Fræser: a Framework for Estimating Errors-in-Variables Systems

Fræser is a framework for estimating the parameters of static and dynamic errors-in-variables systems with the opportunity to compare various errors-in-variables parameter estimation algorithms via simulations. It features a graphical user interface and several examples for simultaneously estimating model and noise parameters.

The framework incorporates the following linear and nonlinear estimation methods for static and dynamic systems:

• model parameter estimation for static systems
• Koopmans method
• linear model and noise parameter estimation for dynamic systems
  • (extended) instrumental variables method (XIV)
  • bias-compensating least-squares method (BCLS)
  • Frisch scheme (FS)
  • generalized Koopmans-Levin method (GKL)
• nonlinear model parameter estimation for static systems
  • nonlinear Koopmans method (NK)
  • approximated maximum likelihood method (AML)
• nonlinear model and noise parameter estimation for dynamic systems
  • bias-compensated least squares method (BCLS)
  • nonlinear Koopmans-Levin method (NKL)
  • nonlinear extennonlinear extension to generalized Koopmans-Levin method (NGKL)

Besides errors-in-variables estimation algorithms, the framework contains a range of MatLab user interface component extensions:

• Property browser: a sheet control for editing properties of a handle (reference) or value object. It comes in two flavors: MatLabPropertySheet has a pure MatLab implementation, while SwingPropertySheet is a Java Swing-enabled implementation for improved editing capabilities (visual clues for bringing up a dialog to edit compound values, expandable and contractable nested properties, check boxes and combo boxes for constrained selection, etc.) SwingPropertyBrowser is an integration of L2FProd.com Common Components 7.3 into MatLab.
• Object list: a list box control for storing object references, with possibility of  adding new items using prototype templates, removing items and changing the order of items (move up and down).
• Object graph: a dialog for visualizing and changing object properties and dependencies.
• Progress indicator: a progress indicator modal dialog that combines the MatLab function waitbar with automatic disposal on error with a mechanism similar to that provided by onCleanup.
• Other features include control docking and unit conversion.

Prerequisites

• MatLab release 2008a (version 7.6) or later
• Optimization Toolbox (optional)
• Symbolic Math Toolbox (optional)

Download

The framework is under active development. The latest (beta) version is available for download. See read me first documentation (also included within compressed package) for installation and usage instructions.

Distribution	File
Development snapshot release for 32-bit Windows
Development snapshot release for 64-bit Windows
Development snapshot source code
freeglut 2.6.0 RC1 DLL for 32-bit Windows
freeglut 2.6.0 RC1 DLL for 64-bit Windows

How to cite

The Koopmans-Levin family of algorithms are described in the following brief paper and references within:

Levente Hunyadi and István Vajk, Identification of errors-in-variables systems using data clustering in Gregor Rozinaj, Jozef Čepko, Peter Trúchly, Ján Vrabec and Juraj Vojtko (eds.), Proc. of the 15th International Conference on Systems, Signals and Image Processing, pages 197-200, Bratislava, Slovakia, June 2008.

@INPROCEEDINGS{hunyadi:iwssip08eivcluster,
author = {Levente Hunyadi and Istv\'{a}n Vajk},
title = {Identification of errors-in-variables systems using data clustering},
year = {2008},
month = {June},
day = {25--28},
booktitle = {Proc. of the 15th International Conference on Systems, Signals and Image Processing (IWSSIP)},
editor = {Gregor Rozinaj and Jozef \v{C}epko and Peter Tr\'{u}chly and J\'{a}n Vrabec and Juraj Vojtko},
address = {Bratislava, Slovakia},
isbn = {978-80-227-2856-0},
pages = {197--200}
}

Copyright

Copyright © 2007-2009 Levente Hunyadi hunyadi [@at@] aut [.dot.] bme [.dot.] hu
Copyright © 2007-2009 BME Department of Automation and Applied Informatics

This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.