Adaptive identification of continuous time MIMO state-space models

Chouaib Afri (1) , Laurent Bako (2) , Vincent Andrieu (1) , Pascal Dufour (1)

1 LAGEP - Laboratoire d'automatique et de génie des procédés
2 MIS - Ampère, Département Méthodes pour l'Ingénierie des Systèmes

This paper studies the problem of identifyinglinear continuous-time state-space models from input-outputmeasurements. An adaptive identifier is developed for the onlineestimation of both the state and the model parameters in adeterministic framework. Our approach is a non parametricone in the sense that it provides an arbitrary realization ofthe system. It relies on ideas from the subspace identificationliterature and adaptive observer.

Domaines: Automatique / Robotique

Chouaib Afri, Laurent Bako, Vincent Andrieu, Pascal Dufour. Adaptive identification of continuous time MIMO state-space models. 54th IEEE CDC, Dec 2015, Osaka, Japan. pp.5677-5682, ⟨10.1109/CDC.2015.7403110⟩. ⟨hal-01265057⟩