Intelligent Fault Diagnosis in Nonlinear Systems

Fault diagnosis in nonlinear systems is a challenging and very active research area. One of the difficulties to detect and isolate faults in nonlinear systems via observer-based methods is the design of a residual generator. In this work an integrated procedure combining conventional decoupling methods and Fuzzy Takagi-Sugeno observers for fault diagnosis in nonlinear systems is proposed. The advantage of the proposed AI based approach is that the design condition for the observers is relaxed in contrast with conventional approaches. Furthermore the potential use of decoupling techniques is reinforced. The design methodology is shown using a three tank system.