General Interpolated Fast Fourier Transform: A New Tool for Diagnosing Large Rotating Machinery

Abstract

Vibration monitoring and diagnosis of rotating machinery is an important part of a predictive maintenance program to reduce operating and maintenance costs. In order to improve the efficiency and accuracy of diagnosis, the general interpolated fast Fourier transform (GIFFT) is introduced in this paper. In comparison to present interpolated fast Fourier transform, this new approach can deal with any type of window functions and possesses high accuracy and robust performance, especially coping with a small number of sampling points. Then, for the purpose of rotating machinery diagnosis, the harmonic vibration ellipse and orbit is reconstructed based on the GIFFT to extract the features of faults and remove the interference from environmental noise and some irrelevant components. This novel scheme is proving to be very effective and reliable in diagnosing several types of malfunctions in gas turbines and compressors and characterizing of the transient behavior of rotating machinery in the run-up stage.