Variable Structure Control Based Soft Sensor for Continuous Bioreactor

Abstract

Bioreactors are the vessels used for the processes involved with microbes, plant and animal cells and components of cells such as enzymes. These are mainly used to manufacture new products or to destroy harmful wastes in case of waste treatment plants. Feedback control is necessary to ensure satisfactory performance of continuous bioreactors in the presence of external disturbances, changes in operational requirements and uncertainty in parameters. Non-availability of reliable sensors is a major hindrance for the implementation of effective control strategies. Online estimator has been proved as an attractive solution for online measurement for various types of bioreactors. These “soft sensors” play an important role in control of the bioreactor. Bioprocesses are nonlinear and time variant. Dynamics of bioprocess is under-modeled and uncertain. Design of the observer for such processes is quite challenging. Variable Structure Control (VSC) approach is an attractive solution to handle parametric uncertainty. This concept has been extended to the design of the soft sensor. The proposed observer is designed by the method of Lyapunov for Sliding Mode Observer (SMO) as suggested by Walcott and Zak. Simulation results are presented for fermentation of Single Cell Protein (SCP). The double limiting substrate model in unstructured and unsegregated pattern for continuous mode of operation of bioreactor has been considered. The robustness of the soft sensor is tested. Chattering problem, which is usually associated with VSC technique, has been addressed.