Rotary machine fault feature extraction method based on fractional order holographic principle

Abstract

The invention discloses a rotary machine fault feature extraction method based on a fractional order holographic principle. the rotary machine fault feature extraction method comprehensively utilizes the ability of fractional order Fourier transform for processing non-stationary signals and the ability of holospectrum for accurately reflecting the status of a machine set, and is applied to information extraction and fault diagnosis in the stopping process of a rotary machine. According to the method, the fractional order Fourier transform and the holospectrum are fused from data layers, a two-dimensional holospectrum and a three-dimensional holospectrum are reconstructed, and a holographic waterfall figure is applied to processing non-stationary information of cycling start and stopping; therefore, various feature parameters can be extracted for start and stopping signals, the application scope of the traditional holospectrum is greatly expanded, the traditional limitation of only carrying out Bode diagram analysis is broken through, the level and range of monitor and diagnosis of rotary machine faults are widened, and powerful support is provided for healthy and reliable operation of rotary machines.

Description

technical field [0001] The invention belongs to the field of fault diagnosis of mechanical equipment, and in particular relates to a method for extracting fault features of rotating machinery based on the principle of fractional Fourier transform-holographic spectrum. Background technique [0002] The start and stop of rotating machinery is an accident-prone stage during the operation of the unit. The unit has experienced an unsteady process in the process of starting and stopping, and the process of starting and stopping contains a wealth of information about the state of the unit, which cannot be obtained in the normal steady-state operation. At present, the unbalance information of rotary machinery is basically extracted from the steady state process, and the research on extracting unbalance information from the vibration signal of the start-stop process is relatively rare. How to make full use of the vibration information during the start-stop process, and analyze the B...

AI Technical Summary

Problems solved by technology

[0004] However, holographic spectrum technology also has some shortcomings: the main analysis object is the steady-state signal of fixed speed, it is difficult to analyze the speed fluctuation and the non-stationary signal in the process of speed up and down, which greatly limits the scope of application

In the past, the traditional order ratio analysis method had a complicated process when extracting the multiplier component, and required high accuracy for the calculation of the rotational speed, and it was difficult to determine the parameters such as the number of filter stages and bandwidth. Fractional Fourier transform can be more simple and effective Processing of Chirp Signals

Images