Nonparametric estimation and filtering of uncertain nonlinear processes application to a nitrification process

The aim of this paper is to present some contributions of a nonparametric statistical approach to the identification of uncertain dynamical systems. This method is particularly well adapted to biotechnological processes. We developed a semi-parametric filtering algorithm, able to estimate both non measured state variables and functional kinetic parameters without any a priori assumption on the modelling of these parameters. This new technique is described using the example of a nitrification process.