PARAMETER-ESTIMATION OP A MONOD-TYPE MODEL .2. USE OF A NEURAL-NETWORK WITH ASSOCIATIVE MEMORIES

On the basis of a previous study related with parameter identifiabilit y and sensitivity analysis of a Monod-type model, a parameter estimati on method based on Artificial Neural Networks (ANN) with Associative M emories (AMs) is presented. A combination of an iterative procedure an d a convergence index given by AMs allows to confirm the nature of rel ations existing between state variables and parameters which were foun d in the first part of the study. The convergence criterion is particu larly well adapted to showing various influences of state variables on parameter estimation of such a model.