Fault identification method and device of rotating mechanism, computer equipment and rotating mechanism

Abstract

The invention discloses a fault identification method and device for a rotating mechanism, computer equipment and the rotating mechanism. The method comprises the following steps: acquiring rotating speed information and a time domain vibration signal of the rotating mechanism; converting the time domain vibration signal into an angular domain vibration signal based on the rotating speed information; performing feature extraction on the angular domain vibration signal based on the autocorrelation matrix of the angular domain vibration signal and the filter coefficient to update the filter coefficient and obtain a target filter coefficient; and determining fault characteristics of the rotating mechanism based on a target vibration signal obtained by processing the angular domain vibration signal based on the target filtering coefficient. According to the method provided by the invention, the fault features of the unsteady vibration signals of the rotating mechanism can be effectively extracted, so that reliable fault diagnosis of the rotating mechanism is realized.

Description

Technical Field

[0001] This application relates to the field of equipment fault analysis technology, specifically to a fault identification method, device, computer equipment, and rotating mechanism for a rotating mechanism. Background Technology

[0002] In actual working conditions such as ports, rotating mechanisms such as bearings in hoisting and trolley equipment are in a complex working environment. Their vibration signals are affected by the combined effects of speed change and load change, exhibiting non-steady-state characteristics.

[0003] However, this unsteady characteristic causes the potential fault features of the bearing to be submerged in the complex vibration signal, making it difficult for fault identification methods in related technologies to accurately analyze the fault components from the unsteady vibration signal of the rotating mechanism in order to achieve fault diagnosis of the equipment. Summary of the Invention

[0004] This application provides a fault identification method, apparatus, computer equipment, and rotating mechanism for a rotating mechanism, aiming to solve the problem that rotating mechanisms in complex working environments are difficult to diagnose faults through their unsteady vibration signals.

[0005] In a first aspect, this application provides a fault identification method for a rotating mechanism, comprising: Obtain the rotational speed information and time-domain vibration signal of the rotating mechanism; Based on the rotational speed information, the time-domain vibration signal is converted into an angular-domain vibration signal that varies with the angle. Based on the autocorrelation matrix of the angular domain vibration signal and the preset initial filtering coefficients, feature extraction is performed on the angular domain vibration signal, and the filtering coefficients are updated based on the feature extraction results until the target filtering coefficients are obtained. The fault characteristics of the rotating mechanism are determined based on the target vibration signal obtained by processing the angular domain vibration signal using the target filtering coefficients.

[0006] In one embodiment of this application, converting the time-domain vibration signal into an angular-domain vibration signal that varies with angle based on the rotational speed information includes: Integrate the rotational speed information to obtain the angle information corresponding to the rotational speed information, and generate an isoangular grid of the angle information; Based on the isotropic grid and the rotational speed information, determine the timestamp information corresponding to each angle in the isotropic grid; The time-domain vibration signal is resampled based on the timestamp information to obtain the angular domain vibration signal.

[0007] In one embodiment of this application, the step of extracting features from the angular domain vibration signal based on the autocorrelation matrix of the angular domain vibration signal and preset initial filter coefficients, and updating the filter coefficients based on the feature extraction results, includes: The angular domain vibration signal is filtered according to the preset initial filtering coefficients to obtain the angular domain filtered signal; Based on preset frequency parameters, the periodic components in the angular domain filtered signal are matched to obtain the periodic signal components in the angular domain filtered signal; Based on the autocorrelation matrix and the periodic signal components, the weighted autocorrelation matrix obtained by weighting the autocorrelation matrix is ​​subjected to eigenvalue decomposition to obtain eigenvectors; and The eigenvectors are used as updated filter coefficients to process the angular domain vibration signal again.

[0008] In one embodiment of this application, the step of matching the periodic components in the angular domain filtered signal based on preset frequency parameters to obtain the periodic signal components in the angular domain filtered signal includes: The angular domain filtered signal is transformed by a complex exponential matrix composed of preset frequency parameters to obtain an angular domain transformed signal. The angular domain transformed signal is subjected to conjugate transpose and matrix operation to obtain the angular domain feature signal; The corner feature signal is filtered based on the threshold obtained by statistical analysis of the corner feature signal to obtain the periodic signal component in the corner filtered signal.

[0009] In one embodiment of this application, the step of using the feature vector as updated filter coefficients to process the angular domain vibration signal again includes: If the preset conditions are not met, the feature vector is used as the updated filter coefficient to process the angular domain vibration signal again. Under the condition that the preset conditions are met, the feature vector is used as the target filtering coefficient; The preset conditions include the difference between the feature value of the feature vector and the feature value of the feature vector in the previous iteration being less than a preset difference threshold, and / or the number of iterations of the filter coefficients exceeding a preset number threshold.

[0010] In one embodiment of this application, determining the fault characteristics of the rotating mechanism based on the target vibration signal obtained by processing the angular domain vibration signal using the target filtering coefficients includes: Perform a moving average operation on the target vibration signal and calculate the difference between the vibration signal after the moving average operation and the target vibration signal to obtain the residual signal; The residual signal is subjected to an envelope operation to obtain an envelope signal; The envelope signal is subjected to Fourier transform to obtain a frequency domain signal. Based on the transformed frequency domain result, the fault characteristics of the rotating mechanism are determined, wherein the fault characteristics include fault type and / or fault location.

[0011] In one embodiment of this application, the time-domain vibration signal of the rotating mechanism is obtained through the following steps: The initial vibration signal of the rotating mechanism is collected by a sensor installed on the rotating mechanism; The initial vibration signal is subjected to bandpass filtering to obtain the filtered vibration signal; The filtered vibration signal is subjected to power processing to enhance the impact component in the vibration signal, thereby obtaining the time-domain vibration signal.

[0012] Secondly, this application also provides a fault identification device for a rotating mechanism, comprising: The acquisition module is used to acquire the rotational speed information and time-domain vibration signal of the rotating mechanism; The sampling module is used to convert the time-domain vibration signal into an angular-domain vibration signal that varies with the angle based on the rotational speed information. The extraction module is used to extract features from the angular domain vibration signal based on the autocorrelation matrix of the angular domain vibration signal and the preset initial filtering coefficients, and update the filtering coefficients based on the feature extraction results until the target filtering coefficients are obtained. The processing module is used to determine the fault characteristics of the rotating mechanism based on the target vibration signal obtained by processing the angular domain vibration signal according to the target filtering coefficient.

[0013] Thirdly, this application also provides a computer device, the computer device comprising: One or more processors; Memory; and One or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the processor to implement the fault identification method for the rotating mechanism described in any of the preceding claims.

[0014] Fourthly, this application also provides a rotating mechanism, which is provided with a fault identification device for the rotating mechanism as described above, or is communicatively connected to the computer equipment described above, so as to realize fault identification of the rotating mechanism through the fault identification method of the rotating mechanism as described in any of the above claims.

[0015] The fault identification method, apparatus, computer equipment, and rotating mechanism provided in this application convert time-domain vibration signals into angle-domain vibration signals that vary with angle by using rotational speed information, thereby eliminating the influence of speed variation factors. To further eliminate the interference of rotational frequency characteristics on the fault frequency, feature extraction is performed on the angle-domain vibration signal using an autocorrelation matrix and filtering coefficients. The filtering coefficients are iteratively updated until an optimal target filtering coefficient that meets certain conditions is obtained. This ensures that the target vibration signal obtained through iterative processing accurately retains the fault components in the signal, facilitating subsequent processing to determine the fault of the rotating mechanism. The method provided in this application can effectively process the unsteady-state vibration signals of rotating mechanism bearings under the combined effects of speed variation and load variation in actual working conditions such as ports, accurately extracting fault features and providing a reliable technical means for equipment fault diagnosis. This helps to promptly detect potential equipment faults, ensure the safe and stable operation of equipment, and reduce economic losses and safety risks. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1a A schematic diagram illustrating the spectral analysis results obtained by transforming a non-stationary time-domain signal to the frequency domain, providing insights for related technologies. Figure 1b A schematic diagram illustrating the spectral analysis results in a specific frequency domain for related technologies; Figure 2 A flowchart illustrating the steps of a fault identification method for a rotating mechanism provided in an embodiment of this application; Figure 3 This application provides a schematic flowchart of steps for processing vibration signals to obtain time-domain vibration signals according to an embodiment of the present application. Figure 4 This application provides a schematic flowchart of a step for converting a time-domain vibration signal into an angular-domain vibration signal, as shown in the embodiments of the present application. Figure 5a This is a schematic diagram illustrating an embodiment of the present application based on converting data sampled at equal time intervals into resampled data sampled at equal angles. Figure 5b A schematic diagram of order resampling data within a specific order is provided as an embodiment of this application; Figure 6aThis is a schematic diagram illustrating the effect of directly performing envelope demodulation on a angular domain vibration signal according to an embodiment of this application. Figure 6b This is a schematic diagram illustrating the effect of the final extracted fault frequency features achieved through the solution provided in the embodiments of this application; Figure 7 This application provides a schematic flowchart of a step for iteratively updating filter coefficients based on a cyclic deconvolution operation, as illustrated in an embodiment of the present application. Figure 8 This is a schematic flowchart illustrating the steps for identifying periodic components in an angular domain filtered signal to obtain periodic signal components, as provided in an embodiment of this application. Figure 9 This application provides a schematic flowchart of steps for determining fault characteristics based on target vibration signals in an embodiment of the present application. Figure 10 A flowchart illustrating the completion analysis of a fault identification method for rotating equipment provided in this application embodiment; Figure 11 A schematic diagram of the structure of a fault identification device for a rotating mechanism provided in an embodiment of this application; Figure 12 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0018] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0019] In this application, the term "exemplary" is used to mean "used as an example, illustration, or description." Any embodiment described as "exemplary" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use this application. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that this application can be made without using these specific details. In other instances, well-known structures and processes are not described in detail to avoid obscuring the description of this application with unnecessary detail. Therefore, this application is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0020] To facilitate understanding of the fault identification method for rotating mechanisms provided in this application, the relevant application scenarios of the method provided in this application will be described below. Specifically, fault identification of rotating mechanisms mainly refers to analyzing and processing the vibration signals of rotating mechanisms to obtain fault components that can reflect faults in the rotating mechanisms, thereby achieving fault identification of the rotating mechanisms. However, in some complex working environments, such as in ports, rotating mechanisms like bearings in hoisting and trolley mechanisms are often affected by the combined effects of speed changes and load changes, resulting in non-steady-state characteristics in the actual collected vibration signals. These non-steady-state characteristics can cause potential bearing fault characteristics to be submerged in the complex vibration signals, making it difficult for related fault identification methods to accurately analyze the fault components from the non-steady-state vibration signals of such rotating mechanisms, thus hindering equipment diagnosis. For example, for easier understanding, please refer to... Figure 1a and Figure 1b , Figure 1a and Figure 1b This diagram illustrates a spectral analysis result obtained by transforming a non-stationary time-domain signal to the frequency domain, providing insights into related technologies. Figure 1a The results of the spectral analysis in the full frequency domain are shown, while Figure 1b Then, for a specific frequency domain within the frequency domain (corresponding to) Figure 1a A magnified schematic diagram of the spectral analysis results within the frequency domain (within the dashed area). It can be seen that the frequency band obtained through the above spectral analysis is ambiguous (which can be understood as having certain impact characteristics within a certain bandwidth of the frequency domain), rather than possessing typical impact characteristics. This causes some difficulties for subsequent feature extraction, making it impossible to accurately separate fault components and affecting the subsequent fault identification effect on this type of rotating mechanism.

[0021] To address the aforementioned technical problems, this application provides a fault identification method for rotating mechanisms, applicable to fault identification of such mechanisms subject to the combined effects of speed and load variations. By specifically processing the generated unsteady vibration signals, the fault characteristics of the rotating mechanism can be accurately extracted. Specifically, for ease of understanding, specific embodiments will be described below.

[0022] Please see Figure 2 , Figure 2 This application provides a flowchart illustrating the steps of a fault identification method for a rotating mechanism, specifically including steps S210 to S240: S210 acquires the rotational speed information and time-domain vibration signal of the rotating mechanism.

[0023] In the embodiments of this application, a rotating mechanism generally refers to a structural component capable of rotation, such as common examples including bearings and gears. Specifically, in one embodiment of this application, a rotating mechanism can also refer to a specific structural component that is affected by changes in speed and load due to complex working conditions; for example, it could be a bearing in hoisting or trolley equipment in a port setting. This application does not limit the definition of a rotating mechanism, but for ease of description, subsequent embodiments will use the aforementioned bearing in hoisting or trolley equipment in a port setting as an example.

[0024] Furthermore, the rotational speed information of the rotating mechanism can be obtained in various ways. For example, in one possible implementation, it can be acquired through contact or non-contact speed sensors. For instance, encoder-type sensors, such as incremental encoders that measure rotational speed and angle by pulse counting, or absolute encoders that read absolute angular positions, can be used to obtain the rotational angle information of the rotating mechanism and further determine its rotational speed. Alternatively, proximity switch-type sensors, such as eddy current sensors that detect key phase marks, Hall effect sensors that detect magnetic marks, or photoelectric sensors that detect reflective marks, can be used to obtain the periodic frequency of a specific mark and determine the rotational speed information. Of course, besides capturing the rotational speed information of the rotating mechanism through sensors, since the rotational frequency characteristics of the rotating mechanism are hidden in its vibration signal, the rotational speed can also be determined by analyzing the vibration signal of the rotating mechanism to extract the rotational frequency components. This application does not limit the implementation method for obtaining the rotational speed information of the rotating mechanism. However, considering that the frequency characteristics in the vibration signal can also be affected by fault components, in one embodiment of this application, the rotation speed information of the rotating mechanism is directly acquired by a speed sensor, thereby avoiding the influence of fault components.

[0025] Building upon the aforementioned foundation, in addition to capturing the rotational speed information of the rotating mechanism using a speed sensor, this application also employs a vibration sensor to collect the vibration signal of the rotating mechanism. Specifically, a vibration sensor refers to a sensor capable of capturing minute displacement changes on the bearing. It can be an acceleration sensor, a velocity sensor, or even a displacement sensor. This application does not limit the type of sensor used to collect the bearing's vibration signal; any sensor capable of acquiring the bearing's displacement over time and thus obtaining the bearing's vibration signal is within the scope of protection claimed in this application. In particular, in this application's embodiments, the vibration signal is typically a time-domain vibration signal, describing the bearing's vibration information over time.

[0026] Furthermore, it should be noted that in one embodiment of this application, the time-domain vibration signal used for subsequent fault identification can be the raw vibration signal directly acquired by the sensor. Of course, to further improve the identification effect of subsequent fault identification, the time-domain vibration signal can also be processed by performing certain targeted processing on the raw vibration signal to appropriately filter out high-frequency and low-frequency noise unrelated to the fault characteristics and appropriately enhance the fault characteristic signals in the vibration signal. For details, please refer to [link to relevant documentation]. Figure 3 , Figure 3 This application provides a flowchart illustrating the steps for processing a vibration signal to obtain a time-domain vibration signal, specifically including steps S310 to S330: S310, the initial vibration signal of the rotating mechanism is collected by a sensor installed on the rotating mechanism.

[0027] In one embodiment of this application, the initial vibration signal usually refers to the raw data directly collected by a sensor set on the rotating mechanism. It retains all the vibration characteristics of the rotating mechanism. In addition to the fault component used to indicate the fault information of the rotating mechanism and the frequency component used to indicate the rotation characteristics of the rotating mechanism, it often also contains some noise characteristics, such as high-frequency noise or low-frequency noise.

[0028] S320, the initial vibration signal is subjected to bandpass filtering to obtain the filtered vibration signal.

[0029] In one embodiment of this application, by performing bandpass filtering on the initial vibration signal, high-frequency and low-frequency noise that is irrelevant to the fault characteristics can be effectively filtered out, making the filtered vibration signal more correlated with the fault characteristic signal and thus more targeted.

[0030] S330, the filtered vibration signal is subjected to power processing to enhance the impact component in the vibration signal, thereby obtaining the time-domain vibration signal.

[0031] In one embodiment of this application, by performing power processing on the filtered vibration signal, that is, by exponentially processing the signal data, the impact component in the vibration signal can be effectively enhanced, that is, the intensity of the fault characteristic signal can be strengthened, making the fault characteristic signal in the obtained time-domain vibration signal easier to separate.

[0032] Of course, the above-described solution is merely one possible implementation of obtaining a time-domain vibration signal from the initial vibration signal for subsequent fault separation and identification. In fact, based on practical needs, bandpass filtering or exponential processing of the initial vibration signal to obtain the subsequent time-domain vibration signal for fault separation and identification is also feasible. Furthermore, other conventional preprocessing operations performed on the initial vibration signal to obtain the subsequent time-domain vibration signal for fault separation and identification, based on the solutions provided in the embodiments of this application, should also be considered within the scope of protection claimed in this application.

[0033] S220, based on the rotational speed information, the time-domain vibration signal is converted into an angular-domain vibration signal that varies with the angle.

[0034] Based on the rotational speed information and time-domain vibration signal of the rotating mechanism obtained through the solutions provided in any of the foregoing embodiments, in order to further eliminate the influence of speed change on the vibration signal, in the embodiments of this application, the time-domain vibration signal is converted into an angular domain vibration signal that varies with the angle based on the rotational speed information. That is, the original unsteady vibration signal that varies with time is converted into a signal that varies with the angle, thereby eliminating the influence of speed change and making the obtained angular domain vibration signal easier to analyze and process.

[0035] Specifically, converting time-domain vibration signals based on rotational speed information to obtain angle-varying angular-domain vibration signals typically requires establishing a time-angle mapping relationship using rotational speed information. This transforms time-varying unsteady vibration signals into angle-varying signals. For further details, please refer to [link to relevant documentation]. Figure 4 , Figure 4 This application provides a flowchart illustrating the steps for converting a time-domain vibration signal into an angular-domain vibration signal, specifically including steps S410 to S430: S410, Integrate the rotational speed information to obtain the angle information corresponding to the rotational speed information, and generate an isoangular grid of the angle information.

[0036] In one embodiment of this application, it can be understood that the rotation angle is the integral of angular velocity over time, and the rotational speed information collected by the rotational speed sensor is usually in the form of angular velocity. That is, by integrating the rotational speed information, i.e., by integrating the rotational speed data to accumulate angles, the rotational speed information can be converted into the corresponding angle information through integration. Of course, the above explanation is based on the rotational speed information being angular velocity. If the rotational speed information is linear velocity, the linear velocity can also be processed into angular velocity first through the correlation between linear velocity and angular velocity (determined based on the radius of the rotating mechanism), and then the angular velocity can be integrated to obtain the corresponding angle information. Based on this, based on a given angle interval, an equiangular grid with equal angle intervals can be generated. For example, the angles corresponding to some grid points on the equiangular grid can be (π, 2π, 3π), etc. Of course, based on actual needs, the angle interval of the equiangular grid can be set even smaller, such as 0.5π or 0.1π, etc. It can be understood that the smaller the angle interval, the richer the feature information retained by the obtained angular domain vibration signal, but the more computation is required.

[0037] S420, based on the isoangular grid and the rotational speed information, determine the timestamp information corresponding to each angle in the isoangular grid.

[0038] Based on the aforementioned foundation, and using the equiangular grid and rotational speed information, the timestamp information corresponding to each angle in the equiangular grid can be calculated in reverse, such as the time when the angle is 0.5π and the time when the angle is π. It can be understood that due to the influence of speed change, the timestamp information corresponding to these angles is not strictly equal in time interval.

[0039] S430, the time-domain vibration signal is resampled based on the timestamp information to obtain the angular domain vibration signal.

[0040] In this embodiment, by resampling the time-domain vibration signal using the aforementioned timestamp information, vibration information at each angle in the equiangular grid can be obtained, thereby further obtaining the angular-domain vibration signal that varies with the angle. Here, the time-domain vibration signal can be the initial vibration signal directly acquired by the sensor, or it can be the vibration signal after bandpass filtering or exponentiation processing; these details will not be repeated here.

[0041] In this embodiment, the above processing effectively eliminates the influence of speed variation on the time-domain vibration signal, thus eliminating the ambiguity of the spectrum caused by rotational speed fluctuations. This allows the "order" component, which is a fixed multiple of the rotational speed, to become a true single spectral line in the spectrum. This makes the resulting angular-domain vibration signal, which varies with angle, easier to analyze and process subsequently. For a clearer understanding of the above, please refer to [link to relevant documentation]. Figure 5a and Figure 5b , Figure 5a and Figure 5b This illustration shows the effect of converting data sampled at equal time intervals into order-resampled data sampled at equal angular intervals, i.e., the order spectral lines of angular domain vibration signals, as provided in this application embodiment. Figure 5a The complete order spectrum of the angular domain vibration signal is shown, while Figure 5b It provides access within a specific order domain ( Figure 5a A magnified schematic diagram of the order spectral lines in the dashed area (in the diagram).

[0042] It can be seen that, compared to Figure 1a and Figure 1b The spectrum shown is in Figure 5a and Figure 5b The order spectrum diagram shown is clearer, and the sideband characteristics on both sides of the center frequency can be clearly identified, which greatly improves the accuracy of analysis under unsteady conditions.

[0043] S230, feature extraction is performed on the angular domain vibration signal based on the autocorrelation matrix of the angular domain vibration signal and the preset initial filtering coefficients, and the filtering coefficients are updated based on the feature extraction results until the target filtering coefficients are obtained.

[0044] In one embodiment of this application, after processing the time-domain vibration signal using the aforementioned method to obtain the angular-domain vibration signal, the angular-domain vibration signal often contains fault characteristic information and frequency conversion characteristic information. For example, please refer to... Figure 6a , Figure 6a This is a schematic diagram illustrating the effect of directly performing envelope demodulation on angular domain vibration signals, as provided in an embodiment of this application.

[0045] exist Figure 6aIn the diagram, the horizontal axis represents frequency (Hz), and the vertical axis represents amplitude. It can be seen that after directly performing envelope demodulation on the angular domain vibration signal, the rotational frequency characteristics of the signal are very obvious, while the fault frequency characteristics are submerged within these characteristics, making them difficult to detect. Therefore, to more accurately identify the fault frequency characteristics in the angular domain vibration signal, one embodiment of this application provides a method that performs a cyclic deconvolution operation on the angular domain vibration signal based on the autocorrelation matrix and filter coefficients to identify the optimal filter coefficients that meet certain conditions for filtering the angular domain vibration signal, thereby retaining more significant fault characteristic signals for analyzing the fault characteristics of the rotating mechanism. For details, please refer to [link to relevant documentation]. Figure 7 , Figure 7 This application provides a flowchart illustrating the steps for iteratively updating filter coefficients based on a cyclic deconvolution operation, specifically including steps S710-S740: S710, the angular domain vibration signal is filtered according to the preset initial filtering coefficient to obtain the angular domain filtered signal.

[0046] The preset initial filter coefficients can be initialized filter coefficients, which are usually randomly generated. The goal of the angular domain filtered signal obtained by filtering the angular domain vibration signal using the filter coefficients is to retain as many significant fault characteristic signals as possible, that is, to make the angular domain filtered signal closer to the signal components indicating the fault characteristic signals. Therefore, in the embodiments of this application, the filter coefficients will be continuously optimized through iterative calculations until the final target filter coefficients are obtained.

[0047] S720, based on preset frequency parameters, match the periodic components in the angular domain filtered signal to obtain the periodic signal components in the angular domain filtered signal.

[0048] In one embodiment of this application, after filtering the angular domain vibration signal based on the filtering coefficient to obtain the angular domain filtered signal, the angular domain filtered signal will be matched and identified based on the preset frequency parameters, thereby identifying the periodic components in the angular domain filtered signal as the periodic signal components in the angular domain filtered signal.

[0049] For a clearer understanding of the above, please refer to [link / reference]. Figure 8 , Figure 8 This application provides a flowchart illustrating a process for identifying periodic components in an angular domain filtered signal to obtain periodic signal components. Specifically, it includes steps S810 to S830: S810, the angular domain filtered signal is transformed based on a complex exponential matrix composed of preset frequency parameters to obtain an angular domain transformed signal.

[0050] In this embodiment, the complex exponential matrix corresponding to the preset frequency parameter α is used. This transforms the angular domain filtered signal to match the periodic components within it, thereby highlighting fault-related features. Here, j is the imaginary unit.

[0051] S820, the angular domain transformation signal is processed by conjugate transpose and matrix operation to obtain the angular domain feature signal.

[0052] Building upon the aforementioned foundation, further processing of the angle-domain transformed signal, including conjugate transpose and matrix operations, can be performed to extract angle-domain feature signals related to the fault. For example, in one possible implementation, the angle-domain feature signals can be extracted through eigenvalue decomposition of the covariance matrix.

[0053] S830, the corner feature signal is filtered based on the threshold obtained by statistical analysis of the corner feature signal to obtain the periodic signal component in the corner filtered signal.

[0054] In one embodiment of this application, after processing the corner domain feature signal through the aforementioned solution, the corner domain feature signal is further filtered based on the threshold obtained from the statistics of the corner domain feature signal, so that the periodic signal component that is significant in the signal and used to refer to the fault feature signal can be retained.

[0055] For example, in one possible implementation of this application, the significant fault characteristic signal, i.e., the periodic signal component, can be retained by using statistical methods, such as using the mean of the signal plus twice the standard deviation as a threshold, and filtering out components of the feature vector below the threshold as noise components.

[0056] The periodic component analysis module provided in this application can obtain a stationary source signal through filtering, and then match and identify the periodic components in the signal based on frequency parameters to highlight fault-related features. By combining thresholds related to statistical methods, the feature signals can be filtered, and significant fault feature signals can be retained, which facilitates further processing of the signal to update and optimize the filter coefficients.

[0057] S730, based on the autocorrelation matrix and the periodic signal components, the weighted autocorrelation matrix obtained by weighting the autocorrelation matrix is ​​subjected to eigenvalue decomposition to obtain the eigenvector.

[0058] In one embodiment of this application, after obtaining the periodic signal component used to refer to the fault characteristic signal in the signal, the corresponding eigenvector can be obtained by performing a generalized eigenvalue solution, i.e. eigenvalue decomposition, based on the autocorrelation matrix of the angular domain vibration signal and the weighted autocorrelation matrix obtained by weighting the autocorrelation matrix with the weighting coefficients corresponding to the periodic signal component.

[0059] Specifically, autocorrelation refers to the degree of similarity between a deterministic or random signal and itself at different time delays τ, measured by its autocorrelation function (ACF). This application does not limit the specific steps for solving the autocorrelation matrix of angular domain vibration data. Specifically, the autocorrelation matrix reflects the inherent correlation of the signal, providing a basis for the identification and extraction of periodic components.

[0060] Specifically, in the embodiments of this application, the periodic components, i.e. periodic signal components, are extracted and used as weighting coefficients. The influence of the periodic components related to the fault can be highlighted through weighting operations, thereby obtaining the corresponding weighted autocorrelation matrix.

[0061] Building upon the aforementioned foundation, by solving for the generalized eigenvalues ​​from the autocorrelation matrix and the weighted autocorrelation matrix, the corresponding eigenvectors can be obtained. In particular, the corresponding eigenvalues ​​can usually also be obtained. These eigenvectors and eigenvalues ​​reflect key feature information in the signal.

[0062] Specifically, the generalized eigenvalues ​​can be determined using the following formula: A * h = λ * B * h Where A is the weighted autocorrelation matrix, B is the autocorrelation matrix, h is the eigenvector, and λ is the eigenvalue.

[0063] S740, the feature vector is used as the updated filter coefficients to process the angular domain vibration signal again.

[0064] Based on the aforementioned scheme, the obtained eigenvectors can be used as updated filter coefficients to process the diagonal domain vibration signal again. This includes filtering, periodic component identification and matching, threshold screening, and weighted determination of the updated weighted autocorrelation matrix provided in the aforementioned embodiments. This process further solves for new eigenvectors and eigenvalues, which are then used as new filter coefficients. This process continues until the target filter coefficient that meets the conditions is finally obtained. This target filter coefficient can be considered as the optimal filter coefficient that can effectively retain the fault feature components in the signal.

[0065] Specifically, in one embodiment, using the feature vector as a new filter coefficient to perform a new round of iterative calculation, or determining whether it can be used as the optimal filter coefficient that can effectively preserve the fault feature components in the signal, can be determined by preset conditions. That is, when certain conditions are met, the obtained feature vector can be used as the target filter coefficient. In another embodiment, using the feature vector as the updated filter coefficient to process the angular domain vibration signal again includes: If the preset conditions are not met, the feature vector is used as the updated filter coefficient to process the angular domain vibration signal again. Under the condition that the preset conditions are met, the feature vector is used as the target filtering coefficient.

[0066] In one possible implementation, the preset conditions include the difference between the eigenvalue of the feature vector and the eigenvalue of the feature vector in the previous iteration being less than a preset difference threshold, and / or the number of iterations of the filtering coefficients exceeding a preset number threshold. That is, the difference between the eigenvalue of the feature vector and the eigenvalue of the feature vector in the previous iteration, or the rate of change of the eigenvalue, is less than a certain value, for example, in... When the iteration is complete, then, The preset error threshold is used to measure whether the change in the feature value is small enough. At this time, the feature vector can be used as the target filtering coefficient after optimization.

[0067] Of course, in addition to the methods mentioned above, it is also possible to determine whether the number of iterations exceeds a preset threshold, that is, to determine the number of iterations. , A preset maximum number of iterations is set to prevent iterations from failing to terminate. Once the maximum iteration count is reached, the iteration stops, and the current feature vector is used as the optimized target filter coefficient.

[0068] Of course, in addition to the above, it is also feasible to set other termination conditions based on actual needs to end the update of the filter coefficients and obtain the final target filter coefficients. Furthermore, through the iterative method provided above, based on the operational idea of ​​cyclic deconvolution and using cyclic stationarity theory, an iterative optimization strategy, namely maximizing the generalized Rayleigh quotient, is adopted to enhance and extract fault features from the angular domain vibration signal. That is, by "focusing" and "enhancing" the periodic statistical characteristics of the signal in the cyclic frequency domain, a blind deconvolution method is implemented, thereby accurately obtaining the optimal filter coefficients that can be used to separate fault feature signals from the frequency features of the angular domain feature signal for subsequent fault component analysis.

[0069] S240, based on the target filtering coefficients, the target vibration signal obtained by processing the angular domain vibration signal determines the fault characteristics of the rotating mechanism.

[0070] After obtaining the target filter coefficients using the aforementioned scheme, the target vibration signal obtained by processing the angular domain vibration signal based on these target filter coefficients can be used to characterize the fault features of the rotating mechanism. That is, the iteratively obtained feature vector is used as filter coefficients to filter the relevant signals. This filter can selectively extract signal components related to the fault. Therefore, based on this, the fault characteristics of the rotating mechanism can be accurately determined using relevant signal analysis and processing methods.

[0071] For details, please refer to Figure 9 , Figure 9 This application provides a flowchart illustrating steps for determining fault characteristics based on target vibration signals, specifically including steps S910-S930: S910, perform a moving average operation on the target vibration signal, and calculate the difference between the vibration signal after the moving average operation and the target vibration signal to obtain the residual signal.

[0072] In one embodiment of this application, the vibration signal can be smoothed by performing a moving average operation, which reduces the influence of random noise. Then, by subtracting the result of the moving average from the original filtered result, the resulting residual signal can further highlight the changes in fault characteristics.

[0073] S920, perform an envelope operation on the residual signal to obtain an envelope signal.

[0074] In some embodiments of this application, since fault characteristics are often reflected in changes in the envelope, the envelope of the vibration signal, especially the envelope information of the residual signal, can be further extracted to obtain an envelope signal reflecting the fault characteristics. Specifically, the envelope signal of the residual signal can be obtained by performing envelope demodulation on the residual signal.

[0075] S930, the envelope signal is subjected to Fourier transform to obtain a frequency domain signal, and the fault characteristics of the rotating mechanism are determined based on the transformed frequency domain result.

[0076] Since fault characteristics are usually manifested in the frequency domain as specific frequency components, making them easier to analyze and identify, in one embodiment of this application, the fault characteristics of the rotating mechanism can be determined based on the transformed frequency domain signal by performing a Fourier transform on the envelope signal in the time domain.

[0077] Specifically, fault characteristics typically include information such as fault type and / or fault location, which are usually related to a specific frequency domain of the frequency domain signal (e.g., a frequency with impulse signal characteristics) and its amplitude. These correlations can often be obtained through experimental testing in advance, and this application does not limit this.

[0078] Of course, the aforementioned solution is merely one approach to analyze fault characteristics based on fault-related signal components, i.e., the target vibration signal. In fact, based on practical needs, it is also feasible to directly envelop the target vibration signal and convert it into a frequency domain signal for fault characteristic analysis. Furthermore, based on the solution provided in this application, other related technologies for analyzing and processing fault characteristic signals to determine fault characteristics are also within the scope of protection claimed in this application. However, the solution provided above allows for more accurate fault identification of rotating mechanisms. For details, please refer to... Figure 6b , Figure 6b This is a schematic diagram illustrating the effect of the final extracted fault frequency features achieved through the solution provided in the embodiments of this application. Figure 6b In the diagram, the horizontal axis represents frequency (in Hz), and the vertical axis represents amplitude.

[0079] and Figure 6a In contrast, the above method can clearly identify fault characteristics and their harmonics, and there are obvious characteristics with the frequency as a sideband around the fault frequency.

[0080] To clearly understand the complete steps of the fault identification method provided in the embodiments of this application, please refer to... Figure 10 , Figure 10 This is a flowchart illustrating the completion analysis of a fault identification method for rotating equipment provided in an embodiment of this application.

[0081] The specific details of the flowchart above can be found in the descriptions of the aforementioned embodiments. Specifically, the integral accumulation angle refers to obtaining the angle by accumulating and integrating the rotational speed data, especially the angular velocity. Generating an isotropic mesh refers to dividing the obtained angles into several isotropic sections, such as several angles with an angle interval of 0.1π. The inverse timestamp calculation is based on the rotational speed information to determine the time information corresponding to each grid vertex in the isotropic mesh.

[0082] Based on this, bandpass filtering is applied to the time-domain vibration signal directly acquired by the sensor to filter out high and low frequency noise. Then, power-law processing is used to enhance the noise characteristic signal components in the signal. The processed time-domain vibration signal is sampled using the time information obtained above, so as to convert the data sampled at equal time intervals into data sampled at equal angular intervals, thereby eliminating the interference of speed change on the signal and facilitating the subsequent analysis and processing of the angular domain vibration signal.

[0083] Based on this, the angular vibration signal is processed through a series of steps using the initialized filter coefficients to determine the periodic components in the signal and to weight the autocorrelation matrix. This facilitates the subsequent generalized eigenvalue calculation using the autocorrelation matrix and weighted autocorrelation matrix of the angular vibration signal. If specific conditions are not met, the currently obtained eigenvector is used as the updated filter coefficient to reprocess the angular vibration signal until the preset conditions are met. Then, the currently obtained eigenvector is used as the optimal target filter coefficient to filter the angular vibration signal, thereby selectively extracting the fault-related signal components in the angular vibration signal.

[0084] After obtaining the target vibration signal by filtering the filter coefficients, the residual signal is obtained based on the moving average. Then, the envelope signal of the residual signal is obtained by envelope demodulation, and finally transformed into the frequency range by Fourier transform. The fault characteristics of the rotating mechanism can then be analyzed by the spectral characteristics.

[0085] The fault identification method, apparatus, computer equipment, and rotating mechanism provided in this application convert time-domain vibration signals into angle-domain vibration signals that vary with angle by using rotational speed information, thereby eliminating the influence of speed variation factors. To further eliminate the interference of rotational frequency characteristics on the fault frequency, feature extraction is performed on the angle-domain vibration signal using an autocorrelation matrix and filtering coefficients. The filtering coefficients are iteratively updated until an optimal target filtering coefficient that meets certain conditions is obtained. This ensures that the target vibration signal obtained through iterative processing accurately retains the fault components in the signal, facilitating subsequent processing to determine the fault of the rotating mechanism. The method provided in this application can effectively process the unsteady-state vibration signals of rotating mechanism bearings under the combined effects of speed variation and load variation in actual working conditions such as ports, accurately extracting fault features and providing a reliable technical means for equipment fault diagnosis. This helps to promptly detect potential equipment faults, ensure the safe and stable operation of equipment, and reduce economic losses and safety risks.

[0086] Based on the aforementioned fault identification method for rotating mechanisms, this application also provides a fault identification device for rotating mechanisms. For details, please refer to [link to relevant documentation]. Figure 11 , Figure 11This application provides a schematic diagram of the structure of a fault identification device for a rotating mechanism, specifically including: The acquisition module 1110 is used to acquire the rotational speed information and time-domain vibration signal of the rotating mechanism; The sampling module 1120 is used to convert the time-domain vibration signal into an angular-domain vibration signal that varies with the angle based on the rotational speed information. The extraction module 1130 is used to extract features from the angular domain vibration signal based on the autocorrelation matrix of the angular domain vibration signal and the preset initial filtering coefficients, and update the filtering coefficients based on the feature extraction results until the target filtering coefficients are obtained. The processing module 1140 is used to determine the fault characteristics of the rotating mechanism based on the target vibration signal obtained by processing the angular domain vibration signal according to the target filtering coefficient.

[0087] In one embodiment of this application, the sampling module 1120 is further configured to integrate the rotational speed information to obtain the angle information corresponding to the rotational speed information, and generate an isoangular grid of the angle information; Based on the isotropic grid and the rotational speed information, determine the timestamp information corresponding to each angle in the isotropic grid; The time-domain vibration signal is resampled based on the timestamp information to obtain the angular domain vibration signal.

[0088] In one embodiment of this application, the extraction module 1130 is further configured to filter the angular domain vibration signal according to a preset initial filtering coefficient to obtain an angular domain filtered signal. Based on preset frequency parameters, the periodic components in the angular domain filtered signal are matched to obtain the periodic signal components in the angular domain filtered signal; Based on the autocorrelation matrix and the periodic signal components, the weighted autocorrelation matrix obtained by weighting the autocorrelation matrix is ​​subjected to eigenvalue decomposition to obtain eigenvectors; and The eigenvectors are used as updated filter coefficients to process the angular domain vibration signal again.

[0089] In one embodiment of this application, the extraction module 1130 is further configured to transform the angular domain filtered signal based on a complex exponential matrix composed of preset frequency parameters to obtain an angular domain transformed signal. The angular domain transformed signal is subjected to conjugate transpose and matrix operation to obtain the angular domain feature signal; The corner feature signal is filtered based on the threshold obtained by statistical analysis of the corner feature signal to obtain the periodic signal component in the corner filtered signal.

[0090] In one embodiment of this application, the extraction module 1130 is further configured to use the feature vector as updated filtering coefficients to process the angular domain vibration signal again if the preset conditions are not met. Under the condition that the preset conditions are met, the feature vector is used as the target filtering coefficient; The preset conditions include the difference between the feature value of the feature vector and the feature value of the feature vector in the previous iteration being less than a preset difference threshold, and / or the number of iterations of the filter coefficients exceeding a preset number threshold.

[0091] In one embodiment of this application, the processing module 1140 is further configured to perform a moving average operation on the target vibration signal and calculate the difference between the vibration signal after the moving average operation and the target vibration signal to obtain a residual signal; The residual signal is subjected to an envelope operation to obtain an envelope signal; The envelope signal is subjected to Fourier transform to obtain a frequency domain signal. Based on the transformed frequency domain result, the fault characteristics of the rotating mechanism are determined, wherein the fault characteristics include fault type and / or fault location.

[0092] In one embodiment of this application, the processing module 1140 is further configured to process the target resource image according to the distribution characteristics of similar pixels in the target resource image, using a processing strategy corresponding to the distribution characteristics, to obtain a processed target resource image.

[0093] In one embodiment of this application, the acquisition module 1110 is further configured to acquire the initial vibration signal of the rotating mechanism through a sensor disposed on the rotating mechanism; The initial vibration signal is subjected to bandpass filtering to obtain the filtered vibration signal; The filtered vibration signal is subjected to power processing to enhance the impact component in the vibration signal, thereby obtaining the time-domain vibration signal.

[0094] The fault identification device for rotating mechanisms provided in this application converts time-domain vibration signals into angle-domain vibration signals that vary with angle using rotational speed information, thereby eliminating the influence of speed variation. To further eliminate interference from rotational frequency characteristics on fault frequencies, feature extraction is performed on the angle-domain vibration signals using an autocorrelation matrix and filtering coefficients. The filtering coefficients are iteratively updated until an optimal target filtering coefficient that meets certain conditions is obtained. This ensures that the target vibration signal obtained through iterative processing accurately retains the fault components in the signal, facilitating subsequent processing to determine the faults in the rotating mechanism. The device provided in this application can effectively process unsteady-state vibration signals of rotating mechanism bearings under the combined effects of speed variation and load variation in actual working conditions such as ports, accurately extracting fault features and providing a reliable technical means for equipment fault diagnosis. This helps to promptly detect potential equipment faults, ensure the safe and stable operation of equipment, and reduce economic losses and safety risks.

[0095] In some embodiments of this application, the fault identification method for the rotating mechanism can be implemented as a computer program, which can be implemented in, for example... Figure 12 It runs on the computer device shown. The computer device's memory can store the various program modules that make up the management device of the resource file, for example, Figure 11 The diagram shows an acquisition module 1110, a sampling module 1120, an extraction module 1130, and a processing module 1140. The computer program comprised of these modules causes the processor to execute the steps in the fault identification method for the rotating mechanism of the various embodiments of this application described in this specification.

[0096] For example, Figure 12 The computer equipment shown can be used as follows Figure 11 The acquisition module 1110 in the fault identification device for the rotating mechanism shown executes step S210. The computer device can execute step S220 via the sampling module 1120. The computer device can execute step S230 via the extraction module 1130. The computer device can execute step S240 via the processing module 1140. The computer device includes a processor, memory, and a network interface connected via a system bus. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface of the computer device is used to communicate with external computer devices via a network connection. When the computer program is executed by the processor, it implements a fault identification method for a rotating mechanism.

[0097] Those skilled in the art will understand that Figure 12 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0098] In some embodiments of this application, a computer device is provided, including one or more processors; memory; and one or more application programs, wherein the one or more application programs are stored in the memory and configured to be executed by the processor to perform the following steps: Obtain the rotational speed information and time-domain vibration signal of the rotating mechanism; Based on the rotational speed information, the time-domain vibration signal is converted into an angular-domain vibration signal that varies with the angle. Based on the autocorrelation matrix of the angular domain vibration signal and the preset initial filtering coefficients, feature extraction is performed on the angular domain vibration signal, and the filtering coefficients are updated based on the feature extraction results until the target filtering coefficients are obtained. The fault characteristics of the rotating mechanism are determined based on the target vibration signal obtained by processing the angular domain vibration signal using the target filtering coefficients.

[0099] In some embodiments of this application, a computer-readable storage medium is provided, storing a computer program that is loaded by a processor, causing the processor to perform the following steps: Obtain the rotational speed information and time-domain vibration signal of the rotating mechanism; Based on the rotational speed information, the time-domain vibration signal is converted into an angular-domain vibration signal that varies with the angle. Based on the autocorrelation matrix of the angular domain vibration signal and the preset initial filtering coefficients, feature extraction is performed on the angular domain vibration signal, and the filtering coefficients are updated based on the feature extraction results until the target filtering coefficients are obtained. The fault characteristics of the rotating mechanism are determined based on the target vibration signal obtained by processing the angular domain vibration signal using the target filtering coefficients.

[0100] In one embodiment of this application, a rotating mechanism is also provided, which is provided with a fault identification device for the rotating mechanism as provided in any of the above embodiments. Alternatively, the rotating mechanism is communicatively connected to a computer device as described in any of the above embodiments, so that the computer device can identify faults in the rotating mechanism through the fault identification method for the rotating mechanism as described in any of the above embodiments.

[0101] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0102] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0103] The above provides a detailed description of the fault identification method, apparatus, computer equipment, and rotating mechanism provided in the embodiments of this application. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A fault identification method for a rotating mechanism, characterized in that, include: Obtain the rotational speed information and time-domain vibration signal of the rotating mechanism; Based on the rotational speed information, the time-domain vibration signal is converted into an angular-domain vibration signal that varies with the angle. Based on the autocorrelation matrix of the angular domain vibration signal and the preset initial filtering coefficients, feature extraction is performed on the angular domain vibration signal, and the filtering coefficients are updated based on the feature extraction results until the target filtering coefficients are obtained. The fault characteristics of the rotating mechanism are determined based on the target vibration signal obtained by processing the angular domain vibration signal using the target filtering coefficients.

2. The method according to claim 1, characterized in that, The process of converting the time-domain vibration signal into an angular-domain vibration signal that varies with angle based on the rotational speed information includes: Integrate the rotational speed information to obtain the angle information corresponding to the rotational speed information, and generate an isoangular grid of the angle information; Based on the isotropic grid and the rotational speed information, determine the timestamp information corresponding to each angle in the isotropic grid; The time-domain vibration signal is resampled based on the timestamp information to obtain the angular domain vibration signal.

3. The method according to claim 1, characterized in that, The step of extracting features from the angular domain vibration signal based on the autocorrelation matrix of the angular domain vibration signal and preset initial filter coefficients, and updating the filter coefficients based on the feature extraction results, includes: The angular domain vibration signal is filtered according to the preset initial filtering coefficients to obtain the angular domain filtered signal; Based on preset frequency parameters, the periodic components in the angular domain filtered signal are matched to obtain the periodic signal components in the angular domain filtered signal; Based on the autocorrelation matrix and the periodic signal components, the weighted autocorrelation matrix obtained by weighting the autocorrelation matrix is ​​subjected to eigenvalue decomposition to obtain eigenvectors; and The eigenvectors are used as updated filter coefficients to process the angular domain vibration signal again.

4. The method according to claim 3, characterized in that, The step of matching the periodic components in the angular domain filtered signal based on preset frequency parameters to obtain the periodic signal components in the angular domain filtered signal includes: The angular domain filtered signal is transformed by a complex exponential matrix composed of preset frequency parameters to obtain an angular domain transformed signal. The angular domain transformed signal is subjected to conjugate transpose and matrix operation to obtain the angular domain feature signal; The corner feature signal is filtered based on the threshold obtained by statistical analysis of the corner feature signal to obtain the periodic signal component in the corner filtered signal.

5. The method according to claim 3, characterized in that, The step of using the feature vector as updated filter coefficients to process the angular domain vibration signal again includes: If the preset conditions are not met, the feature vector is used as the updated filter coefficient to process the angular domain vibration signal again. Under the condition that the preset conditions are met, the feature vector is used as the target filtering coefficient; The preset conditions include the difference between the feature value of the feature vector and the feature value of the feature vector in the previous iteration being less than a preset difference threshold, and / or the number of iterations of the filter coefficients exceeding a preset number threshold.

6. The method according to claim 1, characterized in that, The method of determining the fault characteristics of the rotating mechanism based on the target vibration signal obtained by processing the angular domain vibration signal according to the target filtering coefficient includes: Perform a moving average operation on the target vibration signal and calculate the difference between the vibration signal after the moving average operation and the target vibration signal to obtain the residual signal; The residual signal is subjected to an envelope operation to obtain an envelope signal; The envelope signal is subjected to Fourier transform to obtain a frequency domain signal. Based on the transformed frequency domain result, the fault characteristics of the rotating mechanism are determined, wherein the fault characteristics include fault type and / or fault location.

7. The method according to any one of claims 1 to 6, characterized in that, The time-domain vibration signal of the rotating mechanism is obtained through the following steps: The initial vibration signal of the rotating mechanism is collected by a sensor installed on the rotating mechanism; The initial vibration signal is subjected to bandpass filtering to obtain the filtered vibration signal; The filtered vibration signal is subjected to power processing to enhance the impact component in the vibration signal, thereby obtaining the time-domain vibration signal.

8. A fault identification device for a rotating mechanism, characterized in that, include: The acquisition module is used to acquire the rotational speed information and time-domain vibration signal of the rotating mechanism; The sampling module is used to convert the time-domain vibration signal into an angular-domain vibration signal that varies with the angle based on the rotational speed information. The extraction module is used to extract features from the angular domain vibration signal based on the autocorrelation matrix of the angular domain vibration signal and the preset initial filtering coefficients, and update the filtering coefficients based on the feature extraction results until the target filtering coefficients are obtained. The processing module is used to determine the fault characteristics of the rotating mechanism based on the target vibration signal obtained by processing the angular domain vibration signal according to the target filtering coefficient.

9. A computer device, characterized in that, The computer device includes: One or more processors; Memory; and One or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the processor to implement the fault identification method for the rotating mechanism according to any one of claims 1 to 7.

10. A rotating mechanism, characterized in that, The rotating mechanism is provided with a fault identification device for the rotating mechanism as described in claim 8, or is communicatively connected to a computer device as described in claim 9, so as to realize fault identification of the rotating mechanism through the fault identification method for the rotating mechanism as described in any one of claims 1 to 7.

Images