A multivariable adaptive control strategy for a recycled depollution bioprocess

This paper presents the design and analysis of a nonlinear multivariable adaptive control strategy for a depollution wastewater bioprocess that is carried out in a recycle bioreactor. The adaptive control structure is based on the non-linear process model and is derived by combining a linearizing control law with an asymptotic state observer and a parameter estimator, which plays the role of the software sensor for on-line estimation of the bioprocess unknown kinetics. The effectiveness and performance of both estimation and control algorithms are illustrated by simulations performed for a recycled depollution process for which kinetic dynamics are highly nonlinear, time-varying and incompletely known.