 |
    |
Biomass Inferential Sensor Based on Ensemble of Models Generated by Genetic Programming
Arthur Kordon*1, Elsa Jordaan2, Lawrence Chew3, Guido Smits2, Torben Bruck3, Keith Haney3, and Annika Jenings3
1The Dow Chemical Company, Corporate R&D, 2301 N Brazosport Blvd., Freeport, TX, 77541, USA
2Dow Benelux BV, Corporate R&D, 5 Herbert H Dowweg, Terneuzen, The Netherlands
3The Dow Chemical Company, Biotechnology, 5501 Oberlin Dr., San Diego, CA 92121, USA
Abstract. A successful industrial application of a novel type biomass estimator based on Genetic Programming (GP) is described in the paper. The biomass is inferred from other available measurements via an ensemble of nonlinear functions, generated by GP. The models are selected on the Pareto front of performance-complexity plane. The advantages of the proposed inferential sensor are: direct implementation into almost any process control system, rudimentary self-assessment capabilities, better robustness toward batch variations, and more effective maintenance. The biomass inferential sensor has been applied in high cell density microbial fermentations at The Dow Chemical Company.
*Corresponding author, akordon@dow.com.
LNCS 3103, p. 1078 ff.
lncs@springer.de
© Springer-Verlag Berlin Heidelberg 2004