Data informativity
What is data informativity?
In system identification, data informativity is the property that one can distinguish between two models (with different dynamics) using the experimental data. It is related to the amount of information about the system dynamics carried by the data. If this property is not guaranteed, the resulting identified model may be completely inaccurate!
This property has been widely studied for single-input single-output linear systems, partially for multiple-inputs multiple-outputs linear systems and never for nonlinear systems. I have contributed in developing new results with the aim of conpleting the missing cases.
Related journal papers
Xavier Bombois, Kévin Colin, Paul M. J. Van den Hof, Håkan Hjalmarsson. On the informativity on direct identification experiments in dynamical networks. Automatica, vol. 148, 110742. Open access: https://hal.archives-ouvertes.fr/hal-03329586. 2023
Kévin Colin, Xavier Bombois, Laurent Bako, Federico Morelli. Closed loop Identification of MIMO Systems in the Prediction Error Framework Data informativity Analysis. Automatica, vol. 121, 109171. Open access: https://hal.archives-ouvertes.fr/hal-02351669. 2020
Kévin Colin, Xavier Bombois, Laurent Bako, Federico Morelli. Data informativity for the open loop identification of MIMO systems in the prediction error framework. Automatica, vol. 117, 109000. Open access: https://hal.archives-ouvertes.fr/hal-02305057. 2020
Related conference papers
Kévin Colin, Laurent Bako, Xavier Bombois. Data Informativity for the Closed Loop Identification of MISO ARX. Presented at 19th Symposium on System Identification (SYSID), virtual conference, Padova, Italy. Published on IFAC-PapersOnLine, vol. 54, issue 7, pp 779-784. Open access: https://hal.archives-ouvertes.fr/hal-02971657. 2021
Kévin Colin, Xavier Bombois, Laurent Bako, Federico Morelli. Data Informativity for the Identification of Particular Hammerstein Systems. Presented at 21th IFAC World Congress (1st Virtual IFAC World Congress), virtual conference, Berlin, Germany. Published on IFAC-PapersOnLine, vol. 53, issue 2, pp 1102-1107. Open access: https://hal.archives-ouvertes.fr/hal-02392113. 2020
Kévin Colin, Xavier Bombois, Laurent Bako, Federico Morelli. Data informativity for the identification of MISO FIR system with filtered white noise excitation. Presented at IEEE 58th Conference on Decision and Control (CDC), Nice, France. Published in the proceedings pp 1734-1739. Open access: https://hal.archives-ouvertes.fr/hal-02161598. 2019
Kévin Colin, Xavier Bombois, Laurent Bako, Federico Morelli, Informativity: how to get just sufficiently rich for the Identification of MISO FIR Systems with Multisine Excitation? Presented at IEEE 18th European Control Conference (ECC), Napoli, Italy. Published in the proceedings pp 351-356. Open access https://hal.archives-ouvertes.fr/hal-02070880v2. 2019