Comparison between two interval observers for wastewater treatment models

This paper deals with the design of interval observers for wastewater treatment plants under the conditions of uncertain input and unknown kinetics. We will focus on the influence of the a priori knowledge of the process on the estimation results. We build two interval observers: the first, based on a simple model including substrate, biomass and recycled biomass, is designed to reconstruct the upper and lower bounds of biomasses from substrate measurements. The second one based on a more complex model including in addition the dissolved oxygen concentration, is applied for the on-line estimation of the interval of variation of biomasses and substrate from dissolved oxygen data. We compare the performances of the observers with respect to the asymptotic interval errors.