Motor bearing fault diagnosis method and device based on weighted sparse subspace clustering

Abstract

The invention discloses a motor bearing fault diagnosis method and device based on weighted sparse subspace clustering. The motor bearing fault diagnosis method based on weighted sparse subspace clustering comprises the steps of obtaining a sample set, wherein the sample set comprises vibration signals of bearings under different fault types; acquiring a vibration acceleration signal of the to-be-diagnosed bearing; performing feature extraction on the vibration acceleration signal of the to-be-diagnosed bearing so as to obtain a to-be-diagnosed feature vector; obtaining a weighted item according to the vibration acceleration signal of the bearing to be diagnosed and the sample set; according to the weighted item, the to-be-diagnosed feature vector and the sample set, clustering is carried out through a weighted sparse subspace clustering method, and a clustering result is obtained. According to the method, the similarity degree between the data is introduced into the algorithm as a weighted item, so that the bearing fault can be accurately diagnosed.

Description

Technical Field

[0001] The present application relates to the technical field of motor bearing fault diagnosis, and in particular to a motor bearing fault diagnosis method based on weighted sparse subspace clustering and a motor bearing fault diagnosis device based on weighted sparse subspace clustering. Background Art

[0002] With the continuous improvement of the degree of automation in industrial production, the safe and stable operation of motors as core power equipment is of great significance to the entire production process. Bearings, as important supporting components of motors, are easily threatened by long-term operation under high load conditions. Therefore, studying the fault diagnosis technology of motor bearings is of great significance to ensure the normal operation of motors.

[0003] Bearing fault diagnosis usually detects the sound, temperature, smell, vibration and other information of the bearing under different working conditions. When the bearing fault is relatively minor, it is difficult to judge based on its temperature and sound signals, but the vibration signal of the bearing contains rich fault information and can reflect the movement state of the bearing over time. Therefore, effective analysis of the vibration signal is currently an important method for diagnosing bearing faults. The analysis process is generally divided into signal acquisition, signal processing, feature extraction, and decision diagnosis. Among them, signal processing and feature extraction generally select time-frequency domain methods to extract the hidden features of the signal, and decision diagnosis generally requires solving pattern recognition or clustering problems.

[0004] In recent years, algorithms such as sparse subspace clustering have been widely used in bearing fault diagnosis. Although the traditional sparse subspace clustering algorithm can divide high-dimensional data features into different low-dimensional subspaces, its solution process is based on the global expression between data, which makes the algorithm susceptible to noise, resulting in low fault diagnosis accuracy and unsuitable for engineering applications under strong noise background. Summary of the invention

[0005] The object of the present invention is to provide a motor bearing fault diagnosis method based on weighted sparse subspace clustering to solve at least one of the above-mentioned technical problems.

[0006] One aspect of the present invention provides a motor bearing fault diagnosis method based on weighted sparse subspace clustering, the motor bearing fault diagnosis method based on weighted sparse subspace clustering comprising: Acquire a sample set, wherein the sample set includes vibration signals of the bearing under different fault types; Obtaining a vibration acceleration signal of the bearing to be diagnosed; Performing feature extraction on the vibration acceleration signal of the bearing to be diagnosed, thereby obtaining a feature vector to be diagnosed; Acquire a weighted item according to the vibration acceleration signal of the bearing to be diagnosed and the sample set; According to the weighted items, the feature vector to be diagnosed and the sample set, clustering is performed by a weighted sparse subspace clustering method to obtain a clustering result.

[0007] Optionally, the step of extracting features from the vibration acceleration signal of the bearing to be diagnosed, thereby obtaining a feature vector to be diagnosed, comprises: Performing three-layer wavelet packet decomposition and reconstruction on the vibration acceleration signal; The decomposed and reconstructed low-frequency signal is subjected to singular value decomposition to obtain the feature vector to be diagnosed.

[0008] Optionally, performing singular value decomposition on the decomposed and reconstructed low-frequency signal to obtain a feature vector to be diagnosed includes: Select the low-frequency signal among the decomposed and reconstructed signals to form the initial feature matrix; The initial feature matrix is ​​subjected to singular value decomposition to obtain the feature vector to be diagnosed.

[0009] Optionally, the selecting of low-frequency signals among the decomposed and reconstructed signals to form an initial feature matrix comprises: Select the first five low-frequency node signals among the node signals of the third-layer wavelet packet transform; The five low-frequency node signals are combined to form an initial feature matrix.

[0010] Optionally, the clustering and obtaining the clustering result by a weighted sparse subspace clustering method according to the weighted item, the feature vector to be diagnosed and the sample set includes: Constructing a Gaussian similarity matrix between signals according to the feature vector to be diagnosed and the sample set; Solving the weighted sparse representation model according to the Gaussian similarity matrix to obtain a representation coefficient matrix; Calculate a similarity matrix based on the obtained representation coefficient matrix; The spectral clustering algorithm is applied to the similarity matrix to obtain the clustering results.

[0011] Optionally, applying a spectral clustering algorithm to the similarity matrix to obtain a clustering result includes: Apply the normalized cut method to the similarity matrix for spectral clustering; The clustering results of the vibration acceleration signal of the bearing to be diagnosed are obtained according to the spectral clustering results.

[0012] Optionally, the vibration signals of the bearing under different fault types include bearing health vibration signals, rolling element fault vibration signals, inner ring fault vibration signals, and outer ring fault vibration signals.

[0013] The present application also provides a motor bearing fault diagnosis device based on weighted sparse subspace clustering, characterized in that the motor bearing fault diagnosis device based on weighted sparse subspace clustering comprises: A sample set acquisition module, the sample set acquisition module is used to acquire a sample set, the sample set includes vibration signals of bearings under different fault types; A vibration acceleration signal acquisition module, wherein the vibration acceleration signal acquisition module is used to acquire a vibration acceleration signal of a bearing to be diagnosed; A module for acquiring feature vectors to be diagnosed, wherein the module is used for extracting features of the vibration acceleration signal of the bearing to be diagnosed, thereby acquiring feature vectors to be diagnosed; A sample set acquisition weighted item acquisition module, wherein the sample set acquisition weighted item acquisition module is used to acquire a weighted item according to the vibration acceleration signal of the bearing to be diagnosed and the sample set; The clustering result acquisition module is used to perform clustering and acquire clustering results by using a weighted sparse subspace clustering method according to the weighted items, the feature vector to be diagnosed and the sample set.

[0014] Beneficial effects:

[0015] The motor bearing fault diagnosis method based on weighted sparse subspace clustering proposed in this application extracts signal features by performing wavelet packet transform and singular value decomposition on the vibration acceleration signal of the bearing to be diagnosed, establishes a weighted sparse subspace using the characteristic parameters of the vibration acceleration signal of the bearing to be diagnosed and the sample set, and clusters using the representation coefficient clustering method, thereby constructing a dynamic deep model and finally realizing the identification of the fault type of the test set. Compared with the traditional sparse subspace clustering diagnosis method, the present invention introduces the similarity between data as a weighting term in the algorithm, which is conducive to accurate diagnosis of bearing faults. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a flow chart of a motor bearing fault diagnosis method based on weighted sparse subspace clustering in one embodiment of the present invention; Figure 2 is a schematic diagram of a three-layer wavelet packet decomposition process in one embodiment of the present invention; Figure 3 is a flow chart of an algorithm for weighted sparse subspace clustering in one embodiment of the present invention; Figure 4 is a vibration signal waveform diagram collected in one embodiment of the present invention; Figure 5 It is a three-dimensional effect diagram after clustering the fault characteristics of vibration signals in one embodiment of the present invention; Figure 6It is an exemplary structural diagram of an electronic device capable of implementing a motor bearing fault diagnosis method based on weighted sparse subspace clustering provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical scheme and advantages of the implementation of this application clearer, the technical scheme in the embodiment of this application will be described in more detail below in conjunction with the drawings in the embodiment of this application. In the drawings, the same or similar reference numerals throughout represent the same or similar elements or elements with the same or similar functions. The described embodiments are part of the embodiments of this application, not all of them. The embodiments described below with reference to the drawings are exemplary and are intended to be used to explain this application, and should not be construed as limitations on this application. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application. The embodiments of this application are described in detail below in conjunction with the drawings.

[0018] like Figure 1 The motor bearing fault diagnosis method based on weighted sparse subspace clustering shown includes: Step 1: Acquire a sample set, wherein the sample set includes vibration signals of bearings under different fault types; Step 2: Obtain the vibration acceleration signal of the bearing to be diagnosed; Step 3: extracting features of the vibration acceleration signal of the bearing to be diagnosed, thereby obtaining a feature vector to be diagnosed; Step 4: obtaining a weighted item according to the vibration acceleration signal of the bearing to be diagnosed and the sample set; Step 5: Based on the weighted items, the feature vector to be diagnosed and the sample set, clustering is performed by a weighted sparse subspace clustering method to obtain a clustering result.

[0019] The motor bearing fault diagnosis method based on weighted sparse subspace clustering proposed in this application extracts signal features by performing wavelet packet transform and singular value decomposition on the vibration acceleration signal of the bearing to be diagnosed, establishes a weighted sparse subspace using the characteristic parameters of the vibration acceleration signal of the bearing to be diagnosed and the sample set, and clusters using the representation coefficient clustering method, thereby constructing a dynamic deep model and finally realizing the identification of the fault type of the test set. Compared with the traditional sparse subspace clustering diagnosis method, the present invention introduces the similarity between data as a weighting term in the algorithm, which is conducive to accurate diagnosis of bearing faults.

[0020] In this embodiment, the step of extracting features from the vibration acceleration signal of the bearing to be diagnosed, thereby obtaining a feature vector to be diagnosed, includes: Performing three-layer wavelet packet decomposition and reconstruction on the vibration acceleration signal; The decomposed and reconstructed low-frequency signal is subjected to singular value decomposition to obtain the feature vector to be diagnosed.

[0021] In this embodiment, performing singular value decomposition on the decomposed and reconstructed low-frequency signal to obtain a feature vector to be diagnosed includes: Select the low-frequency signal among the decomposed and reconstructed signals to form the initial feature matrix; The initial feature matrix is ​​subjected to singular value decomposition to obtain the feature vector to be diagnosed.

[0022] In this embodiment, the low-frequency signal in the selected decomposed and reconstructed signal to form an initial feature matrix includes: Select the first five low-frequency node signals among the node signals of the third-layer wavelet packet transform; The five low-frequency node signals are combined to form an initial feature matrix.

[0023] In this embodiment, clustering and obtaining clustering results by using a weighted sparse subspace clustering method according to the weighted items, the feature vector to be diagnosed and the sample set includes: Constructing a Gaussian similarity matrix between signals according to the feature vector to be diagnosed and the sample set; Solving the weighted sparse representation model according to the Gaussian similarity matrix to obtain a representation coefficient matrix; Calculate a similarity matrix based on the obtained representation coefficient matrix; The spectral clustering algorithm is applied to the similarity matrix to obtain the clustering results.

[0024] In this embodiment, applying the spectral clustering algorithm to the similarity matrix to obtain a clustering result includes: Apply the normalized cut method to the similarity matrix for spectral clustering; The clustering results of the vibration acceleration signal of the bearing to be diagnosed are obtained according to the spectral clustering results.

[0025] In this embodiment, the vibration signals of the bearing under different fault types include a bearing health vibration signal, a rolling element fault vibration signal, an inner ring fault vibration signal, and an outer ring fault vibration signal.

[0026] In this embodiment, the similarity matrix is ​​calculated using the following formula: , Among them, the matrix U is the representation coefficient matrix of the weighted sparse representation model, and For the matrix U and The model, W is the similarity matrix, Represents the transposed matrix.

[0027] In this embodiment, the wavelet packet decomposition algorithm includes: , in, d i,j,m For the j The mth sub-band of the layer wavelet packet decomposition i coefficients ( i Indicates the index number of a frequency band. j represents the number of layers of the current decomposition, and m represents the sub-band index number); h(k) and g(k) are the wavelet packet filter coefficients, Express k The sum of variables, Indicates adjusting the signal position.

[0028] In this embodiment, the wavelet packet reconstruction algorithm includes: , in, d i,j,m For the j The first layer of wavelet packet decomposition m The sub-band i coefficients; h(k) and g(k) are the wavelet packet filter coefficients.

[0029] In this embodiment, the singular value decomposition formula is: , Among them, the matrix A yes m×n The initial feature matrix of dimension U yes m×m dimensional left singular value matrix, matrix yes n×n The right singular value matrix of yes m×n dimensional diagonal matrix, the elements on the main diagonal are the singular values ​​of the matrix A.

[0030] In this embodiment, the signal feature vector is: , in, is the signal feature vector, A matrix arranged from large to small The values ​​of the elements on the main diagonal, express The transpose of .

[0031] In this embodiment, the Gaussian similarity function formula is: , in, Represents two signals d i and d j Gaussian similarity of ; exp represents the natural exponential function; d i and d j are the eigenvectors of different groups of signals, Indicates signal d i and d j The square of the Euclidean distance between is the standard deviation of the Gaussian distribution. The distance is mapped to the similarity through the Gaussian function. The smaller the distance, the higher the similarity. The closer it is to 1.

[0032] In this embodiment, the Gaussian similarity matrix between the signals is for: , in, is the Gaussian similarity matrix, is the element in the matrix, representing the signal d i and d j Gaussian similarity.

[0033] In this embodiment, the weighted sparse representation model is: , in, is the modulus of the sparse vector, st It means it is restricted by, that is, the constraint condition, d represents the signal feature vector, D is the dataset processed by the algorithm, u To represent the coefficients, the model is subject to the condition that the data set is self-representative and cannot be represented by itself alone, .

[0034] In this embodiment, the present application first extracts signal features from the collected test set vibration signals through wavelet packet transform and singular value decomposition, then calculates the similarity between the test set and all data in the sample set through a Gaussian similarity function as a weighted term, introduces the weighted term into the sparse subspace clustering algorithm, and finally completes the fault diagnosis of the bearing through the clustering results of the algorithm.

[0035] Therefore, it can strengthen the connection between data with high similarity according to the similarity between signal features, which is conducive to clustering signal features under the same fault type. Compared with the traditional sparse subspace clustering diagnosis method, the present invention introduces the similarity between data as a weighted item in the algorithm, which is conducive to accurate diagnosis of bearing faults.

[0036] In this embodiment, in order to refine the frequency segmentation and improve the signal analysis effect, the collected vibration acceleration signal of the bearing to be diagnosed is decomposed and reconstructed by three layers of wavelet packets. That is, the original signal is first divided into 23 = 8 frequency bands by using the wavelet packet decomposition algorithm, and then the stray signals of each frequency band are reconstructed by using the wavelet packet reconstruction algorithm, thereby completing the refinement of the original signal into 8 frequency bands.

[0037] Wavelet packet decomposition has a strong ability to analyze spectrum refinement. K-layer wavelet packet decomposition can divide the original frequency band into 2 k There are sub-bands, and each band does not overlap or miss anything, so the wavelet packet decomposition can be regarded as a filter bank. The schematic diagram of the three-layer wavelet packet decomposition is as follows: Figure 2 As shown, A represents the low frequency part, D represents the high frequency part, and the original signal is divided into 8 frequency bands.

[0038] In this embodiment, the reconstructed signals of each frequency band need to be screened to extract the characteristic values ​​of the signals, and the first five low-frequency node signals with rich fault information among the eight node signals are selected to form an initial characteristic matrix for singular value decomposition.

[0039] Through the above method, it is possible to effectively extract the small fault features that are easily submerged in the noise. However, it is difficult to achieve such an extraction effect by directly performing time domain or frequency domain analysis on the vibration fault impact signal of the bearing. Therefore, the method proposed in the embodiment of the present invention can effectively extract the fault features of the signal, which is conducive to improving the overall diagnostic accuracy.

[0040] In this embodiment, the sample set can be obtained by the following method: Select several groups of bearing vibration signals under different fault types from the database, including several groups of vibration acceleration signals under the working conditions of bearing health, rolling element fault, inner ring fault, and outer ring fault to form a sample set. Calculate the Gaussian similarity function value between all data in the test set and the sample set. The Gaussian similarity function formula is: , in, Represents two signals d i and d j Gaussian similarity of ; exp represents the natural exponential function; di and d j are the eigenvectors of different groups of signals, Indicates signal d i and d j The square of the Euclidean distance between is the standard deviation of the Gaussian distribution. The distance is mapped to the similarity through the Gaussian function. The smaller the distance, the higher the similarity. The closer it is to 1.

[0041] Afterwards, the calculation results are combined into a Gaussian similarity matrix : , in, is the Gaussian similarity matrix, is the element in the matrix, representing the signal d i and d j Gaussian similarity.

[0042] Solve the weighted sparse representation model to obtain the solution matrix U : , in, is the modulus of the sparse vector, st It means it is restricted by, that is, the constraint condition, d represents the signal feature vector, D is the dataset processed by the algorithm, u To represent the coefficients, the model is subject to the condition that the data set is self-representative and cannot be represented by itself alone, .

[0043] Using the obtained expression coefficient matrix U Solving the similarity matrix W : , Among them, the matrix U is the representation coefficient matrix of the weighted sparse representation model, and For the matrix U and The model, W is the similarity matrix, Represents the transposed matrix.

[0044] By applying the spectral clustering algorithm to the solved similarity matrix W, the clustering results of the data can be obtained.

[0045] Figure 3 This is the algorithm flow chart of the above method.

[0046] The weighted sparse subspace clustering algorithm proposed in this application can establish a stronger connection between data with high similarity based on the similarity between the data, thereby increasing the representation coefficient between similar data, which is beneficial to improving the algorithm's ability to cluster similar data.

[0047] The present application is further described in detail below by way of examples. It should be understood that the examples do not constitute any limitation to the present application.

[0048] The acceleration sensor model for collecting vibration acceleration signals is HK8100, with an output sensitivity of 50mV / g. The vibration signals are collected by a data acquisition card and stored in a computer, with a sampling frequency of 10kHz. The test was conducted on a bearing with an inner ring fault that was confirmed to have pitting or spalling. The bearing was a cylindrical roller bearing of model BC1B326441A / HB1 produced by SKF. The motor speed was controlled to 18Hz and run for a period of time to collect vibration signals. Figure 4 The waveform diagram of the vibration signal of the inner ring fault bearing collected is shown in Figure 1. The method proposed in this application is programmed using Matlab software.

[0049] In order to verify the effectiveness of the method proposed in this application, the collected vibration signal is first transformed into a three-layer wavelet packet using the db4 wavelet. Then, the first five nodes of the eight nodes in the third layer are selected to form the initial fault feature matrix of the bearing, and the initial feature matrix is ​​subjected to singular value decomposition to extract the feature vector to be diagnosed of the group of vibration acceleration signals as follows: , Similarly, several groups of data are selected from the database to form a sample set, which includes 5 groups of vibration signals under bearing health, rolling element failure, inner ring failure, and outer ring failure, for a total of 20 groups. The fault feature vectors of the signals are also extracted by wavelet packet transform and singular value decomposition for the 20 groups of vibration signals in the sample set, and the fault feature matrix of the sample set is formed as follows. The matrix has 20 rows, each of which represents a feature vector extracted from a group of vibration signals.

[0050]

[0051] Calculate the Gaussian similarity function between all data and and Together we perform weighted sparse subspace clustering. Figure 5The three-dimensional clustering results of weighted sparse subspace clustering are shown. In the figure, the red circle represents the clustering results of healthy bearings in the sample set, the green triangle represents the clustering results of rolling element faulty bearings in the sample set, the blue star represents the clustering results of inner ring faulty bearings in the sample set, the black diamond represents the clustering results of outer ring faulty bearings in the sample set, and the purple rectangle represents the clustering results of the test set. Because the final step of the weighted sparse subspace clustering algorithm is to perform K-means clustering, the three axes Vector 1, Vector 2, and Vector 3 of the three-dimensional graph respectively represent the first three elements of the feature vector calculated before the final K-means clustering of each data point, that is, the coordinates of the points in the figure are the sizes of the first three values ​​of the feature vector taken out of each data point under the premise of K-means clustering. The spatial relative positions of the points in the three-dimensional graph show the clustering effect of the algorithm. Although the three-dimensional graph cannot display all clustering information, it can reflect the data processing capability and clustering effect of the weighted sparse subspace clustering algorithm to a certain extent. Figure 5 It can be observed that the test set data is clustered into the category of inner ring fault bearings, which is consistent with the actual situation and successfully verifies the effectiveness of the method proposed in this invention. Figure 5 An ideal classification effect is presented, that is, the distances between data of the same fault type are close, while the distances between data of different fault types are far.

[0052] The present application also provides a motor bearing fault diagnosis device based on weighted sparse subspace clustering, the motor bearing fault diagnosis device based on weighted sparse subspace clustering includes a sample set acquisition module, a vibration acceleration signal acquisition module, a feature vector acquisition module to be diagnosed, a sample set acquisition weighted item acquisition module and a clustering result acquisition module, wherein: The sample set acquisition module is used to acquire a sample set, wherein the sample set includes vibration signals of bearings under different fault types; The vibration acceleration signal acquisition module is used to acquire the vibration acceleration signal of the bearing to be diagnosed; The module for acquiring the feature vector to be diagnosed is used for extracting the feature of the vibration acceleration signal of the bearing to be diagnosed, so as to obtain the feature vector to be diagnosed; The sample set acquisition weighted item acquisition module is used to acquire the weighted item according to the vibration acceleration signal of the bearing to be diagnosed and the sample set; The clustering result acquisition module is used to perform clustering and acquire clustering results by using a weighted sparse subspace clustering method according to the weighted items, the feature vector to be diagnosed and the sample set.

[0053] The motor bearing fault diagnosis method based on weighted sparse subspace clustering provided by the present application first extracts signal features from the collected test set vibration signal through wavelet packet transformation and singular value decomposition, then calculates the similarity of signal features between the test set and the sample set through the Gaussian similarity function as a weighted term, introduces the weighted term into the sparse subspace clustering algorithm, and finally completes the fault diagnosis of the bearing through the clustering result of the algorithm. Moreover, the present invention introduces the similarity between data as a weighted term in the algorithm, which is conducive to accurate diagnosis of bearing faults.

[0054] It should be noted that the above explanations of the method embodiment are also applicable to the device of this embodiment and will not be repeated here.

[0055] The present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor. When the processor executes the computer program, the above-mentioned motor bearing fault diagnosis method based on weighted sparse subspace clustering is implemented.

[0056] The present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it can implement the above-mentioned motor bearing fault diagnosis method based on weighted sparse subspace clustering.

[0057] Figure 6 It is an exemplary structural diagram of an electronic device capable of implementing a motor bearing fault diagnosis method based on weighted sparse subspace clustering provided according to an embodiment of the present application.

[0058] like Figure 6 As shown, the electronic device includes an input device 501, an input interface 502, a central processor 503, a memory 504, an output interface 505, and an output device 506. The input interface 502, the central processor 503, the memory 504, and the output interface 505 are interconnected through a bus 507, and the input device 501 and the output device 506 are connected to the bus 507 through the input interface 502 and the output interface 505, respectively, and then connected to other components of the electronic device. Specifically, the input device 501 receives input information from the outside, and transmits the input information to the central processor 503 through the input interface 502; the central processor 503 processes the input information based on the computer executable instructions stored in the memory 504 to generate output information, temporarily or permanently stores the output information in the memory 504, and then transmits the output information to the output device 506 through the output interface 505; the output device 506 outputs the output information to the outside of the electronic device for use by the user.

[0059] That is to say, Figure 6The electronic device shown can also be implemented as comprising: a memory storing computer executable instructions; and one or more processors, which can implement the combination when executing the computer executable instructions. Figure 1 A motor bearing fault diagnosis method based on weighted sparse subspace clustering is described.

[0060] In one embodiment, Figure 6 The electronic device shown can be implemented to include: a memory 504 configured to store executable program code; and one or more processors configured to run the executable program code stored in the memory 504 to execute the motor bearing fault diagnosis method based on weighted sparse subspace clustering in the above-mentioned embodiment.

[0061] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0062] Memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0063] Computer readable media include permanent and non-permanent, removable and non-removable, and the media can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disk read-only memory (CD-ROM), data versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device.

[0064] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and a module, a program segment or a part of a code includes one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes identified in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or the total flow chart can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0065] The processor referred to in this embodiment may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0066] The memory can be used to store computer programs and / or modules. The processor realizes various functions of the device / terminal equipment by running or executing the computer programs and / or modules stored in the memory and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data created according to the use of the mobile phone (such as audio data, phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0067] In this embodiment, if the module / unit integrated in the device / terminal equipment is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device that can carry computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. Although the present application is disclosed as above with preferred embodiments, it is not intended to limit the present application. Any person skilled in the art may make possible changes and modifications without departing from the spirit and scope of the present application.

[0068] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0069] Furthermore, it is obvious that the word "comprising" does not exclude other elements or steps.

[0070] Although the present invention has been described in detail above with general descriptions and specific embodiments, it is obvious to those skilled in the art that some modifications or improvements may be made thereto based on the present invention. Therefore, these modifications or improvements made without departing from the spirit of the present invention all fall within the scope of protection claimed by the present invention.

Claims

1. A motor bearing fault diagnosis method based on weighted sparse subspace clustering, characterized in that: The motor bearing fault diagnosis method based on weighted sparse subspace clustering includes: Acquire a sample set, wherein the sample set includes vibration signals of the bearing under different fault types; Obtaining a vibration acceleration signal of the bearing to be diagnosed; Performing feature extraction on the vibration acceleration signal of the bearing to be diagnosed, thereby obtaining a feature vector to be diagnosed; Acquire a weighted item according to the vibration acceleration signal of the bearing to be diagnosed and the sample set; According to the weighted items, the feature vector to be diagnosed and the sample set, clustering is performed by a weighted sparse subspace clustering method to obtain a clustering result.

2. The motor bearing fault diagnosis method based on weighted sparse subspace clustering according to claim 1, characterized in that: The step of extracting features from the vibration acceleration signal of the bearing to be diagnosed, thereby obtaining a feature vector to be diagnosed, comprises: Performing three-layer wavelet packet decomposition and reconstruction on the vibration acceleration signal; The decomposed and reconstructed low-frequency signal is subjected to singular value decomposition to obtain the feature vector to be diagnosed.

3. The motor bearing fault diagnosis method based on weighted sparse subspace clustering according to claim 2, characterized in that: The step of performing singular value decomposition on the decomposed and reconstructed low-frequency signal to obtain the feature vector to be diagnosed comprises: Select the low-frequency signal among the decomposed and reconstructed signals to form the initial feature matrix; The initial feature matrix is ​​subjected to singular value decomposition to obtain the feature vector to be diagnosed.

4. The motor bearing fault diagnosis method based on weighted sparse subspace clustering according to claim 3, characterized in that: The low-frequency signal in the selected decomposed and reconstructed signal is used to form an initial feature matrix, which includes: Select the first five low-frequency node signals among the node signals of the third-layer wavelet packet transform; The five low-frequency node signals are combined to form an initial feature matrix.

5. The motor bearing fault diagnosis method based on weighted sparse subspace clustering according to claim 4, characterized in that: The step of clustering and obtaining a clustering result by a weighted sparse subspace clustering method according to the weighted item, the feature vector to be diagnosed and the sample set comprises: Constructing a Gaussian similarity matrix between signals according to the feature vector to be diagnosed and the sample set; Solving the weighted sparse representation model according to the Gaussian similarity matrix to obtain a representation coefficient matrix; Calculate a similarity matrix based on the obtained representation coefficient matrix; The spectral clustering algorithm is applied to the similarity matrix to obtain the clustering results.

6. The motor bearing fault diagnosis method based on weighted sparse subspace clustering according to claim 5, characterized in that: The step of applying a spectral clustering algorithm to the similarity matrix to obtain a clustering result comprises: Apply the normalized cut method to the similarity matrix for spectral clustering; The clustering results of the vibration acceleration signal of the bearing to be diagnosed are obtained according to the spectral clustering results.

7. The motor bearing fault diagnosis method based on weighted sparse subspace clustering according to claim 6, characterized in that: The vibration signals of the bearing under different fault types include a bearing health vibration signal, a rolling element fault vibration signal, an inner ring fault vibration signal, and an outer ring fault vibration signal.

8. A motor bearing fault diagnosis device based on weighted sparse subspace clustering, characterized in that: The motor bearing fault diagnosis device based on weighted sparse subspace clustering comprises: A sample set acquisition module, the sample set acquisition module is used to acquire a sample set, the sample set includes vibration signals of bearings under different fault types; A vibration acceleration signal acquisition module, wherein the vibration acceleration signal acquisition module is used to acquire a vibration acceleration signal of a bearing to be diagnosed; A module for acquiring feature vectors to be diagnosed, wherein the module is used for extracting features of the vibration acceleration signal of the bearing to be diagnosed, thereby acquiring feature vectors to be diagnosed; A sample set acquisition weighted item acquisition module, wherein the sample set acquisition weighted item acquisition module is used to acquire a weighted item according to the vibration acceleration signal of the bearing to be diagnosed and the sample set; The clustering result acquisition module is used to perform clustering and acquire clustering results by weighted sparse subspace clustering method according to the weighted items, the feature vector to be diagnosed and the sample set.

Images