Des nouvelles approches de commande et d’estimation non linéaires robustes dédiées aux entraînements électriques.

Abstract

The purpose of the research presented in this thesis is to propose a methodology for the control and observation of the induction motor (IM) based on the algorithms using the mean value theorem (MVT) and the transformation by sector non-linearity approach. In the first step, the different control techniques of electric drives were identified and analyzed. A robust state and estimation feedback control approach is then developed with variable parameters. In the field of low power, the removal of the mechanical speed sensor can be of economic interest and improve operational safety. We have presented two categories of methods that allow reconstructing and controlling the rotor speed with desired quantities under field-oriented control of the IM’s machine, the MVT observer and the robust MVT controller respectively. All the solutions have been validated by numerical simulation and affirmed by experimental tests to compare the accuracy and dynamics characteristics of the different methods with the MVT control. Finally, new robust control and estimation approaches with a novel representation for uncertain systems with varying parameters based on the MVT and sector nonlinear addressed to control the IM ‘s machine with FOC control. The results of the various simulation tests and the different experimental trials put into evidence the robustness and the success properties of the proposed algorithms. The thesis ends with a review of our contribution in terms of research.