VSS learning based intelligent control of a bioreactor system

Abstract

Intelligent control of a class of biochemical reactor systems is discussed. For this purpose, an Adaptive Neuro-fuzzy Inference System (ANFIS) architecture, whose parameters are updated using a VSS based learning algorithm, is utilized as controller. Incorporation of this new update mechanism guarantees stability and robustness of the learning dynamics under the existence of parametric and nonparametric uncertainties. Furthermore, the proposed approach does not require measurement of cell concentration that is very hard to do in practical applications. Simulation results presented demonstrate the efficacy of the control architecture.