Signal processing and fuzzy cluster based online fault diagnosis

The aim of this paper is to explain the application of signal processing to online fault diagnosis using fuzzy cluster. Wavelet transform is used as a signal processing. Wavelet transform characterizes the local regularity of signals by decomposing signals. In this study, fuzzy logic controller is used for three tank system control. Five component faults are examined on the system. For fault diagnosis firstly wavelet transform is applied. Secondly, fuzzy cluster is constructed using the detail coefficients. Detail coefficients contain the high frequency information are used to find faults. Fuzzy cluster gives a fault decision which occurs in the system. If fault is detected and identified, the system is stopped. As a result, fault diagnosis algorithm obtains safety of the system.