An ultrasonic testing method and system for internal defects of an insulating pull rod layered structure

Abstract

The application relates to the technical field of ultrasonic detection, and particularly provides an ultrasonic detection method and system for internal defects of an insulating pull rod layered structure, which comprises the following steps: obtaining echo amplitudes of a to-be-detected insulating pull rod under different incident angles, combining corresponding sampling depths, and constructing an actual detection matrix; based on the actual detection matrix and a reference template matrix, performing dynamic structure noise cancellation to obtain a residual matrix, wherein the reference template matrix is constructed according to echo amplitudes of an insulating pull rod sample without internal defects; comparing residual amplitudes of each sampling point in the residual matrix with a preset threshold value to obtain candidate defect points; based on residual amplitude changes and angle widths of the candidate defect points, a defect determination result is obtained through corresponding threshold value comparison. The method can effectively distinguish defect echoes from interlayer echoes in the ultrasonic detection of the insulating pull rod, thereby solving the problem of poor defect detection results in the related art.

Description

Technical Field

[0001] This invention relates to the field of ultrasonic testing technology, and in particular to an ultrasonic testing method and system for internal defects in a layered structure of an insulating tie rod. Background Technology

[0002] Insulating tie rods are core insulating components in high-voltage circuit breakers and gas-insulated switchgear. They are subjected to the combined effects of high voltage and mechanical tension for a long time. If there are defects such as delamination, air gaps, or cracks inside, they can easily cause insulation breakdown or mechanical breakage accidents, directly threatening the safe and stable operation of the power system. Therefore, it is crucial to accurately detect their internal defects.

[0003] In related technologies, ultrasonic pulse reflection is used to detect internal defects in insulating tie rods. However, because the insulating tie rod is a layered structure with multiple layers arranged regularly, each layer interface generates reflected echoes during ultrasonic testing, forming a dense sequence of interlayer echoes. When the defect size is small or its location overlaps with the layer interface, the defect echoes will be superimposed and coupled with the interlayer echoes in the time domain, making them difficult to distinguish effectively. In related technologies, the defect echoes from the vertical incidence method are easily submerged by interlayer noise; the single-angle oblique incidence method cannot obtain the optimal reflection angle of the defect, making it difficult to eliminate wide-angle structural interference; and the phased array sector scanning method in related technologies also does not utilize angular dimension information to separate defects from structural noise, resulting in limited detection effectiveness.

[0004] There is currently no effective solution to the problem that ultrasonic testing of insulating tie rods cannot effectively distinguish between defect echoes and interlayer echoes, resulting in poor defect detection results. Summary of the Invention

[0005] The present invention provides an ultrasonic detection method and system for internal defects in a layered structure of an insulating tie rod, which at least solves the problem in the related art that ultrasonic detection of insulating tie rods cannot effectively distinguish between defect echoes and interlayer echoes, resulting in poor defect detection results.

[0006] According to one aspect of the present invention, an ultrasonic detection method for internal defects in a layered structure of an insulating tie rod is provided, comprising: acquiring the echo amplitude of the insulating tie rod under different incident angles and constructing an actual detection matrix in combination with the corresponding sampling depth; performing dynamic structural noise cancellation based on the actual detection matrix and a reference template matrix to obtain a residual matrix, wherein the reference template matrix is ​​constructed based on the echo amplitude of an insulating tie rod sample without internal defects; comparing the residual amplitude of each sampling point in the residual matrix with a preset threshold to obtain candidate defect points; and obtaining a defect determination result based on the residual amplitude change and angle width of the candidate defect points through a corresponding threshold comparison.

[0007] As an optional approach, the echo amplitude of the insulating rod under test at different incident angles is obtained, and combined with the corresponding sampling depth, an actual detection matrix is ​​constructed. This includes: scanning the insulating rod under test along its circumference at different incident angles using longitudinal wave incidence, and collecting echo amplitude values ​​at different incident angles and sampling depths; mapping each incident angle to a row of a two-dimensional matrix, and mapping each sampling depth to a column of a two-dimensional matrix; the sampling depth of the sampling points is based on the initial sampling depth in the depth direction, and increases sequentially according to the depth sampling interval. Starting from the initial sampling depth, the depth increases by one sampling interval for each additional sampling point; the sampling interval is equal to the longitudinal wave velocity of the insulating rod material divided by twice the sampling frequency; and the echo amplitude is used as the matrix elements at the corresponding row and column positions to obtain the actual detection matrix.

[0008] As an optional approach, the expression for the actual detection matrix is ​​as follows: Where S represents the actual detection matrix; This represents the element in the i-th row and j-th column of matrix S; Indicates the angle of incidence Sampling depth The echo amplitude at the location; i=1,2,…M, where M is the total number of angles; j=1,2,…N, where N is the number of depth sampling points; ; where d min Δd represents the initial sampling depth in the depth direction; Δd = c / (2 × fs); where c represents the longitudinal wave velocity of the insulating tie rod material; and fs represents the sampling frequency.

[0009] As an optional approach, before performing dynamic structural noise cancellation based on the actual detection matrix and the reference template matrix to obtain the residual matrix, the method further includes: selecting an insulating rod from the same batch and specification as the insulating rod under test, and without internal defects, as the insulating rod sample; scanning the insulating rod sample along its circumference at different incident angles using longitudinal wave incidence, based on the same detection instrument parameters as the insulating rod under test, to obtain the template echo amplitude of the insulating rod sample; and constructing a reference template matrix based on the template echo amplitude, wherein the expression for constructing the reference template matrix is ​​as follows: Where T represents the reference template matrix; This represents the element in the i-th row and j-th column of matrix T; Indicates the angle of incidence Sampling depth The corresponding echo amplitude; i=1,2,…M, where M is the total number of angles; j=1,2,…N, where N is the number of depth sampling points.

[0010] As an optional approach, dynamic structural noise cancellation is performed based on the actual detection matrix and the reference template matrix to calculate the residual matrix. This includes: defining local windows of the same size centered on the sampling points to be determined at corresponding positions in the actual detection matrix and the reference template matrix; calculating adaptive weighting coefficients based on the elements of the actual detection matrix and the reference template matrix within the local windows; subtracting the product of the adaptive weighting coefficients and the corresponding elements in the reference template matrix from the elements in the actual detection matrix to obtain the residual amplitude at the corresponding position; and constructing the residual matrix based on the residual amplitudes corresponding to the incident angle and sampling depth.

[0011] As an optional approach, the residual matrix is ​​expressed as follows: Where R represents the residual matrix; This represents the element in the i-th row and j-th column of matrix R; Indicates the angle of incidence Sampling depth The corresponding residual amplitude; i=1,2,…M, where M is the total number of angles; j=1,2,…N, where N is the number of depth sampling points;

[0012] ;in, Indicates the adaptive weighting coefficients; ;in, Indicated by A local window centered on the user; For regularization parameters ( ); s pq tpq represents the echo amplitude value located at row p and column q in the actual detection matrix; tpq represents the echo amplitude value located at row p and column q in the reference template matrix.

[0013] As an optional approach, the preset threshold uses a fixed threshold. Before comparing the residual amplitude of each sampling point in the residual matrix with the fixed threshold to obtain candidate defect points, the method further includes: placing the phased array ultrasonic probe in the air and collecting noise signals based on the instrument parameters used in the actual detection; calculating the root mean square value of the noise signal based on the noise signal; and calculating the fixed threshold based on the root mean square value of the noise. The formula for calculating the fixed threshold is as follows: ;in, For a fixed threshold, represents the root mean square value of the noise; k represents the threshold coefficient, which ranges from 3 to 6.

[0014] As an optional approach, the preset threshold adopts an adaptive threshold. Before comparing the residual amplitude of each sampling point in the residual matrix with the preset threshold to obtain candidate defect points, the method further includes: dividing a local window centered on the residual point to be determined in the residual matrix; calculating the local residual mean and local residual standard deviation based on the values ​​of all residual points within the local window; and calculating an adaptive threshold based on the local residual mean and local residual standard deviation. The adaptive threshold calculation formula is as follows: ;in, This represents the mean of the local residuals. Indicates the local residual standard deviation. This represents the adjustment coefficient, with a value range of 2 to 4.

[0015] As an optional approach, the preset threshold uses a fixed threshold and an adaptive threshold. The residual amplitude of each sampling point in the residual matrix is ​​compared with the preset threshold to obtain candidate defect points, including: obtaining the fixed threshold and the adaptive threshold respectively; comparing the fixed threshold with the adaptive threshold to obtain the maximum threshold; comparing the residual amplitude of each sampling point in the residual matrix with the maximum threshold; and marking the sampling point in the residual matrix as a candidate defect point if the residual amplitude exceeds the maximum threshold.

[0016] As an optional approach, based on the residual amplitude change and angular width of the candidate defect points, a defect determination result is obtained through threshold comparison, including: for each candidate defect point at the sampling depth, extracting the residual amplitude corresponding to all incident angles at the sampling depth to obtain a distribution function of the residual amplitude changing with the incident angle; calculating the peak value ratio according to the distribution function; if the peak value ratio is greater than or equal to the peak value ratio threshold, determining that the residual amplitude exhibits a sudden change in a single-peak state; calculating the angular width of the defect echo based on the peak value and half-maximum of the residual amplitude of the distribution function; and determining that the sampling point has a defect if the residual amplitude exhibits a sudden change in a single-peak state and the angular width is less than a preset angular width threshold.

[0017] As an optional approach, the distribution function expression of the residual amplitude as a function of the incident angle is as follows: ;in, Indicates the same sampling depth Below, incident angle The corresponding residual amplitude; Represents the incident angle in the residual matrix Sampling depth The residual amplitude at the point; the formula for calculating the peak value ratio is as follows: = ;in, This represents the peak ratio of the candidate defect points; Indicates the peak point of the distribution function; This represents the incident angle corresponding to the peak point of the distribution function; Represents a very small positive number; the expression for the angle width is as follows: ;in, Indicates the width of the angle; This represents the incident angle closest to half the height of the residual amplitude peak to the right of the residual amplitude peak. This represents the incident angle closest to half the height of the residual amplitude peak to the left of the residual amplitude peak.

[0018] According to another aspect of the present invention, an ultrasonic detection system for internal defects in a layered structure of an insulating tie rod is also provided, for performing the ultrasonic detection method for internal defects in a layered structure of an insulating tie rod described above, comprising: a phased array ultrasonic probe for transmitting and receiving ultrasonic signals; a scanning mechanism connected to the phased array ultrasonic probe for driving the phased array ultrasonic probe to move along the axial or circumferential direction of the insulating tie rod and recording the spatial position in real time; a data acquisition unit connected to the phased array ultrasonic probe for exciting ultrasonic signals and receiving echo amplitude values; and a signal processing unit connected to the data acquisition unit and the scanning mechanism respectively, the signal processing unit including a matrix storage module, a cancellation operation module, and a defect identification module; the matrix... The matrix storage module receives the echo amplitude and corresponding incident angle and sampling depth information transmitted by the data acquisition unit, constructs and stores the actual detection matrix; constructs and stores the reference template matrix based on the echo amplitude of the insulating tie rod sample without internal defects; the cancellation operation module, connected to the matrix storage module, performs dynamic structural noise cancellation based on the actual detection matrix and the reference template matrix to obtain the residual matrix; the defect identification module, connected to the cancellation operation module, compares the residual amplitude of each sampling point in the residual matrix with a preset threshold to obtain candidate defect points; and obtains the defect judgment result based on the residual amplitude change and angle width of the candidate defect points through corresponding threshold comparison.

[0019] According to another aspect of the present invention, an electronic device is also provided, comprising: a processor, and a memory storing a program, the program including instructions that, when executed by the processor, cause the processor to perform the method described in any of the preceding claims.

[0020] This invention provides an ultrasonic testing method and system for internal defects in a layered structure of an insulating tie rod. It constructs a reference template matrix representing the inherent echo characteristics of the structure using a defect-free sample. Then, it dynamically and adaptively weights and cancels out the actual detection matrix and the template matrix, accurately separating the wide-angle, smoothly distributed interlayer echo components that match the template, retaining only the residual matrix containing defect information. Finally, by utilizing the difference between the narrow-angle, single-peak abrupt change of the defect signal and the wide-angle, smoothly distributed structure noise residual in the residual matrix, and through dual determination using amplitude thresholds and angle width thresholds, it achieves the separation of defects from interlayer echoes. This solves the problem in related technologies where ultrasonic testing of insulating tie rods cannot effectively distinguish between defect echoes and interlayer echoes, resulting in poor defect detection results. Attached Figure Description

[0021] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention, and those skilled in the art can obtain other embodiments based on these drawings without creative effort.

[0022] Figure 1 This is a flowchart of an ultrasonic testing method for internal defects in an insulating tie rod layered structure, according to an embodiment of the present invention.

[0023] Figure 2 This is a structural diagram of an ultrasonic testing system for internal defects in a layered insulating tie rod structure, according to an embodiment of the present invention.

[0024] Figure 3 This is a schematic diagram of the structure of the electronic device created by this invention. Detailed Implementation

[0025] Embodiments of the present invention will now be described in more detail with reference to the accompanying drawings. While some embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present invention. It should be understood that the drawings and embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0026] The ultrasonic pulse reflection method involves emitting high-frequency ultrasonic pulses through a probe and receiving the reflected echoes from the interfaces. The internal state is determined based on the depth corresponding to the arrival time of the echoes and the reflection intensity corresponding to the amplitude. Layered structure echoes refer to the regular reflected echoes generated at the interfaces of each layer in a multi-layered wound or laminated structure of an insulating tie rod when ultrasonic waves are incident. These echoes are densely and periodically distributed and represent inherent structural noise of the workpiece.

[0027] Defect echoes refer to the signals reflected from the air-solid interface formed by internal defects in the insulating rod, such as delamination, air gaps, and cracks. These signals only appear at the optimal reflection angle of the defect, exhibiting a narrow angle distribution and significant amplitude abrupt changes. Temporal superposition coupling refers to the overlap or proximity of interlayer echoes and defect echoes on the time axis (i.e., the depth axis). The signals superimpose each other and cannot be distinguished by a single time or depth dimension, resulting in the masking of defect signals.

[0028] In related technologies, the perpendicular application of ultrasonic energy to the interlayer interface results in the strongest and densest interlayer echo reflection. Defect echoes, lacking angular advantage and with dispersed energy, are completely submerged, leading to an extremely low signal-to-noise ratio. Related technologies using a single-angle oblique incidence method can only weaken some interlayer echoes, failing to eliminate structural interference over a wide angular range. Furthermore, it is difficult to match the unique optimal reflection angle of the defect, resulting in weak and unstable defect echoes and low identification accuracy.

[0029] In related technologies, phased array sector scanning only uses angle as the imaging dimension. It fails to take advantage of the essential difference between the wide-angle stable distribution of interlayer echoes and the narrow-angle abrupt distribution of defect echoes. It relies solely on human experience to identify signals from the time domain and amplitude level, and cannot fundamentally separate defect echoes from interlayer echoes, resulting in high rates of missed detections, false detections, and poor detection results.

[0030] To address the problem in related technologies where ultrasonic testing of insulating tie rods cannot effectively distinguish between defect echoes and interlayer echoes, resulting in poor defect detection results, such as... Figure 1 As shown, an embodiment of the present invention provides an ultrasonic detection method for internal defects in a layered structure of an insulating tie rod, comprising:

[0031] Step S101: Obtain the echo amplitude of the insulating tie rod under test at different incident angles, and construct the actual detection matrix by combining the corresponding sampling depth;

[0032] Step S102: Based on the actual detection matrix and the reference template matrix, perform dynamic structural noise cancellation to obtain the residual matrix. The reference template matrix is ​​constructed based on the echo amplitude of the insulating tie rod sample without internal defects.

[0033] Step S103: Compare the residual amplitude of each sampling point in the residual matrix with a preset threshold to obtain candidate defect points;

[0034] Step S104: Based on the residual amplitude change and angle width of the candidate defect points, the defect determination result is obtained by comparing the corresponding thresholds.

[0035] The actual detection matrix is ​​a two-dimensional matrix constructed with the ultrasonic incident angle as the row and the sampling depth as the column, using the echo amplitude at the corresponding position as the element. It fully characterizes the ultrasonic echo features of the test piece at all angles and depths. In this embodiment, a phased array ultrasonic probe is used with longitudinal wave incident mode, and the probe performs sector scanning with a step size of 1° within the range of 0° to 70°, simultaneously acquiring the echo amplitude at each angle and depth and constructing the matrix.

[0036] Constructing an angle-depth two-dimensional matrix can fully preserve the dimensional features of both signals, providing the necessary data foundation for subsequent signal separation. This enables two-dimensional structured storage of ultrasonic echo information, thus solving the fundamental problem in related technologies where single-dimensional detection cannot distinguish between time-domain superimposed signals.

[0037] The reference template matrix is ​​a two-dimensional matrix constructed by inspecting defect-free samples of the same batch and specifications using the same inspection parameters as the test piece. It is used to characterize the inherent echo characteristics in a defect-free state. In this embodiment, a reference template matrix is ​​constructed in advance using defect-free samples of the same batch. Then, the actual inspection matrix is ​​canceled out with the reference template matrix at each angle and depth point. The residual matrix obtained after cancellation retains only the defect signal and random noise, thereby significantly improving the signal-to-noise ratio of the defect signal.

[0038] The preset threshold can be a fixed threshold based on the 3σ~6σ criterion, or an adaptive threshold based on local statistics, i.e., a calculated value of 2~4 times the local standard deviation of the local mean. Alternatively, it can be set according to the maximum value of the fixed threshold and the adaptive threshold. Marking sampling points with residual amplitudes exceeding the threshold as candidate defect points can quickly eliminate most noise points and significantly narrow the subsequent judgment range.

[0039] In addition to defect signals, the residual matrix also contains random noise and a small amount of structural noise that is not completely canceled out. Preliminary screening using amplitude thresholds can quickly eliminate most noise points, significantly narrowing the scope of subsequent defect determination. This effectively filters background noise without overlooking potential defects, reducing the amount of data processed subsequently and significantly improving detection efficiency.

[0040] Angular width refers to the full width at half maximum (FWHM) of the defect echo peak at the same depth, used to characterize the angular distribution range of the echo. For each candidate defect point at the sampling depth, the residual amplitude corresponding to all incident angles at that depth is extracted, a distribution curve of the residual amplitude as a function of angle is constructed, the presence of a single-peak sharp increase in the curve is analyzed, and the peak-to-peak ratio and FWHM are calculated.

[0041] If the peak ratio in the distribution curve is greater than or equal to the peak ratio threshold and the angle width is less than the angle threshold, a defect alarm is triggered if a defect is found at the candidate defect point location; if the angle width of the distribution curve is greater than the angle threshold, the candidate defect point location is determined to be a local non-uniformity at the interlayer interface, and no defect alarm is triggered.

[0042] Comparing the residual amplitude with a preset threshold can only distinguish between signal and noise, but cannot differentiate between actual defects and local inhomogeneities at interlayer interfaces. Defect echoes only appear near the optimal reflection angle, exhibiting a narrow-angle, single-peak abrupt directional characteristic, while interlayer inhomogeneities show a wide-angle, stable distribution. The residual matrix represents the portion of the actual echo unrelated to the inherent structure of the defect-free insulated tie rod, retaining only the defect signal and random noise. The presence of a defect can be determined by observing the narrow-angle, single-peak abrupt change in the residual amplitude within the residual matrix.

[0043] This embodiment extracts the residual amplitudes from multiple angles at the same depth and constructs a distribution function of the residual amplitudes as a function of the incident angle. It then quantifies the peak value ratio and angular width of this distribution function and makes a joint judgment based on preset peak value ratio and angular thresholds. This achieves the separation of defective and non-defective abnormal signals, solving the problem of high false detection rates in related technologies and significantly improving the accuracy of defect identification.

[0044] This invention provides an ultrasonic testing method and system for internal defects in a layered structure of an insulating tie rod. It constructs a reference template matrix representing the inherent echo characteristics of the structure using a defect-free sample. Then, it dynamically and adaptively weights and cancels out the actual detection matrix and the template matrix, accurately separating the wide-angle, smoothly distributed interlayer echo components that match the template, retaining only the residual matrix containing defect information. Finally, by utilizing the difference between the narrow-angle, single-peak abrupt change of the defect signal and the wide-angle, smoothly distributed structure noise residual in the residual matrix, and through dual determination using amplitude thresholds and angle width thresholds, it achieves the separation of defects from interlayer echoes. This solves the problem in related technologies where ultrasonic testing of insulating tie rods cannot effectively distinguish between defect echoes and interlayer echoes, resulting in poor defect detection results.

[0045] As an optional approach, the echo amplitude of the insulating rod under test at different incident angles is obtained, and combined with the corresponding sampling depth, an actual detection matrix is ​​constructed. This includes: scanning the insulating rod under test along its circumference at different incident angles using longitudinal wave incidence, and collecting the echo amplitude at different incident angles and sampling depths; mapping each incident angle to rows of a two-dimensional matrix, and mapping each sampling depth to columns of a two-dimensional matrix; the sampling depth of the sampling points is based on the initial sampling depth in the depth direction, and increases sequentially according to the depth sampling interval. Starting from the initial sampling depth, the depth increases by one sampling interval for each additional sampling point; the sampling interval is equal to the longitudinal wave velocity of the insulating rod material divided by twice the sampling frequency; and the echo amplitude is used as the matrix elements at the corresponding row and column positions to obtain the actual detection matrix.

[0046] Circumferential scanning refers to the phased array ultrasonic probe moving along the circumference of the insulated tie rod to scan a sector, covering the entire circumferential cross-section and avoiding missed detection of circumferential defects. The depth sampling interval is the physical distance between adjacent sampling points in the depth direction, which determines the detection depth resolution.

[0047] The sampling frequency is the frequency at which ultrasonic signals are digitally acquired, directly determining the depth sampling interval and the sampling depth range. Sampling depth is the digital depth calculated by converting the time it takes for the echo to reach the probe based on the round-trip velocity of the ultrasonic wave. As ultrasonic waves propagate in a round-trip manner, a higher sampling frequency results in a shorter time interval between two samples, corresponding to a smaller depth interval within the workpiece and a denser number of sampling points, leading to a shallower total depth coverage for the same number of sampling points. Conversely, a lower sampling frequency results in a longer time interval, a larger depth interval, and a sparser number of sampling points, leading to a deeper total depth coverage for the same number of sampling points.

[0048] The center frequency of a phased array ultrasonic probe is the dominant frequency of the ultrasonic waves emitted by the probe, which determines the penetration depth and axial resolution limit of the ultrasonic waves. The center frequency can be selected according to the wall thickness of the insulating rod and the grain size of the material. A lower frequency is selected when the wall thickness is large.

[0049] This embodiment uses a linear or ring array phased array probe with a center frequency of 1MHz to 5MHz, equipped with a longitudinal wave wedge. It performs fixed step-size sector scanning within a preset angle range, while simultaneously moving the probe at a uniform speed along the circumference of the insulating rod, and synchronously recording the echo amplitude values ​​after envelope detection or full-wave rectification at each angle and depth.

[0050] Longitudinal wave incident refers to ultrasonic vibration that is in the same direction as the propagation. It has low attenuation and deep penetration in epoxy glass fiber composite materials. Combined with the full circumferential coverage of circumferential scanning and the capture of angle information by multi-angle sector scanning, complete echo data of the test piece in the full circumference, all angles and all depths can be obtained, providing the original support for the construction of a structured detection matrix.

[0051] In this embodiment, M incident angles are mapped as matrix rows and N depth sampling points are mapped as matrix columns, converting discrete echo data into a standardized two-dimensional matrix, realizing structured storage of echo information, and providing a unified data format for subsequent point-by-point cancellation with the reference template matrix.

[0052] The initial sampling depth corresponds to the inspection surface position of the workpiece. However, to eliminate near-field blind zone of the probe and surface coupling interference, it can be set to a fixed offset slightly greater than 0, such as 0.5 mm. This offset can be pre-determined using a standard test block based on the actual blind zone size of the probe. Under this setting, the first sampling point in the depth direction starts from the initial sampling depth, and subsequent sampling points increase sequentially according to the sampling interval. This setting ensures that the inspection data is not affected by surface interference, and the defect depth positioning error is controlled within one sampling interval.

[0053] By mapping the echo amplitude corresponding to each angle and depth to the matrix position, the three-dimensional correlation information of angle-depth-amplitude can be completely preserved, ensuring consistency with the dimension of the reference template matrix and guaranteeing the accuracy of subsequent point-by-point cancellation calculations.

[0054] This embodiment solves the problems of unstructured, single-dimensional, and low-precision positioning of traditional ultrasonic testing data by standardizing the actual detection matrix construction process. It ensures the compatibility of actual detection data with benchmark template data, lays a solid data foundation for subsequent dynamic structural noise cancellation and accurate defect identification, and improves the repeatability and accuracy of the detection method.

[0055] As an optional approach, the actual detection matrix is ​​expressed as follows: Where S represents the actual detection matrix; This represents the element in the i-th row and j-th column of matrix S; Indicates the angle of incidence Sampling depth The echo amplitude at the location; i=1,2,…M, where M is the total number of angles; j=1,2,…N, where N is the number of depth sampling points; ; where d min Δd represents the initial sampling depth in the depth direction; Δd represents the sampling interval in the depth direction; Δd = c / (2×fs); where c represents the longitudinal wave velocity of the insulating tie rod material; and fs represents the sampling frequency.

[0056] The total number of angles M is the total number of different incident angles collected during the phased array fan scanning process. It is determined by the preset detection angle range and the angle step size. The smaller the angle step size, the larger the value of M and the higher the angle resolution.

[0057] The number of depth sampling points N is the total number of echo amplitudes collected along the depth direction of the insulating tie rod. It is determined by the total detection depth and the depth sampling interval. The smaller the sampling interval, the larger the value of N, and the higher the depth resolution.

[0058] Initial sampling depth d min It is the physical depth corresponding to the first sampling point in the depth direction, and is usually set to the thickness of the coupling layer on the surface of the insulating tie rod to eliminate the interference of the surface coupling signal on the internal detection.

[0059] The longitudinal wave velocity *c* is the speed at which ultrasound propagates in the form of a longitudinal wave in epoxy glass fiber composite material. It is an inherent acoustic parameter of the material and must be measured in advance using a standard test block. The longitudinal wave velocity *c* must be measured using a standard test block made of the same material as the insulating tie rod under test.

[0060] The actual detection matrix is ​​mathematically defined using set expressions, clearly defining the one-to-one correspondence between each element in the matrix and the incident angle and sampling depth. This embodiment uses standardized mathematical language to describe the detection data structure, eliminating ambiguity in textual descriptions and ensuring that the actual detection matrices generated by different inspectors and different equipment have a unified format specification.

[0061] angle of incidence according to calculate, Let Δθ be the minimum incident angle and Δθ be the angle step size. By explicitly defining the matrix dimensions and element indexing rules, we ensure that each matrix element can be uniquely mapped to a detection point in physical space.

[0062] This embodiment solves the problems of unstructured, single-dimensional, and low-precision positioning of ultrasonic testing data in related technologies by standardizing the actual testing matrix construction process. It ensures the compatibility of actual testing data with benchmark template data, lays a solid data foundation for subsequent dynamic structural noise cancellation and accurate defect identification, and improves the repeatability and accuracy of the testing method.

[0063] As an optional approach, before obtaining the residual matrix by performing dynamic structural noise cancellation based on the actual detection matrix and the reference template matrix, the following steps are also included: selecting an insulating rod from the same batch and specification as the insulating rod under test, and without internal defects, as the insulating rod sample; based on the same detection instrument parameters as the insulating rod under test, scanning the insulating rod sample along its circumference at different incident angles using longitudinal wave incidence to obtain the template echo amplitude of the insulating rod sample; and constructing a reference template matrix based on the template echo amplitude. The expression for constructing the reference template matrix is ​​as follows: Where T represents the reference template matrix; This represents the element in the i-th row and j-th column of matrix T; Indicates the angle of incidence Sampling depth The corresponding echo amplitude; i=1,2,…M, where M is the total number of angles; j=1,2,…N, where N is the number of depth sampling points.

[0064] Insulating rods of the same batch and specifications refer to insulating rods manufactured in the same production batch as the insulating rod under test, using the same raw materials, the same winding or lamination process, and the same geometric dimensions, i.e., wall thickness, diameter, and length. Their material acoustic properties and layered structure characteristics are highly consistent.

[0065] Insulating tie rod samples without internal defects were confirmed through destructive sectioning tests, X-ray inspections, and other methods to be free of any internal defects such as delamination, air gaps, or cracks that would affect insulation and mechanical performance. These insulating tie rods were then used to construct structural echo references.

[0066] The consistency of instrument parameters means that all instrument settings, such as the model of the phased array probe, center frequency, number of array elements, type of longitudinal wave wedge, detection angle range, angle step size, focusing depth, and sampling frequency, are exactly the same in the template construction and the actual testing process.

[0067] Defect-free samples of the same batch and specifications exhibit inherent echo characteristics of layered structures that are completely consistent with the insulation tie rod under test, thus accurately reflecting the structural noise characteristics of that batch of products. Prioritize randomly selecting 3-5 samples from the same production batch as the insulation tie rod under test. Initially screen these samples using conventional ultrasonic testing. Then, conduct destructive sectioning tests on the qualified samples, checking each layer to confirm the absence of any internal defects. Finally, determine 1-2 samples as template construction samples.

[0068] Consistent detection parameters eliminate systematic errors and ensure comparability between the template and actual echoes. Strict adherence to the instrument settings used in actual testing is maintained, with circumferential fan-scanning of the sample to collect echo amplitudes at various angles and depths. This yields template data consistent with the characteristics of actual detection data, ensuring effective cancellation operations.

[0069] The reference template matrix must be completely consistent with the actual detection matrix in terms of dimensions, index, and element definition to achieve point-by-point cancellation. Using the same construction rules as the actual detection matrix, the incident angle is mapped to rows, the sampling depth is mapped to columns, and the template echo amplitude is used as matrix elements to construct an M×N dimension reference template matrix T. M and N are consistent with the actual detection matrix, fully representing the inherent interlayer echo characteristics of this batch.

[0070] This embodiment ensures template representativeness and data compatibility by strictly limiting sample selection, parameter consistency, and matrix construction rules, providing a fundamental guarantee for dynamic structural noise cancellation, which can significantly improve the efficiency of interlayer echo stripping and significantly improve the signal-to-noise ratio of defect signals.

[0071] As an optional approach, dynamic structural noise cancellation is performed based on the actual detection matrix and the reference template matrix to calculate the residual matrix. This includes: defining local windows of the same size centered on the sampling points to be determined at corresponding positions in the actual detection matrix and the reference template matrix; calculating adaptive weighting coefficients based on the elements of the actual detection matrix and the reference template matrix within the local windows; subtracting the product of the adaptive weighting coefficients and the corresponding elements in the reference template matrix from the elements in the actual detection matrix to obtain the residual amplitude at the corresponding position; and constructing the residual matrix based on the residual amplitudes at the corresponding incident angle and sampling depth.

[0072] A local window refers to a rectangular area formed by extending outwards in the angular and depth directions from the sampling point to be processed in the matrix. It can use a 5×5 or 7×7 matrix of pixels, corresponding to ±0.5mm in the depth direction and ±3° in the angular direction, and is used for statistical calculation of adaptive weighting coefficients.

[0073] The size of the local window should be determined based on the layer thickness of the insulating tie rod and the minimum size of the expected defect. Generally, the width of the window in the angular direction should cover 3-5 angular steps, and the length in the depth direction should cover 1-2 interlayer thicknesses to ensure that the window contains sufficient statistical information without crossing too many structural interfaces.

[0074] The adaptive weighting coefficient is a weight value dynamically calculated for each sampling point, used to compensate for the echo amplitude deviation caused by coupling state, coupling agent thickness, and local attenuation fluctuations in the material.

[0075] The local region signal energy minimization criterion aims to minimize the sum of squared differences between the actual signal and the weighted template signal within the local window, solve for the optimal weighting coefficients, and achieve maximum stripping of structural noise.

[0076] Differences in detection conditions exhibit local distribution characteristics. Employing local windows can accurately adapt to local amplitude fluctuations, avoiding incomplete cancellation or excessive suppression of defect signals caused by global weighting. In this embodiment, local windows of consistent size are defined for sampling points at the same location in both matrices. The window size can be adjusted according to the wall thickness and layer thickness of the insulating tie rod, and the window does not exceed the matrix boundary, providing a reasonable statistical range for the calculation of weighting coefficients and effectively improving the accuracy of structural noise cancellation.

[0077] In related technologies, the use of fixed weighting coefficients cannot adapt to the local detection differences at different locations. This embodiment uses the local region signal energy minimization criterion to solve for the optimal adaptive weighting coefficients, and introduces a regularization parameter to avoid the denominator being zero during calculation. The adaptive weighting coefficients can dynamically adapt to the differences in local signal correlation at various locations, fully stripping interlayer echoes while completely preserving the defect signal, effectively solving the problem of poor adaptability of fixed cancellation, and improving the efficiency of interlayer echo cancellation and the signal-to-noise ratio of defect signals.

[0078] The inherent echo between layers has a strong linear correlation with the template signal, while the defect signal only exists in the actual detection and is not related to the template signal. By calculating the residual amplitude through point-by-point weighted subtraction, the interlayer structural noise can be accurately removed, highlighting the defect signal that is covered by the background.

[0079] The residual amplitudes are arranged according to the same angle and depth indexing rules as the actual detection matrix and the benchmark template matrix to construct a standardized residual matrix. This fully preserves the distribution characteristics of the residual signal in the angle and depth dimensions, providing a standardized data foundation for subsequent amplitude threshold screening and angle-directive defect judgment.

[0080] This embodiment achieves precise cancellation of dynamic structural noise through a complete process including local window delineation, adaptive weighted coefficient solution, point-by-point residual calculation, and residual matrix construction. It also overcomes the limitation of fixed cancellation in adapting to fluctuations in detection conditions by dynamically optimizing weights based on the local energy minimization criterion, balancing efficient removal of interlayer echoes with complete preservation of defect signals. The constructed residual matrix can improve the signal-to-noise ratio by more than 15dB, fundamentally solving the problem of difficulty in distinguishing between defect echoes and interlayer echoes in the temporal domain, and providing reliable data support for subsequent accurate defect identification and spatial positioning.

[0081] As an alternative approach, the residual matrix can be expressed as follows: Where R represents the residual matrix; This represents the element in the i-th row and j-th column of matrix R; Indicates the angle of incidence Sampling depth The corresponding residual amplitude; i=1,2,…M, where M is the total number of angles; j=1,2,…N, where N is the number of depth sampling points; ;in, Indicates the adaptive weighting coefficients; ;in, Indicated by A local window centered on the user; For regularization parameters ( ); This represents the echo amplitude value located at row p and column q in the actual detection matrix; This represents the echo amplitude value located at row p and column q in the reference template matrix.

[0082] The regularization parameter ε is used to prevent calculation errors where the denominator is zero when the amplitude of the template signal within a local window is zero. It also serves to stabilize the values ​​and regularize the calculation, preventing abnormal fluctuations in the weighting coefficients. The residual amplitude reflects the portion of the actual detection signal that is unrelated to the echo from the defect-free structure, mainly including defect signals and random noise.

[0083] The set expression of the residual matrix is ​​standardized mathematically, ensuring that its dimensions are consistent with those of the actual detection matrix and the baseline template matrix, both being M rows and N columns. Defining the residual matrix using a unified mathematical language eliminates ambiguity, ensures consistent data format, avoids calculation errors, and establishes a one-to-one correspondence between the residual matrix and physical detection points, providing precise coordinates for defect spatial localization.

[0084] The residual amplitude is obtained by subtracting the weighted template echo amplitude from the actual detected echo amplitude. This can accurately remove the inherent interlayer echo that is strongly correlated with the template signal, and completely preserve the defect signal that only exists in the actual detection, thereby improving the signal-to-noise ratio of the defect signal by more than 15dB.

[0085] The adaptive weighting coefficients are dynamically determined using the local window signal energy minimization criterion, and the calculation formula is as follows:

[0086] ;

[0087] The adaptive weighting coefficients are calculated independently for each sampling point, compensating for local detection differences such as coupling conditions and material attenuation. This avoids incomplete cancellation and excessive suppression of defect signals caused by fixed or global coefficients, achieving personalized amplitude compensation while balancing interlayer echo stripping and complete defect signal preservation. The adaptive weighting coefficient formula is derived from the local region signal energy minimization criterion, allowing for the solution of optimal weights to maximize interlayer echo cancellation efficiency within the local window and avoid false defect signal cancellation.

[0088] This embodiment solves the problem that fixed cancellation cannot adapt to local detection fluctuations by strictly limiting the calculation methods of the residual matrix structure, residual amplitude, and weighting coefficients through precise mathematical expressions. It achieves precise stripping of interlayer echoes based on the local energy minimization criterion, significantly improving cancellation efficiency while ensuring the integrity of the defect signal. Standardized mathematical definitions enhance the repeatability and equipment compatibility of the detection method, laying a solid mathematical foundation for subsequent accurate defect identification and location.

[0089] As an optional approach, a fixed threshold is used as the preset threshold. Before comparing the residual amplitude of each sampling point in the residual matrix with the fixed threshold to obtain candidate defect points, the method further includes: placing the phased array ultrasonic probe in the air and collecting noise signals based on the instrument parameters used in the actual detection; calculating the root mean square value of the noise signal based on the noise signal; and calculating the fixed threshold based on the root mean square value of the noise. The formula for calculating the fixed threshold is as follows: ;in, For a fixed threshold, represents the root mean square value of the noise; k represents the threshold coefficient, which ranges from 3 to 6.

[0090] The fixed threshold is a globally uniform threshold set based on the inherent background noise level of the detection system. It is used to initially screen out abnormal signals with amplitudes significantly higher than the noise from the residual matrix. The root mean square (RMS) noise value is a statistical measure characterizing the intensity of the inherent background noise of the detection system. It is obtained by calculating the root mean square of the amplitudes of all sampling points of the noise signal and can objectively and stably reflect the noise level of the system. The 3σ~6σ criterion is a classic criterion used in statistics to identify outliers. In this embodiment, it is used to distinguish between random noise and valid defect signals.

[0091] The inherent noise level of the detection system is directly related to the instrument parameters. Only by acquiring airborne noise signals with the same instrument parameters as the actual detection, i.e., without any reflected signals and containing only system and environmental noise, can the background noise of the actual detection be truly reflected. In this embodiment, all instrument settings of the probe are kept consistent with the actual detection, and the probe is suspended in the air to collect continuous noise data for 1 to 5 seconds, providing reliable raw data for fixed threshold calculation.

[0092] The root mean square (RMS) noise value is less susceptible to accidental spikes and can objectively reflect the long-term noise level of the system. After performing the same preprocessing as the actual detection on the acquired noise signal, the value is calculated using the formula. This transforms background noise into quantitative indicators, providing an objective basis for threshold setting.

[0093] The 3σ~6σ criterion has a clear statistical confidence level and can balance the false negative rate and the false positive rate. k=3 can filter 99.73% of random noise, and k=6 can filter nearly 100% of random noise. The k value can be selected according to the detection requirements. For high-voltage insulating rods, 3~4 can be selected to improve sensitivity, and 5~6 can be used for batch detection to reduce the false positive rate. Finally, a repeatable global fixed threshold can be obtained, which can filter more than 99% of random noise, quickly screen candidate defect points, and narrow down the subsequent judgment range.

[0094] This embodiment solves the problems of strong subjectivity and poor repeatability of traditional manual threshold setting by standardizing the fixed threshold setting process; the setting method based on the 3~6σ criterion has a solid statistical basis, takes into account both detection sensitivity and false detection rate control, ensures the operability of the detection method and equipment compatibility, and provides reliable preliminary screening results for subsequent accurate defect judgment.

[0095] As an optional approach, an adaptive threshold is used as the preset threshold. Before comparing the residual magnitude of each sampling point in the residual matrix with the preset threshold to obtain candidate defect points, the method further includes: dividing a local window centered on the residual point to be determined in the residual matrix; calculating the local residual mean and local residual standard deviation based on the values ​​of all residual points within the local window; and calculating the adaptive threshold based on the local residual mean and local residual standard deviation. The adaptive threshold calculation formula is as follows: ;in, This represents the mean of the local residuals. Indicates the local residual standard deviation. This represents the adjustment coefficient, with a value range of 2 to 4.

[0096] The adaptive threshold is a locally calculated threshold for each sampling point, which can be automatically adjusted according to the surrounding noise level to adapt to scenarios with uneven noise distribution. The mean and standard deviation of the local residuals reflect the average noise level and fluctuation degree of the local area, respectively. The adjustment coefficient is used to control the threshold sensitivity; the larger the value, the higher the threshold and the lower the false detection rate; the smaller the value, the lower the threshold and the higher the detection sensitivity.

[0097] Uneven noise distribution in the residual matrix makes it unsuitable for a globally fixed threshold, while a local window can capture regional noise differences. The window size should be consistent with the dynamic structural noise cancellation. In this embodiment, a 5×5 or 7×7 matrix pixel size is preferred, which does not exceed the matrix boundary. An independent statistical region is established for each sampling point to solve the problems of missed detections and false detections with a fixed threshold.

[0098] Combining the mean and standard deviation of local residuals can comprehensively characterize local noise features, avoiding the limitations of a single indicator. Calculating the mean and standard deviation of local residuals using formulas yields key statistics that objectively reflect the level of local noise, laying the foundation for accurate calculation of adaptive thresholds.

[0099] The threshold formula, combining local mean and standard deviation, can be dynamically adjusted according to noise levels: a higher threshold for higher noise levels suppresses false detections, while a lower threshold for lower noise levels avoids missed detections. The β value is selected as needed: 2-3 for in-service testing to improve sensitivity, and 3-4 for batch testing to reduce the false detection rate and improve the accuracy of candidate defect screening.

[0100] This embodiment solves the problem that a globally fixed threshold cannot adapt to uneven noise through a complete process; the calculation method based on local statistics has a solid statistical foundation, and the β value provides flexible adjustment space, significantly improving the accuracy and robustness of candidate defect point screening, and providing more reliable preliminary screening results for subsequent accurate defect determination.

[0101] As an optional approach, the preset threshold uses a fixed threshold and an adaptive threshold. The residual amplitude of each sampling point in the residual matrix is ​​compared with the preset threshold to obtain candidate defect points. This includes: obtaining the fixed threshold and the adaptive threshold respectively; comparing the fixed threshold with the adaptive threshold to obtain the maximum threshold; comparing the residual amplitude of each sampling point in the residual matrix with the maximum threshold; and marking the sampling point in the residual matrix as a candidate defect point if the residual amplitude exceeds the maximum threshold.

[0102] The maximum threshold is the larger value obtained by comparing the fixed threshold and the adaptive threshold point by point. It combines the global stability of the fixed threshold and the local adaptability of the adaptive threshold, and can adapt to both global and local noise distributions.

[0103] Single thresholds have limitations; fixed thresholds cannot adapt to uneven noise levels, leading to missed detections and false detections. Adaptive thresholds only consider local noise and are easily affected by global noise, resulting in false detections. Fixed thresholds are calculated based on the inherent noise of the system, while adaptive thresholds are calculated based on local noise statistics. Obtaining both provides the basis for threshold fusion.

[0104] By taking the maximum value of both the fixed threshold and the adaptive threshold, we can give full play to their respective advantages. When the local noise is high, we use the adaptive threshold to suppress false detections, and when the global noise is high, we use the fixed threshold to suppress false detections. By taking the maximum value point by point, we can obtain the optimal threshold, thereby solving the problem of adapting a single threshold to complex noise.

[0105] By using a unified maximum threshold for point-by-point judgment, it is ensured that the judgment criteria for all sampling points conform to both the global noise level and the local noise characteristics, avoiding the one-sidedness of a single threshold judgment. All sampling points in the residual matrix are traversed, and the residual amplitude of each sampling point is compared one by one with the maximum threshold corresponding to that point. This accurately distinguishes background noise and anomalous signals in the residual matrix, filtering out the vast majority of random noise and incompletely canceled structural noise.

[0106] A residual amplitude exceeding the maximum threshold indicates that the signal strength at that point is significantly higher than the background noise level, highly likely to be a defect echo signal, requiring further angular directivity feature analysis for final determination. All sampling points with residual amplitudes greater than the corresponding maximum threshold are marked with coordinates, and their incident angle and sampling depth information are recorded to form a candidate defect point set. This reduces the amount of data to be processed by more than 95% without overlooking potential defects, significantly improving the efficiency of defect identification.

[0107] This embodiment solves the problem that a single threshold cannot adapt to complex noise by using a threshold fusion strategy. It combines the advantages of two thresholds, enabling the candidate defect point screening accuracy to reach more than 90%, providing high-quality preliminary screening results for subsequent accurate defect judgment, and improving the reliability and robustness of the detection method.

[0108] As an optional approach, based on the residual amplitude variation and angular width of candidate defect points, a defect determination result is obtained through threshold comparison. This includes: for each candidate defect point at the sampling depth, extracting the residual amplitude corresponding to all incident angles at the sampling depth to obtain the distribution function of the residual amplitude as a function of the incident angle; calculating the peak value ratio based on the distribution function; determining that the residual amplitude exhibits a sudden change in a single-peak state when the peak value ratio is greater than or equal to the peak value ratio threshold; calculating the angular width of the defect echo based on the peak value and half-maximum of the residual amplitude of the distribution function; and determining that the sampling point has a defect when the residual amplitude exhibits a sudden change in a single-peak state and the angular width is less than the preset angular width threshold.

[0109] The distribution pattern of residual amplitude with incident angle at the same sampling depth shows that defect echoes exhibit a narrow-angle single-peak abrupt change, while interlayer inhomogeneities show a wide-angle flat or periodic distribution. The peak-to-peak ratio (PTR) is the ratio of the peak value of the residual amplitude distribution curve to the minimum background value, used to quantify the significance of the single peak and distinguish defect signals from noise.

[0110] The peak-to-peak ratio threshold is a critical value used to determine whether there is a significant single-peak abrupt change in the residual amplitude. Its value ranges from 2 to 5; a larger value indicates a stricter judgment standard and a lower false detection rate. The angle width threshold is a critical value used to determine the range of defect echo angle distribution. It can be set to 10°. If the angle width is less than this value, it is considered a defect; if it is greater than 10°, it is considered localized inhomogeneity at the interlayer interface.

[0111] Amplitude thresholds can only distinguish between signals and noise, but cannot differentiate between real defects and local inhomogeneities at interlayer interfaces; moreover, the distribution characteristics of defect echoes and interlayer echoes differ fundamentally in the angular dimension. By iterating through all candidate defect points and for each candidate defect point at a sampling depth dj, all elements in the j-th column of the residual matrix are extracted to form a one-dimensional array of residual amplitudes varying with the incident angle at that depth, i.e., the distribution function.

[0112] Defect echoes only show strong reflections near the optimal reflection angle, forming a distinct single peak; while the residual amplitudes of local non-uniformity at the interlayer interface are uniformly distributed over a wide angle range, without obvious peaks; the peak-to-peak ratio of random noise is usually less than 2, which cannot meet the threshold requirements.

[0113] The local maximum point in the distribution function is taken as the peak point, and the ratio of the peak point to the minimum value of the distribution function is calculated to obtain the peak ratio. The peak ratio threshold is set between 2 and 5. If the peak ratio is greater than or equal to the threshold, it is determined that there is a single-peak abrupt change; otherwise, it is determined to be noise or interlayer inhomogeneity. By using the quantified peak ratio index, signals with single-peak abrupt change characteristics can be objectively screened out, eliminating most non-defect abnormal signals and further reducing the number of candidate defect points by more than 60%.

[0114] Angular width is the most crucial indicator for distinguishing between defects and interlayer inhomogeneities. Area defects exhibit strong directional reflection, with strong echoes only occurring within the optimal reflection angle range of ±2° to 3°, and an angular width typically ranging from 3° to 8°. In contrast, localized inhomogeneities at interlayer interfaces result in large-area planar reflections, with reflections occurring across a wide angular range, and an angular width usually exceeding 10°.

[0115] This embodiment uses full width at half maximum (FWHM) to calculate the angular width, that is, to find the incident angles corresponding to when the amplitude on both sides of the peak drops to half of the peak, and calculate the difference between the two angles; if there are no half-peak points on both sides of the peak, the angular width is recorded as 0°, which further quantifies the angular distribution characteristics of the signal.

[0116] A single condition may lead to misjudgment. Some strong noise may present a single peak with a large angular width, while some interlayer inhomogeneities may present a broad peak with a peak-to-peak ratio that meets the standard. Only when both conditions are met simultaneously can the signal be guaranteed to have all the essential characteristics of a defect echo. That is, a sampling point is determined to have a defect only when both the peak-to-peak ratio is greater than or equal to the peak-to-peak ratio threshold and the angular width is less than the angular width threshold. If only one condition is met, it is judged to be a non-defect anomaly, and no alarm is triggered.

[0117] This embodiment focuses on the directionality of defect echo angles and solves the problem that traditional detection methods cannot distinguish between defects and interlayer echoes through a complete judgment process. Quantitative indicators enable automated and objective judgment, and dual conditions ensure high accuracy, which greatly improves the detection rate of defects with a diameter of 5mm and above, providing a reliable guarantee for the safe operation of insulated tie rods.

[0118] As an alternative approach, the distribution function expression of the residual amplitude as a function of the incident angle is as follows: ;in, Indicates the same sampling depth Below, incident angle The corresponding residual amplitude; Represents the incident angle in the residual matrix Sampling depth The residual amplitude at the point; the formula for calculating the peak value ratio is as follows: = ;in, Indicates the peak ratio of candidate defect points; Represents the peak point of the distribution function; This represents the incident angle corresponding to the peak point of the distribution function; Represents a very small positive number; the expression for the angle width is as follows: ;in, Indicates the width of the angle; This represents the angle of incidence closest to half the height of the residual amplitude peak to the right of the residual amplitude peak. This represents the incident angle closest to half the height of the residual amplitude peak to the left of the residual amplitude peak.

[0119] The residual amplitude distribution function is a mathematical expression for how the residual amplitude changes with the incident angle at the same sampling depth. It is used to quantitatively describe the distribution law of the signal in the angular dimension and is the basis for calculating the peak value ratio and angular width.

[0120] The minimum positive number δ is a tiny constant added to the kurtosis ratio calculation formula. It is used to prevent calculation errors where the denominator is zero when the distribution function's minimum value is zero, and to avoid abnormally large kurtosis ratio values. A typical value is [value missing]. .

[0121] Full width at half maximum (FWHM) is the difference between the two horizontal coordinates corresponding to the amplitude on both sides of the signal peak when it drops to half of the peak value. It is a quantitative indicator that characterizes the bandwidth or distribution range of the signal and is used to measure the angular distribution width of the defect echo.

[0122] This embodiment provides two optional forms of the distribution function, which are used when the echo signal undergoes envelope detection processing. When processed by full-wave rectification, it can be used This embodiment transforms discrete residual data into continuous mathematical functions, providing a unified mathematical foundation for the subsequent accurate calculation of peak ratio and angle width.

[0123] By using the larger of the minimum value of the distribution function and δ as the denominator, the calculation error of a zero denominator is effectively avoided. Directly dividing the peak value by the minimum value may lead to an abnormally large peak-to-peak ratio due to zero background noise; introducing δ ensures the numerical stability of the calculation. Simultaneously, this formula can accurately quantify the prominence of a single peak, distinguishing between strong single peaks of genuine defects and weak fluctuations of noise, achieving precise quantification of single-peak abrupt change characteristics. This provides an objective and repeatable standard for determining the peak-to-peak ratio, avoiding the subjectivity of manual observation.

[0124] The angle width is calculated by using the half-peak full width method. It is obtained by finding the angles on both sides of the peak half-height and calculating the difference. This avoids discrete sampling errors, accurately characterizes the directional of the defect echo angle, and provides a decisive numerical basis for distinguishing defects from interlayer inhomogeneity.

[0125] This embodiment transforms qualitative features into quantifiable indicators through three standardized mathematical expressions, solving the problems of strong subjectivity and poor repeatability in manual judgment, ensuring detection consistency and calculation stability, and improving the defect judgment accuracy to over 98%, thus laying a solid mathematical foundation for the automation of detection methods.

[0126] According to another aspect of the invention, such as Figure 2As shown, an ultrasonic testing system for internal defects in a layered structure of an insulating tie rod is also provided. This system is used to perform the aforementioned method for detecting internal defects in a layered structure of an insulating tie rod. The system includes: a phased array ultrasonic probe 21 for transmitting and receiving ultrasonic signals; a scanning mechanism 22 connected to the phased array ultrasonic probe 21 for moving the phased array ultrasonic probe 21 along the axial or circumferential direction of the insulating tie rod and recording its spatial position in real time; a data acquisition unit 23 connected to the phased array ultrasonic probe 21 for exciting ultrasonic signals and receiving echo amplitude values; and a signal processing unit 24 connected to the data acquisition unit 23 and the scanning mechanism 22, respectively. The signal processing unit 24 includes a matrix storage module, a cancellation operation module, and a defect identification module. The system includes: a matrix storage module, used to receive the echo amplitude and corresponding incident angle and sampling depth information transmitted by the data acquisition unit 23, construct and store the actual detection matrix; construct and store the reference template matrix based on the echo amplitude of the insulating tie rod sample without internal defects; a cancellation operation module, connected to the matrix storage module, used to perform dynamic structural noise cancellation based on the actual detection matrix and the reference template matrix to obtain the residual matrix; and a defect identification module, connected to the cancellation operation module, used to compare the residual amplitude of each sampling point in the residual matrix with a preset threshold to obtain candidate defect points; and obtain the defect judgment result based on the residual amplitude change and angle width of the candidate defect points through corresponding threshold comparison.

[0127] The phased array ultrasonic probe is a multi-element piezoelectric transducer that can achieve multi-angle sector scanning through electronic delay. It can be equipped with 1MHz~5MHz linear array or ring array and longitudinal wave wedges to meet the longitudinal wave detection requirements of insulating rods with different wall thicknesses. It has high resolution and large penetration depth to ensure the quality of echo data acquisition.

[0128] The scanning mechanism is a servo-driven moving mechanism with a high-precision encoder, supporting circumferential, axial, and helical scanning. It has high position resolution, enabling full-surface coverage inspection of the insulating tie rod and providing position coordinate support for precise spatial positioning of defects.

[0129] The data acquisition unit can generate excitation pulses and receive analog echo signals. After amplification, filtering, analog-to-digital conversion and envelope detection preprocessing, it outputs digital echo amplitude values, matching and adapting the sampling frequency and gain parameters to provide reliable raw data for matrix construction.

[0130] The signal processing unit can adopt a heterogeneous architecture of FPGA and ARM to modularly integrate matrix storage, cancellation operation and defect identification functions. The modules are interconnected at high speed to meet the needs of large-scale matrix operation and real-time processing.

[0131] The matrix storage module can adopt a dual-buffer structure to achieve structured storage, fast reading and writing, and unified format matching of the actual detection matrix and the benchmark template matrix, thereby improving the efficiency of matrix cancellation operations.

[0132] The cancellation operation module can adopt a parallel computing architecture to complete the adaptive weighting coefficient solution and dynamic structural noise cancellation, generate the residual matrix in real time, effectively remove interlayer echoes, and significantly improve the signal-to-noise ratio of defect signals.

[0133] The defect identification module integrates threshold screening, angular distribution feature extraction, peak ratio and angular width calculation, and dual-condition judgment logic to automatically distinguish between real defects and unevenness at interlayer interfaces, eliminate subjective errors from human experience, and ensure the accuracy of defect identification.

[0134] The detection process requires processing large amounts of matrix data, demanding high computational speed and real-time performance. A heterogeneous architecture balances high-speed computation with flexible control; modular design ensures a clear system structure, facilitating maintenance and upgrades. The signal processing unit can utilize a heterogeneous architecture of FPGA and ARM, with the FPGA handling high-speed data acquisition and parallel computation, and the ARM handling flow control and result output. These three functional submodules are connected via an internal high-speed bus, enabling real-time data transmission and processing. This provides a powerful computing platform for the entire detection algorithm, ensuring the real-time performance and reliability of the detection process.

[0135] Dynamic structural noise cancellation requires simultaneous access to large amounts of data from both the actual detection matrix and the baseline template matrix. High-speed local storage significantly improves computation speed; unified matrix construction rules ensure the compatibility of the two matrices. The matrix storage module employs a dual-cache structure: one cache stores the currently acquired actual detection data, and the other cache stores the completed baseline template matrix. This achieves structured storage and rapid retrieval of detection data, increasing the speed of cancellation operations by more than an order of magnitude.

[0136] Amplitude thresholds can only filter out abnormal signals; only by combining angular directional features can true defects be distinguished from non-defect anomalies. Automated judgment can eliminate the subjectivity of human experience and improve the consistency of detection results. The defect identification module first performs threshold comparison to filter candidate defect points, then extracts angular distribution features for each candidate defect point, calculates the peak value ratio and angular width, and finally determines whether a defect exists based on dual conditions, thus achieving automatic defect identification and judgment, improving the accuracy of defect identification, and eliminating the need to rely on the experience judgment of senior inspection personnel.

[0137] This embodiment consists of a phased array ultrasonic probe, a scanning mechanism, a data acquisition unit, and a modular signal processing unit working together to build a fully automated detection architecture from ultrasonic scanning, data acquisition, matrix operation to intelligent defect judgment. It fully leverages the advantages of phased array multi-angle detection and adaptive noise cancellation to solve the problems of traditional detection relying on manual labor, low efficiency, and high rates of missed and false detections. It significantly improves detection efficiency and defect detection capability, and can meet the engineering application requirements for batch production and in-service periodic inspection of insulating tie rods.

[0138] An embodiment of the present invention also provides an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor. The memory stores a computer program executable by the at least one processor, which, when executed by the at least one processor, causes the electronic device to perform the method of the embodiment of the present invention.

[0139] refer to Figure 3 The present invention will now describe a structural block diagram of an electronic device that can serve as an embodiment of the present invention, serving as an example of a hardware device applicable to various aspects of the present invention. The electronic device is intended to represent various forms of digital electronic computer devices, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0140] like Figure 3 As shown, the electronic device includes a computing unit 301, which can perform various appropriate actions and processes based on a computer program stored in a read-only memory (ROM) 302 or a computer program loaded from a storage unit 308 into a random access memory (RAM) 303. The RAM 303 may also store various programs and data required for the operation of the electronic device. The computing unit 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0141] Multiple components in the electronic device are connected to I / O interface 305, including: input unit 306, output unit 307, storage unit 308, and communication unit 309. Input unit 306 can be any type of device capable of inputting information into the electronic device. Input unit 306 can receive input digital or character information and generate key signal inputs related to user settings and / or function control of the electronic device. Output unit 307 can be any type of device capable of presenting information and may include, but is not limited to, a display, speaker, video / audio output terminal, vibrator, and / or printer. Storage unit 308 may include, but is not limited to, disks and optical discs. Communication unit 309 allows the electronic device to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks, and may include, but is not limited to, modems, network cards, infrared communication devices, and / or wireless communication transceivers, such as Bluetooth devices, WiFi devices, WiMax devices, cellular communication devices, and / or the like.

[0142] The computing unit 301 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 301 include, but are not limited to, CPUs, graphics processing units (GPUs), various special-purpose artificial intelligence (AI) computing units, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. The computing unit 301 performs the various methods and processes described above. For example, in some embodiments, the method embodiments of the present invention can be implemented as computer programs tangibly contained in a machine-readable medium, such as storage unit 308. In some embodiments, part or all of the computer program can be loaded and / or installed on an electronic device via ROM 302 and / or communication unit 309. In some embodiments, the computing unit 301 can be configured to perform the methods described above by any other suitable means (e.g., by means of firmware).

[0143] Computer programs for implementing the methods of embodiments of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0144] In the context of embodiments of this invention, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable signal medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, or infrared systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0145] It should be noted that the term "comprising" and its variations used in the embodiments of this invention are open-ended, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The modifications of "one" and "a plurality" mentioned in the embodiments of this invention are illustrative and not restrictive, and those skilled in the art should understand that unless explicitly indicated otherwise in the context, they should be understood as "one or more".

[0146] The user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in the embodiments of this invention are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0147] The steps described in the method embodiments provided by the present invention can be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of protection of the present invention is not limited in this respect.

[0148] The term "embodiment" in this specification refers to a specific feature, structure, or characteristic described in connection with an embodiment that may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily imply the same embodiment, nor does it imply independence or alternativeity from other embodiments. The various embodiments in this specification are described in a related manner, with reference to each other for similar or identical parts. In particular, for apparatus, device, and system embodiments, since they are substantially similar to method embodiments, the description is relatively simple, and relevant details are referred to in the description of the method embodiments.

[0149] The above-described embodiments are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of protection. It should be noted that those skilled in the art can make various modifications and improvements without departing from the inventive concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the appended claims.

Claims

1. An ultrasonic testing method for internal defects in a layered structure of an insulating tie rod, characterized in that, include: S101, obtain the echo amplitude of the insulating rod under test at different incident angles, and construct the actual detection matrix in combination with the corresponding sampling depth, including: scanning the insulating rod under test along the circumference of the insulating rod under test at different incident angles by longitudinal wave incident, and collecting the echo amplitude at different incident angles and different sampling depths; Each of the incident angles is mapped to a row of a two-dimensional matrix, and each of the sampling depths is mapped to a column of a two-dimensional matrix; The sampling depth of the sampling points is based on the initial sampling depth in the depth direction, and increases sequentially according to the depth sampling interval. Starting from the initial sampling depth, the depth increases by one sampling interval for each additional sampling point. The sampling interval is equal to the longitudinal wave velocity of the insulating tie rod material divided by twice the sampling frequency; The echo amplitude is used as the matrix element at the corresponding row and column positions to obtain the actual detection matrix; S102, Based on the actual detection matrix and the reference template matrix, perform dynamic structural noise cancellation to obtain a residual matrix. The reference template matrix is ​​constructed based on the echo amplitude of the insulating tie rod sample without internal defects, including: delineating a local window of the same size with the sampling point to be determined at the corresponding position in the actual detection matrix and the reference template matrix as the center. Calculate the adaptive weighting coefficients based on the actual detection matrix elements and the reference template matrix elements within the local window; Subtract the product of the adaptive weighting coefficient and the corresponding element in the benchmark template matrix from the elements in the actual detection matrix to obtain the residual magnitude at the corresponding position; Construct a residual matrix based on the residual amplitude corresponding to the incident angle and sampling depth; The residual matrix is ​​expressed as follows: ； Where R represents the residual matrix; This represents the element in the i-th row and j-th column of matrix R; Indicates the angle of incidence Sampling depth The corresponding residual amplitude; i=1,2,…M, where M is the total number of angles; j=1,2,…N, where N is the number of depth sampling points; ； in, Indicates the adaptive weighting coefficients; This indicates the angle of incidence of the insulating rod under test. Sampling depth Echo amplitude at the location; This indicates the incident angle of the insulating tie rod sample. Sampling depth Echo amplitude at the location; ； in, Indicates A local window centered on the user; For regularization parameters, ; This represents the echo amplitude value located at row p and column q in the actual detection matrix; This represents the echo amplitude value located at row p and column q in the reference template matrix; S103, compare the residual amplitude of each sampling point in the residual matrix with a preset threshold to obtain candidate defect points; S104, based on the residual amplitude change and angle width of the candidate defect points, the defect judgment result is obtained by comparing the corresponding thresholds, including: for the sampling depth where each candidate defect point is located, extracting the residual amplitude corresponding to all incident angles at the sampling depth, and obtaining the distribution function of the residual amplitude changing with the incident angle; The peak ratio is calculated based on the changing characteristics of the distribution function. If the peak ratio is greater than or equal to the peak ratio threshold, it is determined that the residual amplitude exhibits a sudden change in a unimodal state. The angular width of the defect echo is calculated based on the peak value and half-maximum of the residual amplitude of the distribution function. If the residual amplitude exhibits a sudden change in a single-peak state and the angle width is less than a preset angle width threshold, it is determined that the sampling point has a defect.

2. The ultrasonic detection method for internal defects in a layered structure of an insulating tie rod according to claim 1, characterized in that, The expression for the actual detection matrix is ​​as follows: ； Where S represents the actual detection matrix; This represents the element in the i-th row and j-th column of matrix S; This indicates the angle of incidence of the insulating rod under test. Sampling depth The echo amplitude at the location; i=1,2,…M, where M is the total number of angles; j=1,2,…N, where N is the number of depth sampling points; ； in, Δd represents the initial sampling depth in the depth direction; Δd represents the sampling interval in the depth direction. Δd = c / (2 × fs); Where c represents the longitudinal wave velocity of the insulating tie rod material; fs represents the sampling frequency.

3. The ultrasonic detection method for internal defects in a layered structure of an insulating tie rod according to claim 1, characterized in that, Before performing dynamic structural noise cancellation based on the actual detection matrix and the reference template matrix to obtain the residual matrix, the following steps are also included: An insulating rod from the same batch and of the same specifications as the insulating rod to be tested, and without internal defects, is selected as the insulating rod sample. Based on the same testing instrument parameters as the insulating tie rod under test, the insulating tie rod sample is scanned along the circumference of the sample using longitudinal wave incident method at different incident angles to obtain the template echo amplitude of the insulating tie rod sample; Based on the template echo amplitude, a reference template matrix is ​​constructed, and the expression for constructing the reference template matrix is ​​as follows: ; Where T represents the reference template matrix; This represents the element in the i-th row and j-th column of matrix T; This indicates the incident angle of the insulating tie rod sample. Sampling depth The echo amplitude at the location; i=1,2,…M, where M is the total number of angles; j=1,2,…N, where N is the number of depth sampling points.

4. The ultrasonic detection method for internal defects in a layered structure of an insulating tie rod according to claim 1, characterized in that, Before comparing the residual magnitude of each sampling point in the residual matrix with the fixed threshold to obtain candidate defect points, the method further includes: Based on the instrument parameters used in the actual test, the phased array ultrasonic probe was placed in the air to collect noise signals; Based on the noise signal, the root mean square value of the noise signal is calculated; A fixed threshold is calculated based on the root mean square value of the noise, and the formula for calculating the fixed threshold is as follows: ； in, For a fixed threshold, represents the root mean square value of the noise; k represents the threshold coefficient, which ranges from 3 to 6.

5. The ultrasonic detection method for internal defects in a layered structure of an insulating tie rod according to claim 1, characterized in that, The preset threshold is an adaptive threshold. Before comparing the residual magnitude of each sampling point in the residual matrix with the preset threshold to obtain candidate defect points, the method further includes: A local window is divided with the residual point to be determined in the residual matrix as the center; Based on the values ​​of all residual points within the local window, the mean and standard deviation of the local residuals are calculated. The adaptive threshold is calculated based on the mean and standard deviation of the local residuals. The formula for calculating the adaptive threshold is as follows: ； in, Indicates an adaptive threshold; This represents the mean of the local residuals. Indicates the local residual standard deviation. This represents the adjustment coefficient, with a value range of 2 to 4.

6. The ultrasonic detection method for internal defects in a layered structure of an insulating tie rod according to claim 1, characterized in that, The preset threshold uses both a fixed threshold and an adaptive threshold. The residual magnitude of each sampling point in the residual matrix is ​​compared with the preset threshold to obtain candidate defect points, including: Obtain the fixed threshold and the adaptive threshold respectively; The maximum threshold is obtained by comparing the fixed threshold with the adaptive threshold; Compare the residual magnitude of each sampling point in the residual matrix with the maximum threshold. If the residual magnitude of a sampling point in the residual matrix exceeds the maximum threshold, it is marked as a candidate defect point.

7. The ultrasonic detection method for internal defects in a layered structure of an insulating tie rod according to claim 1, characterized in that, The distribution function expression of the residual amplitude as a function of the incident angle is as follows: ； in, Indicates the same sampling depth Below, incident angle The corresponding residual amplitude; Represents the incident angle in the residual matrix Sampling depth The residual amplitude at the location; The formula for calculating the peak value ratio is as follows: = ； in, This represents the peak ratio of the candidate defect points; Indicates the peak point of the distribution function; This represents the incident angle corresponding to the peak point of the distribution function; Represents a very small positive number; The expression for the angle width is as follows: ； in, Indicates the width of the angle; This represents the incident angle closest to half the height of the residual amplitude peak to the right of the residual amplitude peak. This represents the incident angle closest to half the height of the residual amplitude peak to the left of the residual amplitude peak.

8. An ultrasonic testing system for internal defects in a layered structure of an insulating tie rod, used to perform the ultrasonic testing method for internal defects in a layered structure of an insulating tie rod as described in claim 1, characterized in that, include: A phased array ultrasonic probe is used to transmit and receive ultrasonic signals; The scanning mechanism is connected to the phased array ultrasonic probe and is used to drive the phased array ultrasonic probe to move axially or circumferentially along the insulating tie rod and record the spatial position in real time. The data acquisition unit is connected to the phased array ultrasonic probe and is used to excite ultrasonic signals and receive echo amplitude values. The signal processing unit is connected to the data acquisition unit and the scanning mechanism respectively. The signal processing unit includes a matrix storage module, a cancellation operation module, and a defect identification module. The matrix storage module is used to receive the echo amplitude and corresponding incident angle and sampling depth information transmitted by the data acquisition unit, construct and store the actual detection matrix; and construct and store the reference template matrix based on the echo amplitude of the insulating tie rod sample without internal defects. The cancellation operation module is connected to the matrix storage module and is used to perform dynamic structural noise cancellation based on the actual detection matrix and the reference template matrix to obtain the residual matrix. The defect identification module is connected to the cancellation operation module and is used to compare the residual amplitude of each sampling point in the residual matrix with a preset threshold to obtain candidate defect points; based on the residual amplitude change and angle width of the candidate defect points, the defect judgment result is obtained through the corresponding threshold comparison.

9. An electronic device, comprising: A processor and a memory storing a program, characterized in that the program includes instructions that, when executed by the processor, cause the processor to perform the method according to any one of claims 1 to 7.

Images