Specific Data Sampling and Filtering Helps to Detect and Isolate Periodic Disturbances

Abstract

The Fast Fourier Transform is the unique tool enabling to engage the frequency domain analysis in detection and isolation of periodic disturbances in industrial processes containing rotating elements. Nevertheless, there exist particular problems where the time-domain examination of oscillations can provide equally or more accurate results with less effort. Specific data sampling, filtering and process modelling are introduced in the paper with the aim to classify sources of oscillations in the process of cold rolling. Although the FFT may not be part of the method itself, its use allows to illustrate the presented research.