Publication Details

Moodie, C., Tieu, A. K., Biddle, S. (2007). De-convolution of time-series data using translational symmetry and eigenstate analysis. 14th International Congress on Sound and Vibration (ICSV14) (pp. 1-8). Australia: International Congress on Sound and Vibration.


Translational Symmetry permits a time invariant matrix (a Ring Matrix) of a time series to be formed. Traditional Eigenvalue/vector and Fast Fourier Transform techniques applied to the Ring Matrix provide views of all the vibration modes, their magnitude and frequency content. The Ring Matrix has many interesting features associated with it that enable it to be used as an additional tool in defining the ‘condition state’ of a bearing. This paper introduces the Ring Matrix as a significant step in the ‘Blind’ deconvolution process. Application of the de-convolution method to vibration data obtained from both a large and small slow speed (1 to 4rpm) slew bearing is presented. The Ring Matrix is being applied to the condition monitoring of a bearing which forms part of an experimental test rig specifically built to monitor slow speed (1 rpm) slew bearings with a view to predicting their remaining useful life. Results, from a Coal Reclaimer which has slow speed (4rpm) slew bearings, are also presented to illustrate the various views of the vibration information.

Link to publisher version (URL)

International Congress on Sound and Vibration

Included in

Engineering Commons