Method and device for judging contact degree of two metal surfaces by using ultrasonic technology

Through multi-parameter analysis and signal processing by ultrasonic technology, combined with support vector machine model, the problems of low efficiency and insufficient accuracy of traditional detection methods are solved, and lossless and high-precision metal surface contact degree detection is achieved, suitable for complex shapes and large-scale Small metal structure.

CN118937493AInactive Publication Date: 2024-11-12STATE GRID JIANGSU ELECTRIC POWER CO LTD NANTONG POWER SUPPLY BRANCH +2
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202411166837.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-23
Publication Date
2024-11-12
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Traditional metal contact detection methods are inefficient and cannot accurately evaluate the entire contact surface. Existing contact resistance testers cannot evaluate the contact situation of the entire plane when some areas are in good contact.

Method used

Ultrasonic technology is used to set ultrasonic emission parameters, clean the metal contact surface and coat the ultrasonic coupling agent, and use the ultrasonic probe to emit and receive reflected signals. Through signal processing and multi-parameter analysis, the contact between the metal surfaces, including the intensity of the signal, Time characteristics and spectrum analysis, establish a support vector machine model for identification.

Benefits of technology

It realizes lossless, high-precision, real-time metal surface contact degree detection, and can detect tiny gaps and incomplete contact areas. It is suitable for complex shapes and large and small metal structures, and will not affect the subsequent service performance and life of metal materials.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118937493A_ABST
    Figure CN118937493A_ABST
Patent Text Reader

Abstract

The invention discloses a method and a device for judging the contact degree of two metal surfaces by using an ultrasonic technology, which mainly focuses on detecting whether defects such as cracks, air holes and inclusions exist in a metal material or not and determining the positions, sizes and shapes of the defects. The characteristics that ultrasonic waves are easy to regulate and control, high in penetrating power, concentrated in energy and the like are mainly utilized, and the ultrasonic technology is applied to evaluation of the close degree, uniformity and the like of attachment or contact of the two metal faces and determination of the connection quality of the two metal faces. According to the method, by transmitting and receiving ultrasonic signals, the reflection and projection characteristics of ultrasonic waves on the contact surfaces and the signal attenuation degree are analyzed, and the contact condition between the metal surfaces is evaluated. The device comprises an ultrasonic probe, an ultrasonic generator and a signal receiving and analyzing system, and the contact degree of two metal surfaces can be efficiently and accurately detected under the condition that equipment is not disassembled.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the field of ultrasonic measurement, in particular to a method and a device for judging the contact degree between two metal surfaces by using ultrasonic technology. Background Art

[0002] In the fields of mechanical manufacturing, aerospace, automobile manufacturing, electrical equipment, etc., the contact quality of metal contact surfaces directly affects the performance and life of the product. Traditional contact detection methods often rely on manual inspection or destructive testing. Such detection methods are inefficient and cannot guarantee detection accuracy. Contact resistance testers are also used, but due to the characteristics of the detection principle, if only some areas of the entire contact surface have good contact, it may be judged that the entire surface has good contact, and the contact condition of the entire plane cannot be evaluated. Therefore, it is of great significance to develop a non-destructive, efficient and accurate contact degree detection method. Summary of the invention

[0003] In view of the deficiencies of the prior art, the present invention provides a method and a device for determining the contact degree between two metal surfaces using ultrasonic technology to solve the above problems.

[0004] To achieve the above objectives, the present invention is implemented through the following technical solutions.

[0005] A method and device for determining the contact degree between two metal surfaces using ultrasonic technology comprises the following steps:

[0006] S1. Set ultrasonic emission parameters, clean metal contact surfaces, apply ultrasonic coupling agent, and place the ultrasonic probe in the detection area;

[0007] S2. The ultrasonic generator generates ultrasonic waves according to the parameters;

[0008] S3. Transmit ultrasonic waves to the metal being tested;

[0009] S4. Detect whether there is a reflected wave;

[0010] S5. Receive reflected signal;

[0011] S6. Signal processing;

[0012] S7. Waveform display, judging the contact status between metal surfaces based on the intensity and time characteristics of the reflected signal;

[0013] S8. Output results, non-invasive detection: Ultrasonic waves can evaluate the degree of contact without destroying or damaging the metal structure being tested, without physically intervening or disassembling the metal surface, high-precision measurement: It can provide high-precision measurement results and accurately reflect the contact status between metal surfaces, including tiny gaps and incomplete contact areas, deep penetration ability: Ultrasonic waves have strong penetration ability and can detect the contact situation deep inside the metal without being restricted by the surface condition, real-time monitoring: It can realize real-time online monitoring and obtain information on changes in the degree of contact between metal surfaces in a timely manner, so as to facilitate timely detection of problems and adjustments, multi-parameter analysis: By analyzing multiple parameters such as ultrasonic propagation speed, amplitude, frequency, etc., the contact characteristics of metal surfaces can be fully understood, and the scope of application is wide: Whether it is a large metal structure or a small precision component, ultrasonic testing technology can be effectively used to evaluate the degree of contact between metal surfaces, strong adaptability to complex shapes: It can detect the contact of metal surfaces with complex shapes and geometric structures without being constrained by shape, non-destructive testing advantages: Compared with some other detection methods, ultrasonic testing will not cause damage to metal materials and will not affect their subsequent performance and life.

[0014] Preferably, step S1 and step S2 include the following steps: adjusting the transmitting frequency of the instrument, wherein the higher the frequency, the higher the resolution, but the shallower the penetration depth; the lower the frequency, the deeper the penetration depth, but the lower the resolution; gain adjustment, where gain refers to amplifying the received signal strength, and the detection sensitivity of the probe can be adjusted. Based on comprehensive consideration of factors such as the thickness of the metal plate, material, signal-to-noise ratio, and size of the defect, the frequency and gain are appropriately adjusted to ensure detection accuracy, so that the height of the reflected wave is not less than 80% of the full scale of the screen, and the instrument has sufficient sensitivity margin.

[0015] Preferably, step S3 includes the following steps: before transmitting ultrasonic waves to the metal to be tested for specific detection, the operator first cleans the metal surface to be bonded, and avoids using it in places with strong magnetic fields, strong vibrations, severe dust, etc. that affect ultrasonic propagation and reception.

[0016] Preferably, step S4 includes the following steps: coating an appropriate amount of coupling agent on the surface to be bonded to improve the transmission efficiency of the ultrasonic signal, ensuring good acoustic contact between the probe and the metal surface, and moving evenly along a predetermined scanning path to comprehensively scan the entire metal surface, starting the ultrasonic generator, and the probe will emit ultrasonic pulses according to pre-set parameters. The emitted ultrasonic pulses propagate downward in the vertical direction of the metal. When the ultrasonic wave reaches the metal contact surface, it will reflect part of the ultrasonic signal back to the probe depending on the contact situation.

[0017] Preferably, step S6 includes the following steps:

[0018] S6.1 Data partitioning: Split the dataset into independent training and test sets;

[0019] S6.2 Basic evaluation: Train the model on the training set and perform preliminary evaluation on the test set;

[0020] S6.3 Cross-validation: Use K-fold cross-validation to reduce the variance of the evaluation results;

[0021] S6.4 Performance measurement: Select appropriate performance measurement indicators to evaluate the model;

[0022] S6.5 Hyperparameter optimization: Use grid search, random search, or Bayesian optimization methods to find the optimal hyperparameter combination;

[0023] S6.6 Model verification: The model is retrained using the optimal hyperparameters and finally verified on the test set. The debonding detection method based on pattern recognition is used to identify pavement debonding with the support vector machine algorithm in machine learning. Specifically:

[0024] Select the support vector machine algorithm to identify pavement gaps, establish the SVM model, set the model input value, kernel function and model parameters, the input is the eigenvalue extracted by wavelet transformation, and the output is the gap degree. SVM determines the optimal hyperplane by maximizing the interval. The interval refers to the distance between the hyperplane and the nearest sample point. It can divide the data points into two categories. Assume that the category labels are +1 and -1.

[0025] The equation of the hyperplane can be expressed as: w×x+b=0;

[0026] Where: ||w|| is the normal vector (normal vector of the hyperplane), which determines the direction of the hyperplane; ||x|| is the input feature vector; ||b|| is the bias term, also called the intercept;

[0027] The goal of SVM is to find ||w|| and ||b|| such that for each training sample (xi,yi), the following conditions hold:

[0028] yi(w×x i+b=0)≥1for i=1,2,…,n;

[0029] Where: ||yi|| is the category label of the sample, ||yi||=+1 or ||yi||=-1; n is the number of training samples;

[0030] The optimization goal of SVM is to minimize the norm of ||w||, thereby maximizing the margin and satisfying the above conditions; this can be expressed as the following convex optimization problem:

[0031] minw,b[1 / 2||w||2+C×Σ(max(0,1-yi(wT×x i+b))]for i=1,2,..,n;

[0032] Where: ||w||2 represents the L2 norm square of the weight vector, C is a regularization parameter that controls the trade-off between margin and classification error, yi is the category label of the i-th sample, and xi is the corresponding feature vector;

[0033] The signal is subjected to six consecutive wavelet decompositions, which means that the signal is decomposed into six parts with different frequency ranges; then the signal is reconstructed by combining the low-frequency parts of the multi-level wavelet decomposition. This reconstruction process can be achieved by adding the low-frequency parts of each level, thus obtaining the low-frequency reconstructed signal after the six-level decomposition;

[0034] Then, the low-frequency reconstructed signal after the six-level decomposition is selected to extract the eigenvalues ​​through changes. Then the experimental data is divided into two parts. The first part selects five samples, a total of fifty groups of data, as the training set for pattern recognition, which is used to train the SVM model; the second part also selects five samples, a total of fifty groups of data, as the test set, which is used to evaluate the performance of the model.

[0035] Preferably, the statistical analysis steps of the overall sample data are as follows:

[0036] Observe the amplitude of the reflected echo. A higher amplitude usually indicates a larger poor contact or defect, while a lower amplitude may indicate better contact or a smaller defect.

[0037] By measuring the arrival time of the echo, the depth of the defect or poor contact can be calculated;

[0038] Study the waveform shape of the echo. Sharp peaks may indicate smaller defects, while wide and gentle peaks may indicate larger defects or uneven contact.

[0039] Perform spectrum analysis on the echo signal to observe the changes in frequency components. Different defects and contact conditions may lead to different spectral characteristics of the echo.

[0040] If there are multiple reflection echoes, analyze their amplitude decay and time interval to understand the overall condition of the contact surface.

[0041] Compare the test data with the standard sample data of known good contacts and evaluate the difference.

[0042] Perform statistical analysis on the data of multiple detection points, such as calculating the mean value, standard deviation, etc., to determine the overall contact quality distribution.

[0043] A device that uses ultrasonic technology to determine the contact degree between two metal surfaces, including an ultrasonic probe, an ultrasonic generator, a signal receiving and analyzing system, and a coupling agent.

[0044] Preferably, the ultrasonic probe, ultrasonic generator, signal receiving and analyzing system, and coupling agent are used in sequence.

[0045] Preferably, the signal receiving and analyzing system includes a signal amplifier, a filter, a data collector, a signal processing module, a key input module, a human-computer interaction module, and a transmitting and receiving control module.

[0046] Preferably, the signal amplifier, filter, data collector, signal processing module, human-computer interaction module, transmitting and receiving control module, and ultrasonic generator are connected to each other in sequence, and the key input module is connected to the human-computer interaction module alone.

[0047] Compared with the prior art, the present invention discloses a method and device for judging the contact degree of two metal surfaces by using ultrasonic technology. According to the establishment of a software model, a large amount of data is drilled, and the data is divided into a training set and a test set to evaluate the performance of the model on unseen data. Then, the database is continuously updated from actual detection cases to improve the accuracy of judgment.

[0048] ① Non-invasive detection: Ultrasonic waves can evaluate the degree of contact without destroying or damaging the metal structure being detected, without physically intervening or disassembling the metal surface;

[0049] ② High-precision measurement: It can provide high-precision measurement results and accurately reflect the contact status between metal surfaces, including tiny gaps and incomplete contact areas;

[0050] ③ Deep penetration ability: Ultrasonic waves have strong penetration ability and can detect contact conditions deep inside the metal without being restricted by surface conditions;

[0051] ④Real-time monitoring: It can realize real-time online monitoring and obtain the change information of the contact degree of metal surfaces in time, so as to find problems and make adjustments in time;

[0052] ⑤Multi-parameter analysis: By analyzing multiple parameters such as ultrasonic propagation speed, amplitude, frequency, etc., the contact characteristics of the metal surface can be fully understood;

[0053] ⑥ Wide range of applications: Whether it is a large metal structure or a small precision component, ultrasonic testing technology can be effectively used to evaluate the contact degree of metal surfaces;

[0054] ⑦ Strong adaptability to complex shapes: It can detect the contact of metal surfaces with complex shapes and geometric structures without being constrained by the shape;

[0055] ⑧Advantages of non-destructive testing: Compared with some other testing methods, ultrasonic testing will not cause damage to metal materials and will not affect their subsequent performance and life. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 This is a flow chart of a method for determining the contact degree between two metal surfaces using ultrasonic technology according to the present invention;

[0057] Figure 2 This is a schematic diagram of ultrasonic signal reflection of the present invention;

[0058] Figure 3 It is a schematic diagram of the structure of the ultrasonic detection device of the present invention. DETAILED DESCRIPTION

[0059] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0060] The method and device for determining the contact degree between two metal surfaces by using ultrasonic technology include the following steps:

[0061] S1. Set ultrasonic emission parameters, clean metal contact surfaces, apply ultrasonic coupling agent, and place the ultrasonic probe in the detection area;

[0062] S2. The ultrasonic generator generates ultrasonic waves according to the parameters;

[0063] S3. emit ultrasonic waves to the metal being tested;

[0064] S4. Detect whether there is a reflected wave;

[0065] S5. Receive reflected signal;

[0066] S6. Signal processing;

[0067] S7. Waveform display, judging the contact status between metal surfaces based on the intensity and time characteristics of the reflected signal;

[0068] S8. Output the results.

[0069] The steps S1 and S2 include the following steps: adjusting the transmitting frequency of the instrument, wherein the higher the frequency, the higher the resolution, but the shallower the penetration depth; the lower the frequency, the deeper the penetration depth, but the lower the resolution; gain adjustment, where gain refers to amplifying the received signal strength, and the detection sensitivity of the probe can be adjusted. According to the comprehensive consideration of factors such as the thickness of the metal plate, the material, the signal-to-noise ratio and the size of the defect, the frequency and gain are appropriately adjusted to ensure the detection accuracy, so that the height of the reflected wave is not less than 80% of the full scale of the screen, and the instrument has sufficient sensitivity margin; the step S3 includes the following steps: before transmitting ultrasonic waves to the metal to be tested for specific detection, the operator first The metal surface to be bonded should be cleaned, and avoid using it in places with strong magnetic fields, strong vibrations, severe dust, etc. that affect ultrasonic propagation and reception. Step S4 includes the following steps: applying an appropriate amount of coupling agent on the surface to be bonded to improve the transmission efficiency of the ultrasonic signal, ensuring good acoustic contact between the probe and the metal surface, and moving evenly along a predetermined scanning path to comprehensively scan the entire metal surface. The ultrasonic generator is started, and the probe will emit ultrasonic pulses according to pre-set parameters. The emitted ultrasonic pulses propagate downward in the vertical direction of the metal. When the ultrasonic wave reaches the metal contact surface, it will reflect part of the ultrasonic signal back to the probe depending on the contact situation.

[0070] The step S6 includes the following steps:

[0071] S6.1 Data partitioning: Split the dataset into independent training and test sets;

[0072] S6.2 Basic evaluation: Train the model on the training set and perform preliminary evaluation on the test set;

[0073] S6.3 Cross-validation: Use K-fold cross-validation to reduce the variance of the evaluation results;

[0074] S6.4 Performance measurement: Select appropriate performance measurement indicators to evaluate the model;

[0075] S6.5 Hyperparameter optimization: Use grid search, random search, or Bayesian optimization methods to find the optimal hyperparameter combination;

[0076] S6.6 Model verification: The model is retrained using the optimal hyperparameters and finally verified on the test set. The debonding detection method based on pattern recognition is used to identify pavement debonding with the support vector machine algorithm in machine learning. Specifically:

[0077] Select the support vector machine algorithm to identify pavement gaps, establish the SVM model, set the model input value, kernel function and model parameters, the input is the eigenvalue extracted by wavelet transformation, and the output is the gap degree. SVM determines the optimal hyperplane by maximizing the interval. The interval refers to the distance between the hyperplane and the nearest sample point. It can divide the data points into two categories. Assume that the category labels are +1 and -1.

[0078] The equation of the hyperplane can be expressed as: w×x+b=0;

[0079] Where: ||w|| is the normal vector (normal vector of the hyperplane), which determines the direction of the hyperplane; ||x|| is the input feature vector; ||b|| is the bias term, also called the intercept;

[0080] The goal of SVM is to find ||w|| and ||b|| such that for each training sample (xi,yi), the following conditions hold:

[0081] yi(w×x i+b=0)≥1for i=1,2,…,n;

[0082] Where: ||yi|| is the category label of the sample, ||yi||=+1 or ||yi||=-1; n is the number of training samples, and the optimization goal of SVM is to minimize the norm of ||w||, thereby maximizing the interval and satisfying the above conditions; this can be expressed as the following convex optimization problem, minw,b[1 / 2||w||2+C×Σ(max(0,1-yi(wT×xi+b))]for i=1,2,..,n;

[0083] Among them: ||w||2 represents the L2 norm square of the weight vector, C is a regularization parameter that controls the trade-off between interval and classification error, yi is the category label of the i-th sample, xi is the corresponding eigenvector, and the signal is subjected to six consecutive wavelet decompositions, which means that the signal will be decomposed into six parts with different frequency ranges; then the signal is reconstructed by merging the low-frequency parts in the multi-level wavelet decomposition. This reconstruction process can be achieved by adding the low-frequency parts of each level to obtain the low-frequency reconstructed signal after the six-level decomposition, and then the low-frequency reconstructed signal after the six-level decomposition is selected to extract the eigenvalue through the change, and then the experimental data is divided into two parts. In the first part, five samples are selected, with a total of fifty groups of data, as the training set for pattern recognition, which is used to train the SVM model; in the second part, five samples are also selected, with a total of fifty groups of data, as the test set, which is used to evaluate the performance of the model. The statistical analysis steps of the overall sample data are as follows: observe the amplitude of the reflected echo. A higher amplitude usually indicates a larger poor contact or defect, while a lower amplitude may indicate better contact or a smaller defect. By measuring the echo Arrival time, the depth of the defect or poor contact position can be calculated, and the waveform shape of the echo can be studied. A sharp peak may indicate a smaller defect, while a wide and gentle peak may indicate a larger defect or uneven contact. The echo signal is spectrally analyzed to observe the changes in frequency components. Different defects and contact conditions may result in different spectral characteristics of the echo. If there are multiple reflected echoes, their amplitude attenuation and time interval are analyzed to understand the overall situation of the contact surface. The test data is compared with the standard sample data of known good contact to evaluate the difference. The data of multiple test points are statistically analyzed, such as calculating the average value, standard deviation, etc., to determine the overall contact quality distribution. A device for judging the degree of contact between two metal surfaces using ultrasonic technology includes an ultrasonic probe, an ultrasonic generator, a signal receiving and analysis system, and a coupling agent. The ultrasonic probe, ultrasonic generator, signal receiving and analysis system, and coupling agent are used in sequence. The signal receiving and analysis system includes a signal amplifier, a filter, a data collector, a signal processing module, a key input module, a human-computer interaction module, and a transmitting and receiving control module.

[0084] The signal amplifier, filter, data collector, signal processing module, human-computer interaction module, transmitting and receiving control module, and ultrasonic generator are connected to each other in sequence, and the key input module is connected to the human-computer interaction module separately. Conventional methods for detecting the contact of metal surfaces are mainly through appearance inspection, feeler gauge measurement or contact resistance measurement. No matter which one has certain limitations, such as appearance inspection and feeler gauge measurement can only detect the contact of the outer edge. If the contact area is large, the central part cannot be accurately evaluated, and the detection of contact resistance will cause the overall detection value to be normal due to good contact in some areas, which affects the judgment of the contact of the entire metal surface. We took advantage of the propagation characteristics of ultrasonic waves in different media. If the two metal surfaces are in good contact and the acoustic impedance is well matched, most of the ultrasonic waves can continue to propagate smoothly through the contact surface, and the reflected ultrasonic energy is less. On the contrary, if there are gaps, inclusions, non-fitting and other poor contact conditions between the metal surfaces, the acoustic impedance will suddenly change, and the ultrasonic wave will be strongly reflected at the interface with poor contact. The reflected ultrasonic wave is received by the probe and converted into an electrical signal for transmission to the detection instrument. The detection instrument can effectively determine the contact conditions between metal surfaces by analyzing the amplitude, time, frequency and other characteristics of the reflected wave, avoiding the defects of traditional methods.

[0085] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0086] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein.

[0087] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for determining the degree of contact between two metal surfaces using ultrasonic technology, characterized in that: The following steps are involved: S1. Set ultrasonic emission parameters, clean metal contact surfaces, apply ultrasonic coupling agent, and place the ultrasonic probe in the detection area; S2. The ultrasonic generator generates ultrasonic waves according to the parameters; S3. emit ultrasonic waves to the metal being tested; S4. Detect whether there is a reflected wave; S5. Receive reflected signal; S6. Signal processing; S7. Waveform display, judging the contact status between metal surfaces based on the intensity and time characteristics of the reflected signal; S8. Output the results.

2. The method for determining the contact degree between two metal surfaces using ultrasonic technology according to claim 1, characterized in that: The steps S1 and S2 include the following steps: adjusting the transmitting frequency of the instrument, wherein the higher the frequency, the higher the resolution, but the shallower the penetration depth; the lower the frequency, the deeper the penetration depth, but the lower the resolution; gain adjustment, where gain refers to amplifying the received signal strength, and the detection sensitivity of the probe can be adjusted. Based on comprehensive consideration of factors such as the thickness of the metal plate, the material, the signal-to-noise ratio, and the size of the defect, the frequency and gain are appropriately adjusted to ensure detection accuracy, so that the height of the reflected wave is not less than 80% of the full scale of the screen, and the instrument has sufficient sensitivity margin.

3. The method for determining the contact degree between two metal surfaces using ultrasonic technology according to claim 1, characterized in that: The step S3 includes the following steps: before transmitting ultrasonic waves to the metal to be tested for specific detection, the operator first cleans the metal surface to be bonded, and avoids using it in places with strong magnetic fields, strong vibrations, severe dust, etc. that affect ultrasonic propagation and reception.

4. The method for determining the contact degree between two metal surfaces using ultrasonic technology according to claim 1, characterized in that: The step S4 includes the following steps: coating a proper amount of coupling agent on the surface to be bonded to improve the transmission efficiency of the ultrasonic signal, ensuring good acoustic contact between the probe and the metal surface, and moving evenly along a predetermined scanning path to comprehensively scan the entire metal surface. The ultrasonic generator is started, and the probe will emit ultrasonic pulses according to pre-set parameters. The emitted ultrasonic pulses propagate downward in the vertical direction of the metal. When the ultrasonic wave reaches the metal contact surface, part of the ultrasonic signal will be reflected back to the probe according to different contact conditions.

5. The method for determining the contact degree between two metal surfaces using ultrasonic technology according to claim 1, characterized in that: The step S6 includes the following steps: S6.1 Data partitioning: Split the dataset into independent training and test sets; S6.2 Basic evaluation: Train the model on the training set and perform preliminary evaluation on the test set; S6.3 Cross-validation: Use K-fold cross-validation to reduce the variance of the evaluation results; S6.4 Performance measurement: Select appropriate performance measurement indicators to evaluate the model; S6.5 Hyperparameter optimization: Use grid search, random search, or Bayesian optimization methods to find the optimal hyperparameter combination; S6.6 Model verification: The model is retrained using the optimal hyperparameters and finally verified on the test set. The debonding detection method based on pattern recognition is used to identify pavement debonding with the support vector machine algorithm in machine learning. Specifically: Select the support vector machine algorithm to identify pavement gaps, establish the SVM model, set the model input value, kernel function and model parameters, the input is the eigenvalue extracted by wavelet transformation, and the output is the gap degree. SVM determines the optimal hyperplane by maximizing the interval. The interval refers to the distance between the hyperplane and the nearest sample point. It can divide the data points into two categories. Assume that the category labels are +1 and -1. The equation of the hyperplane can be expressed as: w×x+b=0; Where: ||w|| is the normal vector (normal vector of the hyperplane), which determines the direction of the hyperplane; ||x|| is the input feature vector; ||b|| is the bias term, also called the intercept; The goal of SVM is to find ||w|| and ||b|| such that for each training sample (xi,yi), the following conditions hold: yi(w×xi+b=0)≥1for i=1,2,…,n; Where: ||yi|| is the category label of the sample, ||yi||=+1 or ||yi||=-1; n is the number of training samples; The optimization goal of SVM is to minimize the norm of ||w||, thereby maximizing the margin and satisfying the above conditions; this can be expressed as the following convex optimization problem: minw,b[1 / 2||w||2+C×Σ(max(0,1-yi(wT×xi+b))]for i=1,2,..,n; Where: ||w||2 represents the L2 norm square of the weight vector, C is a regularization parameter that controls the trade-off between margin and classification error, yi is the category label of the i-th sample, and xi is the corresponding feature vector; The signal is subjected to six consecutive wavelet decompositions, which means that the signal is decomposed into six parts with different frequency ranges; then the signal is reconstructed by combining the low-frequency parts of the multi-level wavelet decomposition. This reconstruction process can be achieved by adding the low-frequency parts of each level, thus obtaining the low-frequency reconstructed signal after the six-level decomposition; Then, the low-frequency reconstructed signal after the six-level decomposition is selected to extract the eigenvalues ​​through changes. Then the experimental data is divided into two parts. The first part selects five samples, a total of fifty groups of data, as the training set for pattern recognition, which is used to train the SVM model; the second part also selects five samples, a total of fifty groups of data, as the test set, which is used to evaluate the performance of the model.

6. The method of using ultrasonic technology to determine the contact degree between two metal surfaces according to claim 1, characterized in that: The statistical analysis steps of the overall sample data are as follows: Observe the amplitude of the reflected echo. A higher amplitude usually indicates a larger poor contact or defect, while a lower amplitude may indicate better contact or a smaller defect. By measuring the arrival time of the echo, the depth of the defect or poor contact can be calculated; Study the waveform shape of the echo. Sharp peaks may indicate smaller defects, while wide and gentle peaks may indicate larger defects or uneven contact. Perform spectrum analysis on the echo signal to observe the changes in frequency components. Different defects and contact conditions may lead to different spectral characteristics of the echo. If there are multiple reflection echoes, analyze their amplitude decay and time interval to understand the overall condition of the contact surface. Compare the test data with the standard sample data of known good contacts and evaluate the difference. Perform statistical analysis on the data of multiple detection points, such as calculating the mean value, standard deviation, etc., to determine the overall contact quality distribution.

7. A device for determining the contact degree between two metal surfaces using ultrasonic technology according to any one of claims 1 to 6, characterized in that: It includes an ultrasonic probe, an ultrasonic generator, a signal receiving and analyzing system, and a coupling agent.

8. The device for determining the contact degree between two metal surfaces by using ultrasonic technology according to claim 7, characterized in that: The ultrasonic probe, ultrasonic generator, signal receiving and analyzing system and coupling agent are used in conjunction with each other in sequence.

9. The device for determining the contact degree between two metal surfaces by using ultrasonic technology according to claim 8, characterized in that: The signal receiving and analyzing system comprises a signal amplifier, a filter, a data collector, a signal processing module, a key input module, a human-computer interaction module, and a transmitting and receiving control module.

10. The device for determining the contact degree between two metal surfaces by using ultrasonic technology according to claim 9, characterized in that: The signal amplifier, filter, data collector, signal processing module, human-computer interaction module, transmitting and receiving control module and ultrasonic generator are connected to each other in sequence, and the key input module is connected to the human-computer interaction module alone.

Citation Information

Patent Citations

  • Phased-array ultrasonic detection method for interfacial de-bonding of composite material

    CN103901108A

  • Ultrasonic detecting method of defects in thick-wall composite tubular structure

    CN105388212A

  • Airfield pavement slab void detection method based on ultrasonic waves

    CN117571822A

  • Precise positioning ultrasonic defect detection system and method

    CN117825516A

  • METHOD FOR DETECTING DEFECTS IN ADHESIVE JOINTS AND A DEVICE FOR ITS IMPLEMENTATION

    RU2009148401A