Encadrent: M. Darouach et M. Alma
cotutelle Mexique, thèse financée par CONACYT, soutenance à distance.
In this thesis, the design of observers for nonlinear parameter-varying algebro-differential systems and their applications in fault estimation and parameter estimation are studied. Nonlinear algebro-differential parameter-varying systems allow the preservation of nonlinearities and a better understanding of the equations involved in the analyzed model, as well as a broader operating range.
The observer designed in this work is known as the generalized dynamic observer (GDO). The concept involves incorporating dynamics structure to augment its degrees of freedom, with the goal of achieving accuracy in steady-state and improve robustness in estimation error against disturbances and uncertainties in parameters.
The main idea is to use the advantages of this generalized structure to propose various methodologies for nonlinear parameter-varying algebro-differential systems, thus creating a more general framework than those existing in the literature.
Different estimation algorithms are presented to obtain diverse observer structures. The asymptotic stability of the obtained observers is analyzed through Lyapunov approach using Linear Matrix Inequalities (LMI), and the elimination lemma, which is used to transform these LMIs while preserving the generalized structure of the observers.
Finally, to illustrate the effectiveness of the proposed approaches, the performances of the presented observers are evaluated through some engineering applications.