Release of CONTSID 7.5

22/05/2024 - Version 7.5 of the CONTSID toolbox is now released. This new release is compatible with version 2024a of Matlab.
Minor bugs have also been corrected.

Release of CONTSID 7.4

8/07/2021 - Version 7.4 of the CONTSID toolbox is now released. This new release is compatible with versions 2020 and 2021 of Matlab.

It includes a new App that provides a user-friendly environment that will help the user to focus on the result analysis of data-driven modelling rather than on the technicality of the toolbox commands.
It can be executed by typing in the command window:
• contsid_gui
To know more about the new App, see the paper
H. Garnier, M. Gilson, H. Muller, F. Chen, A new Graphical User Interface for the CONTSID toolbox for Matlab, IFAC Symposium on System Identification (SYSID'2021), Padova, Italy, July, 2021.
Updated versions of SRIVC-based estimation schemes for identifying MISO time delay transfer function models are available:
• TFSRIVC
• TFRIVC
A new demo has been added to illustrate the new feature.
• contsid_TFRIVC_MISO

Automatica paper publication

1/09/2019 - V. Pascu, H. Garnier, L. Ljung, A. Janot. Benchmark problems for continuous-time model identification: design aspects, results and perspectives, Automatica, 107, 511-517, Sept. 2019.

Release of CONTSID 7.3

24/07/2018 - Version 7.3 of the CONTSID toolbox is now released. This new release is compatible with versions 2017 and 2018 of Matlab.

The SRIVC-based routine for identifying transfer function models plus delay for white noise has been robustified and renamed:
• TFSRIVC (replaced the previous TDSRIVC routine)
The toolbox now includes a new RIVC-based routine for identifying transfer function models plus delay for colored noise:
• TFRIVC
The toolbox now includes two new frequency domain PEM and SRIVC-based routines for identifying transfer function models plus time-delay from relay feedback tests:
• TFFCOE
• TFFSRIVC
Updated version of recursive parameter estimation schemes for identifying slowly linear time-varying (LTV) continuous-time models are available:
• RLSSVF
• RIVSVF
• RSRIVC
New demos have been added to illustrate the new features.
• See advanced identification menu
The main demonstration program remains
• CONTSID_DEMO

Release of CONTSID 7.2

16/06/2017 - Version 7.2 of the CONTSID toolbox is now released. This new release is compatible with versions 2016 and 2017 of Matlab.

The toolbox now includes updated version of recursive parameter estimation schemes for identifying slowly linear time-varying (LTV) continuous-time models
• RLSSVF: recursive least squares-based SVF
• RIVSVF: recursive instrumental variable-based SVF
• RSRIVC: recursive simplified refined instrumental variable for continuous-time models
The following SRIVC-based parameter estimation routines have been also updated
• PROCSRIVC for identifying simple process models with delay
• TDSRIVC for identifying transfer function models with delay
The toolbox now includes a new PEM (OE)-based parameter estimation routine for identifying transfer function models plus delay
• TDCOE for identifying transfer function models + delay
New demos have been added to illustrate the new features.
• See advanced identification menu
The main demonstration program has been renamed
• IDCDEMO.M has been replaced by CONTSID_DEMO.M
The toolbox now includes mex files to be used on Linux machine in the case of irregularly sampled data

Call for participation - Doctoral School on Data-driven Model identification

05/02/2017 - Registration is open to attend the 1st Doctoral School to be held in Nancy - April 3-6, 2017
'Springschool'

Release of CONTSID 7.1

05/01/2016 - Version 7.1 of the CONTSID toolbox is released. This new release is compatible with versions 2014, 2015 and 2016 of Matlab.

It includes the following major additions:
• it can now identify linear transfer function model from frequency domain data
• it supports additional routines to estimate simple process models
• it includes new methods to identify transfer function + time-delay models
• it provides new efficient routines to estimate time-varying parameter (TVP) models
• it provides routines to estimate PDE models for heat diffusion.

Last but not least, the demonstration programs, which illustrate the use of the toolbox as well as the latest developments, have been entirely rewritten:
• type idcdemo in the command window