< Back to previous page
Project
Identiļ¬cation of block oriented models with parallel structures for nonlinear systems (FWOTM610)
Linear models are very popular to model physical systems. But they fail when a wide operating range needs to be modeled. Then, nonlinear models are mostly needed to obtain a sufficient accuracy. One type of nonlinear models that combines flexibility and parsimonity, are the block-oriented nonlinear models.
We propose a new identification framework for nonlinear systems, based on a parallel-branch block oriented model structure, which combines low model complexity and wide applicability. The identification framework is designed to properly handle the stochastic properties of noise and model errors in a random excitation context. The developed identification framework will be validated on some illustrative applications taken from different application domains (mechanical, physiological, microwave, ...)
We propose a new identification framework for nonlinear systems, based on a parallel-branch block oriented model structure, which combines low model complexity and wide applicability. The identification framework is designed to properly handle the stochastic properties of noise and model errors in a random excitation context. The developed identification framework will be validated on some illustrative applications taken from different application domains (mechanical, physiological, microwave, ...)
Date:1 Oct 2011 → 30 Sep 2015
Keywords:Automatic Measurement Systems, Nonlinear Modelling, Medical Physics, Fibre Optic, Parameter Estimation, Microwaves, Instrumentation, Underwater Acoustics, System Identification, Telecommunications, Electrical Measurements, Nonlinear Measurements, Electromagnetism
Disciplines:Telecommunication and remote sensing, Acoustics and acoustical devices, waves, Signal processing