INTERVAL OBSERVER FOR A BIOLOGICAL REACTOR MODEL

• Main Authors: T. A. Kharkovskaia, A. S. Kremlev, D. M. Sabirova, D. V. Efimov, R. Tarek
• Format: Article
• Language: English
• Published: Saint Petersburg National Research University of Information Technologies, Mechanics and Optics (ITMO University) 2014-05-01
• Series: Naučno-tehničeskij Vestnik Informacionnyh Tehnologij, Mehaniki i Optiki
• Subjects: interval estimation; observer; nonlinear systems; parametric-varying systems; parametric uncertainty; bioreactor
• Online Access: http://ntv.ifmo.ru/file/article/9642.pdf
• ISSN: 2226-1494; 2500-0373

The method of an interval observer design for nonlinear systems with parametric uncertainties is considered. The interval observer synthesis problem for systems with varying parameters consists in the following. If there is the uncertainty restraint for the state values of the system, limiting the initial conditions of the system and the set of admissible values for the vector of unknown parameters and inputs, the interval existence condition for the estimations of the system state variables, containing the actual state at a given time, needs to be held valid over the whole considered time segment as well. Conditions of the interval observers design for the considered class of systems are shown. They are: limitation of the input and state, the existence of a majorizing function defining the uncertainty vector for the system, Lipschitz continuity or finiteness of this function, the existence of an observer gain with the suitable Lyapunov matrix. The main condition for design of such a device is cooperativity of the interval estimation error dynamics. An individual observer gain matrix selection problem is considered. In order to ensure the property of cooperativity for interval estimation error dynamics, a static transformation of coordinates is proposed. The proposed algorithm is demonstrated by computer modeling of the biological reactor. Possible applications of these interval estimation systems are the spheres of robust control, where the presence of various types of uncertainties in the system dynamics is assumed, biotechnology and environmental systems and processes, mechatronics and robotics, etc.