Principal component ANN for modelling and control of baker's yeast production

Modelling of baker's yeast production by the principal component based arti ficial neural networks (ANN) is presented. The models are derived for their application in adaptive control of fermentation by the internal model cont rol (IMC) method. Modelling data are from industrial production in 40 m(3) deep jet bioreactor and from computer simulations. The modelling effort is focused on selection of ANN structure and model verification. Principal com ponent analysis of process variables results in projection of patterns to a space of low dimension, which enables determination of ANN structure, remo ves data colinearity and random components of measurement signals, and mode l degradation by over-training is eliminated. In view of IMC application, t he models for prediction of the controlled variable (ethanol partial pressu re) and the inverse model for manipulative variable (molasses feed rate) ar e determined. The models are tested for their predictability in the time ho rizon from 1 to 20 min. ANN models are derived with average relative errors for untrained patterns in the range from 1 to 10%. (C) 1998 Elsevier Scien ce B.V. All rights reserved.