Maximum likelihood parameter estimation of a hybrid neural-classical structure for the simulation of bioprocesses

This paper proposes a hybrid structure for the modeling of a bioprocess: cl assical (in the form of a priori knowledge describing the mass balances) an d neural (a radial basis function network describing the nonlinear reaction s kinetics within these mass balances). The aim is to build a continuous si mulator capable to reconstruct from initial conditions the trajectory of st ate variables (i.e. the main component concentrations) by considering also an aspect which usually is not taken into account in bioprocess modeling: t he existence of important measurement errors. A clustering strategy is used for placing the Gaussian centers and a maximum likelihood cost function is defined for the estimation of the network weights and initial conditions f or the simulator The structure is tested on batch animal cell cultures for which rare and asynchronous measurements are available: glucose, glutamine, lactate and biomass concentrations. (C) 2000 IMACS/Elsevier Science B.V. A ll rights reserved.