Fault Diagnosis Method for a Class of Rotating Machinery System Based on Adaptive Noise Reduction Algorithm

Abstract

The invention aims at providing an adaptive-noise-reduction-algorithm-based same-class rotary machinery system fault diagnosis method that is mainly used for separating a periodic signal from an impact and noise signal, carrying out simulation again, and extracting a signal parameter capable of reflecting a machine operating situation precisely for follow-up analyses. In order to achieve the objective, the provided method is characterized in that vibration signals are divided into two parts based on an adaptive filter algorithm, wherein the first part signals are periodic signals and the second part signals are pulse signals and noise signals; the second part signals are processed based on a noise reduction algorithm to obtain the noise signals by separation, thereby improving the signal to noise ratio of the pulse signal; characteristic extraction is carried out; and simulation is carried out based on the data after characteristic extraction to establish a vibration signal model. According to the technical scheme, the pulses signals are separated from the periodic signals and the influence on the pulse signals by the noise signals can be reduced, so that the fault analysis can be carried out accurately.

Description

technical field [0001] The invention relates to a mechanical system fault diagnosis method, in particular to a type of rotating mechanical system fault diagnosis method based on an adaptive noise reduction algorithm. Background technique [0002] The structure of rotating machinery is complex, it has high requirements on operating conditions, and it has been in high-speed operation state for a long time, and in the normal operation process, even if there is a hidden danger of aura failure, it is difficult to be detected. But once a fault breaks out, it will bring inestimable losses. If the rotating machinery is stopped for regular maintenance, although accidents can be prevented, the efficiency of production and work is correspondingly reduced. [0003] In view of such problems, the invention patent No. 201310223686.6 "Angle Domain Resampling Method of Rotating Machinery Vibration Signal Based on Instantaneous Frequency Estimation" was designed in the prior art. This patent...

AI Technical Summary

Problems solved by technology

[0004] However, after careful analysis, it can be found that when the comparison file collects the vibration signal of the rotating machinery, due to the influence of the environment and the effect of noise, it cannot ensure that the accurate vibration signal is collected

However, analysis based on inaccurate vibration signals cannot achieve accurate prediction results.

Images