Fuzzy system identification and fault diagnosis of industrial processes

This paper proposes a method for fault diagnosis of dynamic processes using the multiple model approach. The technique presented concerns the identification of a non-linear dynamic system based on Takagi-Sugeno (TS) fuzzy models. It can be shown that any non-linear dynamic process can, in fact, be described as a composition of several TS models selected according to process operating conditions. In particular, this work addresses a method for the identification and the optimal selection of the local TS models from a sequence of noisy input-output data acquired from the process. The diagnostic scheme exploits the TS fuzzy models to generate residuals. The developed technique was applied to the fault diagnosis of the input-output sensors of an industrial gas turbine and the results are also presented.