Rolling bearing fault diagnosis method, system and equipment based on contig, and medium

Abstract

The invention provides a rolling bearing fault diagnosis method, system and device based on a contig, and a medium, and the method comprises the steps: obtaining a vibration signal, and constructing a reweighted contig sparse shrinkage model through frequency slice wavelet transform based on a frequency slice function; then establishing a mathematical model framework taking harmonic spectrum kurtosis HSK and related Thire index CTI as double optimization objectives, and carrying out collaborative optimization on multiple parameters by using a multi-objective particle swarm optimization (MOPSO) algorithm to obtain optimal parameters; and finally, the optimal parameter is substituted into a model to strengthen a frequency domain resonance frequency band, an envelope demodulation spectrum is obtained through calculation, and fault feature frequency is extracted. The method can effectively separate fault impact components and noise interference, enhances fault feature sparse representation capability, improves diagnosis precision through dual-objective optimization, is suitable for fault diagnosis of rolling bearings under strong noise and complex working conditions, and has the advantages of high efficiency, accuracy, high adaptability and the like.

Description

Technical Field

[0001] This document relates to the technical field of rolling bearing fault diagnosis, and in particular to a rolling bearing fault diagnosis method, system, equipment and medium based on overlapping groups. Background Art

[0002] In the field of industrial equipment operation and maintenance, condition monitoring and fault diagnosis technologies are of strategic significance for safeguarding corporate economic benefits and the safety of life and property. As a core functional component of rotating machinery, the reliability of rolling bearings directly impacts the operational stability of the entire system. Due to the widespread strong coupling characteristics of modern industrial equipment, rolling bearing failures can cause systemic reliability degradation, leading to significant safety hazards. This reality places higher demands on the reliability and safety of industrial systems. Therefore, establishing effective early fault detection and anomaly identification mechanisms is crucial for preventing equipment performance degradation and preventing hazardous operating conditions.

[0003] Sparse representation, as a typical method of data dimensionality reduction, effectively extracts the essential features of data through sparse coding of feature space. The core principle of this method is to construct an over-complete dictionary and use a linear combination of a small number of atomic bases and corresponding sparse coefficients to reconstruct the signal, thereby suppressing redundant data information. With its efficient computing characteristics and feature decoupling capabilities, this theory has been successfully applied to many fields such as signal processing, computer vision, and deep learning. In order to further promote sparsity on the basis of sparse representation theory, Chen proposed an overlapping group shrinkage algorithm to denoise signals with clustering properties. It does not need to construct an over-complete dictionary to efficiently achieve feature enhancement, and has achieved excellent results in the field of data processing.

[0004] Ren et al. have developed a sparse representation method based on the identity matrix. By employing a standard orthogonal basis as the sparse representation basis function, this method not only avoids the subjective bias introduced by pre-set dictionary construction but also significantly reduces the computational dimension of the Majorization-Minimization optimization algorithm. Sun et al. have constructed a differentially constrained sparse regularization model, simultaneously imposing sparsity constraints on both the signal spatial domain and its differential domain. This dual-domain coupling mechanism effectively preserves the temporal singularity of the impulse component while suppressing Gaussian noise interference through band correlation. Some have proposed dictionaries based on linear transformation bases, such as the wavelet transform framework and the tunable Q-factor wavelet transform. To further enhance sparsity based on sparse representation theory, Chen et al. proposed an overlapping group shrinkage algorithm for denoising signals with clustering properties, achieving excellent results. He et al. proposed a method for estimating periodic group sparse signals in noise and applied it to vehicle bearing fault diagnosis. Wang et al. used fractional-order spline wavelet transform and overlapping group shrinkage in conjunction with non-convex regularization and convex optimization to diagnose bearing faults. Liu proposed a new impact force identification method based on the sparsity of non-convex overlapping groups, which can locate the impact and restore its time history under limited measurement conditions. Considering the problem that group penalty will lead to signal amplitude compression, Liu proposed a reweighted overlapping group shrinkage method to improve the signal-to-noise ratio of the time domain signal. Liu proposed a convex sparse regularization method in the ADMM framework to reconstruct and locate unknown impact forces and verified it on aircraft composite laminates. Li proposed a reweighted periodic overlapping group lasso method, which successfully extracted the fault characteristics of spiral bevel gears in a high-noise environment. However, the above-mentioned overlapping group sparse methods are mainly used in time domain signal processing and have not been fully studied in the frequency domain with richer information. The problem of extending this method to the frequency domain deserves further study. Summary of the Invention

[0005] According to an embodiment of the present invention, a rolling bearing fault diagnosis method, system, device and medium based on overlapping groups are provided to solve the above problems.

[0006] According to an embodiment of the present invention, a rolling bearing fault diagnosis method based on overlapping groups is provided, which is characterized by comprising:

[0007] S1. Obtain vibration signals and construct a reweighted overlapping group sparse shrinkage model through frequency slicing wavelet transform based on frequency slicing function;

[0008] S2. Establish a mathematical model framework with the harmonic spectrum kurtosis HSK and the correlation Theil index CTI as dual optimization objectives, and use the multi-objective particle swarm optimization algorithm MOPSO to implement collaborative optimization of multiple parameters of the heavily weighted overlapping group sparse shrinkage modulus to obtain the optimal multiple parameters;

[0009] S3. Bring the optimal multi-parameters into the reweighted overlapping group sparse shrinkage model to strengthen the frequency domain resonance band, and calculate to obtain the envelope demodulation spectrum, and then extract the fault characteristic frequency.

[0010] According to an embodiment of the present invention, a rolling bearing fault diagnosis system based on overlapping groups is provided, which is characterized by comprising:

[0011] A model building module is used to obtain vibration signals and construct a reweighted overlapping group sparse shrinkage model through frequency slicing wavelet transform based on frequency slicing function;

[0012] The parameter acquisition module is used to establish a mathematical model framework with the harmonic spectrum kurtosis HSK and the correlation Theil index CTI as dual optimization objectives. The multi-objective particle swarm optimization algorithm MOPSO is used to implement collaborative optimization of multiple parameters of the heavily weighted overlapping group sparse contraction module to obtain the optimal multiple parameters.

[0013] The fault diagnosis module is used to bring the optimal multi-parameters into the reweighted overlapping group sparse shrinkage model to strengthen the frequency domain resonance band, calculate the envelope demodulation spectrum, and then extract the fault characteristic frequency.

[0014] According to an embodiment of the present invention, there is provided an electronic device, including:

[0015] processor; and,

[0016] The memory is arranged to store computer executable instructions, which, when executed, cause the processor to perform the steps of the rolling bearing fault diagnosis method based on contigs.

[0017] According to an embodiment of the present invention, a storage medium is provided for storing computer-executable instructions. When the computer-executable instructions are executed, the steps of the rolling bearing fault diagnosis method based on overlapping groups are implemented.

[0018] This application proposes a rolling bearing fault diagnosis method based on an overlapping group sparse contraction model. A reweighted overlapping group sparse contraction model is constructed through frequency slicing wavelet transform, and harmonic spectral kurtosis (HSK) and correlated Theil index (CTI) are used as dual optimization targets. The model parameters are collaboratively optimized in combination with the multi-objective particle swarm optimization algorithm (MOPSO). This method can effectively separate the fault impact component from the noise interference, significantly improve the signal-to-noise ratio of the frequency domain resonance band, and thus accurately extract the fault characteristic frequency under a strong noise background. Experimental results show that this method has excellent diagnostic performance under complex working conditions, can achieve accurate identification of early weak faults, and provides reliable technical support for the intelligent operation and maintenance of industrial equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate one or more embodiments of this specification or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments recorded in this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0020] Figure 1 Flowchart of a rolling bearing fault diagnosis method based on overlapping groups according to an embodiment of the present invention;

[0021] Figure 2 Schematic diagram of the waveform and spectrum of the vibration signal according to an embodiment of the present invention;

[0022] Figure 3 Schematic diagram of the envelope spectrum of a vibration signal according to an embodiment of the present invention;

[0023] Figure 4 Schematic diagram of the joint frequency domain results of vibration signals according to an embodiment of the present invention;

[0024] Figure 5 Schematic diagram of an optimal parameter selection strategy for a vibration signal according to an embodiment of the present invention;

[0025] Figure 6 Schematic diagram of the waveform and spectrum of the vibration signal processing result according to an embodiment of the present invention;

[0026] Figure 7 Schematic diagram of the envelope spectrum of the vibration signal processing result according to an embodiment of the present invention;

[0027] Figure 8 Schematic diagram of a rolling bearing fault diagnosis system based on overlapping groups according to an embodiment of the present invention. DETAILED DESCRIPTION

[0028] In order to enable those skilled in the art to better understand the technical solutions in one or more embodiments of this specification, the technical solutions in one or more embodiments of this specification will be clearly and completely described below in conjunction with the drawings in one or more embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on one or more embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of this document.

[0029] Method Example

[0030] According to an embodiment of the present invention, a rolling bearing fault diagnosis method based on overlapping groups is provided. Figure 1This is a flow chart of a rolling bearing fault diagnosis method based on overlapping groups according to an embodiment of the present invention. Figure 1 As shown, the rolling bearing fault diagnosis method based on overlapping groups according to an embodiment of the present invention includes:

[0031] S1. Obtain a vibration signal and construct a reweighted overlapping group sparse shrinkage model by frequency slicing wavelet transform based on a frequency slicing function; S1 specifically includes:

[0032] Obtain the vibration signal of the rolling bearing and decompose it into the linear superposition of the fault impact component and the noise interference;

[0033] Perform joint frequency domain analysis on the signal through frequency slicing wavelet transform and extract key frequency domain features using the preset frequency slicing function;

[0034] A reweighted overlapping group sparse shrinkage model is constructed. By introducing a dynamically adjusted weight matrix and regularization constraints, the sparse representation capability of fault features is enhanced while ensuring the accuracy of signal reconstruction.

[0035] Assume that the measured signal is y(t)∈R N , where N is the signal length. The observed signal can be decomposed into the fault impulse component y0(t)∈R N Interference with noise n(t)∈R N Linear superposition of .

[0036] y(t)=y0(t)+n(t);

[0037] Assume that the square integrable signal y(t)∈L 2 (R), and assume that its frequency slicing function p(t) satisfies p(t)∈L 1 (R)∩L 2 (R), then its Fourier transform exists and is continuous. Therefore, the frequency domain mathematical representation of FSWT can be strictly defined as:

[0038]

[0039] where the energy coefficient λ∈R + The scale factor σ≠0 can be defined as a constant parameter, or constructed as a functional mapping relationship between the observation frequency ω, the observation parameter t, and the evaluation frequency u. In particular, is the frequency domain form of FSF, and the superscript * indicates the complex conjugate operation. According to the basic computability conditions that the function design needs to meet and the systematic study of the time-bandwidth product (Δ t Δ ω ), determine the frequency slicing function as:

[0040]

[0041] Under complex working conditions such as high speed and heavy load, the mathematical model of the reweighted overlapping group sparse shrinkage (ROGSS) model can be expressed as:

[0042]

[0043] Where Y is the original measured signal after FSWT transformation, λ>0 is the regularization parameter that balances fidelity and sparsity, and ||·||2 is the l2norm. P(x) is the penalty function that promotes sparsity. It can be expressed as:

[0044]

[0045] On this basis, a reweighted diagonal matrix w is constructed, which contains w=[w1,…,w N+K-1 ],w i Can be defined as:

[0046]

[0047] Where ε is a small positive constant used to avoid division by zero, and the sparse solution performance is enhanced by constructing a reweighted diagonal matrix w.

[0048] S2. Establish a mathematical model framework with the harmonic spectrum kurtosis (HSK) and the correlation Theil index (CTI) as dual optimization objectives, and use the multi-objective particle swarm optimization algorithm (MOPSO) to collaboratively optimize the multiple parameters of the heavily weighted overlapping group sparse contraction modulus to obtain the optimal multiple parameters. S2 specifically includes:

[0049] The harmonic spectrum kurtosis (HSK) and the correlation Theil index (CTI) are defined as dual-objective evaluation indicators to quantify the significance and distribution characteristics of fault features respectively.

[0050] By establishing a multi-objective optimization framework, the key parameters of group size and regularization coefficient are used as decision variables, and a multi-objective particle swarm algorithm is used for collaborative optimization. The optimal parameter combination is iteratively searched as the optimal multi-parameter under preset constraints.

[0051] The kurtosis of the harmonic spectrum is defined as follows:

[0052]

[0053] The mathematical expression of the Theil index follows the standard entropy form in information theory:

[0054]

[0055] On this basis, the relevant Theil index is defined as follows:

[0056]

[0057] Where N is the length of the autocorrelation sequence of the signal segment, is the corresponding point in the autocorrelation sequence, and μ is the average value of the signal autocorrelation sequence.

[0058] When the observed signal contains significant fault features, the response values ​​of HSK and CTI will increase by orders of magnitude. Based on the minimization modeling criterion, the parameter vector is defined as This establishes a multi-objective optimization framework:

[0059] minm(v)=[m1(v),m2(v)];

[0060] The objective function is defined as:

[0061] m1(v)=1 / HSK; m2(v)=1 / CTI;

[0062] subject to:

[0063] K1∈Z + ,K2∈Z + ,0<α<1,λ>0;

[0064] The parameter selection of the multi-objective particle swarm is as follows: population size is 50, archive size is 20, number of iterations is 20, parameter constraints are 1≤K1≤20, 1≤K2≤6, 1≤λ≤20, 0<α<1.

[0065] S3: Bring the optimal multi-parameters into the reweighted overlapping group sparse shrinkage model to strengthen the frequency domain resonance band, and calculate the envelope demodulation spectrum to extract the fault characteristic frequency. S3 specifically includes:

[0066] Substituting the optimal multi-parameters into the reweighted overlapping group sparse shrinkage model, the resonant frequency band in the signal frequency domain is selectively enhanced while irrelevant noise is suppressed.

[0067] After reconstructing the time domain signal through inverse transformation, the envelope spectrum is extracted using envelope demodulation technology, and finally the characteristic frequency and type of rolling bearing fault are identified to achieve fault diagnosis.

[0068] The parameter set obtained from the previous optimization is brought into the ROGSS model. The frequency domain resonance band enhancement strategy is adopted to implement energy enhancement on the resonance frequency band area in the signal spectrum. At the same time, noise suppression is performed on the non-fault related frequency bands, thus significantly purifying the fault characteristics and improving the signal-to-noise ratio.

[0069] The ROGSS-processed signal is reconstructed back to the time domain using the following formula:

[0070]

[0071] Then, the envelope spectrum is obtained through Hilbert envelope demodulation and the fault characteristic frequency is extracted:

[0072]

[0073] Specifically, in a specific implementation of the present invention, the motor speed is 3000 rpm and the sampling frequency is f s =16384Hz, the number of sampling points is N=16384, after calculation, the characteristic frequency of the inner ring fault of the bearing is f i =269Hz, the characteristic frequency of the outer race fault is f o =179Hz. When collecting vibration signals, the rolling bearing with an outer ring fault is first mounted on a bearing test bench and driven by a motor. A data logger is then used to collect test data, which is then transferred to a computer for subsequent data processing and analysis using MATLAB software.

[0074] Perform Fourier transform on the acquired signal. Figure 2 The waveform and spectrum of the acquired fault signal are shown in Figure 2. The vibration signal is interfered by strong background noise, which significantly masks the periodic impulse response in the time domain waveform. This makes it difficult to effectively extract fault features using conventional spectra.

[0075] By using the envelope demodulation method to process the signal spectrum, the envelope spectrum of the signal is obtained, such as Figure 3 As shown, envelope demodulation spectrum analysis shows that f o =179Hz is identifiable, but the high-order harmonic components are still suppressed by noise and cannot be fully revealed. Therefore, the rolling bearing fault diagnosis method based on overlapping groups proposed in an embodiment of the present invention is used to diagnose this vibration signal.

[0076] The present invention will be further described below with reference to the accompanying drawings and examples.

[0077] The specific implementation process of the rolling bearing outer ring fault diagnosis method provided by the present invention is as follows:

[0078] Input such as Figure 2 For the vibration signal shown in the figure, set the frequency slicing wavelet transform parameters, set the time domain resolution to 10, the frequency sampling points to 8197, and the time-frequency factor to 28.87 to obtain the following: Figure 4 The definition of frequency slicing wavelet transform and the choice of frequency slicing function are as follows:

[0079]

[0080]

[0081] Then the ROGSS model was established, and its mathematical model can be expressed as:

[0082]

[0083] On this basis, a reweighted diagonal matrix w is constructed, which contains w=[w1,…,w N+K-1 ],w i Can be defined as:

[0084]

[0085] A mathematical model framework with HSK and CTI indicators as dual optimization objectives is established, and the multi-objective particle swarm optimization algorithm MOPSO is used to implement collaborative optimization of multiple parameters of ROGSS Model.

[0086] The kurtosis of the harmonic spectrum is defined as follows:

[0087]

[0088] The mathematical expression of the Theil index follows the standard entropy form in information theory:

[0089]

[0090] On this basis, the relevant Theil index is defined as follows:

[0091]

[0092] Based on the minimization modeling criterion, define the parameter vector This establishes a multi-objective optimization framework:

[0093] minm(υ)=[m1(v),m2(v)];

[0094] The objective function is defined as:

[0095] m1(v)=1 / HSK; m2(v)=1 / CTI;

[0096] subject to:

[0097] K1∈Z + ,K2∈Z + ,0<α<1,λ>0;

[0098] The parameter selection of the multi-objective particle swarm is as follows: population size is 50, archive size is 20, number of iterations is 20, parameter constraints are 1≤K1≤20, 1≤K2≤6, 1≤λ≤20, 0<α<1, and the results are as follows Figure 5 shown.

[0099] The parameter set obtained from the previous optimization is brought into the ROGSS model. The frequency domain resonance band enhancement strategy is adopted to implement energy enhancement on the resonance frequency band area in the signal spectrum. At the same time, noise suppression is performed on the non-fault related frequency bands, thus significantly purifying the fault characteristics and improving the signal-to-noise ratio.

[0100] The signal after ROGSS processing is reconstructed back to the time domain using the following formula: Figure 6 As shown:

[0101]

[0102] Then the envelope spectrum is obtained by Hilbert envelope demodulation and the fault characteristic frequency is extracted. The results are as follows: Figure 7 As shown:

[0103]

[0104] Therefore, the present invention can enhance the early weak fault characteristics by globally optimizing and denoising the full-band signal using the ROGSS model.

[0105] The following beneficial effects are achieved by adopting the embodiments of the present invention:

[0106] 1. Efficient fault feature extraction capability. Through frequency slicing wavelet transform and reweighted overlapping group sparse shrinkage model, it can effectively separate fault impact components from noise interference, significantly improve the sparse representation capability of fault features, and thus accurately extract weak fault signals in a strong noise background.

[0107] 2. Dual-objective optimization improves diagnostic accuracy. Harmonic spectral kurtosis (HSK) and correlated Theil index (CTI) are used as dual optimization objectives, combined with the multi-objective particle swarm optimization algorithm (MOPSO) to achieve collaborative optimization of key model parameters, ensuring that the significance and distribution characteristics of fault features are optimized simultaneously, further improving the accuracy and robustness of diagnosis.

[0108] 3. Frequency domain resonance band enhancement: Through targeted enhancement and noise suppression of the frequency domain resonance band, the signal-to-noise ratio of the signal is significantly improved, making the fault characteristic frequency more clearly identifiable in the envelope demodulation spectrum, which is especially suitable for the diagnosis of early weak faults.

[0109] 4. Adaptability to complex working conditions. This method performs well under complex working conditions such as high speed and heavy load. It can effectively cope with the strong coupling and high noise environments commonly found in industrial equipment, providing strong technical support for the reliability monitoring of rolling bearings.

[0110] 5. Automation and intelligence: Through modular design and automated parameter optimization, manual intervention is reduced, subjective bias is lowered, and the fault diagnosis process is made more efficient and reliable, making it suitable for large-scale industrial applications.

[0111] 6. Wide applicability: It is not only suitable for the fault diagnosis of the inner and outer rings of rolling bearings, but can also be extended to the fault detection of other rotating mechanical components. It has broad application prospects and promotion value.

[0112] System Example

[0113] According to an embodiment of the present invention, a rolling bearing fault diagnosis system based on overlapping groups is provided. Figure 8 Schematic diagram of a rolling bearing fault diagnosis system based on overlapping groups according to an embodiment of the present invention. Figure 8 As shown, the rolling bearing fault diagnosis system based on overlapping groups according to an embodiment of the present invention includes:

[0114] A model building module 80 is used to obtain a vibration signal and build a reweighted overlapping group sparse shrinkage model through frequency slicing wavelet transform based on a frequency slicing function;

[0115] The model building module 80 is specifically used to:

[0116] Obtain the vibration signal of the rolling bearing and decompose it into the linear superposition of the fault impact component and the noise interference;

[0117] Perform joint frequency domain analysis on the signal through frequency slicing wavelet transform and extract key frequency domain features using the preset frequency slicing function;

[0118] A reweighted overlapping group sparse shrinkage model is constructed. By introducing a dynamically adjusted weight matrix and regularization constraints, the sparse representation capability of fault features is enhanced while ensuring the accuracy of signal reconstruction.

[0119] The parameter acquisition module 82 is used to establish a mathematical model framework with the harmonic spectrum kurtosis HSK and the correlation Theil index CTI as dual optimization objectives, and adopt the multi-objective particle swarm optimization algorithm MOPSO to implement collaborative optimization on multiple parameters of the re-weighted overlapping group sparse shrinkage module to obtain the optimal multiple parameters;

[0120] The parameter acquisition module 82 is specifically used for:

[0121] The harmonic spectrum kurtosis (HSK) and the correlation Theil index (CTI) are defined as dual-objective evaluation indicators to quantify the significance and distribution characteristics of fault features respectively.

[0122] By establishing a multi-objective optimization framework, the key parameters of group size and regularization coefficient are used as decision variables, and a multi-objective particle swarm algorithm is used for collaborative optimization. The optimal parameter combination is iteratively searched as the optimal multi-parameter under preset constraints.

[0123] The fault diagnosis module 84 is used to bring the optimal multiple parameters into the reweighted overlapping group sparse shrinkage model to strengthen the frequency domain resonance band, and calculate the envelope demodulation spectrum to extract the fault characteristic frequency.

[0124] The fault diagnosis module 84 is specifically used to:

[0125] Substituting the optimal multi-parameters into the reweighted overlapping group sparse shrinkage model, the resonant frequency band in the signal frequency domain is selectively enhanced while irrelevant noise is suppressed.

[0126] After reconstructing the time domain signal through inverse transformation, the envelope spectrum is extracted using envelope demodulation technology, and finally the characteristic frequency and type of rolling bearing fault are identified to achieve fault diagnosis.

[0127] Device Example 1

[0128] According to an embodiment of the present invention, there is provided an electronic device, including:

[0129] processor; and,

[0130] A memory is arranged to store computer executable instructions, which, when executed, cause the processor to perform the steps of the above method embodiments.

[0131] Device Example 2

[0132] According to an embodiment of the present invention, a storage medium is provided for storing computer-executable instructions, wherein the computer-executable instructions implement the steps of the above-mentioned method embodiment when executed.

[0133] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A rolling bearing fault diagnosis method based on overlapping groups, characterized in that include: S1. Obtain vibration signals and construct a reweighted overlapping group sparse shrinkage model through frequency slicing wavelet transform based on frequency slicing function; S2. Establish a mathematical model framework with the harmonic spectrum kurtosis HSK and the correlation Theil index CTI as dual optimization objectives, and use the multi-objective particle swarm optimization algorithm MOPSO to implement collaborative optimization of multiple parameters of the heavily weighted overlapping group sparse shrinkage modulus to obtain the optimal multiple parameters; S3. Bring the optimal multi-parameters into the reweighted overlapping group sparse shrinkage model to strengthen the frequency domain resonance band, and calculate to obtain the envelope demodulation spectrum, and then extract the fault characteristic frequency.

2. The method according to claim 1, characterized in that Said S1 comprises: Obtain the vibration signal of the rolling bearing and decompose it into the linear superposition of the fault impact component and the noise interference; Perform joint frequency domain analysis on the signal through frequency slicing wavelet transform and extract key frequency domain features using the preset frequency slicing function; A reweighted overlapping group sparse shrinkage model is constructed. By introducing a dynamically adjusted weight matrix and regularization constraints, the sparse representation capability of fault features is enhanced while ensuring the accuracy of signal reconstruction.

3. The method according to claim 1, characterized in that The S2 includes: The harmonic spectrum kurtosis (HSK) and the correlation Theil index (CTI) are defined as dual-objective evaluation indicators to quantify the significance and distribution characteristics of fault features respectively. By establishing a multi-objective optimization framework, the key parameters of group size and regularization coefficient are used as decision variables, and a multi-objective particle swarm algorithm is used for collaborative optimization. The optimal parameter combination is iteratively searched as the optimal multi-parameter under preset constraints.

4. The method according to claim 1, wherein The S3 includes: Substituting the optimal multi-parameters into the reweighted overlapping group sparse shrinkage model, the resonant frequency band in the signal frequency domain is selectively enhanced while irrelevant noise is suppressed. After reconstructing the time domain signal through inverse transformation, the envelope spectrum is extracted using envelope demodulation technology, and finally the characteristic frequency and type of rolling bearing fault are identified to achieve fault diagnosis.

5. A rolling bearing fault diagnosis system, characterized in that include: A model building module is used to obtain vibration signals and construct a reweighted overlapping group sparse shrinkage model through frequency slicing wavelet transform based on frequency slicing function; The parameter acquisition module is used to establish a mathematical model framework with the harmonic spectrum kurtosis HSK and the correlation Theil index CTI as dual optimization objectives. The multi-objective particle swarm optimization algorithm MOPSO is used to implement collaborative optimization of multiple parameters of the heavily weighted overlapping group sparse contraction module to obtain the optimal multiple parameters. The fault diagnosis module is used to bring the optimal multi-parameters into the reweighted overlapping group sparse shrinkage model to strengthen the frequency domain resonance band, calculate the envelope demodulation spectrum, and then extract the fault characteristic frequency.

6. The system according to claim 5, characterized in that The model building module is specifically used to: Obtain the vibration signal of the rolling bearing and decompose it into the linear superposition of the fault impact component and the noise interference; Perform joint frequency domain analysis on the signal through frequency slicing wavelet transform and extract key frequency domain features using the preset frequency slicing function; A reweighted overlapping group sparse shrinkage model is constructed. By introducing a dynamically adjusted weight matrix and regularization constraints, the sparse representation capability of fault features is enhanced while ensuring the accuracy of signal reconstruction.

7. The system according to claim 5, characterized in that The parameter acquisition module is specifically used for: The harmonic spectrum kurtosis (HSK) and the correlation Theil index (CTI) are defined as dual-objective evaluation indicators to quantify the significance and distribution characteristics of fault features respectively. By establishing a multi-objective optimization framework, the key parameters of group size and regularization coefficient are used as decision variables, and a multi-objective particle swarm algorithm is used for collaborative optimization. The optimal parameter combination is iteratively searched as the optimal multi-parameter under preset constraints.

8. The system according to claim 5, characterized in that The fault diagnosis module is specifically used for: Substituting the optimal multi-parameters into the reweighted overlapping group sparse shrinkage model, the resonant frequency band in the signal frequency domain is selectively enhanced while irrelevant noise is suppressed. After reconstructing the time domain signal through inverse transformation, the envelope spectrum is extracted using envelope demodulation technology, and finally the characteristic frequency and type of rolling bearing fault are identified to achieve fault diagnosis.

9. An electronic device comprising: processor; as well as, A memory arranged to store computer executable instructions, which, when executed, cause the processor to perform the steps of the rolling bearing fault diagnosis method according to any one of claims 1 to 4.

10. A storage medium for storing computer-executable instructions, wherein the computer-executable instructions, when executed, implement the steps of the rolling bearing fault diagnosis method according to any one of claims 1 to 4.

Images