Fan bolt loosening monitoring and identification method and system based on acoustic emission technology

Abstract

The application provides a fan bolt loosening monitoring and identification method and system based on acoustic emission technology. The method comprises: determining the acoustic source coordinates of the loosening bolt by acquiring sensor coordinates and acoustic emission signal transmission time length, determining the fan working state label data according to the acquired acoustic emission signal, performing adaptive noise reduction processing to obtain an acoustic emission noise reduction signal, and performing acoustic-electric coupling analysis to obtain an acoustic emission noise reduction effective signal, then evaluating the effective acoustic emission signal, processing through a preset cross-correlation function to obtain the stress mutation state of adjacent bolts, if it is a stress mutation bolt, acquiring the acoustic emission energy for processing to determine the pre-tightening force change, and performing real-time monitoring through threshold comparison. The application quantifies the cooperative stress relationship of multiple bolts, predicts the failure risk of grouped bolts, constructs dynamic noise reduction, and improves the signal quality in a complex environment, thereby realizing the fan bolt loosening monitoring and identification based on acoustic emission technology.

Description

Technical Field

[0001] This application relates to the field of wind turbine equipment condition monitoring technology, and more specifically, to a method and system for monitoring and identifying loose wind turbine bolts based on acoustic emission technology. Background Technology

[0002] As the foundation for connecting key components such as the tower, hub, and main shaft, the operating status of wind turbine bolts directly affects the safety and reliability of the unit. Traditional bolt failure monitoring methods mostly focus on monitoring individual bolts independently, without considering the mechanical coupling effect between groups of bolts. In actual operating conditions, the loosening of one bolt can lead to a redistribution of load on adjacent bolts, triggering a chain reaction of failures. Furthermore, the operating environment of wind turbines is complex, including blade aerodynamic noise of 10-100kHz, gearbox mechanical vibration noise, and environmental interference caused by weather conditions such as rain and snow. Existing noise reduction methods for wind turbine bolt monitoring, such as wavelet thresholding, are insufficient to dynamically adapt to the noise characteristics under different operating conditions. Therefore, there is an urgent need for an intelligent monitoring method for wind turbine bolts that quantifies the collaborative stress relationship of multiple bolts, predicts the failure risk of groups of bolts, and improves signal quality in complex environments.

[0003] Effective technical solutions are urgently needed to address the above problems. Summary of the Invention

[0004] The purpose of this application is to provide a method and system for monitoring and identifying loose bolts in wind turbines based on acoustic emission technology. By quantifying the collaborative stress relationship of multiple bolts, the failure risk of grouped bolts can be predicted, dynamic noise reduction can be constructed, and signal quality in complex environments can be improved, thereby realizing the monitoring and identification of loose bolts in wind turbines based on acoustic emission technology.

[0005] Firstly, this application provides a method for monitoring and identifying loose bolts in wind turbines based on acoustic emission technology, comprising the following steps:

[0006] The sensor coordinates, acoustic emission signal, and acoustic emission signal transmission duration of the preset sensor are obtained and processed to obtain the sound source coordinates of the loose bolt.

[0007] The acoustic emission signal is processed through a preset wind turbine operating status evaluation model to obtain wind turbine operating status label data;

[0008] The wind turbine operating status tag data is processed using a preset noise reduction method to obtain an acoustic emission noise reduction signal, and then subjected to acoustic-electric coupling analysis to obtain an effective acoustic emission noise reduction signal.

[0009] The effective acoustic emission signal is processed according to the acoustic emission noise reduction signal to obtain an effective acoustic emission signal, including the effective signal of the loose bolt and the effective signal of the adjacent bolt;

[0010] The effective acoustic emission signal is processed through a preset cross-correlation function to obtain the stress change state of adjacent bolts, including bolts without stress change or bolts with stress change.

[0011] If the stress change state is a stress change bolt, then the acoustic emission energy of the loose bolt is obtained;

[0012] The change in preload of the bolt subjected to sudden force change is obtained by processing the acoustic emission energy in combination with the preset mapping table of acoustic emission energy and preload change and the preload change transfer matrix.

[0013] The change in preload force is compared with a preset preload force variable early warning threshold.

[0014] If the change in preload is less than or equal to the preset preload variable warning threshold, no warning will be output.

[0015] If the change in preload exceeds the preset preload variable warning threshold, a warning response will be output.

[0016] Optionally, in the wind turbine bolt loosening monitoring and identification method based on acoustic emission technology described in this application, the step of acquiring the sensor coordinates of a preset sensor, the acoustic emission signal, and the acoustic emission signal transmission duration, and processing them to obtain the sound source coordinates of the loose bolt, includes:

[0017] Obtain the sensor coordinates, acoustic emission signal, and acoustic emission signal transmission duration of the preset sensor;

[0018] The sensor coordinates and acoustic emission signal transmission time are combined with a preset acoustic propagation speed and processed using a nonlinear least squares method to obtain the sound source coordinates of the loose bolt.

[0019] Optionally, in the wind turbine bolt loosening monitoring and identification method based on acoustic emission technology described in this application, the step of processing the acoustic emission signal through a preset wind turbine operating status evaluation model to obtain wind turbine operating status label data includes:

[0020] The acoustic emission signal is processed by short-time Fourier transform to obtain the corresponding acoustic emission time-frequency diagram;

[0021] The acoustic emission time-frequency diagram is input into a preset wind turbine operating status evaluation model to obtain wind turbine operating status label data corresponding to the acoustic emission signal.

[0022] Optionally, in the wind turbine bolt loosening monitoring and identification method based on acoustic emission technology described in this application, the step of processing the wind turbine operating status tag data through a preset noise reduction method to obtain an acoustic emission noise reduction signal, and performing acoustic-electric coupling analysis to obtain an effective acoustic emission noise reduction signal, includes:

[0023] Based on the wind turbine operating status label data, query the preset operating status and noise reduction method mapping table to obtain the corresponding noise reduction method;

[0024] The noise reduction process is performed using the aforementioned noise reduction method to obtain an acoustic emission noise-reduced signal.

[0025] Acquire the micro-current signal corresponding to the acoustic emission signal bolt;

[0026] The acoustic emission noise reduction signal and the microcurrent signal are processed by a preset mutual information value processing method to obtain a mutual information value;

[0027] If the mutual information value is less than or equal to a preset mutual information qualification threshold, the acoustic emission noise reduction signal is discarded.

[0028] If the mutual information exceeds the preset threshold, the acoustic emission noise reduction signal is retained to obtain an effective acoustic emission noise reduction signal.

[0029] Optionally, in the wind turbine bolt loosening monitoring and identification method based on acoustic emission technology described in this application, the step of processing the effective acoustic emission noise reduction signal to obtain an effective acoustic emission signal, including an effective signal of the loose bolt and an effective signal of the adjacent bolt, includes:

[0030] Obtain the total acoustic emission energy corresponding to the loose bolt and the adjacent bolt;

[0031] The effective acoustic emission noise reduction signal is processed by short-time Fourier transform to obtain the preset characteristic frequency band energy spectrum corresponding to the loose bolt and the adjacent bolt;

[0032] The energy spectrum of the preset characteristic frequency band is processed to obtain the preset characteristic frequency band energy values ​​corresponding to the loose bolt and the adjacent bolt;

[0033] The preset characteristic frequency band energy value is compared with the total acoustic emission energy value to obtain the preset characteristic frequency band energy ratio;

[0034] The preset characteristic frequency band energy ratio is compared with the preset effective energy ratio threshold.

[0035] If the effective energy percentage is less than or equal to the preset threshold, the effective acoustic emission noise reduction signal is determined to be an interference signal.

[0036] If the effective energy percentage is greater than the preset threshold, the effective acoustic emission noise reduction signal is determined to be an effective acoustic emission signal, including the effective signal of the loose bolt and the effective signal of the adjacent bolt.

[0037] Optionally, in the wind turbine bolt loosening monitoring and identification method based on acoustic emission technology described in this application, the step of processing the effective acoustic emission signal through a preset cross-correlation function to obtain the stress change state of adjacent bolts, including bolts without stress change or bolts with stress change, includes:

[0038] The effective signals of the loosened bolts and the effective signals of the adjacent bolts are calculated and normalized using a preset cross-correlation function to obtain the corresponding cross-correlation coefficients.

[0039] The cross-correlation coefficient is compared with a preset cross-correlation threshold.

[0040] If the cross-correlation coefficient is less than or equal to the preset cross-correlation threshold, then the stress change state of the adjacent bolts is determined to be a bolt without stress change.

[0041] If the cross-correlation coefficient is greater than the preset cross-correlation threshold, then the stress change state of the adjacent bolts is determined to be a stress change bolt.

[0042] Optionally, in the wind turbine bolt loosening monitoring and identification method based on acoustic emission technology described in this application, the step of processing the acoustic emission energy in conjunction with a preset acoustic emission energy-preload change mapping table and a preload change transfer matrix to obtain the preload change of the bolt subjected to sudden force change includes:

[0043] Construct a three-dimensional model of the bolt group and obtain the bolt preload attenuation value and the corresponding stress change of adjacent bolts;

[0044] Based on the bolt preload attenuation value and stress change, the influence weights of the bolt and its adjacent bolts are obtained through multiple linear regression fitting, and a preload change transfer matrix is ​​constructed.

[0045] The change in preload is obtained by querying a preset mapping table between acoustic emission energy and preload change based on the acoustic emission energy.

[0046] The change in preload is obtained by processing the change in preload in conjunction with the preload change transfer matrix.

[0047] Secondly, this application provides a wind turbine bolt loosening monitoring and identification system based on acoustic emission technology. The system includes a memory and a processor. The memory includes a program for a wind turbine bolt loosening monitoring and identification method based on acoustic emission technology. When the program for the wind turbine bolt loosening monitoring and identification method based on acoustic emission technology is executed by the processor, it performs the following steps:

[0048] The sensor coordinates, acoustic emission signal, and acoustic emission signal transmission duration of the preset sensor are obtained and processed to obtain the sound source coordinates of the loose bolt.

[0049] The acoustic emission signal is processed through a preset wind turbine operating status evaluation model to obtain wind turbine operating status label data;

[0050] The wind turbine operating status tag data is processed using a preset noise reduction method to obtain an acoustic emission noise reduction signal, and then subjected to acoustic-electric coupling analysis to obtain an effective acoustic emission noise reduction signal.

[0051] The effective acoustic emission signal is processed according to the acoustic emission noise reduction signal to obtain an effective acoustic emission signal, including the effective signal of the loose bolt and the effective signal of the adjacent bolt;

[0052] The effective acoustic emission signal is processed through a preset cross-correlation function to obtain the stress change state of adjacent bolts, including bolts without stress change or bolts with stress change.

[0053] If the stress change state is a stress change bolt, then the acoustic emission energy of the loose bolt is obtained;

[0054] The change in preload of the bolt subjected to sudden force change is obtained by processing the acoustic emission energy in combination with the preset mapping table of acoustic emission energy and preload change and the preload change transfer matrix.

[0055] The change in preload force is compared with a preset preload force variable early warning threshold.

[0056] If the change in preload is less than or equal to the preset preload variable warning threshold, no warning will be output.

[0057] If the change in preload exceeds the preset preload variable warning threshold, a warning response will be output.

[0058] Optionally, in the wind turbine bolt loosening monitoring and identification system based on acoustic emission technology described in this application, the step of acquiring the sensor coordinates of a preset sensor, the acoustic emission signal, and the transmission duration of the acoustic emission signal, and processing them to obtain the sound source coordinates of the loose bolt, includes:

[0059] Obtain the sensor coordinates, acoustic emission signal, and acoustic emission signal transmission duration of the preset sensor;

[0060] The sensor coordinates and acoustic emission signal transmission time are combined with a preset acoustic propagation speed and processed using a nonlinear least squares method to obtain the sound source coordinates of the loose bolt.

[0061] Optionally, in the wind turbine bolt loosening monitoring and identification system based on acoustic emission technology described in this application, the step of processing the acoustic emission signal through a preset wind turbine operating status evaluation model to obtain wind turbine operating status label data includes:

[0062] The acoustic emission signal is processed by short-time Fourier transform to obtain the corresponding acoustic emission time-frequency diagram;

[0063] The acoustic emission time-frequency diagram is input into a preset wind turbine operating status evaluation model to obtain wind turbine operating status label data corresponding to the acoustic emission signal.

[0064] As can be seen from the above, the wind turbine bolt loosening monitoring and identification method and system based on acoustic emission technology provided in this application, by quantifying the collaborative force relationship of multiple bolts, predicting the failure risk of grouped bolts, constructing dynamic noise reduction, and improving signal quality in complex environments, thereby realizing wind turbine bolt loosening monitoring and identification based on acoustic emission technology.

[0065] Other features and advantages of this application will be set forth in the following description and will be apparent in part from the description or may be learned by practicing embodiments of this application. The objectives and other advantages of this application may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings. Attached Figure Description

[0066] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0067] Figure 1 A flowchart illustrating the wind turbine bolt loosening monitoring and identification method based on acoustic emission technology provided in this application embodiment;

[0068] Figure 2 A flowchart illustrating the method for obtaining an effective acoustic emission noise reduction signal in the wind turbine bolt loosening monitoring and identification method based on acoustic emission technology provided in this application embodiment;

[0069] Figure 3 The flowchart illustrates the process of obtaining an effective acoustic emission signal for the wind turbine bolt loosening monitoring and identification method based on acoustic emission technology provided in this application embodiment. Detailed Implementation

[0070] 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 a part of the embodiments of this application, and not all of the embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0071] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0072] Please refer to Figure 1 , Figure 1 This is a flowchart of a wind turbine bolt loosening detection and identification method based on acoustic emission technology according to some embodiments of this application. This wind turbine bolt loosening detection and identification method based on acoustic emission technology is used in terminal devices, such as computers and mobile terminals. The wind turbine bolt loosening detection and identification method based on acoustic emission technology includes the following steps:

[0073] S11. Obtain the sensor coordinates, acoustic emission signal, and acoustic emission signal transmission duration of the preset sensor, and process them to obtain the sound source coordinates of the loose bolt.

[0074] S12. The acoustic emission signal is processed through a preset fan operating status evaluation model to obtain fan operating status label data;

[0075] S13. The fan operating status tag data is processed by a preset noise reduction method to obtain an acoustic emission noise reduction signal, and acoustic-electric coupling analysis is performed to obtain an effective acoustic emission noise reduction signal.

[0076] S14. Process the acoustic emission noise reduction effective signal to obtain an effective acoustic emission signal, including the effective signal of the loose bolt and the effective signal of the adjacent bolt;

[0077] S15. The effective acoustic emission signal is processed by a preset cross-correlation function to obtain the stress change state of adjacent bolts, including bolts without stress change or bolts with stress change.

[0078] S16. If the stress change state is a stress change bolt, then obtain the acoustic emission energy of the loose bolt;

[0079] S17. Based on the acoustic emission energy, combined with the preset acoustic emission energy and preload change mapping table and the preload change transfer matrix, the preload change of the bolt subjected to sudden force change is obtained.

[0080] S18. Compare the change in preload with a preset preload variable warning threshold.

[0081] S191. If the change in preload is less than or equal to the preset preload variable warning threshold, no warning will be output.

[0082] S192. If the change in preload is greater than the preset preload variable warning threshold, then a warning response is output.

[0083] It should be noted that, in order to achieve intelligent detection of wind turbine bolts based on acoustic emission signals, firstly, the bolt coordinates are solved based on the acquired acoustic emission signals to identify loose bolts. Then, the impact on the load redistribution of adjacent bolts is evaluated. By quantifying the collaborative force relationship of multiple bolts, the failure risk of the group of bolts is predicted. At the same time, an intelligent noise reduction processing method is used to match a pre-constructed dynamic noise library, and the validity of the acoustic emission signals is determined by combining acoustic-electric coupling. Finally, by comparing the changes in preload of adjacent bolts with thresholds, the impact of loose bolts on the preload of adjacent bolts is determined, thereby realizing the monitoring and identification of loose wind turbine bolts based on acoustic emission technology.

[0084] According to an embodiment of the present invention, the step of obtaining the sensor coordinates, acoustic emission signal, and acoustic emission signal transmission duration of a preset sensor, and processing them to obtain the sound source coordinates of the loose bolt, includes:

[0085] Obtain the sensor coordinates, acoustic emission signal, and acoustic emission signal transmission duration of the preset sensor;

[0086] The sensor coordinates and acoustic emission signal transmission time are combined with a preset acoustic propagation speed and processed using a nonlinear least squares method to obtain the sound source coordinates of the loose bolt.

[0087] It should be noted that piezoelectric sensors are arranged on the fan bolts at preset positions and numbered. After collecting the guided sound emission signal, a 50-500KHz bandpass filter is first used to remove low-frequency mechanical noise. Then, three non-collinear sensors are selected and the source coordinates of the loose bolt are solved according to the time difference positioning method based on the preset sound propagation speed to determine which bolt is loose.

[0088] According to an embodiment of the present invention, processing the acoustic emission signal through a preset wind turbine operating status evaluation model to obtain wind turbine operating status label data includes:

[0089] The acoustic emission signal is processed by short-time Fourier transform to obtain the corresponding acoustic emission time-frequency diagram;

[0090] The acoustic emission time-frequency diagram is input into a preset wind turbine operating status evaluation model to obtain wind turbine operating status label data corresponding to the acoustic emission signal.

[0091] It should be noted that, firstly, the collected acoustic emission signals are subjected to bandpass filtering, outlier removal, and frame segmentation. The preprocessed acoustic emission signals are then subjected to short-time Fourier transform to generate a 128×128 time-frequency diagram, which is then input into a preset wind turbine operating status evaluation model to determine the wind turbine operating status label data corresponding to the acoustic emission signals. The preset wind turbine operating status evaluation model is a convolutional neural network model trained using acoustic emission time-frequency diagrams of a large number of historical acoustic emission signal samples and the corresponding wind turbine operating status label data.

[0092] Please refer to Figure 2 , Figure 2 This is a flowchart illustrating the process of obtaining an effective acoustic emission noise reduction signal in a wind turbine bolt loosening monitoring and identification method based on acoustic emission technology, as described in some embodiments of this application. According to an embodiment of the present invention, the step of processing the wind turbine operating status tag data using a preset noise reduction method to obtain an acoustic emission noise reduction signal, and then performing acoustic-electric coupling analysis to obtain an effective acoustic emission noise reduction signal, includes:

[0093] S21. Query the preset working status and noise reduction method mapping table according to the fan working status label data to obtain the corresponding noise reduction method;

[0094] S22. Noise reduction processing is performed using the aforementioned noise reduction method to obtain an acoustic emission noise-reduced signal;

[0095] S23. Obtain the micro-current signal corresponding to the acoustic emission signal bolt;

[0096] S24. The acoustic emission noise reduction signal and the micro-current signal are processed by a preset mutual information value processing method to obtain a mutual information value;

[0097] S251. If the mutual information value is less than or equal to a preset mutual information qualification threshold, the acoustic emission noise reduction signal is removed.

[0098] S252. If the value is greater than the preset mutual information qualification threshold, the acoustic emission noise reduction signal is retained to obtain an effective acoustic emission noise reduction signal.

[0099] It should be noted that the corresponding noise reduction method is obtained by querying the preset working state and noise reduction method mapping table based on the obtained wind turbine working state label data. The preset working state and noise reduction method mapping table is preset by those skilled in the art based on a large number of historical samples and can be dynamically adjusted. Table 1 shows an example of the mapping between working state and noise reduction method:

[0100] Table 1. Example of mapping between working state and noise reduction method

[0101]

[0102] After noise reduction, to further eliminate residual noise unrelated to bolt stress and ensure a strong correlation between the signal and mechanical damage, a PVDF piezoelectric film sensor is attached to the head of each bolt to acquire the microcurrent signal corresponding to the acoustic emission signal bolt. A 0.1s sliding window is used to segment the noise-reduced acoustic emission signal and microcurrent signal. The mutual information value is calculated once per window using a preset mutual information value processing method. Finally, the obtained mutual information value is compared with a preset mutual information qualification threshold, and only signal segments greater than the preset mutual information qualification threshold are retained. This results in an effective acoustic emission noise reduction signal with triple purification through preprocessing, working condition matching noise reduction, and acoustic-electric coupling screening.

[0103] Please refer to Figure 3 , Figure 3 This is a flowchart illustrating the process of obtaining an effective acoustic emission signal in a wind turbine bolt loosening monitoring and identification method based on acoustic emission technology, as described in some embodiments of this application. According to embodiments of the present invention, the process of processing the acoustic emission noise-reduced effective signal to obtain an effective acoustic emission signal includes an effective signal for loose bolts and effective signals for adjacent bolts, comprising:

[0104] S31. Obtain the total acoustic emission energy corresponding to the loose bolt and the adjacent bolt;

[0105] S32. Perform short-time Fourier transform processing on the effective acoustic emission noise reduction signal to obtain the preset characteristic frequency band energy spectrum corresponding to the loose bolt and adjacent bolts;

[0106] S33. Process the energy spectrum of the preset characteristic frequency band to obtain the preset characteristic frequency band energy values ​​corresponding to the loose bolt and the adjacent bolt;

[0107] S34. Compare the preset characteristic frequency band energy value with the total acoustic emission energy value to obtain the preset characteristic frequency band energy ratio;

[0108] S35. Compare the preset characteristic frequency band energy ratio with the preset effective energy ratio threshold.

[0109] S361. If the effective energy percentage is less than or equal to the preset effective energy percentage threshold, the effective acoustic emission noise reduction signal is determined to be an interference signal.

[0110] S362. If the effective energy percentage is greater than the preset threshold, the effective acoustic emission noise reduction signal is determined to be an effective acoustic emission signal, including the effective signal of the loose bolt and the effective signal of the adjacent bolt.

[0111] It should be noted that the effective acoustic emission noise reduction signal of the 0.1s segment is processed by short-time Fourier transform, and the proportion of energy in the preset characteristic frequency band (100-150kHz characteristic frequency band) to the total signal energy is calculated. Then, a threshold comparison is performed. If it is greater than the preset effective energy proportion threshold, it means that there is an acoustic emission component related to bolt loosening in the signal segment, which is retained to obtain an effective acoustic emission signal. If it is less than or equal to the preset effective energy proportion threshold, it means that the signal segment is interference such as wind noise or mechanical vibration, which is removed.

[0112] According to an embodiment of the present invention, the step of processing the effective acoustic emission signal through a preset cross-correlation function to obtain the stress change state of adjacent bolts includes bolts without stress change or bolts with stress change, including:

[0113] The effective signals of the loosened bolts and the effective signals of the adjacent bolts are calculated and normalized using a preset cross-correlation function to obtain the corresponding cross-correlation coefficients.

[0114] The cross-correlation coefficient is compared with a preset cross-correlation threshold.

[0115] If the cross-correlation coefficient is less than or equal to the preset cross-correlation threshold, then the stress change state of the adjacent bolts is determined to be a bolt without stress change.

[0116] If the cross-correlation coefficient is greater than the preset cross-correlation threshold, then the stress change state of the adjacent bolts is determined to be a stress change bolt.

[0117] It should be noted that, by a person skilled in the art, a preset cross-correlation function is used to process the effective signal of the loosened bolt after positioning and the effective signal of the adjacent bolt, and then normalize the signal to eliminate the influence of the difference in signal amplitude, thereby obtaining the corresponding cross-correlation coefficient.

[0118] According to an embodiment of the present invention, the step of processing the acoustic emission energy in conjunction with a preset mapping table of acoustic emission energy and preload change, and a preload change transfer matrix, to obtain the preload change of the bolt subjected to sudden force change includes:

[0119] Construct a three-dimensional model of the bolt group and obtain the bolt preload attenuation value and the corresponding stress change of adjacent bolts;

[0120] Based on the bolt preload attenuation value and stress change, the influence weights of the bolt and its adjacent bolts are obtained through multiple linear regression fitting, and a preload change transfer matrix is ​​constructed.

[0121] The change in preload is obtained by querying a preset mapping table between acoustic emission energy and preload change based on the acoustic emission energy.

[0122] The change in preload is obtained by processing the change in preload in conjunction with the preload change transfer matrix.

[0123] It should be noted that a three-dimensional model of the bolt group is established in the ANSYS Workbench analysis software to simulate different bolt preload decay (0-30%) conditions, calculate the stress change of each bolt, obtain the preload transfer coefficient matrix through multiple linear regression fitting, and then look up the preset acoustic emission energy and preload change mapping relationship table based on the obtained loose bolt to obtain the preload change. Then, multiply the preload change by the corresponding preload change transfer matrix to obtain the preload change of the bolt with sudden stress change. Finally, the threshold comparison is used to determine whether an early warning needs to be output.

[0124] It is worth mentioning that, according to embodiments of the present invention, it further includes:

[0125] Obtain the root mean square value, kurtosis, and energy entropy value of the effective acoustic emission signal within a preset time period;

[0126] The real-time preload is obtained by processing the change in preload combined with a preset initial preload.

[0127] Obtain runtime data for the fan bolts;

[0128] The root mean square value, kurtosis, energy entropy, and real-time preload and runtime data are input into a preset acoustic emission characteristic damage quantization mapping model for processing to obtain the bolt wear amount.

[0129] It should be noted that sample data of bolts under different damage states were obtained through offline experiments, including: in the thread wear experiment, the wear amount was controlled at 0.01mm-0.1mm (each 0.01mm is a increment), and the corresponding acoustic emission characteristics were collected simultaneously. Then, combined with the real-time preload and running time data of historical samples, a training set was finally formed. An improved random forest algorithm (RF) was used to construct a mapping model to obtain an acoustic emission characteristic damage quantification mapping model. The collected real-time data was input into the model for processing to obtain the corresponding bolt wear amount. The preset acoustic emission characteristic damage quantification mapping model was obtained by training with a large amount of historical sample root mean square value, kurtosis, energy entropy, real-time preload and running time data, as well as the corresponding bolt wear amount.

[0130] It is worth mentioning that, according to embodiments of the present invention, it further includes:

[0131] Acquire acoustic emission characteristic data and micro-strain characteristic data of the thread;

[0132] The acoustic emission characteristic data includes LZv value and fractal dimension;

[0133] The microstrain characteristic data includes strain gradient and strain energy density;

[0134] The LZv value, fractal dimension, strain gradient, and strain energy density are fused to obtain a feature fusion vector.

[0135] The feature fusion vector is input into a preset microscopic loss category recognition model for processing to obtain microscopic damage category feature data;

[0136] The corresponding monitoring strategy is obtained based on the microscopic damage category characteristic data.

[0137] It should be noted that, in order to distinguish the types of microscopic damage to bolts, such as thread wear and fatigue crack initiation, and improve the accuracy of operation and maintenance, firstly, the amplitude sequence of the acoustic emission signal is converted into a binary symbol sequence (e.g., through 0-1 encoding: amplitude greater than the mean is recorded as 1, and amplitude less than the mean is recorded as 0). By counting the number of new subsequences that have never appeared in the sequence, the complexity of the signal is quantified, and the LZv value is obtained. The LZv value is a parameter describing the "frequency of occurrence of new patterns" in the signal sequence from the perspective of information theory; the higher the value, the more irregular the signal. The time-frequency diagram or amplitude curve of the signal is covered by "boxes" of preset side lengths, and the number of boxes required to cover the entire signal is counted. The fractal dimension is calculated using the logarithmic relationship between the scale and the number of boxes. The fractal dimension is a parameter describing the "self-similar structure" of the signal in the time or frequency domain; the higher the value, the more violent the fluctuation and the more complex the structure of the signal. Traditional time-domain / frequency-domain features (such as amplitude and dominant frequency) can only reflect the signal's inherent characteristics. The statistical characteristics of acoustic emission signals cannot capture the nonlinear evolution of bolt damage (such as the sudden nonlinear growth of crack propagation). However, LZv value and fractal dimension can quantify the "irregularity" and "self-similarity" of the signal, enabling early identification of micro-damage. Then, through precise measurement of material deformation, the acoustic emission signal characteristics are used to identify micro-damage. Strain gradient refers to the physical quantity of the degree of drastic change of strain in space, used to identify local stress concentration areas in the bolt. Strain energy density refers to the energy stored per unit volume of material due to deformation, used to assess the degree of damage accumulation in the bolt; the higher the energy, the greater the risk of fatigue damage. Finally, feature fusion is performed, such as the 6-dimensional features of LZv value and fractal dimension and the 4-dimensional features of strain gradient and strain energy density forming a 10-dimensional feature vector, which is input into a preset micro-loss category identification model for processing to obtain micro-damage category feature data, and then obtain the corresponding monitoring strategy, such as adjusting the monitoring time interval.

[0138] This invention also discloses a wind turbine bolt loosening monitoring and identification system based on acoustic emission technology, including a memory and a processor. The memory includes a wind turbine bolt loosening monitoring and identification method program based on acoustic emission technology. When the processor executes the wind turbine bolt loosening monitoring and identification method program based on acoustic emission technology, it performs the following steps:

[0139] The sensor coordinates, acoustic emission signal, and acoustic emission signal transmission duration of the preset sensor are obtained and processed to obtain the sound source coordinates of the loose bolt.

[0140] The acoustic emission signal is processed through a preset wind turbine operating status evaluation model to obtain wind turbine operating status label data;

[0141] The wind turbine operating status tag data is processed using a preset noise reduction method to obtain an acoustic emission noise reduction signal, and then subjected to acoustic-electric coupling analysis to obtain an effective acoustic emission noise reduction signal.

[0142] The effective acoustic emission signal is processed according to the acoustic emission noise reduction signal to obtain an effective acoustic emission signal, including the effective signal of the loose bolt and the effective signal of the adjacent bolt;

[0143] The effective acoustic emission signal is processed through a preset cross-correlation function to obtain the stress change state of adjacent bolts, including bolts without stress change or bolts with stress change.

[0144] If the stress change state is a stress change bolt, then the acoustic emission energy of the loose bolt is obtained;

[0145] The change in preload of the bolt subjected to sudden force change is obtained by processing the acoustic emission energy in combination with the preset mapping table of acoustic emission energy and preload change and the preload change transfer matrix.

[0146] The change in preload force is compared with a preset preload force variable early warning threshold.

[0147] If the change in preload is less than or equal to the preset preload variable warning threshold, no warning will be output.

[0148] If the change in preload exceeds the preset preload variable warning threshold, a warning response will be output.

[0149] It should be noted that, in order to achieve intelligent detection of wind turbine bolts based on acoustic emission signals, firstly, the bolt coordinates are solved based on the acquired acoustic emission signals to identify loose bolts. Then, the impact on the load redistribution of adjacent bolts is evaluated. By quantifying the collaborative force relationship of multiple bolts, the failure risk of the group of bolts is predicted. At the same time, an intelligent noise reduction processing method is used to match a pre-constructed dynamic noise library, and the validity of the acoustic emission signals is determined by combining acoustic-electric coupling. Finally, by comparing the changes in preload of adjacent bolts with thresholds, the impact of loose bolts on the preload of adjacent bolts is determined, thereby realizing the monitoring and identification of loose wind turbine bolts based on acoustic emission technology.

[0150] According to an embodiment of the present invention, the step of obtaining the sensor coordinates, acoustic emission signal, and acoustic emission signal transmission duration of a preset sensor, and processing them to obtain the sound source coordinates of the loose bolt, includes:

[0151] Obtain the sensor coordinates, acoustic emission signal, and acoustic emission signal transmission duration of the preset sensor;

[0152] The sensor coordinates and acoustic emission signal transmission time are combined with a preset acoustic propagation speed and processed using a nonlinear least squares method to obtain the sound source coordinates of the loose bolt.

[0153] It should be noted that piezoelectric sensors are arranged on the fan bolts at preset positions and numbered. After collecting the guided sound emission signal, a 50-500KHz bandpass filter is first used to remove low-frequency mechanical noise. Then, three non-collinear sensors are selected and the source coordinates of the loose bolt are solved according to the time difference positioning method based on the preset sound propagation speed to determine which bolt is loose.

[0154] According to an embodiment of the present invention, processing the acoustic emission signal through a preset wind turbine operating status evaluation model to obtain wind turbine operating status label data includes:

[0155] The acoustic emission signal is processed by short-time Fourier transform to obtain the corresponding acoustic emission time-frequency diagram;

[0156] The acoustic emission time-frequency diagram is input into a preset wind turbine operating status evaluation model to obtain wind turbine operating status label data corresponding to the acoustic emission signal.

[0157] It should be noted that, firstly, the collected acoustic emission signals are subjected to bandpass filtering, outlier removal, and frame segmentation. The preprocessed acoustic emission signals are then subjected to short-time Fourier transform to generate a 128×128 time-frequency diagram, which is then input into a preset wind turbine operating status evaluation model to determine the wind turbine operating status label data corresponding to the acoustic emission signals. The preset wind turbine operating status evaluation model is a convolutional neural network model trained using acoustic emission time-frequency diagrams of a large number of historical acoustic emission signal samples and the corresponding wind turbine operating status label data.

[0158] According to an embodiment of the present invention, the step of processing the wind turbine operating status tag data using a preset noise reduction method to obtain an acoustic emission noise reduction signal, and performing acoustic-electric coupling analysis to obtain an effective acoustic emission noise reduction signal, includes:

[0159] Based on the wind turbine operating status label data, query the preset operating status and noise reduction method mapping table to obtain the corresponding noise reduction method;

[0160] The noise reduction process is performed using the aforementioned noise reduction method to obtain an acoustic emission noise-reduced signal.

[0161] Acquire the micro-current signal corresponding to the acoustic emission signal bolt;

[0162] The acoustic emission noise reduction signal and the microcurrent signal are processed by a preset mutual information value processing method to obtain a mutual information value;

[0163] If the mutual information value is less than or equal to a preset mutual information qualification threshold, the acoustic emission noise reduction signal is discarded.

[0164] If the mutual information exceeds the preset threshold, the acoustic emission noise reduction signal is retained to obtain an effective acoustic emission noise reduction signal.

[0165] It should be noted that the corresponding noise reduction method is obtained by querying the preset working state and noise reduction method mapping table based on the obtained wind turbine working state label data. The preset working state and noise reduction method mapping table is preset by those skilled in the art based on a large number of historical samples and can be dynamically adjusted. Table 1 shows an example of the mapping between working state and noise reduction method:

[0166] Table 1. Example of mapping between working state and noise reduction method

[0167]

[0168] After noise reduction, to further eliminate residual noise unrelated to bolt stress and ensure a strong correlation between the signal and mechanical damage, a PVDF piezoelectric film sensor is attached to the head of each bolt to acquire the microcurrent signal corresponding to the acoustic emission signal bolt. A 0.1s sliding window is used to segment the noise-reduced acoustic emission signal and microcurrent signal. The mutual information value is calculated once per window using a preset mutual information value processing method. Finally, the obtained mutual information value is compared with a preset mutual information qualification threshold, and only signal segments greater than the preset mutual information qualification threshold are retained. This results in an effective acoustic emission noise reduction signal with triple purification through preprocessing, working condition matching noise reduction, and acoustic-electric coupling screening.

[0169] According to an embodiment of the present invention, the process of processing the effective acoustic emission noise reduction signal to obtain an effective acoustic emission signal includes an effective signal of loose bolts and an effective signal of adjacent bolts, comprising:

[0170] Obtain the total acoustic emission energy corresponding to the loose bolt and the adjacent bolt;

[0171] The effective acoustic emission noise reduction signal is processed by short-time Fourier transform to obtain the preset characteristic frequency band energy spectrum corresponding to the loose bolt and the adjacent bolt;

[0172] The energy spectrum of the preset characteristic frequency band is processed to obtain the preset characteristic frequency band energy values ​​corresponding to the loose bolt and the adjacent bolt;

[0173] The preset characteristic frequency band energy value is compared with the total acoustic emission energy value to obtain the preset characteristic frequency band energy ratio;

[0174] The preset characteristic frequency band energy ratio is compared with the preset effective energy ratio threshold.

[0175] If the effective energy percentage is less than or equal to the preset threshold, the effective acoustic emission noise reduction signal is determined to be an interference signal.

[0176] If the effective energy percentage is greater than the preset threshold, the effective acoustic emission noise reduction signal is determined to be an effective acoustic emission signal, including the effective signal of the loose bolt and the effective signal of the adjacent bolt.

[0177] It should be noted that the effective acoustic emission noise reduction signal of the 0.1s segment is processed by short-time Fourier transform, and the proportion of energy in the preset characteristic frequency band (100-150kHz characteristic frequency band) to the total signal energy is calculated. Then, a threshold comparison is performed. If it is greater than the preset effective energy proportion threshold, it means that there is an acoustic emission component related to bolt loosening in the signal segment, which is retained to obtain an effective acoustic emission signal. If it is less than or equal to the preset effective energy proportion threshold, it means that the signal segment is interference such as wind noise or mechanical vibration, which is removed.

[0178] According to an embodiment of the present invention, the step of processing the effective acoustic emission signal through a preset cross-correlation function to obtain the stress change state of adjacent bolts includes bolts without stress change or bolts with stress change, including:

[0179] The effective signals of the loosened bolts and the effective signals of the adjacent bolts are calculated and normalized using a preset cross-correlation function to obtain the corresponding cross-correlation coefficients.

[0180] The cross-correlation coefficient is compared with a preset cross-correlation threshold.

[0181] If the cross-correlation coefficient is less than or equal to the preset cross-correlation threshold, then the stress change state of the adjacent bolts is determined to be a bolt without stress change.

[0182] If the cross-correlation coefficient is greater than the preset cross-correlation threshold, then the stress change state of the adjacent bolts is determined to be a stress change bolt.

[0183] It should be noted that, by a person skilled in the art, a preset cross-correlation function is used to process the effective signal of the loosened bolt after positioning and the effective signal of the adjacent bolt, and then normalize the signal to eliminate the influence of the difference in signal amplitude, thereby obtaining the corresponding cross-correlation coefficient.

[0184] According to an embodiment of the present invention, the step of processing the acoustic emission energy in conjunction with a preset mapping table of acoustic emission energy and preload change, and a preload change transfer matrix, to obtain the preload change of the bolt subjected to sudden force change includes:

[0185] Construct a three-dimensional model of the bolt group and obtain the bolt preload attenuation value and the corresponding stress change of adjacent bolts;

[0186] Based on the bolt preload attenuation value and stress change, the influence weights of the bolt and its adjacent bolts are obtained through multiple linear regression fitting, and a preload change transfer matrix is ​​constructed.

[0187] The change in preload is obtained by querying a preset mapping table between acoustic emission energy and preload change based on the acoustic emission energy.

[0188] The change in preload is obtained by processing the change in preload in conjunction with the preload change transfer matrix.

[0189] It should be noted that a three-dimensional model of the bolt group is established in the ANSYS Workbench analysis software to simulate different bolt preload decay (0-30%) conditions, calculate the stress change of each bolt, obtain the preload transfer coefficient matrix through multiple linear regression fitting, and then look up the preset acoustic emission energy and preload change mapping relationship table based on the obtained loose bolt to obtain the preload change. Then, multiply the preload change by the corresponding preload change transfer matrix to obtain the preload change of the bolt with sudden stress change. Finally, the threshold comparison is used to determine whether an early warning needs to be output.

[0190] It is worth mentioning that, according to embodiments of the present invention, it further includes:

[0191] Obtain the root mean square value, kurtosis, and energy entropy value of the effective acoustic emission signal within a preset time period;

[0192] The real-time preload is obtained by processing the change in preload combined with a preset initial preload.

[0193] Obtain runtime data for the fan bolts;

[0194] The root mean square value, kurtosis, energy entropy, and real-time preload and runtime data are input into a preset acoustic emission characteristic damage quantization mapping model for processing to obtain the bolt wear amount.

[0195] It should be noted that sample data of bolts under different damage states were obtained through offline experiments, including: in the thread wear experiment, the wear amount was controlled at 0.01mm-0.1mm (each 0.01mm is a increment), and the corresponding acoustic emission characteristics were collected simultaneously. Then, combined with the real-time preload and running time data of historical samples, a training set was finally formed. An improved random forest algorithm (RF) was used to construct a mapping model to obtain an acoustic emission characteristic damage quantification mapping model. The collected real-time data was input into the model for processing to obtain the corresponding bolt wear amount. The preset acoustic emission characteristic damage quantification mapping model was obtained by training with a large amount of historical sample root mean square value, kurtosis, energy entropy, real-time preload and running time data, as well as the corresponding bolt wear amount.

[0196] It is worth mentioning that, according to embodiments of the present invention, it further includes:

[0197] Acquire acoustic emission characteristic data and micro-strain characteristic data of the thread;

[0198] The acoustic emission characteristic data includes LZv value and fractal dimension;

[0199] The microstrain characteristic data includes strain gradient and strain energy density;

[0200] The LZv value, fractal dimension, strain gradient, and strain energy density are fused to obtain a feature fusion vector.

[0201] The feature fusion vector is input into a preset microscopic loss category recognition model for processing to obtain microscopic damage category feature data;

[0202] The corresponding monitoring strategy is obtained based on the microscopic damage category characteristic data.

[0203] It should be noted that, in order to distinguish the types of microscopic damage to bolts, such as thread wear and fatigue crack initiation, and improve the accuracy of operation and maintenance, firstly, the amplitude sequence of the acoustic emission signal is converted into a binary symbol sequence (e.g., through 0-1 encoding: amplitude greater than the mean is recorded as 1, and amplitude less than the mean is recorded as 0). By counting the number of new subsequences that have never appeared in the sequence, the complexity of the signal is quantified, and the LZv value is obtained. The LZv value is a parameter describing the "frequency of occurrence of new patterns" in the signal sequence from the perspective of information theory; the higher the value, the more irregular the signal. The time-frequency diagram or amplitude curve of the signal is covered by "boxes" of preset side lengths, and the number of boxes required to cover the entire signal is counted. The fractal dimension is calculated using the logarithmic relationship between the scale and the number of boxes. The fractal dimension is a parameter describing the "self-similar structure" of the signal in the time or frequency domain; the higher the value, the more violent the fluctuation and the more complex the structure of the signal. Traditional time-domain / frequency-domain features (such as amplitude and dominant frequency) can only reflect the signal's inherent characteristics. The statistical characteristics of acoustic emission signals cannot capture the nonlinear evolution of bolt damage (such as the sudden nonlinear growth of crack propagation). However, LZv value and fractal dimension can quantify the "irregularity" and "self-similarity" of the signal, enabling early identification of micro-damage. Then, through precise measurement of material deformation, the acoustic emission signal characteristics are used to identify micro-damage. Strain gradient refers to the physical quantity of the degree of drastic change of strain in space, used to identify local stress concentration areas in the bolt. Strain energy density refers to the energy stored per unit volume of material due to deformation, used to assess the degree of damage accumulation in the bolt; the higher the energy, the greater the risk of fatigue damage. Finally, feature fusion is performed, such as the 6-dimensional features of LZv value and fractal dimension and the 4-dimensional features of strain gradient and strain energy density forming a 10-dimensional feature vector, which is input into a preset micro-loss category identification model for processing to obtain micro-damage category feature data, and then obtain the corresponding monitoring strategy, such as adjusting the monitoring time interval.

[0204] The present invention discloses a method and system for monitoring and identifying loose wind turbine bolts based on acoustic emission technology. By quantifying the collaborative stress relationship of multiple bolts, predicting the failure risk of grouped bolts, constructing dynamic noise reduction, and improving signal quality in complex environments, the method and system achieve the monitoring and identification of loose wind turbine bolts based on acoustic emission technology.

[0205] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0206] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0207] In addition, in the various embodiments of the present invention, each functional unit can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0208] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0209] Alternatively, if the integrated units of this invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a readable storage medium. Based on this understanding, the technical solutions of the embodiments of this invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.

Claims

1. A method for monitoring and identifying loose bolts in wind turbines based on acoustic emission technology, characterized in that, Includes the following steps: The sensor coordinates, acoustic emission signal, and acoustic emission signal transmission duration of the preset sensor are obtained and processed to obtain the sound source coordinates of the loose bolt. The acoustic emission signal is processed through a preset wind turbine operating status evaluation model to obtain wind turbine operating status label data; The wind turbine operating status tag data is processed using a preset noise reduction method to obtain an acoustic emission noise reduction signal, and then subjected to acoustic-electric coupling analysis to obtain an effective acoustic emission noise reduction signal. The effective acoustic emission signal is processed according to the acoustic emission noise reduction signal to obtain an effective acoustic emission signal, including the effective signal of the loose bolt and the effective signal of the adjacent bolt; The effective acoustic emission signal is processed through a preset cross-correlation function to obtain the stress change state of adjacent bolts, including bolts without stress change or bolts with stress change. If the stress change state is a stress change bolt, then the acoustic emission energy of the loose bolt is obtained; The change in preload of the bolt subjected to sudden force change is obtained by processing the acoustic emission energy in combination with the preset mapping table of acoustic emission energy and preload change and the preload change transfer matrix. The change in preload force is compared with a preset preload force variable early warning threshold. If the change in preload is less than or equal to the preset preload variable warning threshold, no warning will be output. If the change in preload exceeds the preset preload variable warning threshold, a warning response will be output. The process of processing the wind turbine operating status tag data using a preset noise reduction method to obtain an acoustic emission noise reduction signal, and then performing acoustic-electric coupling analysis to obtain an effective acoustic emission noise reduction signal, includes: Based on the wind turbine operating status label data, query the preset operating status and noise reduction method mapping table to obtain the corresponding noise reduction method; The noise reduction process is performed using the aforementioned noise reduction method to obtain an acoustic emission noise-reduced signal. A PVDF piezoelectric film sensor is attached to the head of each bolt to acquire the micro-current signal corresponding to the acoustic emission signal bolt; The acoustic emission noise reduction signal and the microcurrent signal are processed by a preset mutual information value processing method to obtain a mutual information value; If the mutual information value is less than or equal to a preset mutual information qualification threshold, the acoustic emission noise reduction signal is discarded. If the mutual information exceeds the preset threshold, the acoustic emission noise reduction signal is retained to obtain an effective acoustic emission noise reduction signal.

2. The method for monitoring and identifying loose wind turbine bolts based on acoustic emission technology according to claim 1, characterized in that, The process of acquiring the sensor coordinates, acoustic emission signal, and acoustic emission signal transmission duration of the preset sensor, and processing them to obtain the sound source coordinates of the loose bolt, includes: Obtain the sensor coordinates, acoustic emission signal, and acoustic emission signal transmission duration of the preset sensor; The sensor coordinates and acoustic emission signal transmission time are combined with a preset acoustic propagation speed and processed using a nonlinear least squares method to obtain the sound source coordinates of the loose bolt.

3. The method for monitoring and identifying loose wind turbine bolts based on acoustic emission technology according to claim 2, characterized in that, The process of processing the acoustic emission signal through a preset wind turbine operating status evaluation model to obtain wind turbine operating status label data includes: The acoustic emission signal is processed by short-time Fourier transform to obtain the corresponding acoustic emission time-frequency diagram; The acoustic emission time-frequency diagram is input into a preset wind turbine operating status evaluation model to obtain wind turbine operating status label data corresponding to the acoustic emission signal.

4. The method for monitoring and identifying loose wind turbine bolts based on acoustic emission technology according to claim 3, characterized in that, The process of processing the effective acoustic emission noise reduction signal to obtain an effective acoustic emission signal includes an effective signal for loose bolts and an effective signal for adjacent bolts, including: Obtain the total acoustic emission energy corresponding to the loose bolt and the adjacent bolt; The effective acoustic emission noise reduction signal is processed by short-time Fourier transform to obtain the preset characteristic frequency band energy spectrum corresponding to the loose bolt and the adjacent bolt; The energy spectrum of the preset characteristic frequency band is processed to obtain the preset characteristic frequency band energy values ​​corresponding to the loose bolt and the adjacent bolt; The preset characteristic frequency band energy value is compared with the total acoustic emission energy value to obtain the preset characteristic frequency band energy ratio; The preset characteristic frequency band energy ratio is compared with the preset effective energy ratio threshold. If the effective energy percentage is less than or equal to the preset threshold, the effective acoustic emission noise reduction signal is determined to be an interference signal. If the effective energy percentage is greater than the preset threshold, the effective acoustic emission noise reduction signal is determined to be an effective acoustic emission signal, including the effective signal of the loose bolt and the effective signal of the adjacent bolt.

5. The method for monitoring and identifying loose wind turbine bolts based on acoustic emission technology according to claim 4, characterized in that, The effective acoustic emission signal is processed through a preset cross-correlation function to obtain the stress change state of adjacent bolts, including bolts without stress change or bolts with stress change, including: The effective signals of the loosened bolts and the effective signals of the adjacent bolts are calculated and normalized using a preset cross-correlation function to obtain the corresponding cross-correlation coefficients. The cross-correlation coefficient is compared with a preset cross-correlation threshold. If the cross-correlation coefficient is less than or equal to the preset cross-correlation threshold, then the stress change state of the adjacent bolts is determined to be a bolt without stress change. If the cross-correlation coefficient is greater than the preset cross-correlation threshold, then the stress change state of the adjacent bolts is determined to be a stress change bolt.

6. The method for monitoring and identifying loose wind turbine bolts based on acoustic emission technology according to claim 5, characterized in that, The step of processing the acoustic emission energy in conjunction with a preset mapping table of acoustic emission energy and preload change, and a preload change transfer matrix, to obtain the preload change of the bolt subjected to sudden force change includes: Construct a three-dimensional model of the bolt group and obtain the bolt preload attenuation value and the corresponding stress change of adjacent bolts; Based on the bolt preload attenuation value and stress change, the influence weights of the bolt and its adjacent bolts are obtained through multiple linear regression fitting, and a preload change transfer matrix is ​​constructed. The change in preload is obtained by querying a preset mapping table between acoustic emission energy and preload change based on the acoustic emission energy. The change in preload is obtained by processing the change in preload in conjunction with the preload change transfer matrix.

7. A wind turbine bolt loosening monitoring and identification system based on acoustic emission technology, characterized in that, The system includes a memory and a processor. The memory contains a program for a method to monitor and identify loose wind turbine bolts based on acoustic emission technology. When the program is executed by the processor, the method performs the following steps: The sensor coordinates, acoustic emission signal, and acoustic emission signal transmission duration of the preset sensor are obtained and processed to obtain the sound source coordinates of the loose bolt. The acoustic emission signal is processed through a preset wind turbine operating status evaluation model to obtain wind turbine operating status label data; The wind turbine operating status tag data is processed using a preset noise reduction method to obtain an acoustic emission noise reduction signal, and then subjected to acoustic-electric coupling analysis to obtain an effective acoustic emission noise reduction signal. The effective acoustic emission signal is processed according to the acoustic emission noise reduction signal to obtain an effective acoustic emission signal, including the effective signal of the loose bolt and the effective signal of the adjacent bolt; The effective acoustic emission signal is processed through a preset cross-correlation function to obtain the stress change state of adjacent bolts, including bolts without stress change or bolts with stress change. If the stress change state is a stress change bolt, then the acoustic emission energy of the loose bolt is obtained; The change in preload of the bolt subjected to sudden force change is obtained by processing the acoustic emission energy in combination with the preset mapping table of acoustic emission energy and preload change and the preload change transfer matrix. The change in preload force is compared with a preset preload force variable early warning threshold. If the change in preload is less than or equal to the preset preload variable warning threshold, no warning will be output. If the change in preload exceeds the preset preload variable warning threshold, a warning response will be output. The process of processing the wind turbine operating status tag data using a preset noise reduction method to obtain an acoustic emission noise reduction signal, and then performing acoustic-electric coupling analysis to obtain an effective acoustic emission noise reduction signal, includes: Based on the wind turbine operating status label data, query the preset operating status and noise reduction method mapping table to obtain the corresponding noise reduction method; The noise reduction process is performed using the aforementioned noise reduction method to obtain an acoustic emission noise-reduced signal. A PVDF piezoelectric film sensor is attached to the head of each bolt to acquire the micro-current signal corresponding to the acoustic emission signal bolt; The acoustic emission noise reduction signal and the microcurrent signal are processed by a preset mutual information value processing method to obtain a mutual information value; If the mutual information value is less than or equal to a preset mutual information qualification threshold, the acoustic emission noise reduction signal is discarded. If the mutual information exceeds the preset threshold, the acoustic emission noise reduction signal is retained to obtain an effective acoustic emission noise reduction signal.

8. The wind turbine bolt loosening monitoring and identification system based on acoustic emission technology according to claim 7, characterized in that, The process of acquiring the sensor coordinates, acoustic emission signal, and acoustic emission signal transmission duration of the preset sensor, and processing them to obtain the sound source coordinates of the loose bolt, includes: Obtain the sensor coordinates, acoustic emission signal, and acoustic emission signal transmission duration of the preset sensor; The sensor coordinates and acoustic emission signal transmission time are combined with a preset acoustic propagation speed and processed using a nonlinear least squares method to obtain the sound source coordinates of the loose bolt.

9. The wind turbine bolt loosening monitoring and identification system based on acoustic emission technology according to claim 8, characterized in that, The process of processing the acoustic emission signal through a preset wind turbine operating status evaluation model to obtain wind turbine operating status label data includes: The acoustic emission signal is processed by short-time Fourier transform to obtain the corresponding acoustic emission time-frequency diagram; The acoustic emission time-frequency diagram is input into a preset wind turbine operating status evaluation model to obtain wind turbine operating status label data corresponding to the acoustic emission signal.

Images