Feature enhancement-based steel wire rope magnetic flux leakage signal defect analysis method and system

Abstract

The invention relates to the technical field of steel wire rope defect identification, and provides a steel wire rope magnetic flux leakage signal defect analysis method and system based on feature enhancement, and the method comprises the steps: constructing a geometric constraint convolution kernel, carrying out the multi-layer spiral decoupling processing of a radial magnetic flux leakage signal and an axial magnetic flux leakage signal of a steel wire rope, and obtaining a purified magnetic flux leakage signal; carrying out coupling processing on a radial component and an axial component in the purified magnetic leakage signal, constructing a magnetic dipole model and carrying out classification processing to obtain a defect characteristic signal; performing magnetic shielding compensation and inversion processing on the defect characteristic signal to obtain a layered defect signal, and amplifying the layered defect signal based on the sectional area ratio of the steel wire rope to obtain an enhanced defect characteristic signal; and performing defect identification on the enhanced defect characteristic signal to obtain a steel wire rope defect analysis result. According to the invention, high-precision steel wire rope defect automatic identification and positioning are realized, and the defect analysis capability of magnetic flux leakage detection under a complex spiral structure is improved.

Description

Technical Field

[0001] This invention relates to the field of wire rope defect identification technology, and in particular to a method and system for analyzing wire rope leakage magnetic signals based on feature enhancement. Background Technology

[0002] Steel wire rope is a high-strength load-bearing device made of multiple strands of steel wire twisted in a specific helical structure. It is widely used in critical engineering fields such as lifting machinery, elevators, cableways, and bridge cables. During long-term use, steel wire ropes are subject to factors such as load fatigue, environmental corrosion, and wear, resulting in various defects such as wire breakage, wear, corrosion, and deformation. Magnetic flux leakage (MFL) testing, as a primary method of non-destructive testing for steel wire ropes, identifies and assesses defect conditions by detecting magnetic field leakage signals at the defects. However, the unique helical structure, layered construction, and complex electromagnetic environment of steel wire ropes present numerous challenges to MFL defect analysis. The key lies in effectively eliminating helical interference, accurately extracting defect features, and achieving precise identification of multi-level defects.

[0003] In existing technologies, the analysis of magnetic leakage signals in wire ropes mainly employs traditional filtering preprocessing and basic signal processing methods, achieving basic defect detection functions. However, existing methods do not adequately consider the inherent physical correlation mechanism between the complex helical structure of the wire rope and the characteristics of the magnetic leakage signal, making it difficult to organically integrate the objective electromagnetic field physical constraints with the actual geometric structural features of the wire rope. This results in the inability to achieve high-precision and high-reliability automatic identification of wire rope defect states. Summary of the Invention

[0004] In view of this, the present invention proposes a defect analysis method and system for wire rope leakage magnetic signal based on feature enhancement. This solves the problem that the existing technology does not adequately consider the inherent physical correlation mechanism between the complex spiral structure of the wire rope and the characteristics of the leakage magnetic signal, and it is difficult to organically integrate the objective electromagnetic field physical constraints with the actual geometric structural characteristics of the wire rope, resulting in the inability to achieve high-precision and high-reliability automatic identification of wire rope defect status.

[0005] The technical solution of this invention is implemented as follows: In a first aspect, this invention provides a method for defect analysis of wire rope leakage magnetic signals based on feature enhancement, comprising the following steps:

[0006] Radial and axial magnetic flux leakage signals of a steel wire rope are collected. A geometrically constrained convolution kernel is constructed based on the geometric parameters of the steel wire rope. The radial and axial magnetic flux leakage signals are then subjected to multi-layer helical decoupling processing using the geometrically constrained convolution kernel to obtain a purified magnetic flux leakage signal.

[0007] Based on Maxwell's equations, a magnetic field vector constraint loss function is established, and a two-component coupled attention mechanism is constructed to couple the radial and axial components in the purified leakage magnetic field signal to obtain coupled radial and coupled axial components. A magnetic dipole model is constructed, and the coupled radial and coupled axial components are classified to obtain defect feature signals.

[0008] A layered magnetoresistive network model of a steel wire rope is established. The defect feature signal is magnetically shielded and compensated using the steel wire rope layered magnetoresistive network model to obtain a compensated defect feature signal. The compensated defect feature signal is then inverted using a layered signal inversion network to obtain a layered defect signal. The layered defect signal is then amplified based on the cross-sectional area ratio of the steel wire rope to obtain an enhanced defect feature signal.

[0009] Defect identification is performed on the enhanced defect feature signals to obtain the wire rope defect analysis results.

[0010] Based on the above technical solutions, preferably, the step of collecting the radial and axial magnetic flux leakage signals of the wire rope, constructing a geometrically constrained convolution kernel based on the wire rope's geometric parameters, and using the geometrically constrained convolution kernel to perform multi-layer spiral decoupling processing on the radial and axial magnetic flux leakage signals to obtain a purified magnetic flux leakage signal includes:

[0011] The geometric parameters of the wire rope include lay length, number of strands, wire diameter, and helix angle. A geometrically constrained convolution kernel is constructed based on the lay length, number of strands, wire diameter, and helix angle. The kernel size of the geometrically constrained convolution kernel is adaptively adjusted according to the helix period.

[0012] The multi-layer spiral decoupling process includes outer spiral decoupling, inner core spiral decoupling, and wire spiral decoupling. The radial leakage magnetic signal and the axial leakage magnetic signal are respectively subjected to the outer spiral decoupling, inner core spiral decoupling, and wire spiral decoupling processes using the geometrically constrained convolution kernel to obtain the decoupled leakage magnetic signal.

[0013] Based on the above technical solutions, preferably, the acquisition of the purification leakage magnetic signal includes:

[0014] The outer strand helix period is determined based on the outer strand lay pitch of the wire rope, and the first kernel size for the outer strand helix decoupling process is set based on the outer strand helix period;

[0015] The inner core helix period is determined based on the inner core lay pitch of the wire rope, and the second kernel size for the inner core helix decoupling process is set based on the inner core helix period using the geometric constraint convolution kernel.

[0016] The spiral period of the steel wire is determined based on its own spiral angle, and the third kernel size for the decoupling process of the steel wire spiral is set according to the spiral period of the steel wire.

[0017] A permeability weight matrix is ​​established, and the decoupled leakage magnetic field signal is weighted using the permeability weight matrix to obtain the purified leakage magnetic field signal.

[0018] Based on the above technical solutions, preferably, the step of establishing a magnetic field vector constraint loss function based on Maxwell's equations, constructing a dual-component coupled attention mechanism, coupling the radial and axial components in the purified magnetic leakage signal to obtain coupled radial and coupled axial components, constructing a magnetic dipole model, and classifying the coupled radial and coupled axial components to obtain defect feature signals, including:

[0019] The magnetic field vector constraint loss function is established based on the divergence theorem and curl theorem of Maxwell's equations. The magnetic field vector constraint loss function includes divergence constraint terms and curl constraint terms.

[0020] The dual-component coupled attention mechanism includes a radial self-attention module and an axial cross-attention module. The radial self-attention module performs self-attention processing on the radial component of the purified magnetic flux leakage signal, and the axial cross-attention module performs cross-attention processing on the axial and radial components of the purified magnetic flux leakage signal.

[0021] A magnetic dipole model including dipole strength parameters and dipole orientation parameters is constructed. The coupled radial component and the coupled axial component are classified for defect type and material inhomogeneity using the dipole strength parameters and the dipole orientation parameters to obtain defect feature signals.

[0022] Based on the above technical solution, preferably, the step of classifying the coupled radial component and the coupled axial component for defect type and material inhomogeneity using the dipole strength parameter and the dipole direction parameter to obtain defect feature signals includes:

[0023] Establish boundary constraints for the magnetic field of the wire rope. Use the divergence constraint term to constrain the coupled radial component and the coupled axial component so that the coupled radial component and the coupled axial component satisfy the magnetic flux continuity condition. Use the curl constraint term to constrain the coupled radial component and the coupled axial component so that the coupled radial component and the coupled axial component satisfy the tangential continuity condition of the magnetic field on the surface of the wire rope.

[0024] Construct a radial component feature correlation matrix and an axial-radial cross feature correlation matrix; generate radial self-attention weights based on the radial component feature correlation matrix; and generate axial cross attention weights based on the axial-radial cross feature correlation matrix.

[0025] A mapping relationship between the physical characteristics of defects and dipole parameters is established. The dipole strength parameter is determined based on the difference in magnetic permeability of defects inside the wire rope. The dipole orientation parameter is determined based on the geometric orientation of defects in the spiral structure of the wire rope. A defect discrimination threshold and a material inhomogeneity discrimination threshold are set. Classification processing is performed based on the defect discrimination threshold and the material inhomogeneity discrimination threshold.

[0026] Based on the above technical solutions, preferably, the step of establishing a layered magnetoresistive network model of the wire rope, using the layered magnetoresistive network model to perform magnetic shielding compensation on the defect feature signal to obtain a compensated defect feature signal, using a layered signal inversion network to perform inversion processing on the compensated defect feature signal to obtain a layered defect signal, and amplifying the layered defect signal based on the cross-sectional area ratio of the wire rope to obtain an enhanced defect feature signal includes:

[0027] A layered magnetoresistive network model of the wire rope is established based on the layered structural characteristics of the wire rope. The layered magnetoresistive network model of the wire rope includes an outer strand magnetoresistive layer, an inner core magnetoresistive layer, and a wire magnetoresistive layer. The defect feature signal is then subjected to layered magnetic shielding compensation processing through the layered magnetoresistive network model of the wire rope to obtain the compensated defect feature signal.

[0028] A multi-level hierarchical signal inversion network is constructed. The compensation defect feature signal is processed by the hierarchical signal inversion network to obtain the defect signal at each level. The defect signal at each level is proportionally amplified according to the cross-sectional area ratio of each layer of the wire rope to obtain the enhanced defect feature signal.

[0029] Based on the above technical solutions, preferably, the step of performing defect identification on the enhanced defect feature signal to obtain the wire rope defect analysis result includes:

[0030] A wire rope defect feature extractor is constructed. The wire rope defect feature extractor is used to perform multi-dimensional feature extraction processing on the enhanced defect feature signal to obtain defect type features and defect geometric features. Based on the defect type features and defect geometric features, defect type identification is performed to obtain defect identification results.

[0031] A spatial coordinate mapping model for the wire rope is established. The defect identification results are then spatially located using the wire rope spatial coordinate mapping model to obtain the spatial location information of the defect. A quantitative evaluation system for the degree of defect is constructed. Based on the spatial location information of the defect and the quantitative evaluation system for the degree of defect, the defect analysis results of the wire rope are generated.

[0032] Secondly, the present invention also provides a defect analysis system for wire rope leakage magnetic signal based on feature enhancement, the system comprising:

[0033] The signal acquisition and helical decoupling module is used to acquire the radial and axial magnetic flux leakage signals of the wire rope, construct a geometrically constrained convolution kernel based on the geometric parameters of the wire rope, and use the geometrically constrained convolution kernel to perform multi-layer helical decoupling processing on the radial and axial magnetic flux leakage signals to obtain the purified magnetic flux leakage signal.

[0034] The dual-component coupling identification module is used to establish a magnetic field vector constraint loss function based on Maxwell's equations, construct a dual-component coupling attention mechanism, couple the radial and axial components in the purified leakage magnetic signal to obtain coupled radial and coupled axial components, construct a magnetic dipole model, classify the coupled radial and coupled axial components, and obtain defect feature signals.

[0035] The layered signal enhancement module is used to establish a layered magnetoresistive network model of the wire rope, use the layered magnetoresistive network model of the wire rope to perform magnetic shielding compensation on the defect feature signal to obtain a compensated defect feature signal, use the layered signal inversion network to perform inversion processing on the compensated defect feature signal to obtain a layered defect signal, and amplify the layered defect signal based on the cross-sectional area ratio of the wire rope to obtain an enhanced defect feature signal.

[0036] The defect analysis output module is used to identify defects in the enhanced defect feature signals and obtain the defect analysis results of the wire rope.

[0037] Thirdly, the present invention also provides an electronic device, comprising: at least one processor, at least one memory, a communication interface, and a bus;

[0038] The processor, memory, and communication interface communicate with each other through the bus. The memory stores program instructions that can be executed by the processor. The processor calls the program instructions to implement the steps of a feature-enhanced method for analyzing defects in wire rope leakage magnetic signals.

[0039] Fourthly, the present invention also provides a computer-readable storage medium storing computer instructions that enable a computer to perform steps such as those in a feature-enhanced method for analyzing defects in wire rope leakage magnetic signals.

[0040] The present invention provides a method and system for defect analysis of wire rope leakage magnetic signals based on feature enhancement, which has the following advantages over the prior art:

[0041] (1) By constructing a geometrically constrained convolution kernel, the spiral interference is effectively eliminated. The magnetic field vector constraint based on Maxwell's equations and the dual-component coupled attention mechanism are combined to make full use of the complementary information of radial and axial leakage magnetic signals. The magnetic shielding effect of the complex layered structure of the wire rope is solved by using a layered magnetoresistive network model and a signal inversion network. This achieves high-precision automatic identification and positioning of wire rope defects and improves the defect analysis capability of leakage magnetic detection under complex spiral structures.

[0042] (2) By constructing a deformable geometrically constrained convolution kernel based on the geometric parameters of the wire rope (lay pitch, number of strands, wire diameter, and helix angle), and setting adaptive kernel sizes for the three-layer structure of outer strand, inner core, and wire respectively, multi-layer helical decoupling is performed to eliminate the periodic interference signal generated by the complex helical structure of the wire rope. At the same time, the decoupling signal is weighted and optimized by combining the permeability weight matrix, which improves the purification quality of leakage magnetic signal.

[0043] (3) By establishing a magnetic field vector constraint loss function based on the divergence theorem and curl theorem of Maxwell's equations, and combining the dual-component coupled attention mechanism to perform deep feature fusion of the purified magnetic leakage signal, a magnetic dipole model is constructed to classify defects based on the mapping relationship between the physical characteristics of defects and the dipole parameters. This ensures the physical rationality of the signal processing process and improves the utilization efficiency of complementary information of radial and axial magnetic leakage signals and the reliability of defect feature identification. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 This is a flowchart of a method for analyzing the leakage magnetic field signal defects of steel wire ropes based on feature enhancement, according to the present invention. Detailed Implementation

[0046] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0047] Please see Figure 1 This invention provides a method for defect analysis of wire rope leakage magnetic flux signal based on feature enhancement, comprising the following steps:

[0048] Radial and axial magnetic flux leakage signals of a steel wire rope are collected. A geometrically constrained convolution kernel is constructed based on the geometric parameters of the steel wire rope. The radial and axial magnetic flux leakage signals are then subjected to multi-layer helical decoupling processing using the geometrically constrained convolution kernel to obtain a purified magnetic flux leakage signal.

[0049] Based on Maxwell's equations, a magnetic field vector constraint loss function is established, and a two-component coupled attention mechanism is constructed to couple the radial and axial components in the purified leakage magnetic field signal to obtain coupled radial and coupled axial components. A magnetic dipole model is constructed, and the coupled radial and coupled axial components are classified to obtain defect feature signals.

[0050] A layered magnetoresistive network model of a steel wire rope is established. The defect feature signal is magnetically shielded and compensated using the steel wire rope layered magnetoresistive network model to obtain a compensated defect feature signal. The compensated defect feature signal is then inverted using a layered signal inversion network to obtain a layered defect signal. The layered defect signal is then amplified based on the cross-sectional area ratio of the steel wire rope to obtain an enhanced defect feature signal.

[0051] Defect identification is performed on the enhanced defect feature signals to obtain the wire rope defect analysis results.

[0052] Specifically, this embodiment effectively eliminates helical interference by constructing a geometrically constrained convolution kernel, fully utilizes the complementary information of radial and axial leakage magnetic flux signals by combining magnetic field vector constraints based on Maxwell's equations and a dual-component coupled attention mechanism, and solves the magnetic shielding effect of the complex layered structure of the wire rope by using a hierarchical magnetoresistance network model and a signal inversion network. This achieves high-precision automatic identification and location of wire rope defects and improves the defect analysis capability of leakage magnetic flux detection under complex helical structures.

[0053] The radial and axial magnetic flux leakage signals of the steel wire rope are collected. A geometrically constrained convolution kernel is constructed based on the steel wire rope's geometric parameters. This kernel is then used to perform multi-layer spiral decoupling processing on the radial and axial magnetic flux leakage signals to obtain a purified magnetic flux leakage signal, including:

[0054] The geometric parameters of the wire rope include lay length, number of strands, wire diameter, and helix angle. A deformable convolution kernel is constructed based on the lay length, number of strands, wire diameter, and helix angle as the geometrically constrained convolution kernel. The kernel size of the geometrically constrained convolution kernel is adaptively adjusted according to the helix period.

[0055] The multi-layer spiral decoupling process includes outer spiral decoupling, inner core spiral decoupling, and wire spiral decoupling. The radial leakage magnetic signal and the axial leakage magnetic signal are respectively subjected to the outer spiral decoupling, inner core spiral decoupling, and wire spiral decoupling processes using the geometrically constrained convolution kernel to obtain the decoupled leakage magnetic signal.

[0056] The outer strand helix period is determined based on the outer strand lay pitch of the wire rope, and the first kernel size for the outer strand helix decoupling process is set based on the outer strand helix period;

[0057] The inner core helix period is determined based on the inner core lay pitch of the wire rope, and the second kernel size for the inner core helix decoupling process is set based on the inner core helix period using the geometric constraint convolution kernel.

[0058] The spiral period of the steel wire is determined based on its own spiral angle, and the third kernel size for the decoupling process of the steel wire spiral is set according to the spiral period of the steel wire.

[0059] A permeability weight matrix is ​​established, and the decoupled leakage magnetic field signal is weighted using the permeability weight matrix to obtain the purified leakage magnetic field signal.

[0060] In one specific embodiment, the geometrically constrained convolution kernel is constructed based on the geometric parameters of the wire rope, and the kernel function is calculated as follows:

[0061]

[0062] Among them, K geo (u,v) is the geometrically constrained convolution kernel function, W def (u,v) is a deformable weight matrix, adaptively adjusted based on the geometric parameters of the wire rope; u and v are the spatial coordinates of the convolution kernel, u is the abscissa, and v is the ordinate; (u c ,v c ) represents the coordinates of the convolution kernel center; σ spiral P represents the characteristic scale parameter of the spiral structure. eff For the effective spiral period; φ twist is the twist phase angle; exp(·) is the exponential function.

[0063] The adaptive adjustment formula for the kernel size of the geometrically constrained convolution kernel is as follows:

[0064]

[0065] in, α is the kernel size of the layer; layer The weighting coefficients for the spiral period of the th layer are: β is the spiral period of the th layer; layer The weighting coefficient for the diameter of the steel wire in the th layer is . γ is the diameter of the steel wire in the first layer; layer The adaptive weight coefficients for the _th layer; The twist pitch of the first layer is... The core diameter of the first layer; The spiral angle of the first layer.

[0066] The weighted calculation formula for the permeability weight matrix is ​​as follows:

[0067]

[0068] Among them, S clean (x,y) represents the magnetic flux leakage signal at position (x,y); L represents the total number of layers in the wire rope. The permeability weighting coefficient for the l-th layer; This represents the decoupled leakage magnetic signal of the l-th layer; Let l be the magnetic flux distribution function of the l-th layer; Let l be the relative permeability of the l-th layer. Let be the relative permeability of the k-th layer; Let l be the cross-sectional area of ​​the l-th layer. Let be the cross-sectional area of ​​the k-th layer.

[0069] Specifically, this embodiment constructs a deformable geometrically constrained convolution kernel based on the geometric parameters of the wire rope (lay pitch, number of strands, wire diameter, and helix angle), and sets adaptive kernel sizes for the three-layer structure of outer strand, inner core, and wire to perform multi-layer helical decoupling processing. This eliminates the periodic interference signal generated by the complex helical structure of the wire rope. At the same time, the decoupling signal is weighted and optimized by combining the permeability weight matrix, thereby improving the purification quality of the leakage magnetic signal.

[0070] The magnetic field vector constraint loss function is established based on Maxwell's equations, and a two-component coupled attention mechanism is constructed to couple the radial and axial components in the purified magnetic leakage signal to obtain coupled radial and coupled axial components. A magnetic dipole model is then constructed to classify the coupled radial and coupled axial components to obtain defect feature signals, including:

[0071] The magnetic field vector constraint loss function is established based on the divergence theorem and curl theorem of Maxwell's equations. The magnetic field vector constraint loss function includes divergence constraint terms and curl constraint terms.

[0072] The dual-component coupled attention mechanism includes a radial self-attention module and an axial cross-attention module. The radial self-attention module performs self-attention processing on the radial component of the purified magnetic flux leakage signal, and the axial cross-attention module performs cross-attention processing on the axial and radial components of the purified magnetic flux leakage signal.

[0073] A magnetic dipole model including dipole strength parameters and dipole orientation parameters is constructed. The coupled radial component and the coupled axial component are classified for defect type and material inhomogeneity using the dipole strength parameters and the dipole orientation parameters to obtain defect feature signals.

[0074] Establish boundary constraints for the magnetic field of the wire rope. Use the divergence constraint term to constrain the coupled radial component and the coupled axial component so that the coupled radial component and the coupled axial component satisfy the magnetic flux continuity condition. Use the curl constraint term to constrain the coupled radial component and the coupled axial component so that the coupled radial component and the coupled axial component satisfy the tangential continuity condition of the magnetic field on the surface of the wire rope.

[0075] Construct a radial component feature correlation matrix and an axial-radial cross feature correlation matrix; generate radial self-attention weights based on the radial component feature correlation matrix; and generate axial cross attention weights based on the axial-radial cross feature correlation matrix.

[0076] A mapping relationship between the physical characteristics of defects and dipole parameters is established. The dipole strength parameter is determined based on the difference in magnetic permeability of defects inside the wire rope. The dipole orientation parameter is determined based on the geometric orientation of defects in the spiral structure of the wire rope. A defect discrimination threshold and a material inhomogeneity discrimination threshold are set. Classification processing is performed based on the defect discrimination threshold and the material inhomogeneity discrimination threshold.

[0077] In one specific embodiment, the formula for calculating the magnetic field vector constraint loss function is:

[0078] L constraint =λ div ·L divergence +λ curl ·L curl +λ boundary ·L boundary ;

[0079]

[0080] Among them, L constraint λ is the loss function for magnetic field vector constraints. div L represents the weighting coefficient of the divergence constraint loss term. divergence λ is the divergence constraint loss term; curl L represents the weighting coefficient for the curl constraint loss term. curl λ is the curl constraint loss term. boundary L represents the weighting coefficients of the boundary constraint loss term. boundary For boundary constraint loss terms; N points This represents the total number of sampling points; Let be the coupled magnetic field strength vector at point i; Let be the vector of the coupled magnetic field strength at point i; Let i be the equivalent current density vector at point i. This is the gradient operator.

[0081] Specifically, this embodiment establishes a magnetic field vector constraint loss function based on the divergence theorem and curl theorem of Maxwell's equations, combines a dual-component coupled attention mechanism to perform deep feature fusion on the purified magnetic flux leakage signal, and constructs a magnetic dipole model to classify defects based on the mapping relationship between the physical characteristics of defects and dipole parameters. This ensures the physical rationality of the signal processing process, improves the utilization efficiency of complementary information of radial and axial magnetic flux leakage signals, and enhances the reliability of defect feature identification.

[0082] The process involves establishing a layered magnetoresistive network model for the wire rope, using this model to perform magnetic shielding compensation on the defect feature signal to obtain a compensated defect feature signal, inverting the compensated defect feature signal using a layered signal inversion network to obtain a layered defect signal, and amplifying the layered defect signal based on the cross-sectional area ratio of the wire rope to obtain an enhanced defect feature signal, including:

[0083] A layered magnetoresistive network model of the wire rope is established based on the layered structural characteristics of the wire rope. The layered magnetoresistive network model of the wire rope includes an outer strand magnetoresistive layer, an inner core magnetoresistive layer, and a wire magnetoresistive layer. The defect feature signal is then subjected to layered magnetic shielding compensation processing through the layered magnetoresistive network model to obtain the compensated defect feature signal.

[0084] In one specific embodiment, the outer strand magnetoresistive layer, the inner core magnetoresistive layer, and the steel wire magnetoresistive layer are constructed according to the magnetic permeability of the outer strand material, the magnetic permeability of the inner core material, and the magnetic permeability of the steel wire material, respectively, and a magnetic shielding attenuation coefficient between each magnetoresistive layer is established;

[0085] A layered magnetic shielding compensation function is established. The layered magnetic shielding compensation function is weighted based on the difference in magnetic resistance of each layer and the distance between layers. The layered magnetic shielding compensation function is used to perform layered compensation and enhancement on the signal components affected by magnetic shielding in the defect feature signal, thereby eliminating the signal attenuation caused by the layered structure of the wire rope and obtaining the compensated defect feature signal.

[0086] A multi-level hierarchical signal inversion network is constructed. The compensation defect feature signal is processed by the hierarchical signal inversion network to obtain the defect signal at each level. The defect signal at each level is proportionally amplified according to the cross-sectional area ratio of each layer of the wire rope to obtain the enhanced defect feature signal.

[0087] In one specific embodiment, the multi-level hierarchical signal inversion network includes an outer strand inversion sub-network, an inner core inversion sub-network, and a steel wire inversion sub-network. The outer strand inversion sub-network, the inner core inversion sub-network, and the steel wire inversion sub-network are used to perform outer strand layer signal inversion, inner core layer signal inversion, and steel wire layer signal inversion on the compensated defect feature signal, respectively.

[0088] Calculate the cross-sectional area ratios of the outer strand layer, the inner core layer, and the wire layer of the wire rope, establish a layered amplification weight matrix, and use the layered amplification weight matrix to perform differentiated amplification processing on the inverted outer strand layer signal, inner core layer signal, and wire layer signal. Then, weightedly fuse the amplified signals of each layer to obtain the enhanced defect feature signal.

[0089] Specifically, this embodiment establishes a magnetoresistive network model (outer strand, inner core, and steel wire three-layer magnetoresistive layer) based on the layered structure characteristics of the wire rope, and constructs a layered magnetic shielding compensation function using the differences in magnetoresistive properties of each layer and the distance between layers to perform magnetic shielding compensation for defect feature signals. Combined with a multi-level layered signal inversion network (outer strand, inner core, and steel wire three-layered inversion sub-networks) and a layered amplification weight matrix based on the cross-sectional area ratio, differential signal enhancement processing is performed. This solves the magnetic field shielding and signal attenuation problems caused by the complex layered structure of the wire rope, and improves the recovery accuracy of defect signals at different levels and the detection capability of small defects.

[0090] The defect identification of the enhanced defect feature signal to obtain the wire rope defect analysis results includes:

[0091] A wire rope defect feature extractor is constructed. The wire rope defect feature extractor is used to perform multi-dimensional feature extraction processing on the enhanced defect feature signal to obtain defect type features and defect geometric features. Based on the defect type features and defect geometric features, defect type identification is performed to obtain defect identification results.

[0092] In one specific embodiment, the wire rope defect feature extractor includes a morphological feature extraction module, a frequency domain feature extraction module, and a statistical feature extraction module. The morphological feature extraction module extracts the signal amplitude, signal width, and signal shape features of the enhanced defect feature signal. The frequency domain feature extraction module extracts the frequency distribution features and harmonic features of the enhanced defect feature signal. The statistical feature extraction module extracts the mean, variance, and kurtosis features of the enhanced defect feature signal.

[0093] A knowledge base for wire rope defect types is established, which includes wire breakage defect patterns, wear defect patterns, corrosion defect patterns, and deformation defect patterns. The extracted multidimensional features are matched with the knowledge base for wire rope defect types to determine the defect type and calculate the matching confidence level, thereby obtaining the defect identification result.

[0094] A spatial coordinate mapping model for the wire rope is established. The defect identification results are then spatially located using the wire rope spatial coordinate mapping model to obtain the spatial location information of the defect. A quantitative evaluation system for the degree of defect is constructed. Based on the spatial location information of the defect and the quantitative evaluation system for the degree of defect, the defect analysis results of the wire rope are generated.

[0095] In one specific embodiment, a spatial coordinate mapping model of the wire rope is established based on the helical geometric parameters of the wire rope and the spatial arrangement parameters of the detection sensors. The spatial coordinate mapping model of the wire rope establishes the correspondence between the signal detection position and the three-dimensional spatial coordinates of the wire rope. The defect identification result is mapped into three-dimensional spatial coordinates including radial position, axial position and circumferential position using the correspondence.

[0096] The defect severity quantitative assessment system establishes a defect severity grading standard based on defect signal strength, defect impact range, and wire rope load-bearing capacity loss. According to the defect severity grading standard, the safety level of the identified defects is assessed. Combined with the defect spatial location information, a comprehensive analysis report containing defect type, defect location, defect severity, and safety rating is generated as the result of the wire rope defect analysis.

[0097] Specifically, this embodiment constructs a wire rope defect feature extractor (morphological, frequency domain, and statistical feature modules) to extract multidimensional features from enhanced defect feature signals. It combines a wire rope defect type knowledge base (broken wire, wear, corrosion, and deformation patterns) for pattern matching and confidence calculation. Based on the wire rope spatial coordinate mapping model, it achieves three-dimensional spatial positioning (radial, axial, and circumferential positions). Through a defect severity quantification assessment system, it establishes severity grading standards and safety level assessment mechanisms, generating a comprehensive analysis report that includes defect type, location, severity, and safety rating. This achieves comprehensive identification, spatial positioning, and safety assessment of wire rope defects.

[0098] This invention also provides a feature-enhanced wire rope leakage magnetic signal defect analysis system, the system comprising:

[0099] The signal acquisition and helical decoupling module is used to acquire the radial and axial magnetic flux leakage signals of the wire rope, construct a geometrically constrained convolution kernel based on the geometric parameters of the wire rope, and use the geometrically constrained convolution kernel to perform multi-layer helical decoupling processing on the radial and axial magnetic flux leakage signals to obtain the purified magnetic flux leakage signal.

[0100] The dual-component coupling identification module is used to establish a magnetic field vector constraint loss function based on Maxwell's equations, construct a dual-component coupling attention mechanism, couple the radial and axial components in the purified leakage magnetic signal to obtain coupled radial and coupled axial components, construct a magnetic dipole model, classify the coupled radial and coupled axial components, and obtain defect feature signals.

[0101] The layered signal enhancement module is used to establish a layered magnetoresistive network model of the wire rope, use the layered magnetoresistive network model of the wire rope to perform magnetic shielding compensation on the defect feature signal to obtain a compensated defect feature signal, use the layered signal inversion network to perform inversion processing on the compensated defect feature signal to obtain a layered defect signal, and amplify the layered defect signal based on the cross-sectional area ratio of the wire rope to obtain an enhanced defect feature signal.

[0102] The defect analysis output module is used to identify defects in the enhanced defect feature signals and obtain the defect analysis results of the wire rope.

[0103] Specifically, this embodiment of a feature-enhanced wire rope magnetic flux leakage signal defect analysis system constructs a complete wire rope defect analysis technology chain through four collaborative functional modules. Among them: the signal acquisition and helical decoupling module is used to eliminate interference from complex helical structures; the dual-component coupling identification module is used to achieve deep fusion of radial and axial signals and accurate defect feature extraction based on physical constraints; the layered signal enhancement module is used to solve the magnetic shielding problem of layered structures and enhance the signal of minute defects; and the defect analysis output module is used to realize comprehensive defect identification and evaluation. The system as a whole forms a fully automated processing system from raw signals to analysis results, improving the accuracy of wire rope magnetic flux leakage detection and its engineering application value.

[0104] The present invention also discloses an electronic device, comprising: at least one processor, at least one memory communication interface and bus; wherein the processor, memory and communication interface communicate with each other through the bus; the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to implement a feature-enhanced method for analyzing defects in wire rope leakage magnetic signals.

[0105] This invention also discloses a computer-readable storage medium storing computer instructions that enable the computer to implement all or part of the steps of the feature-enhanced method for analyzing magnetic leakage signals in steel wire ropes, as described in this embodiment of the invention. The storage medium includes various media capable of storing program code, such as a USB flash drive, portable hard drive, read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0106] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for analyzing defects in magnetic flux leakage signals of steel wire ropes based on feature enhancement, characterized in that, Includes the following steps: Radial and axial magnetic flux leakage signals of a steel wire rope are collected. A geometrically constrained convolution kernel is constructed based on the geometric parameters of the steel wire rope. The radial and axial magnetic flux leakage signals are then subjected to multi-layer helical decoupling processing using the geometrically constrained convolution kernel to obtain a purified magnetic flux leakage signal. Based on Maxwell's equations, a magnetic field vector constraint loss function is established, and a two-component coupled attention mechanism is constructed to couple the radial and axial components in the purified leakage magnetic field signal to obtain coupled radial and coupled axial components. A magnetic dipole model is constructed, and the coupled radial and coupled axial components are classified to obtain defect feature signals. A layered magnetoresistive network model of a steel wire rope is established. The defect feature signal is magnetically shielded and compensated using the steel wire rope layered magnetoresistive network model to obtain a compensated defect feature signal. The compensated defect feature signal is then inverted using a layered signal inversion network to obtain a layered defect signal. The layered defect signal is then amplified based on the cross-sectional area ratio of the steel wire rope to obtain an enhanced defect feature signal. Defect identification is performed on the enhanced defect feature signals to obtain the wire rope defect analysis results.

2. The method for analyzing defects in wire rope leakage magnetic flux signals based on feature enhancement as described in claim 1, characterized in that, The radial and axial magnetic flux leakage signals of the steel wire rope are collected. A geometrically constrained convolution kernel is constructed based on the steel wire rope's geometric parameters. This kernel is then used to perform multi-layer spiral decoupling processing on the radial and axial magnetic flux leakage signals to obtain a purified magnetic flux leakage signal, including: The geometric parameters of the wire rope include lay length, number of strands, wire diameter, and helix angle. A geometrically constrained convolution kernel is constructed based on the lay length, number of strands, wire diameter, and helix angle. The kernel size of the geometrically constrained convolution kernel is adaptively adjusted according to the helix period. The multi-layer spiral decoupling process includes outer spiral decoupling, inner core spiral decoupling, and wire spiral decoupling. The radial leakage magnetic signal and the axial leakage magnetic signal are respectively subjected to the outer spiral decoupling, inner core spiral decoupling, and wire spiral decoupling processes using the geometrically constrained convolution kernel to obtain the decoupled leakage magnetic signal.

3. The method for analyzing defects in wire rope leakage magnetic flux signals based on feature enhancement as described in claim 2, characterized in that, The acquisition of the purified leakage magnetic field signal includes: The outer strand helix period is determined based on the outer strand lay pitch of the wire rope, and the first kernel size for the outer strand helix decoupling process is set based on the outer strand helix period; The inner core helix period is determined based on the inner core lay pitch of the wire rope, and the second kernel size for the inner core helix decoupling process is set based on the inner core helix period using the geometric constraint convolution kernel. The spiral period of the steel wire is determined based on its own spiral angle, and the third kernel size for the decoupling process of the steel wire spiral is set according to the spiral period of the steel wire. A permeability weight matrix is ​​established, and the decoupled leakage magnetic field signal is weighted using the permeability weight matrix to obtain the purified leakage magnetic field signal.

4. The method for analyzing the leakage magnetic field signal defect of wire rope based on feature enhancement as described in claim 1, characterized in that, The magnetic field vector constraint loss function is established based on Maxwell's equations, and a two-component coupled attention mechanism is constructed to couple the radial and axial components in the purified magnetic leakage signal to obtain coupled radial and coupled axial components. A magnetic dipole model is then constructed to classify the coupled radial and coupled axial components to obtain defect feature signals, including: The magnetic field vector constraint loss function is established based on the divergence theorem and curl theorem of Maxwell's equations. The magnetic field vector constraint loss function includes divergence constraint terms and curl constraint terms. The dual-component coupled attention mechanism includes a radial self-attention module and an axial cross-attention module. The radial self-attention module performs self-attention processing on the radial component of the purified magnetic flux leakage signal, and the axial cross-attention module performs cross-attention processing on the axial and radial components of the purified magnetic flux leakage signal. A magnetic dipole model including dipole strength parameters and dipole orientation parameters is constructed. The coupled radial component and the coupled axial component are classified for defect type and material inhomogeneity using the dipole strength parameters and the dipole orientation parameters to obtain defect feature signals.

5. The method for analyzing defects in wire rope leakage magnetic flux signals based on feature enhancement as described in claim 4, characterized in that, The defect characteristic signal is obtained by classifying the coupled radial component and the coupled axial component for defect type and material inhomogeneity using the dipole strength parameter and the dipole direction parameter. Establish boundary constraints for the magnetic field of the wire rope. Use the divergence constraint term to constrain the coupled radial component and the coupled axial component so that the coupled radial component and the coupled axial component satisfy the magnetic flux continuity condition. Use the curl constraint term to constrain the coupled radial component and the coupled axial component so that the coupled radial component and the coupled axial component satisfy the tangential continuity condition of the magnetic field on the surface of the wire rope. Construct a radial component feature correlation matrix and an axial-radial cross feature correlation matrix; generate radial self-attention weights based on the radial component feature correlation matrix; and generate axial cross attention weights based on the axial-radial cross feature correlation matrix. A mapping relationship between the physical characteristics of defects and dipole parameters is established. The dipole strength parameter is determined based on the difference in magnetic permeability of defects inside the wire rope. The dipole orientation parameter is determined based on the geometric orientation of defects in the spiral structure of the wire rope. A defect discrimination threshold and a material inhomogeneity discrimination threshold are set. Classification processing is performed based on the defect discrimination threshold and the material inhomogeneity discrimination threshold.

6. The method for analyzing defects in wire rope leakage magnetic signals based on feature enhancement as described in claim 1, characterized in that, The process involves establishing a layered magnetoresistive network model for the wire rope, using this model to perform magnetic shielding compensation on the defect feature signal to obtain a compensated defect feature signal, inverting the compensated defect feature signal using a layered signal inversion network to obtain a layered defect signal, and amplifying the layered defect signal based on the cross-sectional area ratio of the wire rope to obtain an enhanced defect feature signal, including: A layered magnetoresistive network model of the wire rope is established based on the layered structural characteristics of the wire rope. The layered magnetoresistive network model of the wire rope includes an outer strand magnetoresistive layer, an inner core magnetoresistive layer, and a wire magnetoresistive layer. The defect feature signal is then subjected to layered magnetic shielding compensation processing through the layered magnetoresistive network model of the wire rope to obtain the compensated defect feature signal. A multi-level hierarchical signal inversion network is constructed. The compensation defect feature signal is processed by the hierarchical signal inversion network to obtain the defect signal at each level. The defect signal at each level is proportionally amplified according to the cross-sectional area ratio of each layer of the wire rope to obtain the enhanced defect feature signal.

7. The method for analyzing defects in wire rope leakage magnetic flux signals based on feature enhancement as described in claim 1, characterized in that, The defect identification of the enhanced defect feature signal to obtain the wire rope defect analysis results includes: A wire rope defect feature extractor is constructed. The wire rope defect feature extractor is used to perform multi-dimensional feature extraction processing on the enhanced defect feature signal to obtain defect type features and defect geometric features. Based on the defect type features and defect geometric features, defect type identification is performed to obtain defect identification results. A spatial coordinate mapping model for the wire rope is established. The defect identification results are then spatially located using the wire rope spatial coordinate mapping model to obtain the spatial location information of the defect. A quantitative evaluation system for the degree of defect is constructed. Based on the spatial location information of the defect and the quantitative evaluation system for the degree of defect, the defect analysis results of the wire rope are generated.

8. A feature-enhanced wire rope leakage magnetic signal defect analysis system, used to execute the feature-enhanced wire rope leakage magnetic signal defect analysis method as described in any one of claims 1-7, characterized in that, The system includes: The signal acquisition and helical decoupling module is used to acquire the radial and axial magnetic flux leakage signals of the wire rope, construct a geometrically constrained convolution kernel based on the geometric parameters of the wire rope, and use the geometrically constrained convolution kernel to perform multi-layer helical decoupling processing on the radial and axial magnetic flux leakage signals to obtain the purified magnetic flux leakage signal. The dual-component coupling identification module is used to establish a magnetic field vector constraint loss function based on Maxwell's equations, construct a dual-component coupling attention mechanism, couple the radial and axial components in the purified leakage magnetic signal to obtain coupled radial and coupled axial components, construct a magnetic dipole model, classify the coupled radial and coupled axial components, and obtain defect feature signals. The layered signal enhancement module is used to establish a layered magnetoresistive network model of the wire rope, use the layered magnetoresistive network model of the wire rope to perform magnetic shielding compensation on the defect feature signal to obtain a compensated defect feature signal, use the layered signal inversion network to perform inversion processing on the compensated defect feature signal to obtain a layered defect signal, and amplify the layered defect signal based on the cross-sectional area ratio of the wire rope to obtain an enhanced defect feature signal. The defect analysis output module is used to identify defects in the enhanced defect feature signals and obtain the defect analysis results of the wire rope.

9. An electronic device, characterized in that, include: At least one processor, at least one memory, a communication interface, and a bus; The processor, memory, and communication interface communicate with each other through the bus. The memory stores program instructions that can be executed by the processor. The processor calls the program instructions to implement the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause the computer to perform the method as described in any one of claims 1 to 7.

Images