Abstract
A nonlinear adaptive controller for controlling a nonlinear and time-varying fed-batch
fermentation process involving one limiting substrate for biomass growth and product synthesis is
proposed. More exactly, in the context of a large class of fed-batch bioreactors for which
dynamical kinetics are not exactly known and time-varying and not all the state variables are
measurable, for the regulation of the substrate concentration, based on input-output state feedback
linearization technique, a nonlinear adaptive control algorithm is proposed. A recursive leastsquares
algorithm with forgetting factor for the on-line estimation of the unknown parameters is
proposed. The stability and convergence properties of this algorithm are proved provided that the
regressor matrix is persistently exciting. Computer simulations performed under identical
circumstances are included in order to demonstrate the performances of this controller compared
to an exactly nonlinear linearizing controller.
