Disturbance Distribution Matrix Computation: Numerial Improvement
| Authors: | Uppal Faisel, The University of Hull, United Kingdom
Lesecq Suzanne, Laboratoire d’Automatique de Grenoble, France
Patton Ron, The University of Hull, United Kingdom
Barraud Alain, Laboratoire d’Automatique de Grenoble, France |
|---|
| Topic: | 6.4 Safeprocess |
|---|
| Session: | Fault Diagnosis and Fault Tolerant Control: Theory |
|---|
| Keywords: | Fault Detection and Isolation (FDI), Neuro-fuzzy, multiple-model observer, least-squares (LS) minimisation, optimization |
|---|
Abstract
Prompt detection and diagnosis of process malfunctions are strategically important due to economic and environmental demands required for industries to remain competitive in world markets. In this paper a new formulation of the computation of the disturbance and fault distribution matrices is suggested for Neuro-Fuzzy and De-coupling Fault Diagnosis Scheme (NFDFDS). NFDFDS is a multiple-model fault detection and isolation (FDI) approach of non-linear dynamic systems. In this approach, powerful approximation and reasoning capabilities of neuro-fuzzy models are combined with the de-coupling capabilities of optimal observers to perform reliable fault detection and isolation. For determination of distribution matrices in this case it is shown that a least-squares approach is the most efficient compared with any other non-linear optimization technique.
