A LMI approach for designing robust FDI filters with guaranteed fault sensitivity performance

The problem of designing fault detection and isolation (FDI) filters is considered. The approach provides new techniques for designing filters that maximize fault sensitivity (effects of faults on the residual signal), while guaranteeing robustness to noise, disturbances and modeling errors. The design of the FDI filters is formulated in terms of linear matrix inequalities (LMIs) and involves the generalized structured singular value μ_g.