In the papers published on Fault Detection and Isolation(FDI) using Bond Graph approach, robust FDI in the compressible flow systems has less been developed in the literature, due to the strong nonlinearities in this systems. In this paper, Bond Graph model in LFT form is developed for compressible systems. The results of this research are shown the causal properties of bond graph can be used to obtain the Analytical Redundancy Relations (ARRs) in presence of parameter uncertainties. The advantage of the bond graph model in LFT form is generation of ARRs and decoupling of the nominal part from the uncertain part. The nominal part can be used to calculate of the residual and the uncertain part can be used to obtain of adaptive thresholds and sensitivity analysis. In the following, the developed model is validated by pedagogical example.

Keywords: Bond Graph, Fault Detection and Isolation (FDI), Analytically Redundancy Relations (ARRs), Uncertainties, Compressible fluid.

Full-Text [PDF 1080 kb]   (3655 Downloads)