited 0

An optimal two-stage identification algorithm for Hammerstein-Wiener nonlinear systems

Automatica (Journal of IFAC)
Least Squares Support Vector Machine Classifiers

Neural Processing Letters
A stable one-step-ahead predictive control of non-linear systems

Automatica (Journal of IFAC)
Identification of MIMO Hammerstein models using least squares support vector machines

Automatica (Journal of IFAC)
Pruning error minimization in least squares support vector machines

IEEE Transactions on Neural Networks

Quantified Score

Hi-index	0.00

Visualization

Abstract

A novel LSSVM-ARX Hammerstein model structure is proposed for a continuous stirred tank reactor (CSTR). LSSVM with a radial basis function (RBF) kernel is used to represent the static nonlinear block in the Hammerstein model. The dynamic linear part of the model is realized by a linear autoregression model with exogenous input (ARX). The linear model parameters and the static nonlinearity can be obtained simultaneously by solving a set of linear equations followed by singular value decomposition. Identification results of CSTR indicate that the proposed Hammerstein model has higher prediction accuracy in comparison with the traditional Hammerstein model, and it can approximate the dynamic behavior of the plant efficiently.