Binczak, Stéphane (Univ. of Burgundy), Busvelle, Eric (Univ. of Burgundy), Gauthier, Jean-Paul (Univ. of Burgundy), Jacquir, Sabir (Univ. of Burgundy)

Identification of Unknown Functions in Dynamic Systems

Scheduled for presentation during the Regular Session "Nonlinear Estimation and Identification" (ThP09), Thursday, July 27, 2006, 16:10−16:35, Room J

17th International Symposium on Mathematical Theory of Networks and Systems, July 24-28, 2006, Kyoto, Japan

This information is tentative and subject to change. Compiled on May 19, 2024

Abstract

We consider the problem of representing a complex process by a simple model, in order to perform advanced control for instance. In many cases, the main dynamic of the process is well known and some knowledge-based equations can be written, but some parts of the process are unknown. In this paper, we will present one such application, but we have already encountered many other situations of this kind. Depending on measurements and control variables, it is sometimes possible to identify the unknown part of the model and the unmeasured state variables. We will briefly recall some theoretical results concerning this problem, and we will also present a general methodology to perform this identification. Then we will explain more deeply how we apply this approach to an electronic circuit representing a neuron. We will estimate effectively the unknown function from actual measurements.