Fault Detection and Isolation in Presence of Input and Output Uncertainties Using Bond Graph Approach
Résumé
In this paper, a bond graph model based approach for robust diagnosis in presence of input and output uncertainties is presented. Based on the structural and causal proprieties of the bond graph tool, a procedure of input and output uncertainties modeling is proposed in order to generalize the threshold generation. The proposed procedure is applied to the graphical model in preferred derivative causality, used for analytical redundancy relations. Simulation results are presented in order to validate the proposed procedure of thresholds and residuals generation.