Study of More Efficient Vibration Signal Processing Techniques for Anomalies Detection

The proposed research is aimed to study the most suitable method for a clearer interpretation and estimation of transient phenomena detectable in vibration signals of rotating machinery, when characterized by anomalies. The signals, which are going to be considered here, pertain industrial applications (i.e., machinery and turbo-machinery monitored status and usury). The paper explores the application of time-frequency based analysis techniques for the detection and feature extraction from the change of vibration signals caused by common machinery faults. The purpose here is to provide a more suitable criterion to detect anomalies, through the analysis of vibrational signals, on the basis of the kind of information from the signal itself.