Industrial Process Diagnosis based on Information-based Features

Abstract

Change detection and diagnosis based on information-based features is considered. As benchmark case study, the faults of bearings are selected. Information from the working states, including those generated by faults in bearings, are transformed and carried by the vibration’s signals, which are further processed by advanced techniques of information and signal processing, as e.g., statistical models, Renyi and Tsallis entropies, and other complexity measures.