09 Septembre – Thesis defense - Evgeny Shulga
09 h30 Amphi Jean-Paul Dom - laboratory IMS / buiding A31 (University of Bordeaux - Talence campus)
Development of optimal preview control laws for a drivetrain.
This thesis proposes a methodology for controlling complex systems. A technique using the predictive feedforward command and the robust feedback controller is its main proposal. The feedforward action uses a non-linear MPC technique to perform multi-objective optimization of the control efforts and outputs of its embedded model with or without anticipatory effect. The feedback action uses the CRONE methodology guaranteeing the robustness and stability of the closed-loop. This architecture has been applied to control the air supply systems of an internal combustion engine (monovariable case) and the cathode of a fuel cell (multivariable case). The comparison of the proposed approach with traditional control laws has shown its effectiveness and robustness. A practical application to the control of an academic system "ball on plate" is also presented.