Using rhythmograms to diagnose mechanical systems

Abstract

Abstract The possibility of using a rhythmogram for the diagnosis of mechanical systems is discussed. The rhythmogram is obtained using the method we developed for processing quasiperiodic pulsed signals. The method is based on the principles of the theory of optimal filtration, the theory of wavelet transform and Hermite transform. The efficiency of the diagnostic method is demonstrated by the example of the operation of bearings of a gas turbine engine. A conclusion is given on the further development of the method for assessing the states of mechanical systems.