Motor bearing fault diagnosis method, device, equipment and medium

Abstract

The invention discloses a motor bearing fault diagnosis method, device and equipment and a medium, and relates to the technical field of motor bearing fault diagnosis, and the method comprises the steps: obtaining a vibration signal and a current signal of a motor bearing, and carrying out the feature extraction of the vibration signal and the current signal, obtaining a first time domain feature, a first frequency domain feature and a first energy feature of the vibration signal and a second time domain feature, a second frequency domain feature and a second energy feature of the current signal; carrying out feature fusion and feature dimension reduction on the first time domain feature, the first frequency domain feature, the first energy feature, the second time domain feature, the second frequency domain feature and the second energy feature to obtain a feature after fusion dimension reduction, and taking the feature after fusion dimension reduction as input; and determining whether the motor bearing has a fault and the fault type when the fault exists by using the fault diagnosis model. According to the invention, the fault diagnosis precision can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of motor bearing fault diagnosis, and particularly to a method, device, equipment and medium for motor bearing fault diagnosis. Background Art

[0002] Today, with the rapid development of high-speed train technology, the traction motor, as the core power system of high-speed trains, its reliability and safety are directly related to the operation safety of the entire high-speed train. The bearing is a key component in the traction motor, and its failure will not only lead to a decline in the performance of the traction motor, but may also cause major safety accidents in severe cases. Therefore, accurately diagnosing the fault type of the traction motor bearing is of great significance for ensuring the safe operation of high-speed trains and reducing maintenance costs.

[0003] Traditional motor bearing fault diagnosis methods mainly rely on the analysis of a single signal, such as vibration signal or current signal. However, in a complex working environment, a single signal is often difficult to comprehensively reflect the health status of the motor bearing, and is easily affected by noise and interference, resulting in inaccurate fault diagnosis results. The multi-signal fusion technology has attracted extensive attention in the field of fault diagnosis. By integrating multiple sensor signals, it can provide more comprehensive and reliable fault information, effectively improving the accuracy and robustness of fault diagnosis. However, currently, when performing multi-signal fusion, multiple signals are often directly used as the model input for fault diagnosis through the model, resulting in low fault diagnosis accuracy. Summary of the Invention

[0004] The purpose of the present application is to provide a method, device, equipment and medium for motor bearing fault diagnosis, which can improve the fault diagnosis accuracy.

[0005] To achieve the above object, the present application provides the following solutions.

[0006] In the first aspect, the present application provides a method for motor bearing fault diagnosis, and the method for motor bearing fault diagnosis includes: Obtain the vibration signal and current signal of the motor bearing; Respectively perform feature extraction on the vibration signal and the current signal to obtain the first time-domain feature, first frequency-domain feature and first energy feature of the vibration signal, and the second time-domain feature, second frequency-domain feature and second energy feature of the current signal; both the first energy feature and the second energy feature are energy entropy; Perform feature fusion and feature dimensionality reduction on the first time-domain feature, the first frequency-domain feature, the first energy feature, the second time-domain feature, the second frequency-domain feature and the second energy feature to obtain the fused and dimension-reduced features; Taking the feature after the fusion and dimension reduction as the input, a fault diagnosis model is used to determine whether there is a fault in the motor bearing and the type of the fault when the fault exists; the fault diagnosis model is a support vector machine, and the hyperparameters of the support vector machine are determined by an improved dung beetle optimization algorithm, and the improved dung beetle optimization algorithm is an algorithm obtained by improving the dung beetle optimization algorithm, and the improvement includes: using the Logistic chaotic map to initialize the population, using the golden sine strategy to update the position update formula of the dung beetle rolling ball behavior to obtain a first update formula, using the multi-objective selection strategy to update the position update formula of the dung beetle reproduction behavior to obtain a second update formula, using the adaptive step size and differential evolution strategy to update the position update formula of the dung beetle foraging behavior to obtain a third update formula, and using the dynamic weight strategy to update the position update formula of the dung beetle stealing behavior to obtain a fourth update formula.

[0007] In a second aspect, the present application provides a motor bearing fault diagnosis device, and the motor bearing fault diagnosis device includes: a vibration sensor, a current sensor, and a processor; The vibration sensor is installed on the motor bearing, and the vibration sensor is used to collect the vibration signal of the motor bearing; The current sensor is installed on the motor bearing, and the current sensor is used to collect the current signal of the motor bearing; The processor is respectively communicatively connected to the vibration sensor and the current sensor, and the processor is used to execute the above-mentioned motor bearing fault diagnosis method.

[0008] In a third aspect, the present application provides a computer device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor executes the computer program to implement the above-mentioned motor bearing fault diagnosis method.

[0009] In a fourth aspect, the present application provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the above-mentioned motor bearing fault diagnosis method is implemented.

[0010] According to the specific embodiments provided by the present application, the present application has the following technical effects: The present application provides a method, device, equipment and medium for diagnosing motor bearing faults. The vibration signal and current signal of the motor bearing are acquired, and feature extraction is respectively performed on the vibration signal and the current signal to obtain the first time-domain feature, the first frequency-domain feature and the first energy feature of the vibration signal, and the second time-domain feature, the second frequency-domain feature and the second energy feature of the current signal. Feature fusion and feature dimensionality reduction are performed on the first time-domain feature, the first frequency-domain feature, the first energy feature, the second time-domain feature, the second frequency-domain feature and the second energy feature to obtain the fused and dimension-reduced feature. Using the fused and dimension-reduced feature as the input, a fault diagnosis model is used to determine whether there is a fault in the motor bearing and the fault type when a fault exists. By introducing feature extraction, feature fusion and feature dimensionality reduction, the present application can fully fuse the features of the vibration signal and the current signal to obtain the fused and dimension-reduced feature. Subsequently, the fused and dimension-reduced feature is input into the fault diagnosis model for fault diagnosis, so as to fully consider the correlation between the vibration signal and the current signal for fault diagnosis, and the fault diagnosis accuracy can be improved. Description of the Drawings

[0011] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0012] Figure 1 It is an application environment diagram of a method for diagnosing motor bearing faults provided in Embodiment 1 of the present application.

[0013] Figure 2 It is a schematic flowchart of a method for diagnosing motor bearing faults provided in Embodiment 1 of the present application.

[0014] Figure 3 It is a schematic technical route diagram of a method for diagnosing motor bearing faults provided in Embodiment 1 of the present application.

[0015] Figure 4 It is a schematic detailed flowchart of a method for diagnosing motor bearing faults provided in Embodiment 1 of the present application.

[0016] Figure 5 It is a time-frequency domain diagram of performing variational mode decomposition on a vibration signal provided in Embodiment 1 of the present application; wherein, Figure 5 a1, a2, a3, a4, a5, a6 in are the time-domain diagrams of 6 intrinsic mode function components obtained by performing variational mode decomposition on the vibration signal, the abscissa is time, and the ordinate is amplitude; Figure 5In it, b1, b2, b3, b4, b5, and b6 are the frequency domain diagrams of six intrinsic mode function components obtained by variational mode decomposition of the vibration signal. The abscissa is the frequency, and the ordinate is the amplitude.

[0017] Figure 6 It is the time-frequency domain schematic diagram of variational mode decomposition of the current signal provided in Embodiment 1 of the present application; among them, Figure 6 In it, a1, a2, a3, a4, a5, a6, and a7 are the time domain diagrams of seven intrinsic mode function components obtained by variational mode decomposition of the current signal. The abscissa is the time, and the ordinate is the amplitude; Figure 6 In it, b1, b2, b3, b4, b5, b6, and b7 are the frequency domain diagrams of seven intrinsic mode function components obtained by variational mode decomposition of the current signal. The abscissa is the frequency, and the ordinate is the amplitude.

[0018] Figure 7 It is the flow schematic diagram of the improved dung beetle optimization algorithm provided in Embodiment 1 of the present application.

[0019] Figure 8 It is the schematic diagram of the classification prediction result of the test set provided in Embodiment 1 of the present application.

[0020] Figure 9 It is the schematic diagram of the classification confusion matrix of the test set provided in Embodiment 1 of the present application.

[0021] Figure 10 It is the structural schematic diagram of a computer device provided in Embodiment 3 of the present application. Detailed implementation manners

[0022] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0023] Embodiment 1.

[0024] The motor bearing fault diagnosis method provided in the embodiment of the present application can be applied to, for example Figure 1In the application environment shown. Among them, the terminal communicates with the server through the network. The data storage system can store the data that the server needs to process. The data storage system can be set separately, integrated on the server, or placed on the cloud or other servers. The terminal can send a fault diagnosis request to be processed to the server. After receiving the fault diagnosis request to be processed, for the fault diagnosis request to be processed, the server acquires the vibration signal and current signal of the motor bearing; respectively extracts features from the vibration signal and the current signal to obtain the first time-domain feature, first frequency-domain feature, and first energy feature of the vibration signal, and the second time-domain feature, second frequency-domain feature, and second energy feature of the current signal; performs feature fusion and feature dimensionality reduction on the first time-domain feature, first frequency-domain feature, first energy feature, second time-domain feature, second frequency-domain feature, and second energy feature to obtain the fused and dimension-reduced feature; uses the fused and dimension-reduced feature as input to determine whether there is a fault in the motor bearing and the fault type when there is a fault by using a fault diagnosis model. The server can feedback to the terminal the fault diagnosis result of whether there is a fault in the motor bearing for the fault diagnosis request and the fault type when there is a fault.

[0025] In addition, in some embodiments, the motor bearing fault diagnosis method can also be implemented by the server or the terminal alone. For example, the terminal can directly process the fault diagnosis request to be processed, or the server can obtain the fault diagnosis request to be processed from the data storage system and process the fault diagnosis request to be processed.

[0026] In an exemplary embodiment, as Figure 2 shown, a motor bearing fault diagnosis method is provided. This method is executed by a computer device, and specifically can be executed alone by a computer device such as a terminal or a server, or jointly executed by the terminal and the server. In the embodiments of the present application, taking this method applied to Figure 1 the server in it as an example for illustration, the method includes the following steps.

[0027] Step S1, acquire the vibration signal and current signal of the motor bearing.

[0028] Step S2, respectively extract features from the vibration signal and the current signal to obtain the first time-domain feature, first frequency-domain feature, and first energy feature of the vibration signal, and the second time-domain feature, second frequency-domain feature, and second energy feature of the current signal; both the first energy feature and the second energy feature are energy entropy.

[0029] Step S3, perform feature fusion and feature dimensionality reduction on the first time-domain feature, the first frequency-domain feature, the first energy feature, the second time-domain feature, the second frequency-domain feature, and the second energy feature to obtain the fused and dimension-reduced feature.

[0030] Step S4: Using the fused and dimension-reduced features as input, determine whether there is a fault in the motor bearing and the type of fault when a fault exists by using a fault diagnosis model; the fault diagnosis model is a support vector machine, and the hyperparameters of the support vector machine are determined by an improved dung beetle optimization algorithm. The improved dung beetle optimization algorithm is an algorithm obtained by improving the dung beetle optimization algorithm. The improvements include: using the Logistic chaotic map to initialize the population, using the golden sine strategy to update the position update formula of the dung beetle's ball-rolling behavior to obtain a first update formula, using the multi-objective selection strategy to update the position update formula of the dung beetle's reproduction behavior to obtain a second update formula, using the adaptive step size and differential evolution strategy to update the position update formula of the dung beetle's foraging behavior to obtain a third update formula, and using the dynamic weight strategy to update the position update formula of the dung beetle's stealing behavior to obtain a fourth update formula.

[0031] Implementing the above steps S1 to S4, this embodiment provides a method for diagnosing faults in a motor bearing. First, extract the features of the vibration signal and the current signal respectively to obtain the first time-domain feature, the first frequency-domain feature, and the first energy feature of the vibration signal, as well as the second time-domain feature, the second frequency-domain feature, and the second energy feature of the current signal. Subsequently, perform feature fusion and feature dimension reduction on the first time-domain feature, the first frequency-domain feature, the first energy feature, the second time-domain feature, the second frequency-domain feature, and the second energy feature to obtain the fused and dimension-reduced features. Finally, using the fused and dimension-reduced features as input, determine whether there is a fault in the motor bearing and the type of fault when a fault exists by using a fault diagnosis model. By introducing feature extraction, feature fusion, and feature dimension reduction, it is possible to fuse the features of the vibration signal and the current signal before inputting the vibration signal and the current signal into the fault diagnosis model, so that when diagnosing faults using the vibration signal and the current signal, the correlation between the vibration signal and the current signal is considered, and the fault diagnosis accuracy can be improved.

[0032] The following, in combination with Figure 3 and Figure 4 , introduce in detail the method for diagnosing faults in the motor bearing used in this embodiment, including the following steps.

[0033] Step 1: Collect the vibration signal and the current signal of the motor bearing.

[0034] In this embodiment, an acceleration sensor and a current sensor are respectively used to collect the vibration signal and current signal of the motor bearing. Specifically, a three-axis acceleration sensor can be used to collect the vibration signal of the motor bearing, and a current transformer can be used to collect the current signal of the motor bearing. Then, time alignment and downsampling processing (i.e., decimation processing) are performed on the collected vibration signal and current signal to balance signal quality and calculation cost, and an original data set is constructed.

[0035] At this time, in this embodiment, the vibration signal and current signal of the motor bearing are obtained.

[0036] Step 2: Extract the time-domain features and frequency-domain features of the vibration signal and current signal, construct a time-frequency domain data set. Decompose the vibration signal and current signal by Variational Mode Decomposition (VMD), and calculate the energy entropy of the Intrinsic Mode Function (IMF) components obtained by the decomposition, and construct an energy entropy data set.

[0037] In this embodiment, by extracting the typical time-domain features and typical frequency-domain features of the collected vibration signal and current signal, the time-domain features of the vibration signal and current signal are obtained respectively, such as mean value, kurtosis, root mean square value and peak factor. At the same time, the frequency-domain features of the vibration signal and current signal are obtained respectively, such as the main frequency of the spectrum, harmonic energy ratio, power spectral density and total harmonic distortion, and a typical time-domain and frequency-domain data set corresponding to the vibration signal (including the time-domain features and frequency-domain features of the vibration signal) and a typical time-domain and frequency-domain data set corresponding to the current signal (including the time-domain features and frequency-domain features of the current signal), and constitute a time-frequency domain data set.

[0038] In this embodiment, by using a dual-index comprehensive evaluation method based on reconstruction accuracy and modal independence, the most suitable VMD decomposition layer numbers for the vibration signal and current signal are analyzed respectively. Based on these VMD decomposition layer numbers, the vibration signal and current signal are respectively decomposed by VMD to obtain multiple Intrinsic Mode Function components corresponding to the vibration signal and current signal respectively. Then, the energy entropy of the multiple Intrinsic Mode Function components obtained by the decomposition is calculated, and a vibration signal VMD decomposition energy entropy data set (including the energy entropy of the vibration signal) and a current signal VMD decomposition energy entropy data set (including the energy entropy of the current signal), and constitute an energy entropy data set.

[0039] The energy entropy H The calculation expression is: (1); In formula (1), K is the number of intrinsic mode function components; is the energy proportion of the -th intrinsic mode function component, which is equal to the ratio of the energy of the -th intrinsic mode function component to the total energy, and the total energy is the sum of the energies of each intrinsic mode function component.

[0040] The method for finding the most suitable VMD decomposition layer number is a dual-index comprehensive evaluation method based on reconstruction accuracy and modal independence. In this dual-index comprehensive evaluation method based on reconstruction accuracy and modal independence, the root mean square error (RMSE) and the mean cross-correlation coefficient (MCC) are combined to select the VMD decomposition layer number K , specifically by traversing different VMD decomposition layer numbers K , calculating the root mean square error RMSE between the reconstructed signal and the original signal and the mean cross-correlation coefficient MCC between each IMF component for each K value, and selecting the optimal K value according to the comprehensive performance of these two indicators. The specific steps for using the dual-index comprehensive evaluation method based on reconstruction accuracy and modal independence to find the most suitable VMD decomposition layer number are as follows.

[0041] (1) Set the search range of the VMD decomposition layer number K , for example, from 2 to 20.

[0042] (2) Take multiple values within this search range at a preset step size to obtain multiple K values. For each K value, at this K value, perform VMD decomposition on the original signal to obtain K IMF components , add these K IMF components to reconstruct the original signal, and obtain the reconstructed signal .

[0043] Among them, the calculation formula for the reconstructed signal is: (2); In formula (2), is the -th IMF component.

[0044] (3) Calculate the reconstructed signal and the original signal to obtain the RMSE at this K value of RMSE.

[0045] Among them, the calculation formula of the root mean square error RMSE is: (3); In formula (3), is the K RMSE at this value; is the signal length, that is, the number of time points of the original signal; is the original signal at the time point value; is the reconstructed signal at the time point value.

[0046] (4) Calculate the correlation between each IMF component using the average cross-correlation coefficient MCC to obtain the MCC at this K value of MCC to measure whether the modes are redundant.

[0047] Among them, the calculation formula of the average cross-correlation coefficient MCC is: (4); In formula (4), is the K MCC at this value; is the th IMF component and the th IMF component Pearson correlation coefficient between.

[0048] The calculation formula of the Pearson correlation coefficient is: (5); In formula (5), is the th IMF component and the th IMF component covariance between, describing the joint change trend between the th IMF component and the th IMF component ; is the th IMF component standard deviation of, describing the fluctuation degree of the th IMF component ; is the The standard deviation of an IMF component describes the degree of fluctuation of the th IMF component.

[0049] The calculation formulas for covariance and standard deviation are as follows: (6); In formula (6), is the value of the th IMF component at time point ; is the mean value of the th IMF component, ; is the value of the th IMF component at time point ; is the mean value of the th IMF component, .

[0050] (5) Construct a weighted comprehensive scoring function to calculate the comprehensive score for each K value, and select the K value with the minimum comprehensive score as the most suitable VMD decomposition layer number.

[0051] The weighted comprehensive scoring function is: (7); In formula (7), is the comprehensive score at the K value; is the first weight; is K the normalized RMSE at the value; is the second weight; K is the normalized MCC at the and are determined according to user requirements. When more emphasis is placed on reconstruction accuracy, increase ; when more emphasis is placed on modal independence, increase .

[0052] Normalize RMSE and MCC respectively to obtain the normalized RMSE and the normalized MCC. The calculation formulas are: (8); In formula (8), is the minimum value of RMSE for all K values; is for all KThe maximum value of RMSE at the value; For all K The minimum value of MCC at the value; For all K The maximum value of MCC at the value.

[0053] As an emerging signal processing technology, the core idea of variational mode decomposition is: to decompose a given signal into a series of intrinsic mode function components with specific center frequencies, while minimizing the sum of the estimated bandwidths of each intrinsic mode function component.

[0054] To ensure the stability of the decomposition result, a constrained variational model is introduced, expressed as follows: (9); In Equation (9), Represents the set composed of each intrinsic mode function component; Represents the set composed of the center frequencies of each intrinsic mode function component; K Is the number of VMD decomposition layers, that is, the number of intrinsic mode function components; Is the partial derivative with respect to time Indicates the gradient; Is the unit impulse function, used to define the Hilbert transform; j Is the imaginary unit, satisfying , used for complex number operations; Is the value of the th intrinsic mode function component at the time point ; Is the complex exponential modulation term, used to modulate the analytic signal to the position where the base frequency is zero, Is the th center frequency of the intrinsic mode function component; Is the given signal.

[0055] To solve the constrained variational problem described in Equation (9) above, the alternating direction multiplier method is used to solve it. An augmented Lagrangian function is introduced, and the constrained variational problem is transformed into an unconstrained variational problem by using the Lagrangian operator and the quadratic penalty operator. The augmented Lagrangian function Is as follows: (10); In Equation (10), Is the Lagrange multiplier, used to enforce the constraint condition; Is the regularization parameter, used to control the strength of the bandwidth constraint.

[0056] Continuously update the intrinsic mode function component And the center frequency To minimize the spectral overlap between modes and to satisfy the iterative termination condition through alternating iteration with Lagrange multipliers.

[0057] After obtaining the most suitable number of VMD decomposition layers, the vibration signal and the current signal are subjected to VMD decomposition, and the time-domain and frequency-domain diagrams after decomposition are as Figure 5 and Figure 6 shown.

[0058] At this time, in this embodiment, the vibration signal and the current signal are respectively subjected to feature extraction to obtain the first time-domain feature, the first frequency-domain feature, and the first energy feature of the vibration signal, as well as the second time-domain feature, the second frequency-domain feature, and the second energy feature of the current signal.

[0059] Among them, the first time-domain feature and the second time-domain feature both include: mean, kurtosis, root mean square value, and peak factor. The first frequency-domain feature and the second frequency-domain feature both include: spectral main frequency, harmonic energy ratio, power spectral density, and total harmonic distortion. The first energy feature and the second energy feature are both energy entropy.

[0060] The determination method of the first energy feature includes: performing variational mode decomposition on the vibration signal to obtain a plurality of first intrinsic mode function components, calculating the energy of each first intrinsic mode function component, and calculating the first energy feature based on the energies of all first intrinsic mode function components, which is specifically calculated by Equation (1). The determination method of the second energy feature includes: performing variational mode decomposition on the current signal to obtain a plurality of second intrinsic mode function components, calculating the energy of each second intrinsic mode function component, and calculating the second energy feature based on the energies of all second intrinsic mode function components, which is specifically calculated by Equation (1).

[0061] Among them, performing variational mode decomposition on the vibration signal to obtain a plurality of first intrinsic mode function components specifically includes: using a dual-index comprehensive evaluation method based on reconstruction accuracy and modal independence to determine the first decomposition layer used for variational mode decomposition of the vibration signal, and performing variational mode decomposition on the vibration signal according to the first decomposition layer to obtain a plurality of first intrinsic mode function components.

[0062] Among them, performing variational mode decomposition on the current signal to obtain a plurality of second intrinsic mode function components specifically includes: using a dual-index comprehensive evaluation method based on reconstruction accuracy and modal independence to determine the second decomposition layer used for variational mode decomposition of the current signal, and performing variational mode decomposition on the current signal according to the second decomposition layer to obtain a plurality of second intrinsic mode function components.

[0063] Among them, a dual-index comprehensive evaluation method based on reconstruction accuracy and modal independence is used to determine the first decomposition layer number used for variational mode decomposition of vibration signals, which specifically includes: designing multiple first initial decomposition layer numbers; for each first initial decomposition layer number, performing variational mode decomposition on the vibration signal according to the first initial decomposition layer number to obtain multiple first initial intrinsic mode function components, calculating the sum value of all the first initial intrinsic mode function components to obtain a reconstructed vibration signal, calculating the first root mean square error at the first initial decomposition layer number based on the vibration signal and the reconstructed vibration signal, calculating the first average cross-correlation coefficient at the first initial decomposition layer number based on all the first initial intrinsic mode function components, calculating the first comprehensive score at the first initial decomposition layer number based on the first root mean square error and the first average cross-correlation coefficient; selecting the first initial decomposition layer number with the smallest first comprehensive score as the first decomposition layer number used for variational mode decomposition of the vibration signal.

[0064] Among them, a dual-index comprehensive evaluation method based on reconstruction accuracy and modal independence is used to determine the second decomposition layer number used for variational mode decomposition of current signals, which specifically includes: designing multiple second initial decomposition layer numbers; for each second initial decomposition layer number, performing variational mode decomposition on the current signal according to the second initial decomposition layer number to obtain multiple second initial intrinsic mode function components, calculating the sum value of all the second initial intrinsic mode function components to obtain a reconstructed current signal, calculating the second root mean square error at the second initial decomposition layer number based on the current signal and the reconstructed current signal, calculating the second average cross-correlation coefficient at the second initial decomposition layer number based on all the second initial intrinsic mode function components, calculating the second comprehensive score at the second initial decomposition layer number based on the second root mean square error and the second average cross-correlation coefficient; selecting the second initial decomposition layer number with the smallest second comprehensive score as the second decomposition layer number used for variational mode decomposition of the current signal.

[0065] Step 3: Fuse and reduce the dimension of the obtained time-frequency domain dataset and energy entropy dataset, and output a multi-modal dataset.

[0066] This embodiment passes through ( , ), ( , In the pairwise combination method, the Canonical Correlation Analysis (CCA) algorithm is used to fuse the constructed time-frequency domain dataset and energy entropy dataset respectively to obtain the fused dataset. In order to select appropriate feature dimensions, improve model performance, reduce the risk of overfitting, and improve computational efficiency, the Principal Component Analysis (PCA) algorithm is used to reduce the dimensionality of the feature dimensions of the fused dataset to construct a multi-modal dataset.

[0067] In this embodiment, the features of vibration signals and current signals are fused through canonical correlation analysis, effectively extracting the correlation information between the two and improving the accuracy of fault diagnosis.

[0068] Canonical correlation analysis aims to find the linear combinations of two sets of variables such that the correlation between these two linear combinations is maximized. The specific principle and formula of canonical correlation analysis are as follows: For two datasets X and Y , find two projection vectors and , and represent the linear combinations of X and Y as: (11); In formula (11), is the result of the linear combination of dataset X under the projection vector ; is the result of the linear combination of dataset Y under the projection vector .

[0069] For and two projection vectors, make the correlation coefficient maximum, and its correlation coefficient is defined as: (12); To ensure that the results are interpretable and numerically stable, its constraint condition is: (13); In the above formula, is the covariance of and ; is the variance of ; is the variance of ; The superscript T represents transpose.

[0070] In this embodiment, principal component analysis is used to reduce the dimension of the fused dataset, reduce the risk of overfitting, improve the computational efficiency, and obtain a multi-modal dataset.

[0071] Principal component analysis is a statistical method widely used in data dimensionality reduction and feature extraction. Its core idea is to project the original high-dimensional data into a low-dimensional space through a linear transformation while retaining as much variance information of the original high-dimensional data as possible.

[0072] Suppose the original data matrix is , is the number of samples, is the number of features. The purpose of principal component analysis is to find a transformation matrix , is the number of features after dimensionality reduction, such that the new feature matrix retains as much variance information of the original data matrix as possible. Through eigenvalue decomposition, the covariance matrix can be expressed as: (14); In Equation (14), U is the eigenvector matrix of the covariance matrix Σ.

[0073] The feature extraction, feature fusion, and feature dimensionality reduction of this embodiment have the following advantages.

[0074] (1) Improve information integrity through multi-dimensional feature fusion: By combining time-frequency domain features with the energy entropy obtained by VMD decomposition, which is an energy feature, it can not only reflect the basic statistical characteristics and frequency distribution of the signal but also capture its non-linear dynamic characteristics, realizing a comprehensive description of the signal characteristics.

[0075] (2) Enhance the ability to analyze complex signals: Time-frequency domain features are suitable for depicting the overall trend of the signal, and the energy entropy obtained by VMD decomposition, which is an energy feature, can quantify the local non-linear modal complexity. The combination of the two can accurately characterize the global and local characteristics of the signal, significantly improving the analysis accuracy of complex signals.

[0076] (3) Improve the robustness of the system: Multi-dimensional feature fusion effectively reduces the sensitivity of a single method to noise, reduces the performance degradation caused by signal variation or interference, and thus enhances the noise resistance and stability of the system.

[0077] (4) Optimize the model performance: Improve the model's expression ability to capture subtle changes in the signal; enhance the ability to distinguish different types of signals, reducing the risks of misclassification and missed classification; and still maintain high accuracy in the case of noise interference or signal anomalies, ensuring reliable analysis results.

[0078] In summary, in this embodiment, more comprehensive and in-depth signal description is achieved through multi-dimensional feature fusion, significantly improving the parsing ability, discrimination ability, and robustness of the model.

[0079] At this time, in this embodiment, feature fusion and feature dimensionality reduction are performed on the first time-domain feature, the first frequency-domain feature, the first energy feature, the second time-domain feature, the second frequency-domain feature, and the second energy feature to obtain the fused and dimension-reduced feature.

[0080] Among them, performing feature fusion and feature dimensionality reduction on the first time-domain feature, the first frequency-domain feature, the first energy feature, the second time-domain feature, the second frequency-domain feature, and the second energy feature to obtain the fused and dimension-reduced feature specifically includes: using the canonical correlation analysis algorithm to perform feature fusion on the first time-domain feature, the first frequency-domain feature, the second time-domain feature, and the second frequency-domain feature to obtain the first fused feature; using the canonical correlation analysis algorithm to perform feature fusion on the first energy feature and the second energy feature to obtain the second fused feature; using the principal component analysis algorithm to perform feature dimensionality reduction on the first fused feature and the second fused feature to obtain the fused and dimension-reduced feature.

[0081] Step 4, improve the Dung Beetle Optimizer (DBO) algorithm, and use the improved Dung Beetle Optimizer (abbreviated as IDBO) to optimize the hyperparameters of the Support Vector Machine (SVM) to construct the IDBO-SVM fault diagnosis model.

[0082] SVM has become an important tool for fault diagnosis due to its excellent performance in processing small samples, non-linear, and high-dimensional data. However, the classification performance of SVM is strongly dependent on its hyperparameters. Traditional parameter optimization methods such as grid search and genetic algorithms have high computational complexity when dealing with high-dimensional feature data and are prone to falling into local optima. The DBO algorithm has become a powerful tool for SVM parameter optimization due to its excellent global search ability and convergence speed. However, the DBO algorithm still has problems such as insufficient initial population diversity and being prone to falling into local optima. Therefore, in this embodiment, the DBO algorithm is improved, combined with a dynamic weight strategy and an efficient search mechanism, and combined with a multi-modal feature fusion technology, providing a new and efficient solution for motor bearing fault diagnosis.

[0083] The dung beetle optimization algorithm finds the global optimal solution by mathematically modeling the behaviors of dung beetles rolling balls, reproducing, foraging, and stealing.

[0084] Dung beetles use celestial cues for navigation to achieve the behavior of rolling dung balls in a straight line. Assuming that the dung beetle moves in a specific direction in the entire search space, during this process, the position of the dung beetle will be continuously updated, and the position update formula is: (15); In Equation (15), is the position information of the th dung beetle at the +1-th iteration; is the position information of the th dung beetle at the th iteration; is a natural coefficient assigned -1 or 1, indicating no deviation, indicating deviation from the original direction; is a constant used to control the weighting coefficient, usually between 0 and 1, and specifically 0.1 can be taken in this embodiment; is the position information of the th dung beetle at the -1-th iteration; is a constant used to control the weighting coefficient, usually between 0 and 1, and specifically 0.3 can be taken in this embodiment; is a parameter used to simulate the intensity change of light; is the global worst position.

[0085] The dung beetle optimization algorithm proposes a boundary selection strategy to simulate the area where dung beetles lay eggs in a safe environment, which is defined as follows: (16); In Equation (16), is the lower limit of the spawning area; is the local best position; R is a dynamic parameter that changes with the number of iterations; is the lower limit of the optimization problem; is the upper limit of the spawning area; is the upper limit of the optimization problem; is the maximum number of iterations.

[0086] Female dung beetles produce only one egg in each iteration. Dung beetles will hide the eggs in the brood balls. The position of the brood balls is defined as follows: (17); In Equation (17), is the position information of the th brood ball at the +1-th iteration; and are two independent random vectors with a size of 1×Dim, where Dim is the dimension of the optimization problem; is the position information of the th brood ball at the th iteration.

[0087] The boundary of the optimal foraging area is defined as follows: (18); In formula (18), is the lower limit of the optimal foraging area; is the global best position; is the upper limit of the optimal foraging area.

[0088] The position update formula for the small dung beetle to forage is as follows: (19); In formula (19), is the position information of the th small dung beetle at the +1 - th iteration; is the position information of the th small dung beetle at the th iteration; is a random number following a normal distribution; is a random vector belonging to (0, 1).

[0089] The position update formula for the thief dung beetle is as follows: (20); In formula (20), is the position information of the th thief dung beetle at the +1 - th iteration; S is a constant value; g is a random vector of size 1×Dim, following a normal distribution; is the position information of the th thief dung beetle at the th iteration.

[0090] The process of the dung beetle optimization algorithm is specifically as follows.

[0091] (1) Initialize the population and parameters: Set parameters such as the initial population size and the maximum number of iterations, and randomly generate the initial positions of the dung beetle individuals.

[0092] (2) Calculate the fitness value: Calculate the fitness value of each dung beetle individual according to the objective function. In this embodiment, the objective function is defined as the cross - validation accuracy of SVM.

[0093] (3) Simulate the behavior of dung beetles: Rolling behavior: Update the position of the dung beetle using Equation (15) to simulate the process of rolling the ball. Reproductive behavior: Define the position of the brood ball using Equation (17) to simulate the reproductive process. Foraging behavior: Update the position of the young dung beetles using Equation (19) to simulate the foraging process. Thieving behavior: Update the position of the thieving dung beetle using Equation (20) to simulate the thieving behavior.

[0094] (4) Boundary selection strategy: Determine the upper and lower limits of the spawning area using Equation (16) to ensure that the dung beetles move within a safe area.

[0095] (5) Update the best position: Record the current global best position and global worst position for the next iteration.

[0096] (6) Judge the termination condition: Check whether the maximum number of iterations is reached or other termination conditions are met. If not, return to (2) to continue the iteration; otherwise, output the final solution.

[0097] The improved dung beetle optimization algorithm is constructed using the idea as Figure 7 shown, and the dung beetle optimization algorithm is improved using Logistic chaotic mapping, golden sine strategy, multi-objective selection strategy, adaptive step size, differential evolution strategy and dynamic weight strategy. Use the improved dung beetle optimization algorithm to find the hyperparameters of the support vector machine and construct the IDBO-SVM fault diagnosis model.

[0098] (1) The population initialization method using Logistic chaotic mapping generates a sequence with good uniformity and randomness, which can make the initial population evenly distributed in the search space. The expression of Logistic chaotic mapping is: (21); In Equation (21), is the chaotic sequence value of the -th iteration; is the control parameter, usually taking the value of 4; is the chaotic sequence value of the -th iteration. The chaotic sequence value is used to initialize the position of individuals in the population to ensure that the population is evenly distributed in the search space.

[0099] (2) Introduce the golden sine strategy in the position update of the dung beetle's rolling behavior to narrow the search space and conduct a full search in the local area, so that the global search ability and local development ability of the algorithm reach a good balance. The position update formula using the golden sine strategy is as follows: (22); In Equation (22), and are the golden section coefficients; is the golden ratio, ; is the position information of the th dung beetle at the +1-th iteration; is the position information of the th dung beetle at the th iteration; is a random number within ; is a random number within ; is the position information of the th dung beetle at the -1-th iteration.

[0100] (3) Introduce a multi-objective selection strategy in the reproductive behavior of dung beetles. By combining the influence of local optimal solutions and global elite solutions, the selection of the reproductive area is made more flexible. The position update formula using the multi-objective selection strategy is as follows: (23); In formula (23), is the position information of the th brood ball at the +1-th iteration; is the weight coefficient used to balance the influence of local optimal solutions and global elite solutions. Usually, ranges from [0, 1]; is the local optimal solution, which needs to be ensured to be between and ; is the global elite solution, which needs to be ensured to be between and ; is the perturbation intensity coefficient used to control the magnitude of random perturbation; represents a random perturbation term that follows a normal distribution with a mean of 0 and a variance of 1.

[0101] If the calculation result exceeds the boundary, boundary correction is performed: (24); In formula (24), in the min function is calculated using formula (23).

[0102] Within the limited area, the multi-objective selection strategy can better utilize the existing information, avoid blind search, balance exploration and exploitation better in the limited space by weighing the influence of local optimal solutions and global elite solutions, and can explore potential fault features more comprehensively to improve diagnostic accuracy.

[0103] (4) Introduce an adaptive step size and differential evolution strategy into the foraging behavior of dung beetles. The adaptive step size dynamically adjusts the step size according to the distance between the current solution and the global optimal solution, balancing exploration and exploitation. The differential evolution strategy generates new solutions using multiple individuals in the population, enhancing population diversity and global search ability. The improved position update formula is: (25); In formula (25), is the position information of the -th dung beetle at the +1-th iteration; are three different individuals randomly selected from the current population; is the first scaling factor, controlling the intensity of differential evolution; is the second scaling factor, used to control the maximum value of the step size; is a regulation parameter, used to control the sensitivity of the step size change with distance; is the global optimal solution; is the -th dung beetle at the -th iteration; is the upper bound of the spawning area; is the lower bound of the spawning area.

[0104] By improving the foraging behavior of the dung beetle optimization algorithm and introducing an adaptive step size and differential evolution strategy, the global search ability and population diversity of the algorithm can be significantly improved without destroying the basic framework of the original algorithm. These improvements not only enhance the performance of the algorithm but also improve its application effect in fault diagnosis.

[0105] (5) Introduce a dynamic weight strategy into the stealing behavior of dung beetles. The position update formula for introducing the dynamic weight strategy is: (26); In formula (26), is the position information of the -th thief dung beetle at the +1-th iteration; and are both weight parameters. Set to a larger value in the early stage of iteration, so that the dung beetle explores a better area near the position of the best food, increasing the global optimization ability of the algorithm. In the later stage of iteration gradually increases, enabling the dung beetle to have the ability to jump out of the local optimum.

[0106] To improve the DBO algorithm, in this embodiment, during the population initialization stage, the Logistic chaotic mapping is used to generate the initial population. This method has good uniformity and randomness, making the population resources more evenly distributed in the search space, thus improving the initial exploration ability of the algorithm. During the position update stage of the dung beetle ball-rolling behavior, the golden sine strategy is introduced. By shrinking the search space and conducting a full search in the local area, a good balance between the global search ability and the local development ability is achieved. During the position update stage of the dung beetle reproduction behavior, a multi-objective selection strategy is adopted. Combining the influence of the local optimal solution and the global elite solution makes the selection of the reproduction area more flexible. At the same time, the rationality of the search range is ensured through boundary correction, further enhancing the exploration and development effects of the algorithm. During the position update stage of the dung beetle foraging behavior, an adaptive step size and differential evolution strategy are introduced. The step size is dynamically adjusted to balance exploration and development, and new solutions are generated using multiple individuals in the population to enhance population diversity and global search ability. During the position update stage of the dung beetle stealing behavior, a dynamic weight strategy is adopted. The global optimization ability of the algorithm is increased in the early stage of iteration, while the local search ability is gradually enhanced in the later stage of iteration, enabling the algorithm to have the ability to jump out of the local optimum. The DBO algorithm is improved in the above way, and the improved DBO algorithm is used to find the hyperparameters of the SVM to construct the IDBO-SVM fault diagnosis model.

[0107] Finding the hyperparameters of the SVM using the improved DBO algorithm includes: defining the optimization objective and initializing, setting the hyperparameters to be optimized in the SVM, setting the parameters of the improved DBO algorithm, setting the population size of the algorithm, initializing the population position and fitness value, the maximum number of iterations, etc.; executing the improved DBO algorithm, using the cross-validation accuracy of the SVM as the fitness value, calculating and updating the position and fitness value of the dung beetles, and continuously updating the fitness value, the individual optimal position, and the global optimal position of each individual; using the two globally optimal positions as the values of the penalty parameter and the kernel function parameter to construct the IDBO-SVM fault diagnosis model.

[0108] The support vector machine is a machine learning model widely used in classification tasks. Its core idea is to find an optimal hyperplane in the feature space. This optimal hyperplane can not only correctly classify all training samples but also maximize the distance between the positive and negative samples and the decision boundary.

[0109] The optimal hyperplane is jointly defined by the weight vector w and the bias vector : (27); In formula (27), is the classification result, is the input vector. For the data points of binary classification , is the th input feature, which is , is the category label corresponding to the th input feature, which is

[0110] For a linearly separable dataset, its basic optimization problem (hard margin case) can be formulated as: (28); In equation (28), is the number of samples.

[0111] For a non-linearly separable dataset, SVM uses the kernel trick to map the data points in the original feature space to a high-dimensional space, where the data may become linearly separable, and thus find an effective separating hyperplane. The kernel function K is defined as follows: (29); In equation (29), is the th input feature; is a mapping function whose main function is to map the data from a low dimension to a high dimension and perform calculations in the high-dimensional space.

[0112] In practical applications, the data is often not completely linearly separable. Therefore, the concept of soft margin is introduced to allow some sample points to slightly violate the margin condition, and the optimization problem becomes: (30); In equation (30), w is the weight vector; is the bias vector; is the slack variable, which allows the sample points to slightly violate the margin condition; C is the penalty parameter, which is a regularization parameter that controls the degree of penalty; is the number of samples; is the th slack variable; is the category label corresponding to the th input feature; is the

[0113] Step 5: Based on the obtained multi-modal dataset, use the IDBO-SVM fault diagnosis model to diagnose the faults of the motor bearing and determine the fault type of the motor bearing.

[0114] Numerically encode the fault type labels, align the fault features (i.e., the features after fusion and dimensionality reduction) with the corresponding fault type labels, divide the training set and the test set according to the ratio of 7:3, use the training set to train the IDBO-SVM fault diagnosis model, input the test set into the trained IDBO-SVM fault diagnosis model, predict the fault type of the motor bearing, and compare the obtained diagnosis results with the actual classification results, calculate the performance indicators of this fault diagnosis model, and evaluate this fault diagnosis model, such as Figure 8 and Figure 9 shown Figure 8 The 10 fault types and their corresponding labels in

[0115] are: pitted inner ring, pitted outer ring, engraved inner ring of engraving machine, engraved outer ring of engraving machine, indented outer ring, drilled outer ring, pitted inner ring + outer ring, indented inner ring + outer ring, engraved inner ring + outer ring of engraving machine, drilled inner ring + outer ring of engraving machine, and the labels are 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 respectively.

[0116] At this time, in this embodiment, the features after fusion and dimensionality reduction are used as the input, and the fault diagnosis model is used to determine whether there is a fault in the motor bearing and the fault type when there is a fault.

[0117] Among them, the fault diagnosis model is a support vector machine.

[0118] The hyperparameters of the support vector machine are determined by an improved dung beetle optimization algorithm. The improved dung beetle optimization algorithm is an algorithm obtained by improving the dung beetle optimization algorithm. The improvements include: using the Logistic chaotic map to initialize the population, that is, Equation (21), using the golden sine strategy to update the position update formula of the dung beetle's ball-rolling behavior to obtain the first update formula, that is, Equation (22), using the multi-objective selection strategy to update the position update formula of the dung beetle's reproduction behavior to obtain the second update formula, that is, Equation (23), using the adaptive step size and differential evolution strategy to update the position update formula of the dung beetle's foraging behavior to obtain the third update formula, that is, Equation (25), and using the dynamic weight strategy to update the position update formula of the dung beetle's stealing behavior to obtain the fourth update formula, that is, Equation (26).

[0118] The hyperparameters of the support vector machine include the penalty parameter and the kernel function parameter.

[0119] This embodiment can effectively integrate multi-source heterogeneous data, overcome the problem of incomplete characterization of single-modal feature information, and significantly improve the discrimination and robustness of fault features. A dual-index comprehensive evaluation method based on reconstruction accuracy and modal independence is used to dynamically determine the number of VMD decomposition layers, which solves the limitation of the traditional method that the number of decomposition layers depends on empirical settings and avoids feature distortion caused by over-decomposition or under-decomposition. Through the improvement of the DBO algorithm by Logistic chaotic mapping, golden sine strategy, multi-objective selection strategy, adaptive step size, differential evolution strategy and dynamic weight strategy, the global search and local development capabilities are effectively balanced, and the problems of traditional optimization algorithms being prone to falling into local optima and slow convergence speed are solved, enabling the hyperparameters of SVM to adaptively match the characteristics of fault data distribution, and significantly improving the accuracy of the fault diagnosis model.

[0120] This embodiment discloses a method for diagnosing motor bearing faults based on improved DBO-SVM, including: collecting vibration signals and current signals of the motor bearing; using VMD decomposition on the original vibration signals and current signals to obtain a certain number of intrinsic mode function components and calculating energy features; calculating typical features of the original vibration signals and current signals in the time-frequency domain; performing dimensionality reduction and fusion on the extracted features through CCA and PCA to form a multi-modal data set; improving the DBO algorithm to enhance robustness and search efficiency; using the improved DBO algorithm to optimize the parameters of SVM; and inputting the multi-modal data set into the improved DBO-SVM model to diagnose the fault type of the motor bearing. This embodiment improves the accuracy of diagnosis and provides a new solution for motor bearing fault diagnosis.

[0121] Embodiment 2.

[0122] This embodiment provides a motor bearing fault diagnosis device, which includes: a vibration sensor, a current sensor, and a processor.

[0123] The vibration sensor is installed on the motor bearing, and the vibration sensor is used to collect the vibration signal of the motor bearing.

[0124] The current sensor is installed on the motor bearing, and the current sensor is used to collect the current signal of the motor bearing.

[0125] The processor is communicatively connected to the vibration sensor and the current sensor respectively, and the processor is used to execute the motor bearing fault diagnosis method described in Embodiment 1.

[0126] Embodiment 3.

[0127] In an exemplary embodiment, a computer device is provided. The computer device can be a server or a terminal, and its internal structure diagram can be as Figure 10As shown in the figure. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals through a network connection. When the computer program is executed by the processor, it implements a method for diagnosing motor bearing faults.

[0128] Those skilled in the art can understand that Figure 10 the structure shown in the figure is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0129] In an exemplary embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, it implements the method for diagnosing motor bearing faults in Embodiment 1.

[0130] Embodiment 4.

[0131] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by the processor, it implements the method for diagnosing motor bearing faults in Embodiment 1.

[0132] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with relevant regulations.

[0133] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combinations of these technical features do not conflict, they should all be considered as the scope described in this specification.

[0134] In this text, specific examples are used to illustrate the principles and implementation manners of the present application. The descriptions of the above embodiments are only used to help understand the method and its core idea of the present application; at the same time, for those of ordinary skill in the art, according to the idea of the present application, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to the present application.

Claims

1. A method for diagnosing motor bearing faults, characterized in that, The motor bearing fault diagnosis method includes: Obtaining the vibration signal and current signal of the motor bearing; Respectively performing feature extraction on the vibration signal and the current signal to obtain the first time-domain feature, first frequency-domain feature, and first energy feature of the vibration signal, and the second time-domain feature, second frequency-domain feature, and second energy feature of the current signal; both the first energy feature and the second energy feature are energy entropy; Performing feature fusion and feature dimensionality reduction on the first time-domain feature, the first frequency-domain feature, the first energy feature, the second time-domain feature, the second frequency-domain feature, and the second energy feature to obtain the fused and dimension-reduced feature; Using the fused and dimension-reduced feature as the input, and using a fault diagnosis model to determine whether there is a fault in the motor bearing and the fault type when there is a fault; the fault diagnosis model is a support vector machine, and the hyperparameters of the support vector machine are determined by an improved dung beetle optimization algorithm. The improved dung beetle optimization algorithm is an algorithm obtained by improving the dung beetle optimization algorithm. The improvements include: using the Logistic chaotic map to initialize the population, using the golden sine strategy to update the position update formula of the dung beetle's ball-rolling behavior to obtain the first update formula, using the multi-objective selection strategy to update the position update formula of the dung beetle's reproduction behavior to obtain the second update formula, using the adaptive step size and differential evolution strategy to update the position update formula of the dung beetle's foraging behavior to obtain the third update formula, and using the dynamic weight strategy to update the position update formula of the dung beetle's stealing behavior to obtain the fourth update formula.

2. The motor bearing fault diagnosis method according to claim 1, characterized in that Both the first time-domain feature and the second time-domain feature include: mean, kurtosis, root mean square value, and peak factor; Both the first frequency-domain feature and the second frequency-domain feature include: spectral main frequency, harmonic energy ratio, power spectral density, and total harmonic distortion; The method for determining the first energy feature includes: performing variational mode decomposition on the vibration signal to obtain a plurality of first intrinsic mode function components, calculating the energy of each first intrinsic mode function component, and calculating the first energy feature based on the energies of all the first intrinsic mode function components; The method for determining the second energy feature includes: performing variational mode decomposition on the current signal to obtain a plurality of second intrinsic mode function components, calculating the energy of each second intrinsic mode function component, and calculating the second energy feature based on the energies of all the second intrinsic mode function components.

3. The motor bearing fault diagnosis method according to claim 2, characterized in that, Performing variational mode decomposition on the vibration signal to obtain a plurality of first intrinsic mode function components specifically includes: using a dual-index comprehensive evaluation method based on reconstruction accuracy and modal independence to determine the first decomposition layer used for variational mode decomposition of the vibration signal, and performing variational mode decomposition on the vibration signal according to the first decomposition layer to obtain a plurality of first intrinsic mode function components; Perform variational mode decomposition on the current signal to obtain multiple second intrinsic mode function components, specifically including: determining the second decomposition layer number used for variational mode decomposition of the current signal by using a dual-index comprehensive evaluation method based on reconstruction accuracy and mode independence, and performing variational mode decomposition on the current signal according to the second decomposition layer number to obtain multiple second intrinsic mode function components.

4. The motor bearing fault diagnosis method according to claim 3, characterized in that, Determine the first decomposition layer number used for variational mode decomposition of the vibration signal by using a dual-index comprehensive evaluation method based on reconstruction accuracy and mode independence, specifically including: designing multiple first initial decomposition layer numbers; for each of the first initial decomposition layer numbers, performing variational mode decomposition on the vibration signal according to the first initial decomposition layer number to obtain multiple first initial intrinsic mode function components, calculating the sum value of all the first initial intrinsic mode function components to obtain a reconstructed vibration signal, calculating the first root mean square error at the first initial decomposition layer number based on the vibration signal and the reconstructed vibration signal, calculating the first average cross-correlation coefficient at the first initial decomposition layer number based on all the first initial intrinsic mode function components, calculating the first comprehensive score at the first initial decomposition layer number based on the first root mean square error and the first average cross-correlation coefficient; selecting the first initial decomposition layer number with the smallest first comprehensive score as the first decomposition layer number used for variational mode decomposition of the vibration signal; Determine the second decomposition layer number used for variational mode decomposition of the current signal by using a dual-index comprehensive evaluation method based on reconstruction accuracy and mode independence, specifically including: designing multiple second initial decomposition layer numbers; for each of the second initial decomposition layer numbers, performing variational mode decomposition on the current signal according to the second initial decomposition layer number to obtain multiple second initial intrinsic mode function components, calculating the sum value of all the second initial intrinsic mode function components to obtain a reconstructed current signal, calculating the second root mean square error at the second initial decomposition layer number based on the current signal and the reconstructed current signal, calculating the second average cross-correlation coefficient at the second initial decomposition layer number based on all the second initial intrinsic mode function components, calculating the second comprehensive score at the second initial decomposition layer number based on the second root mean square error and the second average cross-correlation coefficient; selecting the second initial decomposition layer number with the smallest second comprehensive score as the second decomposition layer number used for variational mode decomposition of the current signal.

5. The motor bearing fault diagnosis method according to claim 1, wherein Perform feature fusion and feature dimensionality reduction on the first time-domain feature, the first frequency-domain feature, the first energy feature, the second time-domain feature, the second frequency-domain feature, and the second energy feature to obtain the fused and dimension-reduced feature, specifically including: Perform feature fusion on the first time-domain feature, the first frequency-domain feature, the second time-domain feature, and the second frequency-domain feature by using the canonical correlation analysis algorithm to obtain the first fused feature; Feature fusion is performed on the first energy feature and the second energy feature by using the canonical correlation analysis algorithm to obtain a second fused feature; Feature dimensionality reduction is performed on the first fused feature and the second fused feature by using the principal component analysis algorithm to obtain a fused and dimension-reduced feature.

6. The motor bearing fault diagnosis method according to claim 1, wherein The hyperparameters of the support vector machine include a penalty parameter and a kernel function parameter.

7. A motor bearing fault diagnosis device, characterized in that, The motor bearing fault diagnosis device includes: a vibration sensor, a current sensor, and a processor; The vibration sensor is installed on the motor bearing, and the vibration sensor is used to collect the vibration signal of the motor bearing; The current sensor is installed on the motor bearing, and the current sensor is used to collect the current signal of the motor bearing; The processor is communicatively connected to the vibration sensor and the current sensor respectively, and the processor is used to execute the motor bearing fault diagnosis method according to any one of claims 1-6.

8. A computer device, comprising: A memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the computer program to implement the motor bearing fault diagnosis method according to any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, the motor bearing fault diagnosis method according to any one of claims 1-6 is implemented.

Images