An Extended Super-Twisting Algorithm for simultaneous estimation of reaction rates and input disturbance in bioprocess

Abstract

The present contribution deals with the simultaneous estimation of the reaction rates and input disturbances in a bioprocess. Online estimation of just the reaction rates is possible by sliding mode observer techniques or high-gain observers, in finite-time or with exponential convergence, respectively. However, these techniques are not robust against unpredictable input bioreactor disturbances and are not readily adaptable to treat external disturbances. Motivated by these facts, we propose a novel extended super-twisting algorithm for a class of nonlinear systems, such that the estimation of uncertain parameters, unknown internal dynamics, and external disturbances converge in a finite time or a neighborhood near its nominal values. The effectiveness of the developed algorithm is tested through simulations in an anaerobic digestion process using a batch and a continuous bioreactor. Simulation results show good robustness performance of the extended super-twisting algorithm.