Eddy current defect signal reconstruction enhancement method and system based on second-order abrupt change constraint

Abstract

The application provides a kind of based on the second order mutation constraint eddy current defect signal reconstruction enhancement method and system, it is related to nondestructive testing and detection signal processing technical field, method includes: obtaining eddy current detection signal to obtain smooth signal and then obtain first order change signal;According to the second order mutation signal calculated from first order change signal and then constructs second order mutation evaluation quantity, to construct robust threshold according to second order mutation evaluation quantity, and construct second order mutation weight in combination with second order mutation evaluation quantity;First order change signal is respectively carried out strong filtering and weak filtering to obtain strong filtering after first order signal and weak filtering after first order signal, and adaptive fusion is carried out in combination with second order mutation weight to obtain first order adaptive filtering signal of second order mutation constraint;Again, integral reconstruction is carried out to obtain initial reconstruction signal, and then enhanced defect signal is obtained according to initial reconstruction signal.The application improves the extraction ability and detection reliability of weak defect signal under complex background.

Description

Technical Field

[0001] This invention relates to the field of nondestructive testing and detection signal processing technology, and in particular to a method and system for reconstructing and enhancing eddy current defect signals based on second-order mutation constraints. Background Technology

[0002] Fillet welds in small connecting pipes are typically used to connect auxiliary structures such as pressure inlet pipes, drain pipes, instrument pipes, vent pipes, and bypass pipes, and are critical connection points in long-distance and pressure pipeline systems. These welds are located in confined spaces with dramatic curvature changes, complex weld reinforcement, and intricate surface textures. Defects such as cracks, lack of fusion, incomplete penetration, porosity, corrosion, and fatigue damage may be distributed on the weld surface, inside, or at the root. Due to the complex structural morphology and limitations imposed by the inspection posture, the detection signal is easily interfered with by factors such as structural background, lift-off variations, mechanical vibration, and environmental noise, making it difficult to accurately extract weak defect responses.

[0003] Eddy current testing has advantages such as being non-contact, having a fast response speed, and being suitable for the inspection of metal components, showing great application potential in the inspection of complex welds and space-constrained structures. However, in actual inspection processes, weld scale patterns, variations in weld reinforcement height, probe lift-off fluctuations, sensor posture changes, mechanical noise, and environmental electromagnetic noise can all be superimposed on the detection signal, making the defect response easily overwhelmed by background signals and random noise.

[0004] Existing eddy current signal processing methods mostly employ filtering, normalization, envelope analysis, wavelet decomposition, empirical mode decomposition, or single feature extraction for defect identification. While these methods can improve the signal-to-noise ratio to some extent, they still have the following shortcomings: First, the filtering parameters are usually fixed, making it difficult to adapt to both smooth background regions and abrupt defect changes simultaneously; second, high-frequency abrupt changes at defect edges are easily weakened during strong filtering, resulting in defect feature loss; third, simply using differential enhancement can easily amplify random noise, while direct integral reconstruction is prone to baseline drift and amplitude distortion. Therefore, there is an urgent need for an eddy current signal enhancement method that can adaptively adjust the filtering intensity according to the degree of local defect abrupt changes and restore the defect response morphology through reconstruction, in order to improve the extraction capability and detection reliability of weak defect signals in complex backgrounds. Summary of the Invention

[0005] Based on this, the purpose of this invention is to provide a method and system for reconstructing and enhancing eddy current defect signals based on second-order mutation constraints. This method can adaptively adjust the filtering intensity according to the degree of local defect mutation and restore the defect response morphology through reconstruction, thereby improving the extraction capability and detection reliability of weak defect signals in complex backgrounds.

[0006] This invention provides a method for reconstructing and enhancing eddy current defect signals based on second-order mutation constraints, comprising: The eddy current detection signal is acquired to obtain a smoothed signal, and a first-order change signal is obtained based on the smoothed signal. A second-order mutation signal is calculated based on the first-order change signal, and a second-order mutation evaluation quantity is constructed based on the second-order mutation signal. A robust threshold is constructed based on the second-order mutation evaluation quantity, and a second-order mutation weight is constructed based on the second-order mutation evaluation quantity and the robust threshold. The first-order change signal is subjected to strong filtering and weak filtering respectively to obtain a strongly filtered first-order signal and a weakly filtered first-order signal, and then adaptively fused with the second-order mutation weight to obtain a first-order adaptive filtered signal with second-order mutation constraint. An initial reconstructed signal is obtained by integrating and reconstructing the first-order adaptive filter signal of the second-order mutation constraint, and then an enhanced defect signal is obtained based on the initial reconstructed signal.

[0007] The above-mentioned eddy current defect signal reconstruction and enhancement method based on second-order mutation constraints describes the changing trend of the eddy current detection signal using a first-order variation signal, and uses a second-order mutation signal to describe the local curvature change and mutation degree. The first-order signal after strong filtering and the first-order signal after weak filtering are adaptively fused by the second-order mutation weight. The smooth background region is strongly filtered to suppress noise, and the defect mutation region is weakly filtered to preserve details. Finally, the first-order adaptively filtered signal with second-order mutation constraints obtained by adaptive fusion is integrated and reconstructed to obtain an enhanced defect signal in which the background noise is suppressed and the local defect response is preserved.

[0008] In addition, the eddy current defect signal reconstruction and enhancement method based on second-order mutation constraint according to the present invention may also have the following additional technical features: Furthermore, in the step of integrating and reconstructing the first-order adaptive filtered signal under the second-order abrupt change constraint to obtain the initial reconstructed signal, the functional expression of the integral reconstruction method is: ; In the formula, r0(n) represents the initial reconstructed signal; d1a( i ) represents the first-order adaptive filtering signal with second-order mutation constraint; Δt represents the sampling time interval between adjacent sampling points. i This represents the i-th discrete sampling point in the integral reconstruction process; n This indicates the sequence number of the sampling point to be reconstructed.

[0009] Furthermore, the step of obtaining the enhanced defect signal based on the initial reconstructed signal includes: The initial reconstructed signal is subjected to linear drift compensation, mean correction and amplitude scaling correction to obtain the enhanced defect signal; The formula for calculating the enhanced defect signal is as follows: r e (n) = K · [r0(n) l(n)] + μ; In the formula, r e (n) represents the enhanced defect signal; l(n) represents the integral drift compensation term, K represents the amplitude scale correction coefficient, μ represents the mean compensation term; r0(n) represents the initial reconstructed signal.

[0010] Furthermore, the calculation formula for the first-order adaptive filter signal under second-order mutation constraint is as follows: d1a(n) = [1 w(n)]·d1s(n) + w(n)·d1m(n); In the formula, d1a(n) represents the first-order adaptive filtered signal with second-order mutation constraint; w(n) represents the second-order mutation weight; d1s(n) represents the first-order signal after strong filtering; and d1m(n) represents the first-order signal after weak filtering. The formula for calculating the first-order signal after strong filtering is: d1s(n) = F strong [d1(n)]; The formula for calculating the first-order signal after weak filtering is: d1m(n) = F mild [d1(n)]; In the formula, d1s(n) represents the first-order signal after strong filtering; d1m(n) represents the first-order signal after weak filtering; d1(n) represents the first-order changing signal; F strong [﹒] denotes a strong filtering operator; F mild [﹒ ] Weak filtering operator.

[0011] Furthermore, the formula for calculating the second-order mutation weight is as follows: w(n) = c(n) / [c(n) + T c + ε]; In the formula, w(n) represents the second-order mutation weight; c(n) represents the second-order mutation evaluation value at the nth sampling point; T c The robust threshold is represented by ε, which is a non-zero small constant used to avoid zero denominators. The formula for calculating the robustness threshold is: T c = median(c) + λ · MAD(c); In the formula, T cλ represents the robust threshold; MAD(c) represents the median absolute deviation of the second-order mutation evaluation, where c represents the second-order mutation evaluation sequence, which is composed of the second-order mutation evaluations of each sampling point, c={ c(1),c(2),…, c(N)}; median(c) represents the median of all elements in the second-order mutation evaluation sequence c. The formula for calculating the second-order mutation evaluation metric is as follows: c(n) = M W {|d2(n)|}; In the formula, c(n) represents the second-order mutation evaluation metric at the nth sampling point; M W { } represents an operator that performs smoothing, energy, or statistical calculations on the input sequence within a local window of length W corresponding to the nth sampling point; W represents the length of the local window; d2(n) represents the second-order abrupt change signal at the nth sampling point; n represents the discrete sampling point index; where: the formula for calculating the second-order abrupt change signal is: d2(n) = [d1(n+1) d1(n 1)] / (2Δt); In the formula, d2(n) represents the second-order abrupt change signal at the nth sampling point; d1(n+1) represents the first-order change signal at the (n+1)th sampling point; d1(n 1) represents the nth The first-order change signal at one sampling point; Δt represents the sampling time interval between adjacent sampling points; n represents the discrete sampling point number.

[0012] Furthermore, the formula for calculating the first-order changing signal is: d1(n) = [r s (n+1) r s (n 1)] / (2Δt); In the formula, d1(n) represents the first-order changing signal; r s (n+1) represents the smoothed signal at the (n+1)th sampling point; r s (n 1) represents the nth The smoothed signal at one sampling point; Δt represents the sampling time interval between adjacent sampling points; n represents the discrete sampling point number.

[0013] Furthermore, the step of acquiring the eddy current detection signal to obtain a smoothed signal includes: Acquire eddy current detection signals and perform preprocessing to obtain the signal to be processed; The signal to be processed is smoothed to obtain a smoothed signal; The preprocessing methods include: First, the background of the eddy current detection signal is estimated to obtain the background signal. Then, the eddy current detection signal is subtracted from the background signal to obtain the residual signal, which is the signal to be processed. The formula for calculating the smoothed signal is as follows: r s (n) = F smooth [r(n)]; In the formula, r s (n) represents a smoothed signal; F smooth [﹒ ] denotes the smoothing operator; r(n) denotes the signal to be processed, where r(n) = x(n). b(n) and x(n) represent the eddy current detection signal; b(n) represents the background signal.

[0014] Another aspect of the present invention provides a signal reconstruction and enhancement system for eddy current defects based on second-order mutation constraints, the system comprising: The acquisition module is used to acquire the eddy current detection signal to obtain a smoothed signal, and to obtain a first-order change signal based on the smoothed signal; The calculation module is used to calculate a second-order mutation signal based on the first-order change signal, construct a second-order mutation evaluation quantity based on the second-order mutation signal, construct a robust threshold based on the second-order mutation evaluation quantity, and construct a second-order mutation weight based on the second-order mutation evaluation quantity and the robust threshold. The fusion module is used to perform strong filtering and weak filtering on the first-order change signal to obtain a strongly filtered first-order signal and a weakly filtered first-order signal, and to perform adaptive fusion with the second-order mutation weight to obtain a first-order adaptive filtered signal with second-order mutation constraints. The reconstruction module is used to perform integral reconstruction on the first-order adaptive filter signal of the second-order mutation constraint to obtain an initial reconstruction signal, and then obtain an enhanced defect signal based on the initial reconstruction signal.

[0015] In another aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method for reconstructing and enhancing eddy current defect signals based on second-order mutation constraints.

[0016] In another aspect, the present invention provides a data processing device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the above-described method for reconstructing and enhancing eddy current defect signals based on second-order mutation constraints. Attached Figure Description

[0017] Figure 1 This is a flowchart of the eddy current defect signal reconstruction and enhancement method based on second-order mutation constraints in the first embodiment of the present invention; Figure 2 This is a schematic diagram of the surface opening crack of the No. 20 carbon steel weld in the first embodiment of the present invention; Figure 3 This is a schematic diagram of the original eddy current detection signal obtained by sampling in the first embodiment of the present invention; Figure 4 This is a schematic diagram of the signal to be processed in the first embodiment of the present invention; Figure 5 This is a schematic diagram of a first-order changing signal in the first embodiment of the present invention; Figure 6 This is a schematic diagram of a second-order mutation signal in the first embodiment of the present invention; Figure 7 This is a schematic diagram of the second-order mutation weights in the first embodiment of the present invention; Figure 8 This is a comparison diagram of the original signal to be processed and the enhanced defect signal in the first embodiment of the present invention; The following detailed description, in conjunction with the accompanying drawings, will further illustrate the present invention. Detailed Implementation

[0018] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Several embodiments of the invention are illustrated in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0020] To provide an eddy current signal enhancement method that can adaptively adjust the filtering intensity according to the degree of local defect mutation and restore the defect response morphology through reconstruction, thereby improving the extraction capability and detection reliability of weak defect signals in complex backgrounds, this invention provides an eddy current defect signal reconstruction and enhancement method and system based on second-order mutation constraints. The method uses a first-order variation signal to describe the changing trend of the eddy current detection signal, and a second-order mutation signal to describe the local curvature change and mutation degree. The first-order signals after strong filtering and weak filtering are adaptively fused using second-order mutation weights. Strong filtering is applied to smooth background regions to suppress noise, while weak filtering is applied to defect mutation regions to preserve details. Finally, the adaptively fused first-order adaptively filtered signal with second-order mutation constraints is integrally reconstructed to obtain an enhanced defect signal where background noise is suppressed and the local defect response is preserved.

[0021] Specifically, the eddy current defect signal reconstruction and enhancement method and system based on second-order mutation constraint provided by this invention can be applied to conventional eddy current detection, pulsed eddy current detection, resonant eddy current detection, array eddy current detection, swept-frequency eddy current detection and frequency-selective eddy current detection systems. It is especially suitable for weak defect response extraction, denoising and enhancement of structures such as complex welds, welds with surface textures, space-constrained welds and small pipe fillet welds.

[0022] Furthermore, the technical solution provided by this invention can suppress noise in a smooth background area, retain local response characteristics in a defect abrupt change area, improve the extraction capability of weak defect eddy current signals in a complex background, and is suitable for defect detection of structures such as small pipe fillet welds, welds with fish scale patterns and complex curved surface welds.

[0023] To facilitate understanding of the present invention, several embodiments are given below. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that the disclosure of the present invention will be more thorough and complete.

[0024] Example 1 Please see Figure 1 The figure shows a method for reconstructing and enhancing eddy current defect signals based on second-order mutation constraints in the first embodiment of the present invention. The method includes steps S101 to S104: S101. Obtain the eddy current detection signal to obtain a smoothed signal, and obtain a first-order change signal based on the smoothed signal.

[0025] As a specific example, the eddy current detection signal is acquired and preprocessed to obtain the signal to be processed; the signal to be processed is then smoothed to obtain a smoothed signal. Specifically, an eddy current sensor is used to scan and detect the structure under test to obtain the eddy current detection signal. The eddy current detection signal can be the receiving coil voltage, impedance change, resonant amplitude, resonant phase, resonant frequency shift, quality factor change, or a multi-channel fused signal. Further, the acquired eddy current detection signal is preprocessed to obtain the signal to be processed. The preprocessing includes one or more of the following: DC removal, normalization, trend term subtraction, or effective detection segment truncation. In this embodiment, the eddy current detection signal is first estimated to obtain a background signal. Then, the eddy current detection signal is subtracted from the background signal to obtain a residual signal. In this embodiment, the residual signal serves as the input signal for subsequent differentiation, filtering, and reconstruction processing, i.e., the signal to be processed.

[0026] Furthermore, the signal to be processed is smoothed to obtain a smoothed signal, and then the smoothed signal is subjected to first-order differentiation or first-order difference to obtain a first-order change signal.

[0027] Specifically, the formula for calculating a smooth signal is: r s (n) = F smooth [r(n)]; In the formula, r s (n) represents a smoothed signal; F smooth [﹒ ] denotes the smoothing operator; r(n) denotes the signal to be processed, where r(n) = x(n). b(n) and x(n) represent eddy current detection signals, where n = 1, 2, ..., N, and N is the number of sampling points; b(n) represents the background signal, which can be obtained by sliding median filtering, sliding mean filtering, low-pass filtering, wavelet decomposition, or empirical mode decomposition.

[0028] Furthermore, the formula for calculating the first-order changing signal is: d1(n) = [r s (n+1) r s (n 1)] / (2Δt); In the formula, d1(n) represents the first-order changing signal, used to characterize the rate of change of the eddy current response with sampling time or scanning position; r s (n+1) represents the smoothed signal at the (n+1)th sampling point; r s (n 1) represents the nth The smoothed signal at one sampling point; n represents the discrete sampling point number; Δt represents the sampling time interval between adjacent sampling points. When the eddy current detection signal is a spatial scanning sequence, Δt can be replaced by the spatial sampling interval Δx.

[0029] S102. Calculate the second-order mutation signal based on the first-order change signal, construct the second-order mutation evaluation quantity based on the second-order mutation signal, construct a robust threshold based on the second-order mutation evaluation quantity, and construct the second-order mutation weight based on the second-order mutation evaluation quantity and the robust threshold.

[0030] As a specific example, the derivative of the first-order change signal is obtained to obtain the second-order mutation signal; then, based on the absolute value, envelope, local mean square value, local energy, or normalized curvature of the second-order mutation signal, a second-order mutation evaluation quantity is constructed; a robust threshold is constructed based on the second-order mutation evaluation quantity, wherein the median and median absolute deviation are used to construct the robust threshold to reduce the influence of outliers and strong noise on the threshold selection; then, the second-order mutation weight is constructed based on the second-order mutation evaluation quantity and the robust threshold.

[0031] Specifically, the formula for calculating the second-order abrupt change signal is: d2(n) = [d1(n+1) d1(n 1)] / (2Δt); In the formula, d2(n) represents the second-order abrupt change signal at the nth sampling point, used to describe the local curvature, edge abrupt changes, and degree of non-stationary anomalies of the detected signal; d1(n+1) represents the first-order change signal at the (n+1)th sampling point; d1(n 1) represents the nth The first-order change signal at one sampling point; Δt represents the sampling time interval between adjacent sampling points; n represents the discrete sampling point number.

[0032] Secondly, the formula for calculating the second-order mutation evaluation metric is: c(n) = M W {|d2(n)|}; In the formula, c(n) represents the second-order mutation evaluation value at the nth sampling point, used to characterize the local mutation intensity at different sampling locations; M W {·} represents an operator that performs smoothing, energy, or statistical calculations on the input sequence within a local window of length W corresponding to the nth sampling point; W represents the length of the local window; d2(n) represents the second-order abrupt change signal at the nth sampling point; n represents the discrete sampling point index; d2(n) represents the second-order abrupt change signal. In addition, the formula for calculating the robustness threshold is: T c = median(c) + λ · MAD(c); In the formula, T c λ represents the robust threshold; MAD(c) represents the median absolute deviation of the second-order mutation evaluation, where c represents the second-order mutation evaluation sequence, which is composed of the second-order mutation evaluations of each sampling point, c={ c(1),c(2),…, c(N)}; median(c) represents the median of all elements in the second-order mutation evaluation sequence c. Furthermore, the formula for calculating the second-order mutation weight is: w(n) = c(n) / [c(n) + T c + ε]; In the formula, c(n) represents the second-order mutation evaluation metric at the nth sampling point; T c ε represents the robust threshold; ε represents a non-zero small constant used to avoid the denominator being zero; w(n) represents the second-order mutation weight, with the value of w(n) ranging from 0 to 1. When w(n) is larger, it indicates that the local mutation at that position is more obvious, and it is more likely to correspond to the defect edge, local defect anomaly, or non-stationary disturbance; when w(n) is smaller, it indicates that the change at that position is more gradual, and it is more likely to correspond to background changes or random noise.

[0033] S103. Perform strong filtering and weak filtering on the first-order changing signal to obtain the strongly filtered first-order signal and the weakly filtered first-order signal, and then perform adaptive fusion with the second-order mutation weight to obtain the first-order adaptive filtered signal with second-order mutation constraint.

[0034] In this embodiment, strong filtering is used to suppress noise in smooth background regions, while weak filtering is used to preserve details in regions with abrupt changes in defects. Strong and weak filtering can employ low-pass filtering, band-pass filtering, wavelet filtering, finite impulse response filtering, or infinite impulse response filtering with different cutoff frequencies, passband widths, or filter orders. Preferably, the cutoff frequency of the strong filter is lower than that of the weak filter.

[0035] Specifically, the formula for calculating the first-order signal after strong filtering is: d1s(n) = F strong [d1(n)]; The formula for calculating the first-order signal after weak filtering is: d1m(n) = F mild [d1(n)]; In the formula, d1s(n) represents the first-order signal after strong filtering; d1m(n) represents the first-order signal after weak filtering; d1(n) represents the first-order changing signal; F strong [﹒] denotes a strong filtering operator; F mild [﹒ ] Weak filtering operator.

[0036] Furthermore, the calculation formula for the first-order adaptive filter signal under second-order mutation constraint is as follows: d1a(n) = [1 w(n)]·d1s(n) + w(n)·d1m(n); In the formula, d1a(n) represents the first-order adaptive filter signal of the second-order mutation constraint; d1s(n) represents the first-order signal after strong filtering; d1m(n) represents the first-order signal after weak filtering; w(n) represents the second-order mutation weight; when w(n) is small, the fusion result mainly adopts the strong filtering result to enhance the suppression of smooth background noise; when w(n) is large, the fusion result mainly adopts the weak filtering result to retain the local mutation information caused by defects such as cracks, lack of fusion, incomplete penetration, and corrosion edges.

[0037] S104. The first-order adaptive filter signal of the second-order mutation constraint is integrally reconstructed to obtain the initial reconstructed signal, and then the enhanced defect signal is obtained based on the initial reconstructed signal.

[0038] The first-order adaptive filtered signal with second-order mutation constraint is integrally reconstructed to obtain the initial reconstructed signal. Since the differentiation and integration process may lead to loss of constant terms, linear trend terms, or endpoint drift, the initial reconstructed signal is subjected to linear drift compensation, mean correction, and amplitude scale correction to finally obtain the enhanced defect signal. Then, defect evaluation features are extracted from the enhanced defect signal. The defect evaluation features include one or more of the following: peak value, peak-to-peak value, root mean square, residual energy, kurtosis, skewness, half-maximum width at half maximum, energy centroid, second-order central moment, signal-to-noise ratio enhancement, imaging contrast, and positioning deviation. Based on the above defect evaluation features, the evaluation results of defect location, defect severity, or defect type are output.

[0039] This enables the present invention to enhance noise suppression in regions with gentle background changes, retain local response characteristics in regions with abrupt defect changes, and restore the overall shape of the defect signal through integral reconstruction, thereby improving the ability to extract weak defect eddy current signals in complex backgrounds.

[0040] Specifically, the formula for calculating the initial reconstructed signal is: ; In the formula, r0(n) represents the initial reconstructed signal; d1a( i ) represents the first-order adaptive filtering signal with second-order mutation constraint; Δt represents the sampling time interval between adjacent sampling points. i This represents the i-th discrete sampling point in the integral reconstruction process; n This indicates the sequence number of the sampling point to be reconstructed.

[0041] Secondly, the formula for calculating the enhanced defect signal is: re (n) = K · [r0(n) l(n)] + μ; In the formula, r e (n) represents the enhanced defect signal; l(n) represents the integral drift compensation term, K represents the amplitude scale correction coefficient, μ represents the mean compensation term; r0(n) represents the initial reconstructed signal.

[0042] To better illustrate the technical solution of the present invention, a specific example is provided below: Taking a No. 20 carbon steel weld test block as an example, such as Figure 2 As shown, Figure 2 This is a schematic diagram of surface-opening cracks in a No. 20 carbon steel weld. The weld area of ​​this test block has a weld reinforcement, and surface-opening cracks E1 to E5 with varying widths are arranged along the weld centerline. Each crack is arranged every 20 mm along the weld centerline, with a crack length and depth of 2 mm and a crack width gradually increasing from 0.35 mm to 1.00 mm. Specific dimensional parameters are shown in Table 1.

[0043] Table 1:

[0044] The original eddy current detection signal is obtained by scanning along the weld seam area using an eddy current sensor. In this embodiment, an eddy current probe is used as the eddy current sensor. Figure 3 The image shows a schematic diagram of the original eddy current detection signal obtained from sampling. Due to the influence of weld scale patterns, variations in weld reinforcement height, and local lift-off fluctuations, the original eddy current detection signal contains significant slowly varying background and random noise, making it susceptible to interference when directly extracting features such as peak values ​​and peak-to-peak values. Therefore: First, background estimation is performed on the original eddy current detection signal, and the signal to be processed, r(n), is constructed, such as... Figure 4 The diagram shows the signal to be processed. The signal r(n) is then smoothed to obtain a smoothed signal, and the first-order change signal d1(n) is calculated based on the smoothed signal, as shown below. Figure 5 As shown, Figure 5 This is a schematic diagram of a first-order variation signal. Specifically, the first-order variation signal d1(n) can highlight the variation trend of the eddy current response in the scanning direction, but it also amplifies local noise. To further distinguish between defect mutations and background disturbances, the derivative of the first-order variation signal d1(n) is taken to obtain the second-order mutation signal d2(n), and the second-order mutation evaluation quantity c(n) is obtained through local smoothing or local energy calculation. Specifically, as shown... Figure 6 The diagram shown is a schematic of a second-order mutation signal.

[0045] Based on the second-order mutation evaluation metric c(n) and the robustness threshold T cConstruct the second-order mutation weight w(n); such as Figure 7 The diagram illustrates the second-order abrupt change weighting. For relatively smooth background areas caused by weld reinforcement, fish-scale patterns, or slow lift-off changes, w(n) is lower, and the fusion process uses stronger filtering results more often. For local abnormal areas caused by cracks, lack of fusion, incomplete penetration, or corrosion pit edges, w(n) is higher, and the fusion process uses weaker filtering results more often. The resulting first-order adaptive filtering signal for the second-order abrupt change constraint can reduce background noise while preserving local defect abrupt change information.

[0046] Finally, the first-order adaptive filtered signal with second-order abrupt change constraints is integrally reconstructed, and linear drift compensation, endpoint constraint correction, mean correction, and amplitude scaling correction are performed to obtain the enhanced defect signal. Compared with the original signal to be processed, the final enhanced defect signal obtained after reconstruction can reduce random noise and background interference, preserve the defect response morphology, and facilitate subsequent defect localization and feature extraction. Figure 8 The image shown is a comparison between the original signal to be processed and the enhanced defect signal.

[0047] In summary, the eddy current defect signal reconstruction and enhancement method based on second-order mutation constraints in the above embodiments of the present invention describes the changing trend of the eddy current detection signal using a first-order variation signal, describes the local curvature change and mutation degree using a second-order mutation signal, and adaptively fuses the strongly filtered first-order signal and the weakly filtered first-order signal using second-order mutation weights. This allows for strong filtering in smooth background regions to suppress noise and weak filtering in defect mutation regions to preserve details. Finally, the first-order adaptively filtered signal with second-order mutation constraints obtained by adaptive fusion is integrated and reconstructed to obtain an enhanced defect signal in which background noise is suppressed and the local defect response is preserved.

[0048] Example 2 The second embodiment of the present invention provides an eddy current defect signal reconstruction and enhancement system based on second-order mutation constraints, the system comprising: The acquisition module is used to acquire the eddy current detection signal to obtain a smoothed signal, and to obtain a first-order change signal based on the smoothed signal; The calculation module is used to calculate a second-order mutation signal based on the first-order change signal, construct a second-order mutation evaluation quantity based on the second-order mutation signal, construct a robust threshold based on the second-order mutation evaluation quantity, and construct a second-order mutation weight based on the second-order mutation evaluation quantity and the robust threshold. The fusion module is used to perform strong filtering and weak filtering on the first-order change signal to obtain a strongly filtered first-order signal and a weakly filtered first-order signal, and to perform adaptive fusion with the second-order mutation weight to obtain a first-order adaptive filtered signal with second-order mutation constraints. The reconstruction module is used to perform integral reconstruction on the first-order adaptive filter signal of the second-order mutation constraint to obtain an initial reconstruction signal, and then obtain an enhanced defect signal based on the initial reconstruction signal.

[0049] In summary, the eddy current defect signal reconstruction and enhancement system based on second-order mutation constraints in the above embodiments of the present invention describes the changing trend of the eddy current detection signal using a first-order variation signal, describes the local curvature change and mutation degree using a second-order mutation signal, and adaptively fuses the strongly filtered first-order signal and the weakly filtered first-order signal using second-order mutation weights. This allows for strong filtering in smooth background regions to suppress noise and weak filtering in defect mutation regions to preserve details. Finally, the first-order adaptively filtered signal with second-order mutation constraints obtained by adaptive fusion is integrated and reconstructed to obtain an enhanced defect signal in which background noise is suppressed and the local defect response is preserved.

[0050] Furthermore, embodiments of the present invention also provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the methods described above.

[0051] Furthermore, embodiments of the present invention also propose a data processing device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the methods described above.

[0052] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0053] More specific examples (a non-exhaustive list) of computer-readable media include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which programs can be printed, because programs can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0054] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0055] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0056] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A method for reconstructing and enhancing eddy current defect signals based on second-order catastrophe constraints, characterized in that, include: The eddy current detection signal is acquired to obtain a smoothed signal, and a first-order change signal is obtained based on the smoothed signal. A second-order mutation signal is calculated based on the first-order change signal, and a second-order mutation evaluation quantity is constructed based on the second-order mutation signal. A robust threshold is constructed based on the second-order mutation evaluation quantity, and a second-order mutation weight is constructed based on the second-order mutation evaluation quantity and the robust threshold. The first-order change signal is subjected to strong filtering and weak filtering respectively to obtain a strongly filtered first-order signal and a weakly filtered first-order signal, and then adaptively fused with the second-order mutation weight to obtain a first-order adaptive filtered signal with second-order mutation constraint. An initial reconstructed signal is obtained by integrating and reconstructing the first-order adaptive filter signal of the second-order mutation constraint, and then an enhanced defect signal is obtained based on the initial reconstructed signal.

2. The eddy current defect signal reconstruction and enhancement method based on second-order mutation constraint according to claim 1, characterized in that, In the step of integrally reconstructing the first-order adaptive filtered signal under the second-order abrupt change constraint to obtain the initial reconstructed signal, the functional expression of the integral reconstruction method is: ； In the formula, r0(n) represents the initial reconstructed signal; d1a( i ) represents the first-order adaptive filtering signal with second-order mutation constraint; Δt represents the sampling time interval between adjacent sampling points. i This represents the i-th discrete sampling point in the integral reconstruction process; n This indicates the sequence number of the sampling point to be reconstructed.

3. The eddy current defect signal reconstruction and enhancement method based on second-order mutation constraint according to claim 1, characterized in that, The steps for obtaining the enhanced defect signal based on the initial reconstructed signal include: The initial reconstructed signal is subjected to linear drift compensation, mean correction and amplitude scaling correction to obtain the enhanced defect signal; The formula for calculating the enhanced defect signal is as follows: r e (n) = K · [r0(n) l(n)] + μ; In the formula, r e (n) represents the enhanced defect signal; l(n) represents the integral drift compensation term, K represents the amplitude scale correction coefficient, μ represents the mean compensation term; r0(n) represents the initial reconstructed signal.

4. The eddy current defect signal reconstruction and enhancement method based on second-order mutation constraint according to claim 1, characterized in that, The formula for calculating the first-order adaptive filter signal under second-order mutation constraints is: d1a(n) = [1 w(n)]d1s(n) + w(n)d1m(n); In the formula, d1a(n) represents the first-order adaptive filtered signal with second-order mutation constraint; w(n) represents the second-order mutation weight; d1s(n) represents the first-order signal after strong filtering; and d1m(n) represents the first-order signal after weak filtering. The formula for calculating the first-order signal after strong filtering is: d1s(n) = F strong [d1(n)]; The formula for calculating the first-order signal after weak filtering is: d1m(n) = F mild [d1(n)]; In the formula, d1s(n) represents the first-order signal after strong filtering; d1m(n) represents the first-order signal after weak filtering; d1(n) represents the first-order changing signal; F strong [﹒] denotes a strong filtering operator; F mild [﹒ ] Weak filtering operator.

5. The eddy current defect signal reconstruction and enhancement method based on second-order mutation constraint according to claim 1, characterized in that, The formula for calculating the second-order mutation weight is: w(n) = c(n) / [c(n) + T c + ε]; In the formula, w(n) represents the second-order mutation weight; c(n) represents the second-order mutation evaluation value at the nth sampling point; T c The robust threshold is represented by ε, which is a non-zero small constant used to avoid zero denominators. The formula for calculating the robustness threshold is: T c = median(c) + λ · MAD(c); In the formula, T c λ represents the robust threshold; MAD(c) represents the median absolute deviation of the second-order mutation evaluation, where c represents the second-order mutation evaluation sequence, which is composed of the second-order mutation evaluations of each sampling point, c={ c(1), c(2),…, c(N)}; median(c) represents the median of all elements in the second-order mutation evaluation sequence c. The formula for calculating the second-order mutation evaluation metric is as follows: c(n) = M W {|d2(n)|}; In the formula, c(n) represents the second-order mutation evaluation metric at the nth sampling point; M W { } represents an operator that performs smoothing, energy, or statistical calculations on the input sequence within a local window of length W corresponding to the nth sampling point; W represents the length of the local window; d2(n) represents the second-order abrupt change signal at the nth sampling point; n represents the discrete sampling point index; where: the formula for calculating the second-order abrupt change signal is: d2(n) = [d1(n+1) d1(n 1)] / (2Δt); In the formula, d2(n) represents the second-order abrupt change signal at the nth sampling point; d1(n+1) represents the first-order change signal at the (n+1)th sampling point; d1(n 1) represents the nth The first-order change signal at one sampling point; Δt represents the sampling time interval between adjacent sampling points; n represents the discrete sampling point number.

6. The eddy current defect signal reconstruction and enhancement method based on second-order mutation constraint according to claim 1, characterized in that, The formula for calculating a first-order changing signal is: d1(n) = [r s (n+1) r s (n 1)] / (2Δt); In the formula, d1(n) represents the first-order changing signal; r s (n+1) represents the smoothed signal at the (n+1)th sampling point; r s (n 1) represents the nth The smoothed signal at one sampling point; Δt represents the sampling time interval between adjacent sampling points; n represents the discrete sampling point number.

7. The eddy current defect signal reconstruction and enhancement method based on second-order mutation constraint according to claim 1, characterized in that, The steps for obtaining a smoothed signal from an eddy current detection signal include: Acquire eddy current detection signals and perform preprocessing to obtain the signal to be processed; The signal to be processed is smoothed to obtain a smoothed signal; The preprocessing methods include: First, the background of the eddy current detection signal is estimated to obtain the background signal. Then, the eddy current detection signal is subtracted from the background signal to obtain the residual signal, which is the signal to be processed. The formula for calculating the smoothed signal is as follows: r s (n) = F smooth [r(n)]; In the formula, r s (n) represents a smoothed signal; F smooth [﹒ ] denotes the smoothing operator; r(n) denotes the signal to be processed, where r(n) = x(n). b(n) and x(n) represent the eddy current detection signal; b(n) represents the background signal.

8. A signal reconstruction and enhancement system for eddy current defects based on second-order mutation constraints, characterized in that, The system includes: The acquisition module is used to acquire the eddy current detection signal to obtain a smoothed signal, and to obtain a first-order change signal based on the smoothed signal; The calculation module is used to calculate a second-order mutation signal based on the first-order change signal, construct a second-order mutation evaluation quantity based on the second-order mutation signal, construct a robust threshold based on the second-order mutation evaluation quantity, and construct a second-order mutation weight based on the second-order mutation evaluation quantity and the robust threshold. The fusion module is used to perform strong filtering and weak filtering on the first-order change signal to obtain a strongly filtered first-order signal and a weakly filtered first-order signal, and to perform adaptive fusion with the second-order mutation weight to obtain a first-order adaptive filtered signal with second-order mutation constraints. The reconstruction module is used to perform integral reconstruction on the first-order adaptive filter signal of the second-order mutation constraint to obtain an initial reconstruction signal, and then obtain an enhanced defect signal based on the initial reconstruction signal.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the eddy current defect signal reconstruction and enhancement method based on second-order mutation constraints as described in any one of claims 1-7.

10. A data processing apparatus, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the eddy current defect signal reconstruction and enhancement method based on second-order mutation constraints as described in any one of claims 1-7.

Images