Ultrasonic defect signal automatic detection method based on instantaneous scanning signal statistics

Abstract

The invention provides an ultrasonic defect signal automatic detection method based on instantaneous a scanning signal statistics. The ultrasonic defect signal automatic detection method comprises the following steps: step 1, acquiring all ultrasonic a scanning signals [U] t1 of a detected material at a moment t; 2, acquiring a signal statistical table [D] t1 at the moment t; 3, all ultrasonic signal types in the signal statistical table [D] t1 are extracted and analyzed; 4, the analysis result serves as reconstruction data and is included in a reconstruction signal statistical table [D] t2 at the t moment; 5, reconstructing all the ultrasonic a scanning signals [U] t1 at the t moment into [U] t2 according to the mapping relation on the physical positions of the signal statistical table [D] t1 at the t moment and the reconstructed signal statistical table [D] t2; and step 6, repeating the steps 1-5, and reconstructing and superposing the ultrasonic scanning signals a at all moments to obtain a reconstructed ultrasonic detection result. According to the method, automatic processing of ultrasonic data can be realized, defect distribution information can be directly obtained after original data are imported, and multi-field technology integration and development based on an ultrasonic nondestructive testing technology are facilitated.

Description

Technical Field

[0001] The invention relates to an automatic detection method for ultrasonic defect signals based on instantaneous A-scan signal statistics, and belongs to the technical field of ultrasonic non-destructive testing. Background Art

[0002] At present, non-destructive testing technology is widely used in the processing of various instruments and equipment and the maintenance and support of in-service equipment. Ultrasonic flaw detection is a non-destructive testing method that uses the difference in acoustic properties of materials and their defects to reflect the ultrasonic propagation waveform and the energy change of penetration time to detect internal defects of materials. It has the advantages of large detection thickness, high sensitivity, fast speed, low cost, harmless to the human body, and the ability to locate and quantify defects. It is widely used in the field of non-destructive testing.

[0003] At present, ultrasonic non-destructive testing technology has the following defects: 1. Most of the ultrasonic testing equipment on the market is of a-scan type, which is not intuitive in displaying defects and is easily affected by subjective and objective factors; 2. The current more intuitive ultrasonic c-scan technology requires the setting of gate parameters, such as gate width, gate starting point and gate threshold, etc., which requires high non-destructive testing experience of operators, which is not conducive to the continued promotion of ultrasonic non-destructive testing technology; 3. The results obtained from ultrasonic non-destructive testing are not easy to save due to subjective and objective factors such as large changes in gate parameters. Therefore, when multiple non-destructive testing technologies are combined in the future, it is difficult to reuse the ultrasonic testing results, and the ultrasonic testing results have a lot of background noise, requiring experienced inspectors to accurately identify the types of defects.

[0004] In existing research, Hao Yanpeng et al. (patent number CN110320274B) of South China University of Technology proposed a method for reconstructing internal defects of three-pillar insulators based on the principle of ultrasonic scanning, and disclosed a method for reconstructing internal defects of three-pillar insulators based on the principle of ultrasonic scanning, including the following steps: S1, building an ultrasonic detection system; S2, the ultrasonic detection system detects the same positions of the three-pillar insulators and records the ultrasonic reflection echo information of each detection position; S3, using the analogy method to judge the ultrasonic reflection echo information of each detection position and determine the defect position; S4, the ultrasonic detection system performs ultrasonic scanning near the defect position and records the ultrasonic reflection echo information of each detection position; S5, constructing the corresponding relationship between the ultrasonic reflection echo information and the defect depth and defect size, and obtaining a defect reconstruction schematic diagram based on the detection results of each scanning position. The present invention can efficiently, accurately and intuitively identify and locate the internal defects of the three-pillar insulator and quantify the defects. Although this method utilizes the echo difference between the normal area and the defect area of ​​the material, the point-by-point calculation efficiency is low, and it is only applicable to small components such as insulators, and the calculation performance for large-area materials is poor.

[0005] Lin Jing et al. (patent number CN104764804A) of Xi'an Jiaotong University disclosed an ultrasonic Lamb wave local cycle scanning probability reconstruction tomography method. The method first plans the detection area, arranges the sensor array on the detection component, selects the Lamb wave mode and designs the excitation signal, and then collects the reference signal and damage signal of each sensor pair respectively, divides the detection area into several sectors, and uses several adjacent sectors as detection sub-areas. After each sub-area detection, the sub-area is rotated clockwise by one sector as the next cycle detection sub-area, and then data selection preprocessing is performed. The effective data length is calculated according to the distance between the sensor pairs and the selected Lamb mode speed, and then the local cycle probability reconstruction tomography calculation is performed, and finally data fusion and image splicing are performed. The invention has strong anti-interference ability and high precision, and can effectively eliminate the influence of interference such as multiple faults on the method, and improve the detection resolution of multiple fault damage. The method realizes anti-interference detection of multi-fault damage through regional detection, but the size of the detection area needs to be manually controlled, which is not conducive to the automatic output and transcription of detection data when multiple systems are linked.

[0006] Yu Xiaonan et al. of Jiangnan University (patent number CN109087246B) disclosed an ultrasonic image reconstruction method based on sparse reconstruction, which relates to the field of image processing technology. The method includes: using an ultrasonic probe to obtain the ultrasonic echo time domain signal of the sample to be tested, constructing an overcomplete dictionary by referring to the echo signal, using the overcomplete dictionary to sparsely decompose the ultrasonic echo time domain signal of the sample to be tested to obtain sparse decomposition coefficients, reconstructing the ultrasonic echo time domain signal according to the sparse decomposition coefficients and the overcomplete dictionary, and then combining the reconstructed ultrasonic echo time domain signals according to the spatial position to form a reconstructed ultrasonic image of the sample to be tested; the horizontal and vertical resolution of the image can be improved, and the position, size and distribution of micro defects inside the sample can be quickly and effectively observed. Although this method can effectively improve the horizontal and vertical resolution of the ultrasonic image, the construction of the dictionary based on the Gabor function requires a large amount of data for fitting, and its applicability is poor in the ultrasonic non-destructive testing environment with limited data. Summary of the invention

[0007] The present invention aims to solve the technical problem that gate parameters need to be manually set during ultrasonic testing, and further proposes an automatic detection method for ultrasonic defect signals based on instantaneous a-scan signal statistics.

[0008] The technical solution adopted by the present invention to solve the above problems is: the present invention proposes an automatic detection method of ultrasonic defect signals based on instantaneous a-scan signal statistics, comprising:

[0009] Step 1: Obtain all ultrasonic a scanning signals [U]t1 of the material under test at time t;

[0010] Step 2: Count all ultrasonic signals in the ultrasonic a scanning signal [U]t1 to obtain a signal statistics table [D]t1 at time t;

[0011] Step 3: extract all ultrasonic signal types in the signal statistics table [D]t1 and analyze them;

[0012] Step 4: The analysis results are recorded as reconstruction data in the reconstructed signal statistics table [D]t2 at time t;

[0013] Step 5: Reconstruct all ultrasonic a-scan signals [U]t1 at time t into [U]t2 according to the mapping relationship between the physical positions of the signal statistics table [D]t1 at time t and the reconstructed signal statistics table [D]t2, and complete the extraction of defect information in the ultrasonic a-scan signals;

[0014] Step 6: Repeat steps 1-5 to reconstruct and superimpose the ultrasonic a-scan signals at all times to obtain the reconstructed ultrasonic detection results.

[0015] Optionally, the step of obtaining all ultrasonic a scanning signals [U]t1 of the tested material at time t in step 1 includes:

[0016] Step 1.1: Use phased array ultrasonic elements to stimulate the material;

[0017] Step 1.2: Obtain the change of pressure at the material's excited position over time, and convert the sound pressure into a voltage signal through a piezoelectric sensor to form a scan data of a point on the material surface;

[0018] Step 1.3: Repeat steps 1.1 to 1.2 to obtain all a-scan data of the material surface, assemble all a-scan data of the material surface according to physical positions, and obtain a three-dimensional matrix [U]t of the sound pressure signal of the material surface changing with time;

[0019] Step 1.4: Based on the three-dimensional matrix [U]t of the time-varying acoustic pressure signal on the material surface, obtain all ultrasonic a-scanning signals [U]t1 of the material under test at time t.

[0020] The present invention can effectively solve the problem that gate parameters require high experience for inspection personnel and require a long time to accumulate inspection experience, realizes fully automatic processing of ultrasonic A-scan data, and effectively lowers the threshold for popularization and use of ultrasonic non-destructive testing technology.

[0021] Optionally, the step of exciting the material using phased array ultrasonic array elements in step 1.1 includes:

[0022] Step 1.1.1: Perform an ultrasonic a scan on the material to be tested, obtain the propagation path of the phased array ultrasonic sound wave in the wedge block and the material to be tested, and obtain the radiated sound wave line of the single ultrasonic array element based on the propagation path of the phased array ultrasonic sound wave in the wedge block and the material to be tested;

[0023] Step 1.1.2: Model the radiated sound wave line of a single ultrasonic array element to obtain the particle vibration velocity generated in the wedge;

[0024] Step 1.1.3: Approximately calculate the integrand in the particle vibration velocity to obtain a model of a single ultrasonic array element with a length of lx and a width of ly radiating ultrasonic longitudinal waves into a solid medium;

[0025] Step 1.1.4: Divide the ultrasonic array element into m blocks and n blocks in the x direction and y direction respectively, and obtain the three-dimensional sound field model of the single ultrasonic array element in the wedge block based on the high-frequency approximation and steady-phase method and the model of a single ultrasonic array element radiating ultrasonic longitudinal waves into a solid medium;

[0026] Step 1.1.5: Introduce the plane wave transmission coefficient T of the ultrasonic wave passing through the liquid coupling layer between the wedge and the material under test into the three-dimensional sound field model of the single array element in the wedge to obtain the three-dimensional sound field model of the single array element in the material under test;

[0027] Step 1.1.6: Obtain the propagation of ultrasonic waves inside the material under test based on the three-dimensional acoustic field model of a single array element in the material under test. When there are defects in the material, the ultrasonic wave propagation distance changes. The relevant information of the first echo is calculated to complete the excitation of the material.

[0028] The expression for the particle vibration velocity generated in the wedge is:

[0029]

[0030] In formula (1), K p (θ′) is the directivity function of the sound field, r is the length of the path from the center of the array element to point x, θ′ is the angle between the path from point x to a certain point x′ on the array element and the horizontal plane, k p1 is the wave number, p0(ω) is the surface pressure of the array element, κ is the wave velocity ratio, G is the gravitational constant, c p1 is the longitudinal wave velocity, i is the number of incident points, S is the surface area of ​​the array element, d′ p is the unit vector from point x to a point x′ on the array element;

[0031] Sound field directivity function, K p The expression of (θ′) is:

[0032]

[0033] In formula (2), κ is the ratio of the longitudinal wave speed to the transverse wave speed when the sound wave propagates in the solid;

[0034] The expression of the model of a single ultrasonic array element radiating ultrasonic longitudinal waves into a solid medium is:

[0035]

[0036] In formula (3), is the pointing function in the spherical coordinate system;

[0037] Pointing function in spherical coordinates The expression is;

[0038]

[0039] The expression for the information about the first echo is:

[0040]

[0041] Optionally, the step of obtaining the signal statistics table [D]t1 at time t in step 2 includes:

[0042] The types of all ultrasonic signals in all ultrasonic a-scan signals [U]t1 of the tested material at time t are extracted and counted, and a signal statistics table [D]t1 at time t is established. The signal statistics table [D]t1 at time t includes all types of ultrasonic signals and the number of ultrasonic signals of the same type.

[0043] Optionally, the step of extracting and analyzing all ultrasonic signal types in the signal statistics table [D]t1 in step 3 includes:

[0044] Step 3.1: Extract all ultrasonic signal types in the signal statistics table [D]t1 at time t, and fold the repeated signal values;

[0045] Step 3.2: Calculate the difference between each type of folded ultrasonic signal and other types of ultrasonic signals.

[0046] Step 3.3: The difference between each type of ultrasonic signal and other types of ultrasonic signals and the ratio of the mean difference between the nearest 1 / 3 ultrasonic signals of the corresponding type of ultrasonic signal and all other types of ultrasonic signals are used as reconstructed data to complete the analysis of all signal types in the signal statistics table [D]t1;

[0047] By reconstructing the a-scan data, the defect information can be regularized and the background information can be leveled, which is beneficial to the secondary use and development of ultrasonic nondestructive testing data.

[0048] The expression for the folded repeat signal value is;

[0049]

[0050] In formula (6), num is the number of corresponding types of ultrasonic signals, p1, p2, p n For different types of ultrasonic signals, p i is the difference between the ith ultrasonic signal and other types of ultrasonic signals;

[0051] The expression of the difference between each type of ultrasonic signal and other types of ultrasonic signals is:

[0052] D=[d1,K,d n ](7);

[0053] d i =[p i -p j |,num i ],j=1,2,K,n.(8);

[0054] In formulas (7) and (8), D is the difference matrix formed by the difference between each type of ultrasonic signal and other types of ultrasonic signals, d1, d n is the difference sequence between the first and nth ultrasonic signals and other types of ultrasonic signals, d i is the difference between the ith ultrasonic signal and other types of ultrasonic signals, and the calculation method is the ith ultrasonic signal p i and the jth ultrasonic signal p j The difference between

[0055] The calculation formula for reconstructed data is:

[0056]

[0057]

[0058] In formulas (9) and (10), RD(p i ) is the mean difference between the nearest 1 / 3 ultrasonic signals of the corresponding type and all other types of ultrasonic signals, F(p i ) is the ratio of the difference between each type of ultrasonic signal and other types of ultrasonic signals to the mean difference between the nearest 1 / 3 ultrasonic signals of the corresponding type of ultrasonic signal and other types of ultrasonic signals, N 1 / 3 (p i ) is the difference between the closest 1 / 3 ultrasonic signal of the type of ultrasonic signal and all other types of ultrasonic signals.

[0059] Optionally, the step of obtaining the reconstructed signal statistics table [D]t2 at time t in step 4 includes:

[0060] Construct the statistical table of reconstructed signals at time t based on the reconstructed data [D]t2;

[0061] The expression of the reconstructed signal statistics table [D]t2 at time t is:

[0062]

[0063] In formula (11), F(p1), F(p2), M and F(p n ) is the reconstructed data.

[0064] The beneficial effects of the present invention are:

[0065] 1. The present invention can effectively solve the problem that gate parameters require high experience for inspection personnel and require a long time to accumulate inspection experience, realizes the fully automatic processing of ultrasonic A-scan data, and effectively lowers the threshold for popularization and use of ultrasonic non-destructive testing technology.

[0066] 2. The present invention solves the problem of misjudgment caused by background noise interference in the detection results when the ultrasonic detection data is reused for the second time and the need to manually set the gate parameters during ultrasonic detection.

[0067] 3. The present invention can realize the automatic processing of ultrasonic data. After the original data is imported, the defect distribution information can be directly obtained, which is helpful for the integration and development of multi-field technologies based on ultrasonic non-destructive testing technology.

[0068] 4. The present invention can effectively solve the problems of disordered structure of ultrasonic nondestructive testing data and excessive background noise. By reconstructing the a-scan data, the defect information can be regularized and the background information can be leveled, which is conducive to the secondary use and development of ultrasonic nondestructive testing data. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 A flow chart of an automatic ultrasonic defect signal detection method based on instantaneous a-scan signal statistics provided by the present invention;

[0070] Figure 2 A schematic diagram of the radiated sound waves of the ultrasonic array element provided by the present invention;

[0071] Figure 3 A diagram of the ultrasonic phased array detection device provided by the present invention;

[0072] Figure 4 Schematic diagram of the front (a) and back (b) morphology and scanning area of ​​the test sample ES1 provided by the present invention;

[0073] Figure 5 A comparison chart of different processing effects of the same group of a scan data provided by the present invention;

[0074] Figure 6 A graph of the carbon fiber reinforced composite material to be tested provided by the present invention;

[0075] Figure 7 The Doppler phased array C-scan result diagram of adjusting the gate parameters to the optimal value provided by the present invention;

[0076] Figure 8 The present invention provides an automatic ultrasonic defect signal detection method based on instantaneous a-scan signal statistics. Figure 6 A scan data result diagram of the middle scan area. DETAILED DESCRIPTION

[0077] Combination Figure 1 and Figure 2 This embodiment is described as follows. Figure 1 As shown, the steps of the automatic detection method of ultrasonic defect signals based on instantaneous a-scan signal statistics described in this embodiment include:

[0078] S1: Obtain all ultrasonic a scanning signals [U]t1 of the material under test at time t;

[0079] S101: Using the Rayleigh-Sommerfeld integral form Figure 2 The propagation path of the phased array ultrasonic sound wave in the wedge and the specimen as shown in the figure and the radiation sound wave line of the single array element are modeled to obtain the particle vibration velocity generated in the wedge:

[0080] The expression for the particle vibration velocity generated in the wedge is:

[0081]

[0082] In formula (1), K p (θ′) is the directivity function of the sound field, r is the length of the path from the center of the array element to point x, θ′ is the angle between the path from point x to a certain point x′ on the array element and the horizontal plane, k p1 is the wave number, p0(ω) is the surface pressure of the array element, κ is the wave velocity ratio, G is the gravitational constant, c p1 is the longitudinal wave velocity, i is the number of incident points, S is the surface area of ​​the array element, d′ p is the unit vector from point x to a point x′ on the array element;

[0083] Sound field directivity function, K p The expression of (θ′) is:

[0084]

[0085] In formula (2), κ is the ratio of the longitudinal wave speed to the transverse wave speed when the sound wave propagates in the solid;

[0086] S102: Substitute the wave number kp1, longitudinal wave velocity cp1, wedge density ρ and wave velocity ratio κ to obtain the vibration velocity of the ultrasonic wave at a certain point in the wedge. Since the actual array element size is small, when the sound wave transmission distance is much larger than the array element size, Figure 2 The distance r' from the calculation point x in the array to a point x' on the array element can be approximated to r. By approximating the integrand, we can obtain the model of a single array element with length lx and width ly radiating ultrasonic longitudinal waves into a solid medium:

[0087] The expression of the model of a single ultrasonic array element radiating ultrasonic longitudinal waves into a solid medium is:

[0088]

[0089] In formula (3), D l (θ, φ) is the pointing function in the spherical coordinate system;

[0090] Pointing function in spherical coordinates The expression is;

[0091]

[0092] S103: Taking into account the liquid coupling agent between the wedge and the specimen, assuming that the wedge and the specimen interface are in smooth contact, the ultrasonic array element is divided into m and n blocks in the x direction and y direction respectively. After obtaining the three-dimensional acoustic field model of a single array element in the wedge based on the high-frequency approximation and steady-phase method, the plane wave transmission coefficient T of the ultrasonic wave passing through the liquid coupling layer between the wedge and the specimen is introduced to obtain the three-dimensional acoustic field model of a single array element in the specimen. According to the acoustic field model, the propagation of ultrasonic waves inside the material can be obtained. When there are defects in the material, it can be regarded as a change in the ultrasonic wave propagation distance, and the relevant information of the first echo can be calculated;

[0093] The expression for the information about the first echo is:

[0094]

[0095] S104: After exciting the material through the phased array ultrasonic array elements in S101, the pressure variation of the excited position over time is obtained, and the sound pressure is converted into an electrical signal through a piezoelectric sensor to finally form a scan data of a point on the surface of the material. By assembling all the a scan signals according to the physical positions, a three-dimensional matrix [U]t of the sound pressure signal on the surface of the material over time is finally obtained;

[0096] S2: Count all ultrasonic signals in the ultrasonic a scanning signal [U]t1 to obtain a signal statistics table [D]t1 at time t;

[0097] Extract all ultrasonic signal types from all ultrasonic a-scan signals [U]t1 of the tested material at time t and perform statistics, and establish a signal statistics table [D]t1 at time t, wherein the signal statistics table [D]t1 at time t includes all types of ultrasonic signals and the number of ultrasonic signals of the same type;

[0098] S3: extracting and analyzing all ultrasonic signal types in the signal statistics table [D]t1, and recording the analysis results as reconstruction data into the reconstructed signal statistics table [D]t2 at time t;

[0099] S301: Taking the signal matrix [U]t1 at a certain moment in the three-dimensional matrix as an example, firstly, [U]t1 is counted to obtain the number of occurrences of each signal in the matrix, and repeated signal values ​​are folded to reduce the number of subsequent calculation steps:

[0100] The expression for the folded repeat signal value is;

[0101]

[0102] In formula (6), num is the number of corresponding types of ultrasonic signals, p1, p2, M, p n For different types of ultrasonic signals, p i is the difference between the ith ultrasonic signal and other types of ultrasonic signals;

[0103] S302: Calculating the difference between each type of folded ultrasonic signal and other types of ultrasonic signals;

[0104] The expression of the difference between each type of ultrasonic signal and other types of ultrasonic signals is:

[0105] D=[d1,K,d n ](7);

[0106] d i =[p i -p j |,num i ],j=1,2,K,n.(8);

[0107] In formulas (7) and (8), D is the difference matrix formed by the difference between each type of ultrasonic signal and other types of ultrasonic signals, d1, d n is the difference sequence between the first and nth ultrasonic signals and other types of ultrasonic signals, d i is the difference between the ith ultrasonic signal and other types of ultrasonic signals, and the calculation method is the ith ultrasonic signal p i and the jth ultrasonic signal p j The difference between

[0108] S303: taking the difference between each type of ultrasonic signal and other types of ultrasonic signals and the ratio of the average difference between the nearest 1 / 3 ultrasonic signals of the corresponding type of ultrasonic signal and other types of ultrasonic signals as reconstruction data, and constructing a statistical table of reconstructed signals at time t [D]t2;

[0109] The calculation formula for reconstructed data is:

[0110]

[0111]

[0112] In formulas (9) and (10), RD(p i ) is the mean difference between the nearest 1 / 3 ultrasonic signals of the corresponding type and all other types of ultrasonic signals, F(p i ) is the ratio of the difference between each type of ultrasonic signal and other types of ultrasonic signals to the mean difference between the nearest 1 / 3 ultrasonic signals of the corresponding type of ultrasonic signal and other types of ultrasonic signals, N 1 / 3 (p i ) is the difference between the closest 1 / 3 ultrasonic signal of the type of ultrasonic signal and all other types of ultrasonic signals.

[0113] The expression of the reconstructed signal statistics table [D]t2 at time t is:

[0114]

[0115] In formula (11), F(p1), F(p2), M and F(p n ) is the reconstructed data.

[0116] S4: reconstruct all ultrasonic a-scan signals [U]t1 at time t into [U]t2 according to the mapping relationship between the physical positions of the signal statistics table [D]t1 at time t and the reconstructed signal statistics table [D]t2, and complete the extraction of defect information in the ultrasonic a-scan signals;

[0117] S5: Repeat S1-S4 to reconstruct and superimpose the ultrasonic a-scan signals at all times to obtain the reconstructed ultrasonic detection results.

[0118] Example

[0119] Example 1

[0120] Combination Figure 3-5 This embodiment is described. Firstly, this embodiment can reduce the human intervention in the detection process, realize the automation of the detection process, and reduce the requirement for the operator's non-destructive testing experience. Figure 3The ROBUST 32-128(64)_TOFD ultrasonic phased array board system shown is used for scanning with the LLA482 phased array probe;

[0121] For example Figure 4 The phenolic resin-based glass fiber reinforced composite material sample ES1 is scanned, and the scanning area is Figure 4 (a) The inner yellow area. The scanning results are set to the best effect by setting the gate, not setting the gate, and processing the a scan data using the algorithm proposed in this patent to obtain the following results: Figure 5 shown.

[0122] Depend on Figure 5 It can be seen that the algorithm proposed in this embodiment effectively avoids the intervention of manually setting gate parameters in the detection process, and obtains better defect distribution results than the gate method C-scan reconstruction.

[0123] Example 2

[0124] Combination Figure 3 and Figure 6-8 This embodiment is described. Figure 3 The ROBUST32-128(64)_TOFD ultrasonic phased array board system and the LLA482 phased array probe are shown in the figure. Figure 6 The marked area of ​​the carbon fiber reinforced composite material ES2 with prefabricated flat-bottom holes is scanned. To ensure the detection effect, the sample ES2 is immersed in water for detection. After adjusting the gate parameters to the optimal effect, the two-dimensional and three-dimensional effect diagrams of the detection results are shown in the figure below. Figure 7 The obtained a-scan data is processed using the method proposed in this patent, and the resulting two-dimensional and three-dimensional effect diagrams are shown as follows Figure 8 shown.

[0125] like Figure 7 As shown, the C-scan effect obtained after improving the coupling environment is better than that in Example 1. Figure 5 The improvement is great. From the two-dimensional graph, the background noise is greatly alleviated. However, from the three-dimensional effect graph, it is still found that the background signal in the test result is poorly flat, and the edge and center signals of the defect area are distributed on both sides of the background signal. This signal format is only suitable for manual recognition. Due to the disorderly distribution of defect information, automated recognition technologies such as neural networks and machine vision cannot accurately extract defect information from background signals, which poses a great challenge to the secondary use of ultrasonic test results.

[0126] like Figure 8 As shown, the automatic ultrasonic defect detection method proposed in this embodiment can achieve a high smoothness of the background signal without human intervention, and place the defect information on one side of the background signal, which is very convenient for subsequent data fusion or reuse by artificial intelligence.

[0127] In summary, after verification of Example 1 and Example 2, the present invention can achieve the following technical effects:

[0128] 1. This method can effectively solve the problem of high experience requirements for inspection personnel and a long time to accumulate inspection experience caused by gate parameters, and effectively reduce the threshold for the popularization and use of ultrasonic non-destructive testing technology;

[0129] 2. This method can realize the automatic processing of ultrasonic data. After importing the original data, the defect distribution information can be directly obtained, which is conducive to the integration and development of multi-field technologies based on ultrasonic non-destructive testing technology;

[0130] 3. This method can effectively solve the problems of complex structure of ultrasonic nondestructive testing data and excessive background noise. By reconstructing the a-scan data, it can regularize defect information and smooth the background information, which is conducive to the secondary use and development of ultrasonic nondestructive testing data.

[0131] The above is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technician familiar with this profession can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent replacement and improvement made to the above embodiments without departing from the content of the technical solution of the present invention, based on the technical essence of the present invention, within the spirit and principles of the present invention, still fall within the protection scope of the technical solution of the present invention.

Claims

1. An automatic ultrasonic defect signal detection method based on instantaneous a-scan signal statistics, characterized in that: The steps of the automatic detection method of ultrasonic defect signals based on instantaneous a-scan signal statistics include: Step 1: Obtain all ultrasonic a scanning signals [U]t1 of the material under test at time t; Step 2: Count all ultrasonic signals in the ultrasonic a-scan signal [U]t1 to obtain a signal statistics table [D]t1 at time t; Step 3: extract all ultrasonic signal types in the signal statistics table [D]t1 and analyze them; Step 4: The analysis results are recorded as reconstruction data in the reconstructed signal statistics table [D]t2 at time t; Step 5: reconstruct all ultrasonic a-scan signals [U]t1 at time t into [U]t2 according to the mapping relationship between the physical positions of the signal statistics table [D]t1 at time t and the reconstructed signal statistics table [D]t2, and complete the extraction of defect information in the ultrasonic a-scan signals; Step 6: Repeat steps 1-5 to reconstruct and superimpose the ultrasonic a-scan signals at all times to obtain the reconstructed ultrasonic detection results.

2. The automatic detection method of ultrasonic defect signals based on instantaneous a-scan signal statistics according to claim 1 is characterized in that: The step of obtaining all ultrasonic a scanning signals [U]t1 of the material under test at time t in step 1 includes: Step 1.1: Use phased array ultrasonic elements to stimulate the material; Step 1.2: Obtain the change of pressure at the material's excited position over time, and convert the sound pressure into a voltage signal through a piezoelectric sensor to form a scan data of a point on the material surface; Step 1.3: Repeat steps 1.1 to 1.2 to obtain all a-scan data of the material surface, assemble all a-scan data of the material surface according to physical positions, and obtain a three-dimensional matrix [U]t of the sound pressure signal of the material surface changing with time; Step 1.4: Obtain all ultrasonic a-scanning signals [U]t1 of the material under test at time t based on the three-dimensional matrix [U]t of the acoustic pressure signal on the material surface that changes with time.

3. The automatic ultrasonic defect signal detection method based on instantaneous a-scan signal statistics according to claim 2 is characterized in that: The steps of using phased array ultrasonic elements to excite the material in step 1.1 include: Step 1.1.1: Perform an ultrasonic a scan on the material to be tested, obtain the propagation path of the phased array ultrasonic sound wave in the wedge block and the material to be tested, and obtain the radiated sound wave line of the single ultrasonic array element based on the propagation path of the phased array ultrasonic sound wave in the wedge block and the material to be tested; Step 1.1.2: Model the radiated sound wave line of a single ultrasonic array element to obtain the particle vibration velocity generated in the wedge; Step 1.1.3: Approximately calculate the integrand in the particle vibration velocity to obtain a model of a single ultrasonic array element with a length of lx and a width of ly radiating ultrasonic longitudinal waves into a solid medium; Step 1.1.4: Divide the ultrasonic array element into m blocks and n blocks in the x direction and y direction respectively, and obtain the three-dimensional sound field model of the single ultrasonic array element in the wedge block based on the high-frequency approximation and steady-phase method and the model of a single ultrasonic array element radiating ultrasonic longitudinal waves into a solid medium; Step 1.1.5: Introduce the plane wave transmission coefficient T of the ultrasonic wave passing through the liquid coupling layer between the wedge and the material under test into the three-dimensional sound field model of the single array element in the wedge to obtain the three-dimensional sound field model of the single array element in the material under test; Step 1.1.6: Obtain the propagation of ultrasound inside the material under test based on the three-dimensional acoustic field model of a single array element in the material under test. When there are defects in the material, the ultrasonic propagation distance changes. The relevant information of the first echo is calculated to complete the excitation of the material.

4. The automatic ultrasonic defect signal detection method based on instantaneous a-scan signal statistics according to claim 1 is characterized in that: The steps of obtaining the signal statistics table [D]t1 at time t in step 2 include: The types of all ultrasonic signals in all ultrasonic a-scan signals [U]t1 of the tested material at time t are extracted and counted, and a signal statistics table [D]t1 at time t is established, wherein the signal statistics table [D]t1 at time t includes all types of ultrasonic signals and the number of ultrasonic signals of the same type.

5. The automatic detection method of ultrasonic defect signals based on instantaneous a-scan signal statistics according to claim 1 is characterized in that: The steps of extracting and analyzing all ultrasonic signal types in the signal statistics table [D]t1 in step 3 include: Step 3.1: extract all ultrasonic signal types in the signal statistics table [D]t1 at time t, and fold the repeated signal values; Step 3.2: Calculate the difference between each type of folded ultrasonic signal and other types of ultrasonic signals; Step 3.3: The difference between each type of ultrasonic signal and other types of ultrasonic signals and the ratio of the mean difference between the nearest 1 / 3 ultrasonic signals of the corresponding type of ultrasonic signal and all other types of ultrasonic signals are used as reconstructed data to complete the analysis of all signal types in the signal statistics table [D]t1; The expression for the folded repeat signal value is; In formula (1), num is the number of corresponding types of ultrasonic signals, p1, p2, p n are the first, second and nth ultrasonic signals, p i is the ith ultrasonic signal; The expression of the difference between each type of ultrasonic signal and other types of ultrasonic signals is: D=[d1,K,d n ](2); d i =[p i -p j |,num i ],j=1,2,K,n.(3); In formulas (2) and (3), D is the difference matrix formed by the difference between each type of ultrasonic signal and other types of ultrasonic signals, d1, d n is the difference sequence between the first and nth ultrasonic signals and other types of ultrasonic signals, d i is the difference between the ith ultrasonic signal and other types of ultrasonic signals, and the calculation method is the ith ultrasonic signal p i and the jth ultrasonic signal p j The difference between The calculation formula for reconstructed data is: In formulas (4) and (5), RD(p i ) is the mean difference between the nearest 1 / 3 ultrasonic signals of the corresponding type and all other types of ultrasonic signals, F(p i ) is the ratio of the difference between each type of ultrasonic signal and other types of ultrasonic signals to the mean difference between the nearest 1 / 3 ultrasonic signals of the corresponding type of ultrasonic signal and other types of ultrasonic signals, N 1 / 3 (p i ) is the difference between the 1 / 3 ultrasonic signal closest to the corresponding type of ultrasonic signal and all other types of ultrasonic signals.

6. The automatic ultrasonic defect signal detection method based on instantaneous a-scan signal statistics according to claim 1 is characterized in that: The steps of obtaining the reconstructed signal statistics table [D]t2 at time t in step 4 include: Construct a statistical table of reconstructed signals at time t based on the reconstructed data [D]t2; The expression of the reconstructed signal statistics table [D]t2 at time t is: In formula (6), F(p1), F(p2) and F(p n ) is the reconstructed data.

Images