engleski

Adaptive on-line optimization of enzyme reactors

Production in enzyme reactors involve series of reactions with complex kinetics. Complexity involves multisubstrate activation/deactivation, enzyme-coenzyme dependencies, and high sensitivity of enzyme reaction rate and saturation parameters on temperature and pH. Enzyme reaction systems exhibit complex static and dynamic behaviour, such as multiple steady states and appearance of limit cycles Modelling and kinetic parameter estimation can be very time consuming due to mathematical complexities of enzyme kinetics, large setes of interdependent parameters and difficulties in prior selection of appropriate experimental conditions for estimation and their extrapolation to production conditions. The main objective of modelling is optimisation of reactor conditions and process control. An alternative approach to optimisation of reaction systems is based on application of "local " input-output models which are continuously on-line adapted to process conditions(1). The local models have simple linear structure and parameters are estimated by the recursive least squares method. Such models do not reveal reaction mechanism but can account for unmodelled effects and are very efficient for adaptive optimal control. In this work are presented results of optimisation of continuous (R)-mandelic acid production in an enzyme membrane reactor. Applied are the reaction kinetics and mass balances from literature(2) and are introduced additional models for rate dependencies on pH and temperature. Results of computer simulation of adaptive optimisation are presented. 1. J. Strohschein, Z. Kurtanjek, Chem. Biochem. Eng. Q. , 1997 ( in press ) 2. Dj. Vasic-Racki, M. Jonas, C. Wandrey, W. Hummel and M.R. Kula, Appl. Microbiol. Biotechnol., 31 (1989) 215.

adaptive control; enzyme reactor; optimization

nije evidentirano