Fuzzy Model-Based Fault Detection Approach for Nonlinear Control Processes

The key objective of this work is to deal with the fault detection issues for a general type of nonlinear control processes. Firstly, the approximation of the nonlinear process is studied based on a set of Takagi-Sugeno fuzzy systems. Then, a fuzzy controller based on an observer is present to achieve the stability of the nonlinear process, which also lays the basis for the construction of the image representations of the system and nominal system. With the image representations, the input and output signal errors are obtained, which can be used for the setting of the evaluation function. By handling a series of matrix inequalities, the threshold is designed, and a fault detection system for the nonlinear control process is realized with the decision logic. In the end, an illustration example is given to validate the obtained results.