Prognosis of gear health using Gaussian process model

On-line condition monitoring of rotational machinery is a very important part of modern control and supervision system. Various methods are used for dealing with this issue. This paper describes application of Gaussian process model for the modelling of time series describing gear health and the prediction of the critical value of harmonic component feature that indicates the wear of gear. The Gaussian process model is an example of a flexible, probabilistic, nonparametric model with uncertainty predictions. It offers a range of advantages for modelling from data and has been therefore used also for dynamic systems identification and time-series modelling. The paper deals with the issue of covariance function selection for the harmonic component time-series modelling, time-series modelling itself and the assessment of different models for the prediction of the time when the harmonic component feature reaches critical value.