Correlation-based gear signal noise reduction method combining EMD and morphological singular value decomposition

Abstract

The invention discloses a correlation-based gear signal noise reduction method combining EMD and morphological singular value decomposition. Firstly, the original signal is EMD decomposed to obtain several IMF components, and the noise dominant components in the IMF component are screened according to the statistical characteristics of the white noise according to the autocorrelation function, andthe remaining IMF components are divided into useful components and spurious components according to the mutual relation number method. Then the singular value denoising of the noise dominant component is performed and then reconstructed with the useful component. Finally, the morphological filtering is used to extract the gear defect characteristic signal. The method solves the problem that theloss of useful components of the signal after denoising and the defect characteristics are not easily recognized by the modal aliasing phenomenon in the EMD decomposition, which highlights the defectcharacteristics of the original vibration signal and has a strong self-detail adaptability, while also improve the reliability and accuracy of gear defect information extraction.

Description

technical field [0001] The invention relates to the field of signal processing, in particular to a signal noise reduction method for extracting gear defect features. Background technique [0002] As an important part of the mechanical system, gears are one of the parts with a high defect rate. Early detection and investigation of early defects of gears is very important for the maintenance of equipment performance, maintenance of precision, and reduction of mechanical accidents. [0003] The early defect signals of gears are often very weak, strongly disturbed by noise and low in signal-to-noise ratio, which has caused great obstacles to the identification of gear defects. The essential. [0004] At present, the most widely used noise reduction method for gear defect signals is an adaptive signal processing method—EMD method, which can decompose the original signal into the sum of several IMF components and a remainder of local characteristic signals of different time scale...

AI Technical Summary

Problems solved by technology

[0005] The technical problem to be solved by the present invention is to provide a gear signal noise reduction method based on correlation EMD and morphological singular value decomposition for the deficiencies in the prior art; this method not only solves the problem of gear defect signals In the EMD method, the useful components of the signal after denoising are lost and the defect features are not easy to identify due to the modal aliasing phenomenon, and the defect features of the original vibration signal are highlighted, which has strong adaptability and also improves the gear Reliability and accuracy of defect information extraction

Images