Switching nonlinear systems are a particular class of hybrid systems, they are described by several nonlinear modes. If the system's modes are assumed to represent healthy or faulty operation modes, the active mode recognition task can be seen as a fault detection task. This paper deals the problem of faulty mode detection in SISO nonlinear switching system represented by Takagi-Sugeno models in two steps. The first step use an extension of the parity space method for detection of the faulty mode. Identification of local parameters of the multimodel that can represent the faulty mode is performed using RLS method to determinate the original of the fault in the second step.
展开▼
