CN-121980158-A - Gear fault diagnosis method and system based on maximized CGGI deconvolution filtering

CN121980158ACN 121980158 ACN121980158 ACN 121980158ACN-121980158-A

Abstract

The invention discloses a gear fault diagnosis method and a gear fault diagnosis system based on maximized CGGI deconvolution filtering, which relate to the technical field of mechanical fault diagnosis and comprise the steps of collecting vibration signals of a gear transmission system; the method comprises the steps of performing generalized envelope spectrum transformation on a vibration signal to obtain a plurality of groups of generalized envelope spectrums, constructing weight coefficients based on the base indexes corresponding to each generalized envelope spectrum, carrying out weighted summation on the plurality of groups of generalized envelope spectrums to obtain a generalized weighted envelope spectrum, calculating generalized base indexes based on the generalized weighted envelope spectrums, rewriting the generalized base indexes into cyclic embedded generalized base indexes based on a cyclic embedded sparsity measurement method, taking the maximized cyclic embedded generalized base indexes as an objective function of blind deconvolution filtering, solving filter coefficients by adopting a eigenvector algorithm, carrying out deconvolution on the vibration signal, and outputting an optimal deconvolution filtering signal. The stability and consistency of the deconvolution result are improved, and a clearer characteristic expression basis is provided for gear fault identification.

Inventors

YANG JIANWEI
ZHU BIN
YAO DECHEN
Hu Zhongshuo
WANG JINHAI
ZENG HENG
WANG NING
Huo Jiyuan
LI XI
ZHANG SHENGLONG

Assignees

北京建筑大学

Dates

Publication Date: 20260505
Application Date: 20251201

Claims (10)

1. The gear fault diagnosis method based on the maximum CGGI deconvolution filtering is characterized by comprising the following steps of: Collecting vibration signals of a gear transmission system; performing generalized envelope spectrum transformation on the vibration signal to obtain a plurality of groups of generalized envelope spectrums, constructing weight coefficients based on the base-to-noise indexes corresponding to each generalized envelope spectrum, and performing weighted summation on the plurality of groups of generalized envelope spectrums to obtain generalized weighted envelope spectrums; calculating a generalized Raney index based on the generalized weighted envelope spectrum, and rewriting the generalized Raney index into a cyclic embedded generalized Raney index based on a cyclic embedded sparsity measurement method; and taking the maximized cyclic embedded generalized radix index as an objective function of blind deconvolution filtering, solving a filter coefficient by adopting a eigenvector algorithm, performing deconvolution on the vibration signal, and outputting an optimal deconvolution filtering signal.
2. The method for gear fault diagnosis based on the maximum CGGI deconvolution filter according to claim 1, wherein the obtaining a plurality of groups of generalized envelope spectra includes performing generalized envelope spectrum transformation within a set transformation parameter range based on a vibration signal, obtaining corresponding generalized envelope spectra by adjusting different transformation parameters, and forming a plurality of groups of generalized envelope spectra within the transformation parameter range.
3. The method for diagnosing gear faults based on the maximum CGGI deconvolution filtering of claim 2, wherein the obtaining of the generalized weighted envelope spectrum comprises the steps of respectively calculating the basis indexes corresponding to a plurality of groups of generalized envelope spectrum, determining corresponding weight coefficients according to the duty ratio of the basis indexes of each generalized envelope spectrum in all transformation parameter ranges, carrying out weighted summation on the plurality of groups of generalized envelope spectrum according to the weight coefficients, and taking the weighted summation result as the generalized weighted envelope spectrum.
4. The method for diagnosing gear faults based on the maximum CGGI deconvolution filter of claim 3, characterized in that calculating the generalized base index comprises sorting data in a generalized weighted envelope spectrum according to the magnitude of a signal amplitude, respectively constructing a nonlinear weighting function and a nonlinear weighting coefficient based on the sorted data, and carrying out weighted calculation on the sorted data through the nonlinear weighting function and the nonlinear weighting coefficient so as to obtain the corresponding generalized base index.
5. The method for diagnosing gear failure based on the inverse convolution filtering of the maximized CGGI as set forth in claim 4, wherein the rewriting the generalized base index into the cyclic embedded generalized base index includes performing equal-length segmentation processing on the generalized weighted envelope spectrum, calculating the corresponding generalized base index for each segmented signal, and performing power averaging processing on the generalized base index obtained for each segmented, with the power average result being the cyclic embedded generalized base index.
6. The method for diagnosing gear faults based on the maximized CGGI deconvolution filtering is characterized in that the method comprises the steps of initializing filter coefficients and calculating initial filtering signals, constructing a feature vector equation for solving the filter coefficients by adopting a feature vector algorithm, calculating the filter coefficients, increasing the cyclic embedded generalized base-ni index value corresponding to the filtering signals and obtaining the filtering signals, wherein the maximized cyclic embedded generalized base-ni index is used as an objective function of the blind deconvolution filtering; And (3) entering an iteration process of blind deconvolution, repeatedly calculating a filter coefficient and a filter signal, judging whether an optimal deconvolution filter result is achieved, and ending iteration and outputting the filter coefficient and the filter signal at the moment when the numerical value of a cyclic embedded generalized radix index and the characteristic value in the characteristic vector equation reach a convergent tolerance criterion or the iteration number reaches the maximum value.
7. The method for gear fault diagnosis based on maximum CGGI deconvolution filtering as claimed in claim 6, wherein the optimal deconvolution filtered signal is a filter coefficient and a filtered signal at the end of iteration.
8. The gear fault diagnosis system based on the maximized CGGI deconvolution filter is based on the gear fault diagnosis method based on the maximized CGGI deconvolution filter according to any one of claims 1 to 7, and is characterized in that: The data module is used for collecting vibration signals of the gear transmission system; The generalized envelope spectrum conversion module is used for carrying out generalized envelope spectrum conversion on the vibration signal to obtain a plurality of groups of generalized envelope spectrums, constructing weight coefficients based on the base indexes corresponding to the generalized envelope spectrums, and carrying out weighted summation on the plurality of groups of generalized envelope spectrums to obtain generalized weighted envelope spectrums; The generalized base Ni index module calculates generalized base Ni index based on generalized weighted envelope spectrum, and rewrites the generalized base Ni index into cyclic embedded generalized base Ni index based on cyclic embedded sparsity measurement method; And the objective function module takes the maximized cyclic embedded generalized radix index as an objective function of blind deconvolution filtering, adopts a eigenvector algorithm to solve the filter coefficient, performs deconvolution on the vibration signal, and outputs an optimal deconvolution filtering signal.
9. A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the computer program is characterized in that the processor executes the computer program to realize the steps of the gear fault diagnosis method based on the maximum CGGI deconvolution filtering as set forth in any one of claims 1 to 7.
10. A computer readable storage medium having a computer program stored thereon, wherein the computer program when executed by a processor performs the steps of the gear fault diagnosis method based on the maximized CGGI deconvolution filter as set forth in any one of claims 1 to 7.

Description

Gear fault diagnosis method and system based on maximized CGGI deconvolution filtering Technical Field The invention relates to the technical field of mechanical fault diagnosis, in particular to a gear fault diagnosis method and system based on maximized CGGI deconvolution filtering. Background The gear transmission system is used as a core component of mechanical equipment, is widely distributed in various key equipment, bears an important function of transmitting power, and the running state of the gear transmission system is directly related to the performance, efficiency and safety of the whole equipment. However, the gear is extremely easy to be damaged such as tooth surface abrasion, pitting corrosion, tooth breakage, cracks and the like under the complex working condition for a long time. Because of the special closed structure of the gear box, in order to avoid frequent disassembly of the structure, the health monitoring of the gear box mostly depends on modes such as vibration diagnosis, acoustic emission technology, temperature monitoring, oil analysis and the like. Wherein vibration signals are widely applied to fault diagnosis of mechanical parts of a transmission system by virtue of easy acquisition and high conformity with dynamic response. Due to factors such as variable working conditions, variable loads, interaction interference of wheel and rail and the like, gear fault characteristics are difficult to submerge by other vibration components and noise, and certain challenges are brought to fault diagnosis. The signal processing method is always a main stream means for denoising and decomposing vibration components, and has been widely applied to efficient extraction of multi-mode information in multiple fields. Including empirical mode decomposition, variational mode decomposition, singular value mode decomposition, and the like. However, the above method generally has problems of mode aliasing, a priori parameter requirements, and the like. Disclosure of Invention The present invention has been made in view of the above-described problems occurring in the prior art. Therefore, the gear fault diagnosis method and the gear fault diagnosis system based on the maximized CGGI deconvolution filtering solve the problems that the existing gear fault diagnosis method is insufficient in weak impact feature extraction capability under complex working conditions, is strong in dependence on parameter selection, is difficult to simultaneously consider signal sparsity and periodic structure features, and causes that fault features are easily submerged by noise and the stability of a filtering objective function is insufficient. In order to solve the technical problems, the invention provides the following technical scheme: In a first aspect, the invention provides a gear fault diagnosis method based on maximized CGGI deconvolution filtering, comprising the steps of collecting vibration signals of a gear transmission system; performing generalized envelope spectrum transformation on the vibration signal to obtain a plurality of groups of generalized envelope spectrums, constructing weight coefficients based on the base-to-noise indexes corresponding to each generalized envelope spectrum, and performing weighted summation on the plurality of groups of generalized envelope spectrums to obtain generalized weighted envelope spectrums; calculating a generalized Raney index based on the generalized weighted envelope spectrum, and rewriting the generalized Raney index into a cyclic embedded generalized Raney index based on a cyclic embedded sparsity measurement method; and taking the maximized cyclic embedded generalized radix index as an objective function of blind deconvolution filtering, solving a filter coefficient by adopting a eigenvector algorithm, performing deconvolution on the vibration signal, and outputting an optimal deconvolution filtering signal. The gear fault diagnosis method based on the maximum CGGI deconvolution filtering is characterized in that the obtaining of multiple groups of generalized envelope spectrums comprises the steps of performing generalized envelope spectrum transformation in a set transformation parameter range based on a vibration signal, obtaining corresponding generalized envelope spectrums by adjusting different transformation parameters, and forming multiple groups of generalized envelope spectrums in the transformation parameter range. The gear fault diagnosis method based on the maximum CGGI deconvolution filtering is characterized in that the obtaining of the generalized weighted envelope spectrum comprises the steps of respectively calculating the basis indexes corresponding to a plurality of groups of generalized envelope spectrum, determining corresponding weight coefficients according to the duty ratio of the basis indexes of each generalized envelope spectrum in all transformation parameter ranges, carrying out weighted summation on the plurality of groups of g