Abstract EN:

The major axes of research of this thesis, deal with the design of observers, robust output feedback nonlinear control and robust minimal time control for a class of nonlinear bioreactors. In face of partial measurements of the state and uncertainties on the dynamics, dynamical bounds on unmeasured variables are developed (interval observers), when the considered systems are not detectable for unknown in puts and not observable for any input. Exponential observers are also designed for particular cases of such systems. Based on nonlinear “min-max” control and Lyapunov’s techniques, robust nonlinear output controllers are developed for the global stabilization. For the minimal time control problem with partial information, the certainty equivalence principe is fulfilled for a class of bioreactors, then the proposed control law has the property that the worst state is given by an interval observer, instead of the exact unmeasured state. Finally, a real application to a biological process is presented for the illustration of the proposer observers, interval observers and robust control law.