284 articles – 2152 references  [version française]
Detailed view Conference proceedings
Control-of-Industrial-Systems.-'Control-for-the-Future-of-the-Youth'.-Proceedings-volume-form-the-IFAC-Conference, Belfort : France (1997)
Linear and nonlinear modelling for fault detection and identification
E. Simeu1, T. Herve1

The input-output maps of nonlinear systems are approximated with linear in parameter polynomial different equations. In both the linear and nonlinear cases, simple least-square algorithms are used for reference nominal system characterisation and for actual system parameter estimation. An asymptotic approach for change detection based on parameter deviation monitoring via a data-to-model distance is presented.
1:  TIMA - Techniques of Informatics and Microelectronics for integrated systems Architecture
input-output-maps – nonlinear-systems – polynomial-different-equations – linear-model – nonlinear-model – parameter-estimation – fault-detection – parameter-deviation – data-to-model-distance – fault-diagnosis – identification- – least-square-algorithms