|
OPTIMAL SENSOR LOCATION FOR THE IDENTIFICATION OF MOVING HEAT SOURCES
L. Autrique(*), J.P. Leyris(*), N. Ramdani(**)
(*) IMP-CNRS, Tecnosud, rambla de la thermodynamique, 66100 Perpignan,
France, tel: +33 468 662 239, fax: +33 468 672 166, email: autrique@univ-perp.fr
(**) CERTES, Université Paris XII–Val de Marne, IUT Créteil, Avenue Général de
Gaulle, 94010 Créteil cedex, France email: ramdani@univ-paris12.fr
For many thermal processes described by partial differential equations, parametric identification of the model can be performed by solving an inverse problem. Measurements noises occurring on one or several measured characteristic states have to be carefully taken into account in order to provide robust identification procedures. However, spatial inaccuracy of the sensor location is rarely investigated. In this paper, a thermal process is presented in order to study the effect of this class of measurement noises on identification robustness. Then the problem of optimal location of the sensor is stated.
Keywords: Process parameter estimation, Partial differential equation, Sensitivity analysis, Optimal experiment design, Non-linear systems
Session slot T-Th-A02: Modeling of Industrial Processes/Area code 3a : Modelling, Identification and Signal Processing
 |
