Acoustic emission source positioning method and system based on hybrid model

Abstract

The invention provides a sound emission source positioning method and system based on a hybrid model, and relates to the technical field of sound source localization, and the method comprises the steps: calibrating the angle range and sound velocity value of each sound velocity partition; calculating a theoretical receiving time difference between the reference acoustic emission receiver probe and other acoustic emission receiver probes for receiving acoustic emission sources at different grid points; acquiring an actual receiving time difference; calculating a plurality of error values between the actual receiving time difference and theoretical receiving time differences corresponding to different grid point acoustic emission sources; the preset proportion error value is reserved, the reserved error value is clustered, and the position of a first prediction acoustic emission source of the target to be detected is positioned; inputting the current actual receiving time difference into the random forest model, and outputting a second predicted acoustic emission source position of the to-be-detected target; and verifying the comprehensive confidence of the first predicted acoustic emission source position, weighting the first predicted acoustic emission source position and the second predicted acoustic emission source position, and outputting the target predicted acoustic emission source position of the target to be detected.

Description

Technical Field

[0001] The present invention relates to the technical field of sound source localization, and particularly relates to a sound emission source localization method and system based on a hybrid model. Background Art

[0002] A sound emission source refers to the origin of acoustic wave signals generated by deformation, cracks, friction, or other physical processes occurring inside a material or structure. When a material is subjected to an external force, local crack propagation or other deformations occur, releasing acoustic waves, which can be detected by sensors and are called sound emission signals. Sound emission source localization is to determine the specific location where the change occurs by analyzing the propagation path and time difference of these acoustic waves.

[0003] Sound emission source localization is very important in many engineering fields, especially in structural health monitoring. During the use of a material or structure, problems such as cracks and damages will occur due to external forces. Sound emission source localization technology can monitor these potential structural defects in real time, and by accurately locating the defect position, it helps to discover problems in a timely manner. This can effectively avoid serious damage or accidents during the use of materials or structures, provide early warnings, and thus ensure the safety and reliability of engineering structures, which is particularly important in high-risk fields such as aviation, bridges, and oil pipelines. The localization of traditional sound emission sources usually uses localization methods based on wave velocity propagation theory (such as hyperbolic localization method).

[0004] However, the localization method based on wave velocity propagation theory often causes localization errors due to the non-uniformity of sound velocity in anisotropic materials, resulting in insufficient localization accuracy. Summary of the Invention

[0005] In view of the above deficiencies of the prior art, the purpose of the embodiments of the present invention is to provide a sound emission source localization method based on a hybrid model, which can solve the technical problems that the localization method based on wave velocity propagation theory in the prior art often causes localization errors due to the non-uniformity of sound velocity in anisotropic materials and has insufficient localization accuracy.

[0006] In the first aspect of the embodiments of the present invention, a sound emission source localization method based on a hybrid model is proposed, which is applied to a sound emission source localization physical device. The sound emission source localization physical device includes a grid plate with multiple uniform grids and a plurality of sound emission receiver probes located on the grid plate and surrounding the target to be measured. Among them, the grid plate includes a plurality of sound velocity partitions equally divided by the central point; the method includes:

[0007] S1: Calibrate the angular range and sound velocity value of each sound velocity partition;

[0008] S2: According to the calibration result, calculate the theoretical reception time difference between the reference sound emission receiver probe and other sound emission receiver probes for receiving sound emission sources at different grid points;

[0009] S3: Obtain the actual reception time differences of the acoustic emission source of the target to be measured received by each acoustic emission receiver probe;

[0010] S4: Combine the grid search algorithm to calculate multiple error values between the actual reception time differences and the theoretical reception time differences corresponding to the acoustic emission sources at different grid points respectively;

[0011] S5: Retain a preset proportion of the error values, and cluster the retained error values according to the clustering algorithm to locate the first predicted acoustic emission source position of the target to be measured;

[0012] S6: Use the actual reception time differences and the corresponding grid point positions as a training set to train a random forest model with a multi-output regressor;

[0013] S7: Input the current actual reception time difference of the target to be measured into the trained random forest model, and output the predicted grid point position of the target to be measured, that is, the second predicted acoustic emission source position;

[0014] S8: Verify the comprehensive confidence level of the first predicted acoustic emission source position through Gaussian confidence;

[0015] S9: Weight the first predicted acoustic emission source position and the second predicted acoustic emission source position according to the comprehensive confidence level, and output the target predicted acoustic emission source position of the target to be measured.

[0016] In the second aspect of the embodiments of the present invention, an acoustic emission source localization system based on a hybrid model is proposed, including: a processor and a memory;

[0017] The memory stores a program or instruction that can run on the processor. When the program or instruction is executed by the processor, the steps of the acoustic emission source localization method based on the hybrid model as in the first aspect are implemented.

[0018] The beneficial effects brought by the technical solutions provided by the embodiments of the present invention at least include:

[0019] In the embodiments of the present invention, a physical device for acoustic emission source localization and a random forest model are combined to localize the acoustic emission source respectively. By using acoustic velocity zoning and aiming at the anisotropy of the material, the acoustic velocity of each region is independently calibrated, significantly improving the localization accuracy and effectively solving the localization error problem caused by the non-uniformity of acoustic velocity in anisotropic materials in the traditional wave velocity propagation theory. Combining grid search and clustering algorithms optimizes the localization error value, thereby accurately determining the position of the acoustic emission source. Then, Gaussian confidence is used to verify the comprehensive confidence of the first predicted acoustic emission source position, and the comprehensive confidence is weighted for the first predicted acoustic emission source position and the second predicted acoustic emission source position to output the target predicted acoustic emission source position of the target to be measured. The prediction results of physical prediction and machine learning prediction are dynamically corrected in a weighted manner, enhancing the reliability of localization and improving the accuracy of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The drawings are only for the purpose of illustrating specific embodiments and are not considered as a limitation of the present invention. Throughout the drawings, the same reference numerals denote the same components. Obviously, the drawings in the following description are only some embodiments described in the embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0021] Figure 1 is a schematic flowchart of a method for acoustic emission source localization based on a hybrid model provided by an embodiment of the present invention;

[0022] Figure 2 is a schematic structural diagram of a physical device for acoustic emission source localization provided by an embodiment of the present invention;

[0023] Figure 3 is a schematic structural diagram of a hybrid model provided by an embodiment of the present invention;

[0024] Figure 4 is a schematic structural diagram of a system for acoustic emission source localization based on a hybrid model provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0025] In order to enable those skilled in the art to better understand the technical solutions in the embodiments of the present invention, the technical solutions of the present invention will be clearly and completely described below in conjunction with the drawings. Obviously, the described embodiments are some embodiments of the present invention, rather than all embodiments. It should be understood that these descriptions are only exemplary and are not used to limit the scope of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0026] The following will combine the accompanying drawings to describe in detail the method for acoustic emission source localization based on a hybrid model provided by the embodiments of the present invention through specific embodiments and their application scenarios.

[0027] Refer to the accompanying drawings of the specification Figure 1 , which shows a schematic flowchart of a method for acoustic emission source localization based on a hybrid model provided by the embodiments of the present invention.

[0028] The embodiments of the present invention provide a method for acoustic emission source localization based on a hybrid model, which is applied to a physical device for acoustic emission source localization. The physical device for acoustic emission source localization includes a grid plate with a plurality of uniform grids and a plurality of acoustic emission receiver probes located on the grid plate and surrounding the target to be measured. Among them, the grid plate includes a plurality of sound velocity partitions divided at equal angles with the center point.

[0029] Figure 1 An example of an optional arrangement of the physical device for acoustic emission source localization is shown in. Among them, P1 - P4 respectively represent a plurality of acoustic emission receiver probes surrounding the target to be measured, Target. Through the combination of a plurality of acoustic emission receiver probes, the device can capture signals from the acoustic emission source and accurately locate the position of the source by combining the characteristics of different sound velocity partitions. According to actual requirements, technicians can adjust the number and layout of the probes to optimize the localization results.

[0030] Optionally, the number of sound velocity partitions can be set to 12.

[0031] In a possible implementation manner, the distance between the grid points of adjacent grids is 0.5 times the grid width.

[0032] It may include the following steps:

[0033] S1: Calibrate the angular range and sound velocity value of each sound velocity partition.

[0034] Among them, the grid plate of the acoustic emission source localization device is divided into multiple sound velocity zones. Each sound velocity zone corresponds to a specific range of regional angles. This angle range refers to the effective range angle of the sound velocity zone within this zone. For example, a certain sound velocity zone may cover the area from 0° to 30°, indicating that the same sound velocity value is used for all grid points within this area. The delineation of the angle range is determined based on the position of the acoustic emission source and the distribution of the receiver probes, which helps the system understand the acoustic wave propagation characteristics of each area. Within each sound velocity zone, the propagation speed of the acoustic wave may be different, especially in different materials or media. The sound velocity value refers to the speed at which the acoustic wave propagates within this specific area, which depends on the characteristics of the material and the position of the sound velocity zone. During the calibration process, an accurate sound velocity value needs to be determined for each sound velocity zone, and this value is usually obtained through experiments or theoretical calculations. The sound velocity values of different areas are different, ensuring more accurate calculation of the time difference of acoustic wave propagation.

[0035] It should be noted that by calibrating the angle range and the sound velocity value, the device can more accurately calculate the propagation path and time difference of the acoustic emission source signal, thereby improving the accuracy of source localization.

[0036] In a possible implementation manner, S1 specifically includes:

[0037] S101: Initialize the angle range and sound velocity value of each sound velocity zone.

[0038] S102: Determine the regional index of different angles to calibrate the angle range, where the regional index represents different angle ranges:

[0039]

[0040] Among them, % represents the modulo operator, g represents the regional index of the angle range, θ represents the angle value, and s represents the number of sound velocity zones.

[0041] S103: Calculate the interval distance and relative angle between the known acoustic emission source position and each acoustic emission receiver probe.

[0042] S104: Measure the actual time difference between the known acoustic emission source position and two different acoustic emission receiver probes.

[0043] S105: Calculate the predicted time difference between the known acoustic emission source position and two different acoustic emission receiver probes:

[0044]

[0045] Among them, represents the predicted time difference between the known acoustic emission source position and the pth and qth acoustic emission receiver probes, d p and dq represent the distances from the known acoustic emission source location to the p-th and q-th acoustic emission receiver probes respectively, v p and v q represent the sound velocities from the known acoustic emission source location to the p-th and q-th acoustic emission receiver probes respectively.

[0046] S106: Adjust the sound velocity values of each sound velocity partition with the goal of minimizing the time difference between the predicted time difference and the actual time difference.

[0047] S107: Perform interpolation calculation on the adjusted sound velocity values:

[0048]

[0049] where, v g1 and v g2 represent the sound velocity values of region indices g1 and g2 respectively, and v represents the interpolated sound velocity value.

[0050] Specifically, first initialize the angular range and sound velocity value of the sound velocity partition. Next, determine the region index (g) at different angles through S102 to calibrate the angular range, so that the sound velocity partition can be calculated according to different angles. S103 calculates the distance and angle between the known acoustic emission source location and the receiver probe, and measures the actual time difference of the signals received by different probes (S104). Subsequently, S105 calculates the theoretical time difference from the acoustic emission source to the receiver probe. The goal of S106 is to minimize the error between the predicted time difference and the actual time difference, and then adjust the sound velocity value of the sound velocity partition. S107 then performs interpolation calculation on the adjusted sound velocity value, uses the L-BFGS-B algorithm for optimization, optimizes the sound velocity range and outputs the final result, so as to accurately calibrate the sound velocity value of each region, optimize the sound velocity model to improve the positioning accuracy.

[0051] For example, the L-BFGS-B (Limited-memory Broyden-Fletcher-Goldfarb-Shanno with Bound constraints) algorithm can be used for numerical optimization to solve large-scale non-linear optimization problems with boundary constraints. Set the sound velocity range to 20 to 600 grids / ms (about 400 to 12000 m / s), iteratively optimize the sound velocity value, and output the intermediate results during the optimization process. The calibration output of the algorithm: After the optimization is completed, update the sound velocity table, and output the angular range and sound velocity value of each region, such as "Region 0 (0° to 30°): 250.3 grids / ms".

[0052] S2: According to the calibration result, calculate the theoretical reception time difference between the reference acoustic emission receiver probe and other acoustic emission receiver probes for receiving acoustic emission sources at different grid points.

[0053] Among them, the reference acoustic emission receiver probe is a receiver probe that serves as a reference point in the acoustic emission source positioning device and is used for comparison and calculation with other receiver probes. In the grid point acoustic emission source, the acoustic emission source is the position where the acoustic wave signal is generated, and the grid point refers to each subdivided area within the positioning area. In the grid-shaped positioning area, each grid point can be regarded as a potential acoustic emission source position, and the acoustic wave emits from these grid points and is transmitted to the receiver probe. Theoretical reception time difference: According to the propagation speed and distance of the acoustic wave, the theoretical reception time difference refers to the theoretical time difference from the acoustic emission source to each receiver probe.

[0054] It should be noted that the theoretical reception time difference of the acoustic wave reaching the receiver from the acoustic emission sources at different grid points between the reference acoustic emission receiver probe and other receiver probes is calculated. This time difference is calculated based on the theoretical model of acoustic wave propagation and the relative positions of different grid points, providing a basis for subsequent positioning.

[0055] In a possible implementation manner, after S2, it further includes:

[0056] Perform double-layer outlier cleaning on the theoretical reception time difference. The specific cleaning process is as follows:

[0057] Perform a global preliminary screening on the theoretical reception time difference of all grid point acoustic emission sources through the three-standard-deviation criterion.

[0058] Perform a local fine screening on the theoretical reception time difference of each grid point acoustic emission source through the one-standard-deviation criterion.

[0059] It should be noted that first, use the three-standard-deviation criterion to perform a global screening on the theoretical reception time difference of all grid points to remove those outliers that are significantly deviated from the normal range. Then, use the one-standard-deviation criterion to perform a local screening on the theoretical reception time difference of each grid point to further refine and exclude the data points with larger errors. Through these two steps, the accuracy of subsequent calculations can be effectively improved and the interference of abnormal data can be reduced.

[0060] S3: Obtain the actual reception time difference of each acoustic emission receiver probe receiving the acoustic emission source to be measured.

[0061] It should be noted that through each acoustic emission receiver probe actually receiving the acoustic wave signal emitted from the acoustic emission source to be measured, record and calculate the actual reception time difference of these signals reaching different receiver probes. These time differences reflect the time required for the acoustic wave to propagate to each receiver probe and are obtained through actual measurement. Compared with the theoretical reception time difference, the actual reception time difference can better reflect the real situation on site and is used for subsequent positioning calculations.

[0062] S4: Calculate multiple error values between the actual reception time difference and the theoretical reception time differences corresponding to different grid points of the acoustic emission source respectively by combining the grid search algorithm.

[0063] Optionally, to implement the grid search, set the grid interval to 0.5 grid, cover the detection area (e.g., 23×23 grids), and generate all grid point coordinates. Propagation direction calculation: For each grid point, calculate the propagation direction angle from it to the acoustic emission receiver probe:

[0064] θ 、 = arctan2(Δy, Δx)

[0065] where θ represents the propagation direction angle from each grid point to each acoustic emission receiver probe, arctan represents the arctangent function, and Δy, Δx represent the coordinate differences between the grid point and the acoustic emission receiver probe.

[0066] The grid search algorithm is an optimization method that finds the optimal solution by traversing the possible parameter space. In acoustic emission source localization, the grid search algorithm traverses the possible positions of different grid points and calculates the error between the theoretical reception time difference corresponding to these positions and the actual reception time difference to find the optimal sound source position. The error value refers to the difference between the theoretical reception time difference and the actual reception time difference. Each grid point will have an error value, indicating the deviation between this position as the acoustic emission source and the actually received time difference.

[0067] It should be noted that the error values between the theoretical reception time difference and the actual reception time difference of each grid point are calculated by combining the grid search algorithm. By setting the grid interval (e.g., 0.5 grid) to cover the entire detection area (e.g., 23×23 grids), the system calculates the propagation direction angle from each grid point to each acoustic emission receiver probe and uses this information for error calculation. The goal is to find the grid point closest to the actual situation by comparing the errors between the theoretical and actual time differences to help accurately locate the acoustic emission source.

[0068] In a possible implementation manner, the calculation method of the error value is specifically:

[0069]

[0070] where represents the theoretical reception time difference between the reference acoustic emission receiver probe corresponding to the i-th grid point and the k-th acoustic emission receiver probe, k = 1, 2, …, K, and K represents the total number of acoustic emission receiver probes, represents the actual reception time difference between the reference acoustic emission receiver probe corresponding to the i-th grid point and the k-th acoustic emission receiver probe, e 0i represents The error value between the actual reception time difference.

[0071] It should be noted that the calculation of the error value is carried out by comparing the theoretical reception time difference between the reference acoustic emission receiver probe and other receiver probes with the actual reception time difference. The theoretical reception time difference is calculated through the sound speed model, while the actual reception time difference is obtained according to actual measurement. By calculating the difference between the two, the error value can be obtained. This error value is used to reflect the deviation between the theoretical and actual reception time differences, thereby helping to optimize the position prediction of the acoustic emission source.

[0072] S5: Retain a preset proportion of the error values, and cluster the retained error values according to the clustering algorithm to locate the first predicted acoustic emission source position of the target to be measured.

[0073] Specifically, first, a certain proportion of the error values are retained, and these error values represent the difference between the actual reception time difference and the theoretical reception time difference. Then, the system uses a clustering algorithm (such as DBSCAN or K-means) to cluster these retained error values, thereby helping to identify potential acoustic emission source positions. The point with the smallest error in the clustering result is the first predicted acoustic emission source position, which has relatively high accuracy and is provided as a preliminary positioning result.

[0074] It should be noted that those skilled in the art can set the size of the preset proportion according to actual needs, and the present invention does not make any limitations here.

[0075] In a possible implementation manner, the clustering algorithm is the DBSCAN algorithm. S5 specifically includes:

[0076] S501: Use the retained error values as candidate points and cluster each candidate point through the DBSCAN algorithm.

[0077] S502: Select the minimum error value in the cluster corresponding to the maximum candidate point density.

[0078] S503: Output the grid point corresponding to the calculated minimum error value as the first predicted acoustic emission source position.

[0079] Specifically, in step S5, first, the error values are retained as candidate points and clustered through the DBSCAN algorithm. The DBSCAN algorithm clusters the candidate points according to the neighborhood radius and the minimum number of samples to ensure that areas with higher density can be accurately identified. Then, the grid point with the smallest error is selected from the cluster with the maximum density as the first predicted acoustic emission source position. If there is no effective cluster, the system will select the point with the smallest error value as the prediction result. This process improves the accuracy and reliability of the positioning through clustering and error screening.

[0080] For example, the candidate points with the top 10% errors can be screened and then clustering enhancement can be performed. The DBSCAN algorithm (neighborhood radius eps = 1.0, minimum number of samples = 5) is used to cluster the candidate points, and the grid point with the smallest error in the largest cluster is selected as the predicted position. In special cases, if there is no effective clustering, the point with the smallest global error (i.e., the smallest error value among the obtained error values) is directly returned to obtain the predicted position.

[0081] S6: Use the actual reception time difference and the corresponding grid point position as the training set to train a random forest model with a multi-output regressor.

[0082] Among them, the random forest model is an ensemble learning model that improves the prediction accuracy and robustness by constructing multiple decision trees and integrating their output results. Each decision tree uses different samples and features during the training process and finally obtains the result by voting or averaging. Random forests can handle complex non-linear problems and have good tolerance for noise in the data. The system uses the actual reception time difference and the position of the corresponding grid point as training data to train a random forest model with a multi-output regressor. The trained model can predict the specific position of the acoustic emission source based on the actual time difference. The random forest model improves the processing ability for complex positioning problems through the comprehensive judgment of multiple decision trees, and can effectively learn from the training data and improve the prediction accuracy.

[0083] In a possible implementation, S6 specifically includes:

[0084] S601: Compose the feature vector from each actual reception time difference:

[0085]

[0086] Among them, represents the actual reception time difference between the reference acoustic emission receiver probe and the k-th acoustic emission receiver probe, and Y represents the feature vector composed of .

[0087] S602: Standardize the feature vector:

[0088]

[0089] Among them, Y norm represents the standardized feature vector, and μ and σ represent the feature mean and feature standard deviation in the feature vector respectively.

[0090] S603: Input the standardized feature vectors and the corresponding grid point positions into a random forest for training. Use a multi-objective regressor to map the feature vectors to multi-objective data representing the grid point positions, i.e., the acoustic emission source positions, until the prediction accuracy of the grid point positions is greater than a preset prediction accuracy:

[0091]

[0092] where, represents the real number field, and f represents the multi-objective regressor function.

[0093] Specifically, first, the system forms a feature vector consisting of all theoretical reception time differences. This vector reflects the theoretical reception time differences between the reference acoustic emission receiver and other receivers. Then, in S602, the feature vector is standardized by subtracting the mean and dividing by the standard deviation, making the values of the feature vector more uniform and comparable for subsequent analysis. Finally, the standardized feature vectors and the corresponding grid point positions are input into the random forest model for training. The model maps the feature vectors to the positions of the acoustic emission sources through a multi-objective regressor. The model training continues until the prediction accuracy reaches the preset standard. In this way, the system can accurately predict the position of the acoustic emission source based on the theoretical time differences.

[0094] It should be noted that those skilled in the art can set the size of the preset prediction accuracy according to actual needs, and the present invention does not make any limitations in this regard.

[0095] S7: Input the current actual reception time difference of the target to be measured into the trained random forest model, and output the predicted grid point position of the target to be measured, i.e., the second predicted acoustic emission source position.

[0096] It should be noted that the actual reception time difference is used as the input and input into the already trained random forest model. The random forest model, based on the rules learned during the training process and through the comprehensive judgment of multiple decision trees, outputs the position of the second predicted acoustic emission source. This position is predicted based on the actually received time differences, which can further verify or optimize the first predicted position and improve the positioning accuracy.

[0097] S8: Verify the comprehensive confidence of the first predicted acoustic emission source position through Gaussian confidence.

[0098] Among them, the Gaussian confidence is used to measure the credibility of the prediction result through the Gaussian distribution (normal distribution). It calculates the confidence based on the residual of the time difference of arrival and the standard deviation of its distribution. The higher the value, the more reliable the prediction result. Generally, the data points with small errors and conforming to the Gaussian distribution have high confidence. The reliability of the first predicted acoustic emission source position is evaluated by calculating the Gaussian confidence of the position. Specifically, according to the residual of the theoretical time difference of arrival and the actual time difference of arrival, the Gaussian distribution is used to calculate the confidence. If the Gaussian confidence of the first predicted position is high, it indicates that the prediction result is relatively reliable and can be further used to weight and fuse other prediction results.

[0099] In a possible implementation, S8 specifically includes:

[0100] S801: Calculate the respective theoretical time differences of arrival at the first predicted acoustic emission source position:

[0101] S802: Calculate the time difference residual between the respective theoretical time differences of arrival and the actual time differences of arrival at the first predicted acoustic emission source position.

[0102] S803: Calculate the Gaussian confidence of each time difference residual:

[0103]

[0104] Among them, φ() represents the Gaussian density function, represents the time difference residual between the reference acoustic emission receiver probe and the k-th acoustic emission receiver probe, that is, the difference between the theoretical time difference of arrival and the current actual time difference of arrival, C 1k represents the Gaussian confidence of the time difference residual between the reference acoustic emission receiver probe and the k-th acoustic emission receiver probe, and τ represents the standard deviation parameter.

[0105] Optionally, the standard deviation parameter can be set to 0.1.

[0106] S804: Take the average of each Gaussian confidence to obtain the comprehensive confidence:

[0107]

[0108] Among them, C 综合 represents the comprehensive confidence.

[0109] Specifically, first calculate the theoretical reception time difference at the first predicted acoustic emission source position, and then calculate the residuals between these theoretical time differences and the actual reception time differences. Next, the system uses the Gaussian density function to calculate the Gaussian confidence of each residual, representing the reliability of each reception time residual. Finally, all the Gaussian confidences are averaged to obtain a comprehensive confidence, which reflects the overall credibility of all predicted positions and helps to judge the accuracy of the prediction results. If the comprehensive confidence is high, it indicates that the predicted acoustic emission source position is more reliable.

[0110] S9: Weight the first predicted acoustic emission source position and the second predicted acoustic emission source position according to the comprehensive confidence, and output the target predicted acoustic emission source position of the target to be measured.

[0111] It should be noted that according to the comprehensive confidence of the first predicted acoustic emission source position and the second predicted acoustic emission source position, these two predicted positions are weighted. The position with a higher comprehensive confidence will be given a greater weight to ensure that the final target predicted position is more accurate. Through this weighting method, the system can combine two different prediction results to generate a more comprehensive and reliable final acoustic emission source position, thereby improving the positioning accuracy.

[0112] In a possible implementation manner, S9 specifically includes:

[0113] S901: When the comprehensive confidence is greater than the preset comprehensive confidence, weight the first predicted acoustic emission source position and the second predicted acoustic emission source position according to the comprehensive confidence, and enter step 903; otherwise, enter step S902.

[0114] S902: Weight the first predicted acoustic emission source position and the second predicted acoustic emission source position according to the preset fixed weighting value.

[0115] S903: Calculate and output the target predicted acoustic emission source position. The calculation method of the target predicted acoustic emission source position is specifically:

[0116]

[0117] where ε represents the preset fixed weighting value, C represents the intermediate variable representing the weighting value, C0 represents the preset comprehensive confidence, X phy and X ml represent the first predicted acoustic emission source position and the second predicted acoustic emission source position respectively, represents the target predicted acoustic emission source position.

[0118] It should be noted that those skilled in the art can set the magnitudes of the preset fixed weighting value and the preset comprehensive confidence according to actual needs, and the present invention does not limit this here.

[0119] Optionally, the preset fixed weighting value can be set to 0.01.

[0120] Specifically, if the comprehensive confidence level is greater than the preset value, the system will weight the first and second predicted acoustic emission source positions according to the comprehensive confidence level to obtain the target predicted acoustic emission source position. Otherwise, a fixed weighting value is used for the weighting calculation. In the calculation, C is an intermediate variable of the weighting value, and the comprehensive confidence level determines the weight of the weighting process, and finally the position of the target predicted acoustic emission source is calculated. Through this weighting method, the system can fuse the two prediction results to ensure higher accuracy.

[0121] Refer to the attached Figure 3 description, which shows a schematic structural diagram of a hybrid model provided by an embodiment of the present invention.

[0122] Figure 3 It shows a structural example of the modular hybrid model of the present solution. The data input layer is used to obtain the actual reception time difference of the acoustic emission receiver probes for receiving the acoustic emission source of the target to be measured. The data preprocessing module is used to perform double-layer outlier cleaning on the theoretical reception time difference. The physical model branch, that is, the physical device for acoustic emission source localization, is used to locate the first predicted acoustic emission source position of the target to be measured. The machine learning branch, that is, the random forest model, is used to output the second predicted acoustic emission source position of the target to be measured. The model fusion layer is used to verify the comprehensive confidence level of the first predicted acoustic emission source position through Gaussian confidence, and weight the first predicted acoustic emission source position and the second predicted acoustic emission source position based on the comprehensive confidence level, and output the target predicted acoustic emission source position of the target to be measured.

[0123] In the actual application process, first, the device calculates the theoretical reception time difference between the reference receiver and other receivers by calibrating the sound speed partition. Then, the actual reception time difference is obtained, and the error value is calculated through the grid search algorithm, and the positioning result is optimized by combining the clustering algorithm. Next, the random forest model is used to train the theoretical time difference and the grid point position as training data, and then the second position of the acoustic emission source is predicted. Finally, considering the Gaussian confidence levels of the two predicted positions comprehensively, the results are combined by using a weighting method to output the final accurate acoustic emission source position. The whole process combines traditional physical models and modern machine learning technologies, improving the positioning accuracy and reliability.

[0124] The beneficial effects brought by the technical solution provided by the embodiment of the present invention at least include:

[0125] In the embodiments of the present invention, a physical device for acoustic emission source localization and a random forest model are combined to localize the acoustic emission source respectively. By using acoustic velocity zoning and aiming at the anisotropy of the material, the acoustic velocity of each region is independently calibrated, significantly improving the localization accuracy and effectively solving the localization error problem caused by the non-uniformity of acoustic velocity in anisotropic materials in the traditional wave velocity propagation theory. Combining grid search and clustering algorithms optimizes the localization error value, thereby accurately determining the position of the acoustic emission source. Then, the Gaussian confidence is used to verify the comprehensive confidence of the first predicted acoustic emission source position. The comprehensive confidence weights the first predicted acoustic emission source position and the second predicted acoustic emission source position, and outputs the target predicted acoustic emission source position of the target to be measured. The prediction results of physical prediction and machine learning prediction are dynamically corrected in a weighted manner, enhancing the reliability of localization and improving the accuracy of the system.

[0126] Refer to the appended drawings of the specification Figure 4 , which shows a schematic structural diagram of an acoustic emission source localization system based on a hybrid model provided by an embodiment of the present invention.

[0127] An embodiment of the present invention provides an acoustic emission source localization system 20 based on a hybrid model, including: a processor 201 and a memory 202;

[0128] The memory 202 stores programs or instructions that can run on the processor 201. When the programs or instructions are executed by the processor 201, the steps of the above-mentioned acoustic emission source localization method based on the hybrid model are implemented, and the same technical effects can be achieved. To avoid repetition, the present invention will not be elaborated herein again.

[0129] It should be understood that the processor 201 in the embodiments of the present invention may be a central processing unit (CPU), and this processor may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or this processor may also be any conventional processor, etc.

[0130] It should also be understood that the memory 202 in the embodiments of the present invention may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable ROM (PROM), an erasable programmable ROM (EPROM), an electrically erasable programmable ROM (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchlink DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0131] The above embodiments can be implemented in whole or in part by software, hardware (such as circuits), firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present invention are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wired (such as infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that contains one or more collections of available media. The available media can be magnetic media (such as floppy disks, hard disks, magnetic tapes), optical media (such as DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0132] It should be understood that in various embodiments of the present invention, the sequence numbers of the above processes do not imply the order of execution, and the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0133] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professionals can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0134] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the devices, apparatuses, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0135] In several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0136] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0137] In addition, in each embodiment of the present invention, the functional units can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0138] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0139] The embodiments of the present invention provide a readable storage medium, including: programs or instructions stored on the readable storage medium. When the programs or instructions are executed by a processor, the steps of the above-mentioned acoustic emission source localization method based on a hybrid model are implemented, and the same technical effects can be achieved. To avoid repetition, the present invention will not be described in detail again.

[0140] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the embodiments of the present invention, rather than to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention. Any changes or replacements that can be easily thought of by those skilled in the art within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention.

Claims

1. An acoustic emission source localization method based on a hybrid model, characterized in that, Applied to a physical device for acoustic emission source localization, the physical device for acoustic emission source localization includes a grid plate with a plurality of uniform grids and a plurality of acoustic emission receiver probes located on the grid plate and surrounding the target to be measured. Among them, the grid plate includes a plurality of sound velocity partitions divided at equal angles with a central point; the method includes: S1: Calibrate the angular range and sound velocity value of each of the sound velocity partitions; S2: According to the calibration results, calculate the theoretical reception time differences between the reference acoustic emission receiver probe and other acoustic emission receiver probes for receiving acoustic emission sources at different grid points; S3: Obtain the actual reception time differences of each of the acoustic emission receiver probes for receiving the acoustic emission source of the target to be measured; S4: Combine the grid search algorithm to calculate a plurality of error values between the actual reception time differences and the theoretical reception time differences corresponding to acoustic emission sources at different grid points; S5: Retain a preset proportion of the error values, and cluster the retained error values according to the clustering algorithm to locate the first predicted acoustic emission source position of the target to be measured; S6: Use the actual reception time differences and the corresponding grid point positions as a training set to train a random forest model with a multi-output regressor; S7: Input the current actual reception time difference of the target to be measured into the trained random forest model, and output the predicted grid point position of the target to be measured, that is, the second predicted acoustic emission source position; S8: Verify the comprehensive confidence of the first predicted acoustic emission source position through Gaussian confidence; S9: Weight the first predicted acoustic emission source position and the second predicted acoustic emission source position according to the comprehensive confidence, and output the target predicted acoustic emission source position of the target to be measured.

2. The acoustic emission source localization method based on a hybrid model according to claim 1, wherein The interval distance between grid points of adjacent grids is 0.5 times the grid width.

3. The acoustic emission source localization method based on a hybrid model according to claim 1, characterized in that The specific content of S1 includes: S101: Initialize the angular range and sound velocity value of each of the sound velocity partitions; S102: Determine the region index of different angles to calibrate the angular range, where the region index represents different angular ranges: Among them, % represents the modulo operator, g represents the region index of the angular range, θ represents the angular value, and s represents the number of sound velocity partitions; S103: Calculate the interval distance and relative angle between the known acoustic emission source position and each of the acoustic emission receiver probes; S104: Measure the actual time difference between the known acoustic emission source position and two different acoustic emission receiver probes; S105: Calculate the predicted time difference between the known acoustic emission source position and two different acoustic emission receiver probes; Among them, represents the predicted time difference from the known acoustic emission source location to the p-th and q-th acoustic emission receiver probes, d p and d q respectively represent the distances from the known acoustic emission source location to the p-th and q-th acoustic emission receiver probes, v p and v q respectively represent the sound velocities from the known acoustic emission source location to the p-th and q-th acoustic emission receiver probes; S106: Take minimizing the time difference between the predicted time difference and the actual time difference as the goal, and adjust the sound velocity value of each of the sound velocity partitions; S107: Perform interpolation calculation on the adjusted sound velocity value; where, v g1 and v g2 represent the sound velocity values of the regional indices g1 and g2, respectively, and v represents the interpolated sound velocity value.

4. The acoustic emission source localization method based on a hybrid model according to claim 1, wherein After S2, it further includes: Perform double-layer outlier cleaning on the theoretical reception time differences; the cleaning process is specifically as follows: Perform a global preliminary screening on the theoretical reception time differences of all grid point acoustic emission sources through the three-sigma criterion; Perform a local fine screening on the theoretical reception time differences of all grid point acoustic emission sources through the one-sigma criterion.

5. The acoustic emission source localization method based on a hybrid model according to claim 1, wherein The specific calculation method of the error value is as follows: Among them, represents the theoretical reception time difference between the reference acoustic emission receiver probe corresponding to the i-th grid point and the k-th acoustic emission receiver probe, where k = 1, 2, …, K, and K represents the total number of acoustic emission receiver probes. represents the actual reception time difference between the reference acoustic emission receiver probe corresponding to the i-th grid point and the k-th acoustic emission receiver probe, e 0i represents the error value between and the actual reception time difference.

6. The acoustic emission source localization method based on a hybrid model according to claim 1, wherein The clustering algorithm is the DBSCAN algorithm; S5 specifically includes: S501: Use the remaining error values as candidate points, and cluster each of the candidate points through the DBSCAN algorithm; S502: Select the minimum error value in the cluster corresponding to the maximum candidate point density; S503: Output the grid point corresponding to the calculated minimum error value as the first predicted acoustic emission source position.

7. The acoustic emission source localization method based on a hybrid model according to claim 1, wherein S6 specifically includes: S601: Form a feature vector from each of the actual reception time differences; Among them, represents the actual reception time difference between the reference acoustic emission receiver probe and the k-th acoustic emission receiver probe, and Y represents the composed eigenvector; S602: Standardize the feature vector; Among them, Y norm represents the standardized eigenvector, and μ and σ respectively represent the mean value and standard deviation of the eigenvector; S603: Input the standardized feature vector and the corresponding grid point positions into the random forest for training, and use a multi-objective regressor to map the feature vector to multi-objective data representing the grid point positions, i.e., the acoustic emission source positions, until the prediction accuracy of the grid point positions is greater than the preset prediction accuracy; wherein, represents the real number field, and f represents the multi-objective regression function.

8. The acoustic emission source localization method based on a hybrid model according to claim 5, wherein S8 specifically includes: S801: Calculate each theoretical reception time difference at the first predicted acoustic emission source position; S802: Calculate the reception time residuals between each theoretical reception time difference and the actual reception time difference at the first predicted acoustic emission source position; S803: Calculate the Gaussian confidence levels of each of the reception time residuals; where, φ() represents the Gaussian density function, represents the residual of the reception time between the reference acoustic emission receiver probe and the k-th acoustic emission receiver probe, that is, the difference between the theoretical reception time difference and the current actual reception time difference, C 1k represents the Gaussian confidence of the residual of the reception time between the reference acoustic emission receiver probe and the k-th acoustic emission receiver probe, and τ represents the standard deviation parameter; S804: Take the mean of each of the Gaussian confidence levels to obtain the comprehensive confidence level; Among them, C 综合 represents the comprehensive confidence level.

9. The acoustic emission source localization method based on a hybrid model according to claim 1, characterized in that, S9 specifically includes: S901: In the case where the comprehensive confidence level is greater than the preset comprehensive confidence level, weight the first predicted acoustic emission source position and the second predicted acoustic emission source position according to the comprehensive confidence level, and proceed to step 903; otherwise, proceed to step S902; S902: Weight the first predicted acoustic emission source position and the second predicted acoustic emission source position according to a preset fixed weighting value; S903: Calculate and output the target predicted acoustic emission source position. The specific calculation method of the target predicted acoustic emission source position is as follows: Among them, ε represents a preset fixed weighting value, C represents an intermediate variable representing the weighting value, C0 represents a preset comprehensive confidence level, X phy and X ml respectively represent the first predicted acoustic emission source position and the second predicted acoustic emission source position, represents the target predicted acoustic emission source position.

10. An acoustic emission source location system based on a hybrid model, characterized in that, Includes: A processor and a memory; The memory stores programs or instructions that can run on the processor. When the programs or instructions are executed by the processor, the steps of the acoustic emission source localization method based on the hybrid model according to any one of claims 1 to 9 are implemented.

Images