{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

A3-2 - Sedma Nacionalna Konferencija so Me|unarodno U~estvo...

Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon
ON-LINE PARAMETERS ESTIMATION OF THE MODEL FOR GLUCOAMYLASE PRODUCTION USING ASPERGILLUS NIGER CELLS Silvya Popova 1 , Svetla Vasileva 1 , Kolishka Tsekova 2 1 Institute of Control and System Research - Bulgarian Academy of Sciences, Acad. G. Bonchev Str. Bl.2, P.O. Box 79, 1113 Sofia, Bulgaria, [email protected] 2 Institute of Microbiology, Bulgarian Academy of Sciences, [email protected] Abstract – Starch-degrading enzymes of microbial origin have a number of industrial applications. The mould extracellular enzyme, glucoamylase, is of major importance in the starch industry in the commercial production of considerable significance in an industrial context since it offers the advantages of increased reaction rates, decreased viscosity, reduced microbial contamination and better storage stability. Fungal glucoamylase are of industrial importance in production of sugar from starch. Most glucoamylases are produced today by submerged fermentation with Aspergillus niger strains. A general chemostat microbial cultivation model is used. The estimation procedure is based on the extended Kalman filtering method. Thus the necessity of performing experiments for the sake of parameter identification could be successfully avoided. In this case-study a batch process of glucoamylase preparation is studied from a mutant strain Aspergilus niger B-77. The estimation performance is studied under different initial conditions. Keywords batch cultivation, parameters estimation, Kalman filtering method 1. INTRODUCTION There are two main approaches for on-line state and parameter estimation of nonlinear systems such as biotechnological processes – exponential estimators design (based on Kalman filtering method) and asymptotic estimators design (based on a linear algebra results) [1, 2]. The main advantage of the first approach is that the speed of convergence of the estimates towards their true values could be arbitrarily fixed. However, this could be achieved on the expense of complete (or particular) knowledge for the process kinetics. Recently, much effort have been done for developing exponential estimators, accounting for a specific features of the process (or class of processes), thus decreasing the amount of “exact” information needed for estimators design. On the other side, standard non-linear estimation techniques are available, such as Kalman filtering method [2], that lead to exponential observers for which it is possible to arbitrarily fix the speed of convergence of the estimated variables towards their true values. Much of the work concentrated on Kalman filtering approach dealt with a model, the accuracy of which played a major determining factor in the quality of estimation . Deficiencies in the process kinetic model, such as unmodelled process dynamics and model parameters’ variations, have to a certain extent been overcome by applying adaptive estimation techniques, where the process state and (some) model parameters are simultaneously estimated [2,3].
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 2
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}