Ship plate identification processing method and system

Abstract

The invention relates to the technical field of plate identification, in particular to a ship plate identification processing method and system, and the method comprises the steps: matrix characteristic data collection: employing a dual-frequency eddy current detection probe to scan a ship plate with an anti-corrosion coating, and obtaining the conductivity abnormal data of a matrix material under the coating; equivalent defect model generation: generating an equivalent matrix defect model representing the position and depth of the matrix defect according to the spatial distribution and amplitude change of the conductivity abnormal data; and energy parameter optimization machining is conducted, specifically, based on defect depth data in the equivalent matrix defect model, the output power and the focus position of a laser cutting machine are adjusted in real time, and self-adaptive machining is executed. By introducing a double-frequency eddy current detection and conductivity anomaly modeling method, the problem of shielding interference of an anti-corrosion coating on a traditional detection means is effectively solved, high-sensitivity detection of defects of a base material under the coating can be achieved, the processing quality consistency and the operation efficiency are effectively improved, and the method has high engineering practicability and popularization value.

Description

Technical Field

[0001] This invention relates to the field of plate identification technology, and in particular to a method and system for identifying and processing ship plates. Background Technology

[0002] During shipbuilding and repair, a large amount of steel plates are exposed to the marine environment for extended periods, and their surfaces are typically covered with thick anti-corrosion coatings. This presents numerous challenges to the quality inspection and precise processing of the internal base materials. Traditional plate identification and processing mainly rely on manual visual inspection and fixed-power laser cutting technology, which makes it difficult to identify potential material defects beneath the coating, such as corrosion pits, cracks, or impurity inclusions. If these invisible defects are not identified and addressed before processing, they will seriously affect the mechanical properties and service life of the hull structure, and may even pose safety hazards.

[0003] Existing detection technologies, such as ultrasonic or single-frequency eddy current methods, suffer from low signal-to-noise ratios and inaccurate defect localization when faced with thick coating interference, failing to provide high-resolution information on defect depth distribution. Meanwhile, existing laser cutting equipment typically employs fixed-parameter processing methods, unable to dynamically adjust energy output based on internal material defects, easily leading to overheating or incomplete cutting in defect areas. Therefore, there is an urgent need for a method that can penetrate coatings to obtain substrate feature data and achieve adaptive energy parameter control based on defect modeling, in order to improve the accuracy, stability, and intelligence of ship plate processing. Summary of the Invention

[0004] This invention provides a method and system for identifying and processing ship plate materials.

[0005] A method for identifying and processing ship plate materials includes the following steps:

[0006] S1, Matrix Feature Data Acquisition: A dual-frequency eddy current detection probe is used to scan the ship plate with anti-corrosion coating to obtain abnormal conductivity data of the matrix material under the coating;

[0007] S2, Equivalent defect model generation: Based on the spatial distribution and amplitude variation of the conductivity anomaly data, an equivalent matrix defect model characterizing the location and depth of matrix defects is generated;

[0008] S3, Energy Parameter Optimization Processing: Based on the defect depth data in the equivalent matrix defect model, the output power and focal position of the laser cutting machine are adjusted in real time to perform adaptive processing.

[0009] Optionally, S1 includes:

[0010] S11, Dual-frequency excitation signal generation: The dual-frequency eddy current excitation method is adopted to control the output of sinusoidal excitation current of fixed frequency by the low-frequency and high-frequency coils respectively, and the current amplitude ratio is set.

[0011] S12, Spiral Path Scanning: Drives the dual-frequency eddy current detection probe to scan the surface of the plate along a spiral trajectory, with the scanning interval set to be less than one-third of the minimum coil diameter;

[0012] S13, Impedance signal acquisition: During the scanning process of the dual-frequency eddy current detection probe, the complex impedance signals of the low-frequency and high-frequency coils are acquired in real time, and their resistance and reactance components are measured.

[0013] S14, Conductivity decoupling calculation: Extract the reactance increment and introduce the coating coupling coefficient for decoupling calculation to obtain the relative conductivity deviation value reflecting the degree of substrate defects;

[0014] S15, Abnormal Data Extraction: Based on the decoupling calculation results, set an empirical threshold to determine defects. When the relative conductivity deviation value exceeds the set empirical threshold, record the corresponding location and its relative conductivity deviation value.

[0015] Optionally, S13 includes:

[0016] S131, Impedance Response Measurement: During the scanning process of the dual-frequency eddy current detection probe, the complex impedance of the low-frequency coil is measured in real time, including the resistive component and the reactance component.

[0017] S132, High-frequency characteristic acquisition: Synchronously acquire the complex impedance information of the high-frequency coil and record the reactance changes.

[0018] Optionally, S14 includes:

[0019] S141, Reactance increment calculation: Based on the currently measured reactance value and the baseline value of the defect-free area, calculate the increments of high-frequency reactance and low-frequency reactance respectively, and extract abnormal change characteristics;

[0020] S142, Coating Interference Compensation: Introduce a coating thickness coupling coefficient and use low-frequency reactance variation as a compensation term;

[0021] S143, Deviation value calculation: Combine the probe calibration coefficient to correct the decoupling result, and finally calculate the relative conductivity deviation value.

[0022] Optionally, S2 includes:

[0023] S21, Spatial Interpolation Reconstruction: Weighted interpolation is performed on the conductivity deviation data obtained from the scan to construct a global continuous distribution function;

[0024] S22, Depth Inversion Calculation: Based on the magnitude of the conductivity deviation function obtained by interpolation, the depth distribution of defects in the thickness direction of the plate is further estimated;

[0025] S23, Equivalent crack width calculation: Combine the spatial change rate of deviation and depth information to calculate the equivalent crack width in the transverse direction;

[0026] S24, Voxel Model Generation: Divide the plate into small voxel meshes, assign values ​​based on depth and width information, and construct a three-dimensional crack morphology model.

[0027] Optionally, S22 includes:

[0028] S221, Deviation Amplitude Analysis: Based on the interpolated conductivity deviation function, obtain the absolute value of the deviation at each location point;

[0029] S222, Nonlinear Depth Mapping: Combining the maximum detection depth and the sensitivity coefficient determined by calibration tests, the deviation value is converted into the defect depth through an exponential mapping relationship.

[0030] Optionally, S23 includes:

[0031] S231, Spatial gradient extraction: Calculate the rate of change of the conductivity deviation function in space and obtain the gradient value at each location;

[0032] S232, Crack width estimation: Combine the gradient value with the defect depth at the corresponding location and introduce the material property coefficient to estimate the equivalent crack width at that location.

[0033] Optionally, S24 includes:

[0034] S241, Spatial Mesh Generation: Divide the area of ​​the ship plate to be inspected into a uniform voxel mesh in three-dimensional space;

[0035] S242, Voxel assignment calculation: Based on the position of each voxel in the thickness direction, compare it with the defect depth and width at the corresponding position, and assign a value to each voxel according to the set rules;

[0036] S243, Crack Region Labeling: Determine whether the value of each voxel is greater than a set threshold, mark the voxels that meet the conditions as crack regions, and construct a three-dimensional crack morphology model.

[0037] Optionally, S3 includes:

[0038] S31, Dynamic Power Compensation: Calculates the laser power compensation value based on the defect depth at the current processing point;

[0039] S32, Focus position shifted downward: Adjust the focus position according to the defect depth and crack width, and calculate the downward shift amount;

[0040] S33, Scanning speed optimization: Adjust the laser scanning speed according to crack gradient information;

[0041] S34, Real-time processing execution: Based on the above calculation results, the processing parameters of the laser cutting machine are controlled in real time;

[0042] S35, Online Quality Monitoring and Self-Learning Adjustment: Based on infrared thermal imaging monitoring of the cutting seam temperature, the power compensation coefficient is updated when the conditions are met.

[0043] A ship plate identification and processing system, used to implement the above-mentioned ship plate identification and processing method, includes the following modules:

[0044] Substrate feature data acquisition module: Scans the surface of ship plate with anti-corrosion coating using a dual-frequency eddy current detection probe, collects abnormal conductivity data of the substrate material under the coating, and outputs a dataset for subsequent modeling;

[0045] Equivalent defect model generation module: Based on the spatial distribution and amplitude variation of the conductivity anomaly data, construct an equivalent matrix defect model characterizing the location and depth of matrix defects, and output the spatial location parameters and depth feature parameters of the defects;

[0046] Energy parameter optimization processing module: Based on the defect depth data in the equivalent matrix defect model, dynamically calculate the output power and focal position of the laser cutting equipment to achieve adaptive processing for different defect areas.

[0047] The beneficial effects of this invention are:

[0048] This invention effectively overcomes the problem of shielding and interfering with traditional detection methods by introducing dual-frequency eddy current detection and conductivity anomaly modeling methods, enabling highly sensitive detection of defects in the substrate material under the coating. The equivalent defect model generated by combining spatial interpolation and nonlinear inversion algorithms not only has millimeter-level defect depth resolution, but also outputs crack width and location features readable by processing equipment, providing precise support for subsequent process parameter control and significantly improving the intelligence and reliability of ship plate identification.

[0049] This invention constructs a voxel-based adaptive laser energy control strategy. By adjusting the laser power, focal position, and scanning speed in real time, it achieves differentiated processing of different defect areas, avoiding material overheating or incomplete cutting caused by uneven energy input. In addition, combined with infrared thermal imaging monitoring and a power compensation factor self-learning mechanism, the system has online quality feedback and dynamic optimization capabilities, effectively improving processing quality consistency and work efficiency, and has strong engineering practicality and promotion value. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in this 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 for this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0051] Figure 1 This is a flowchart of a method according to an embodiment of the present invention;

[0052] Figure 2 This is a system block diagram of an embodiment of the present invention. Detailed Implementation

[0053] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. It should also be noted that, to make the embodiments more comprehensive, the following embodiments are the best and preferred embodiments, and those skilled in the art can use other alternative methods to implement some well-known technologies; moreover, the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.

[0054] It should be noted that the use of terms such as "an embodiment," "an embodiment," "an exemplary embodiment," and "some embodiments" in the specification indicates that the described embodiment may include a specific feature, structure, or characteristic, but not every embodiment necessarily includes that specific feature, structure, or characteristic. Furthermore, when a specific feature, structure, or characteristic is described in connection with an embodiment, implementing such a feature, structure, or characteristic in conjunction with other embodiments (whether explicitly described or not) should be within the knowledge of those skilled in the art.

[0055] Generally, terms can be understood at least partly from their use in context. For example, depending at least partly on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in a singular sense, or a combination of features, structures, or characteristics in a plural sense. Additionally, the term "based on" can be understood not necessarily to convey an exclusive set of factors, but rather, alternatively, depending at least partly on the context, to allow for the presence of other factors that are not necessarily explicitly described.

[0056] like Figure 1 As shown, a method for identifying and processing ship plate materials includes the following steps:

[0057] S1, Matrix Feature Data Acquisition: A dual-frequency eddy current detection probe is used to scan the ship plate with anti-corrosion coating to obtain abnormal conductivity data of the matrix material under the coating;

[0058] S2, Equivalent defect model generation: Based on the spatial distribution and amplitude variation of the conductivity anomaly data, an equivalent matrix defect model characterizing the location and depth of matrix defects is generated;

[0059] S3, Energy Parameter Optimization Processing: Based on the defect depth data in the equivalent matrix defect model, the output power and focal position of the laser cutting machine are adjusted in real time to perform adaptive processing.

[0060] S1 specifically includes:

[0061] S11, Dual-frequency excitation signal generation: In order to achieve penetration detection of defects at different depths in multi-layer structures, a dual-frequency eddy current excitation method is adopted. The low-frequency coil outputs a sinusoidal excitation current I1 with a frequency of f1 (f1 = 20kHz), and the high-frequency coil outputs a sinusoidal excitation current I2 with a frequency of f2 (f2 = 100kHz). In order to enhance the induction capability of the low frequency to the deep region and suppress the edge effect, the current amplitude relationship is set as I1 = 3I2, and I2 ≥ 0.5A.

[0062] Choosing a lower frequency enhances the penetration depth of electromagnetic waves in metals, enabling the detection of deep defects. According to the classical eddy current diffusion depth formula, the lower the frequency, the greater the detection depth. Therefore, f1 = 20kHz is suitable for penetrating anti-corrosion coatings and identifying deep damage. Higher frequencies enhance resolution and sensitivity in shallow areas, suitable for identifying small defects such as surface microcracks or coating erosion. 100kHz provides sufficient shallow resolution and is less susceptible to environmental noise interference. f2 = 100kHz is a stable and effective high-frequency selection point. To compensate for the weak sensing capability of low-frequency signals, increasing the excitation intensity of the low-frequency coil helps enhance its response to deep defects. At the same time, by setting the low-frequency current to three times that of the high-frequency current, edge effects and false detection rates can be effectively suppressed while maintaining the overall energy consumption balance of the system. The high-frequency excitation current cannot be too low, otherwise it will lead to insufficient excitation signal amplitude and decreased detection sensitivity. When the high-frequency current is not less than 0.5A, reliable detection capability for defects smaller than 0.2mm can be ensured, while avoiding probe heating or nonlinear response caused by excessive current.

[0063] S12, Spiral Path Scanning: To achieve full coverage scanning of the board surface and avoid duplication or omission, the detection probe is driven to move in a spiral trajectory. To ensure that the smallest identification unit of the high-frequency probe (such as a micro-crack) is not missed, the probe is driven to scan the board surface in a spiral trajectory. The spiral spacing d satisfies the following condition:

[0064]

[0065] Where d is the spiral scanning path spacing, Φ1 is the low-frequency coil diameter, and Φ2 is the high-frequency coil diameter, ensuring that the scanning interval of each turn of the probe is less than half of its minimum sensing radius, thereby improving spatial resolution;

[0066] set up The purpose is to ensure that the coverage overlap rate between the scanning paths is high enough, so as to ensure that the coverage area of ​​each scan is continuously overlapped (the smallest identification unit of the probe is usually smaller than the sensing radius of the coil. If the spacing is set too large, a blind zone may be generated between two scanning circles. By setting the spacing to one-third of the minimum diameter, it can be ensured that the sensing area of ​​the high-frequency coil overlaps by at least 50%, effectively avoiding missed detection) and improve the resolution of the scanning image (the "1 / 3" rule makes the sampling points of each circle densely cover the center and edge of the sensing area, improving the ability to distinguish small defects or heterogeneous structures, and is suitable for detecting defect features with high contrast or blurred boundaries).

[0067] S13, Impedance Signal Acquisition: During the probe spiral scanning process, the impedance response of the coil under low-frequency and high-frequency excitation is acquired respectively, specifically including:

[0068] (1) Low-frequency coil impedance Z1:

[0069] Z1 = R1 + jX1;

[0070] (2) High-frequency coil impedance Z2:

[0071] Z2 = R2 + jX2;

[0072] Where R1 and R2 are the real parts (resistance components) of the coil circuit impedance, reflecting energy loss, X1 and X2 are the imaginary parts (reactance components) of the induced impedance, reflecting changes in the magnetic field response, and j is the imaginary unit;

[0073] S14, Conductivity Decoupling Calculation: To eliminate the influence of the anti-corrosion coating on the reactance signal, the relative reactance increment is extracted and decoupled calculation is performed to obtain the relative conductivity deviation value δ of the matrix material, expressed as:

[0074]

[0075] ΔX1=X1-X 10 ;

[0076] ΔX² = X² - ​​X 20 ;

[0077] Where δ is the relative conductivity deviation (a defect quantification index), and X 10 X is the low-frequency reactance reference value for the defect-free region. 20α is the high-frequency reactance reference value of the defect-free area, α is the coating thickness coupling coefficient, which is used to adjust the compensation weight of low-frequency reactance to high-frequency response. α = 0.35. Within the common anti-corrosion coating thickness range, α = 0.35 is the most sufficient to offset high-frequency error. If the coating thickness coupling coefficient is not introduced, the change in surface coating thickness will be misjudged as a material defect, resulting in a large number of false detections. k is the probe calibration coefficient, which is used to compensate for systematic errors, including the systematic deviation of measurement results caused by factors such as probe position offset, coil winding capacitance difference, and circuit amplifier linear drift. k∈[0.8,1.2], based on the empirical range obtained from a large number of engineering measurements, to ensure that a high recognition accuracy can still be maintained under different plate thicknesses, materials and temperature conditions. ΔX1 is the low-frequency reactance increment of the defect-free area, and ΔX2 is the high-frequency reactance increment of the defect-free area.

[0078] S15, Abnormal Data Extraction: Based on the calculated relative conductivity deviation value δ, set an empirical threshold δ. th Defect identification, δ th =0.05. When |δ|≥0.05, it is determined that there is an abnormal conductivity at the scan point. The corresponding coordinate position (x,y) and its δ value are recorded as input data for subsequent defect modeling. th A value of 0.05 can distinguish the boundary between defects and natural fluctuations, effectively avoid background fluctuations, reduce the false detection rate, and is the smallest reliable boundary for distinguishing identifiable real defects from weak disturbances / artifacts. It can also maintain the stability of detection sensitivity under different coating thicknesses, material grades and field environmental interference, and has good robustness, making it easy to promote and apply.

[0079] S2 specifically includes:

[0080] S21, Spatial Interpolation Reconstruction: To construct a continuous conductivity deviation field from discrete scan data, a weighted interpolation method is used to reconstruct the two-dimensional coordinates (x, y) of all scan points. i ,y i and the corresponding conductivity deviation value δ i To reconstruct the function, a globally continuous distribution function δ(x,y) is constructed using an improved Shebert interpolation method, expressed as:

[0081]

[0082] Where, δ i Here, (x, y) represents the conductivity deviation value at the i-th scan point, and (x, y) represents the coordinates of the point to be interpolated. iε is the Euclidean distance between the interpolation point and the i-th scan point, with a value range of [0, 20]. The interpolation function needs to measure the spatial distance between the interpolation point and the surrounding known points. According to the actual application scenario of ship plate inspection, the probe scanning distance is generally about 1mm. Therefore, setting a distance upper limit of 20mm can ensure that enough neighboring points are covered for weighting, forming a reasonable local interpolation area. Too small a range may lead to insufficient data support for the interpolation points, affecting smoothness, while too large a range may introduce distant noise points, weakening the sensitivity to local anomalies. ε is the division-to-zero constant, ε = 0.001. During the interpolation process, the interpolation point may completely coincide with a known scan point. If the reciprocal type weighting function is used directly, it will lead to division-to-zero errors or numerical overflow. To avoid such calculation anomalies, a very small positive constant ε is set to play a role in numerical stability protection. This value is much smaller than the minimum scanning distance (≥1mm) and will not affect the weighting. The physical meaning and result distribution of the calculation: p is the distance decay exponent, p = 2.5. The distance decay exponent controls the decay rate of the weights by distance and is an important factor used to adjust locality in the interpolation algorithm. If the value is too small, the weights of distant points decay slowly, the interpolation function becomes overly smooth, and it may mask small defects. If the value is too large, the interpolation function becomes too dependent on the nearest points and is easily affected by noise. When p = 2.5, a balance is achieved between defect boundary sensitivity and overall smoothness. It is the optimal empirical value in scenarios with uneven plate thickness and complex coating coverage. n is the number of scanning points participating in the interpolation, n ≥ 5, with a recommended value of 8-20. Setting a minimum value of 5 ensures that the interpolation is valid and prevents the interpolation from failing due to insufficient points. The recommended range of 8-20 is optimized based on the defect size and scanning density of common plate areas. It can take into account both local detail capture and global trend preservation, while limiting the computational complexity, which is conducive to the real-time application of the algorithm in industrial fields.

[0083] S22, Depth Inversion Calculation: Based on the amplitude of the conductivity deviation function δ(x,y) obtained by interpolation, the depth distribution of defects in the thickness direction of the plate is further estimated. Considering the nonlinear relationship between conductivity perturbation and defect depth, an exponential inversion model is used to map |δ(x,y)| to a defect depth function, expressed as:

[0084]

[0085] Where δ(x,y) is the interpolated deviation value, and D(x,y) is the defect depth, with a value range of [0,3]. max It is the maximum detectable depth, D max=3mm, determined based on the upper limit of the penetration capability of 20kHz low-frequency eddy current excitation. In marine steel, the skin depth of 20kHz eddy current is about 2.8 to 3.2mm. Considering the probe sensitivity, system noise and the actual stability of the identification boundary, its maximum effective identification depth is reasonably limited to 3mm. This can cover common deep defect types and has good engineering applicability. γ is the depth sensitivity coefficient, used to control the response amplitude between the conductivity deviation value and the defect depth. This value is an empirical coefficient obtained by fitting a large number of calibration experiments to ensure that the model has a reasonable depth resolution capability in different amplitude ranges. γ = 0.12. When |δ| = 0.05 (the lower limit of defect judgment), the corresponding inversion depth is about 0.34mm. When |δ| = 0.12, the depth can reach about 1.9mm, which is about 63% of the maximum detection depth. It can take into account the sensitive identification of shallow micro-defects and the gradual approximation of deep defects, and avoid oversaturation or distortion in depth calculation.

[0086] S23, Equivalent Crack Width Calculation: After obtaining the depth information at each location, the transverse geometric characteristics of the defect are further evaluated. Based on the spatial gradient of the conductivity deviation function and the inverted depth, the equivalent defect width W(x,y) is calculated, expressed as:

[0087]

[0088] in, This is the spatial gradient of conductivity deviation, ranging from 0 to 1. It is determined based on the rate of change of the measured conductivity deviation field in a typical crack edge region. The conductivity change is steepest at the defect boundary, typically reaching 0.5–1.0% / mm. In smooth regions or noisy backgrounds, this gradient value approaches 0. Limiting it to the range of 0-1 not only covers typical defect characteristics but also effectively avoids overestimation caused by abnormal gradients. η is a material property constant. For the crack response characteristics of marine steel, η = 0.15. It is an empirical coefficient obtained by fitting the statistical relationship between crack size and electromagnetic response of various marine steel materials. This value represents the equivalent crack width response under unit conductivity gradient change. By comparing different defect types with the actual opening width distribution, it can effectively characterize the crack edge morphology in engineering applications. W(x,y) is the equivalent crack width, with a value range of 0-0.5. This range can cover micro-initial cracks, fatigue propagation cracks and typical pore boundaries, which is consistent with the actual characteristics of defects in ship structural steel plates and is convenient for subsequent crack morphology modeling and risk assessment.

[0089] S24, Voxel Model Generation: To achieve 3D visualization of defects, the plate material is divided into a detailed voxel mesh structure to construct a 3D crack morphology model, specifically including:

[0090] (1) For each voxel (x) v ,y v ,z v ), based on its position z in the thickness direction v With inversion depth D(x) v ,y v The system assigns conditional values ​​to simulate the characteristic of a defect gradually narrowing with depth, maximizing the surface width, gradually decreasing it inwards, and setting it to 0 after exceeding the defect depth. The specific rules are as follows:

[0091]

[0092] Where V(x) v ,y v ,z v ) is the voxel assignment, representing the equivalent crack width, z v It is the depth coordinate of the voxel in the thickness direction of the sheet material, x v ,y v It is a voxel plane coordinate;

[0093] (2) Based on the obtained voxel assignment V(x) v ,y v ,z v The process of constructing a three-dimensional crack geometry model includes:

[0094] 1) Crack voxel extraction: Set a crack presence threshold V th Voxels that meet the conditions are considered as crack regions, represented as:

[0095]

[0096] in, It is a collection of cracked voxels, V th It is the crack assignment threshold, V th =0.1mm, which is set based on the minimum spatial resolution of high-frequency eddy current detection and the actual crack opening size. This ensures that small cracks with actual structural significance can be identified, and effectively eliminates spurious responses caused by noise fluctuations or non-crack disturbances. It also matches the spatial step size of the voxel model, which facilitates subsequent 3D model reconstruction and isosurface extraction, ensuring the continuity and physical reality of the model structure.

[0097] 2) Crack boundary surface reconstruction: based on voxel value function V(x) v ,y v ,z v A three-dimensional crack geometry model is constructed using a three-dimensional isosurface extraction algorithm (such as Marching Cubes), represented as follows:

[0098] ∑={(x,y,z)|V(x,y,z)=Vth};

[0099] Where Σ is the three-dimensional crack geometry model, and V is the equivalent value. th The set of closed surfaces, V(x,y,z) is the voxel interpolation function, and linear interpolation reconstruction can also be performed based on the discrete voxel set.

[0100] S3 specifically includes:

[0101] S31, Dynamic Power Compensation: To dynamically adjust the laser cutting power according to the defect depth, improve cutting consistency, and avoid ablation, the system performs dynamic power compensation at each processing point (x... c ,y c A compensation mechanism is introduced at point ( ). By normalizing the defect depth and the maximum detection depth, and combining it with the surface crack width flag function in the voxel model, the standard power is corrected up and down, and the laser power compensation value P is calculated. comp , is represented as:

[0102]

[0103] Where P0 is the standard laser power in the defect-free area, its value is closely related to the plate thickness and the material's thermal conductivity, ensuring that the cutting depth reaches the full thickness of the plate while minimizing edge thermal impact. μ is the power compensation coefficient, μ = 0.035, which controls the adjustment range of laser power based on defect depth. It means that for every 1mm increase in defect depth, the power increases by 3.5%, ensuring effective penetration of deep cracks while preventing excessive power increase and ablation caused by shallow defects. D max It is the maximum detectable defect depth, D max =3mm, used as a normalization benchmark, to map the defect depth at any point to the 0–1 interval, which can cover most crack types in marine steel structure inspection, V(x c ,y c ,0) is the equivalent crack width of the corresponding surface. A positive value indicates the presence of a crack, and a zero value indicates the absence of a crack. It is used to determine whether power compensation is required. sgn(·) is the sign function. It takes 1 in the defect area and -1 in the no-defect area. Using this binary function can ensure that the control strategy is simple and efficient. At the same time, an overheat protection mechanism is introduced to adapt to the thermal management requirements of the boundary fluctuation area.

[0104] S32, Focus Position Downward Shift: To ensure that the laser energy can be accurately focused to the target depth in the defect area, the focus position needs to be dynamically downward adjusted. The focus position is adjusted according to the defect depth and crack width, and the downward shift Δf is calculated and expressed as:

[0105]

[0106] Where K is the material absorption coefficient, K = 0.25, characterizing the material's ability to absorb laser light. Marine steel is a medium-absorption material, and experimental results show that its laser energy transmission efficiency is moderate. A value of 0.25 is recommended to achieve appropriate compensation of the focal depth for the crack zone, preventing both insufficient cutting due to a shallow focal depth and ablation of the underlying structure due to an excessively deep focal depth. W th It is the crack width response threshold, W th =0.2mm. When the crack width is below this threshold, it is considered a "micro-crack" area. No focal position compensation is performed to avoid introducing unnecessary energy disturbance. For cracks in ship steel materials, this value corresponds to the critical warning size before fatigue propagation at the crack tip.

[0107] S33, Scanning Speed ​​Optimization: To improve the quality of the cutting boundary and reduce the risk of heat accumulation, the scanning speed needs to be adaptively adjusted according to the change amplitude of the crack edge. Spatial gradient is used as the judgment indicator; a larger gradient indicates a steeper boundary, requiring a reduction in scanning speed. The laser scanning speed v is adjusted based on the crack gradient information, expressed as:

[0108]

[0109] Where v0 is the standard scanning speed, ranging from 800 to 1200 mm / min, and is the recommended scanning speed for defect-free areas. Its value is determined comprehensively based on the plate thickness, laser type, and material type. For 3mm thick ship steel, a commonly used value of 800–1200 mm / min is used as the speed adjustment benchmark. This is the spatial gradient of the crack width, representing the geometric steepness of the crack edge. A larger gradient indicates abrupt structural changes in the cutting path, requiring a reduction in cutting speed to ensure machining accuracy and thermal stability. It is obtained from the voxel model through spatial differentiation at the surface layer. λ is the gradient sensitivity coefficient, λ = 2.0, set to 2.0 to achieve a boundary gradient of 0.45mm. -1 At this time, the scanning speed is reduced to 50% of the standard speed to ensure that heat accumulation at the crack boundary does not cause material deformation or irregular cuts;

[0110] S34, Real-time Processing Execution: After completing calculations for power compensation, focus adjustment, and speed optimization, the control system synchronously updates the parameters of the laser cutting equipment to ensure that the processing process matches the current defect status in real time. Specifically, this includes:

[0111] (1) Output power: P = P0 + P comp ;

[0112] (2) Focal position: Focal point = Standard focal point + Δf;

[0113] (3) Scanning speed: v;

[0114] Where P is the actual output laser power, P0 is the reference laser power, and Δf is the focal point shift.

[0115] S35, Online Quality Monitoring and Self-Learning Adjustment: To further improve processing stability, the system introduces a temperature feedback mechanism, using an infrared thermal imager to monitor the cutting seam temperature in real time. When the actual temperature deviates from the theoretical temperature beyond the tolerance range, the power compensation coefficient μ is dynamically adjusted, forming a self-learning feedback loop, as shown below:

[0116]

[0117] |T-T0|>ΔT th ;

[0118] Where T is the real-time monitored cutting kerf temperature, T0 is the theoretical cutting temperature, T0 = 1650℃, set based on its melting point plus a safety margin. Taking marine steel as an example, its melting point is approximately 1550℃, so setting T0 = 1650℃ ensures optimal laser focusing effect and avoids incomplete material melting. ΔT th It is the temperature tolerance threshold, ΔT th =50, mainly considering instantaneous thermal disturbance during the cutting process, material thermal inertia, and sensor accuracy, effectively balancing processing stability and control sensitivity, μ new This is the power compensation coefficient after self-learning adjustment. When the temperature deviation exceeds the tolerance threshold, it corrects the power compensation result to avoid overheating or failure to fuse, achieving real-time adaptive adjustment of laser energy. μ is the power compensation coefficient, and ε... μ It is the learning rate coefficient, which is related to the material's thermal conductivity, processing redundancy power, safety margin, etc., and represents the system's sensitivity to temperature errors. ε μ =0.02, used to control the response speed of power regulation to temperature errors, ensuring μ without excessively amplifying small temperature disturbances. new It can effectively correct large deviations while balancing response sensitivity and system stability.

[0119] like Figure 2 As shown, a ship plate identification and processing system, used to implement the above-mentioned ship plate identification and processing method, includes the following modules:

[0120] Substrate feature data acquisition module: Scans the surface of ship plate with anti-corrosion coating using a dual-frequency eddy current detection probe, collects abnormal conductivity data of the substrate material under the coating, and outputs a dataset for subsequent modeling;

[0121] Equivalent defect model generation module: Based on the spatial distribution and amplitude variation of conductivity anomaly data, construct an equivalent matrix defect model that characterizes the location and depth of matrix defects, and output the spatial location parameters and depth characteristic parameters of the defects;

[0122] Energy parameter optimization processing module: Based on the defect depth data in the equivalent matrix defect model, it dynamically calculates the output power and focal position of the laser cutting equipment to achieve adaptive processing for different defect areas.

[0123] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the following preferred embodiments; however, those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0124] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A ship plate identification processing method characterized by, Includes the following steps: S1, Matrix Feature Data Acquisition: A dual-frequency eddy current detection probe is used to scan the ship plate with anti-corrosion coating to obtain abnormal conductivity data of the matrix material under the coating; S2, Equivalent defect model generation: Based on the spatial distribution and amplitude variation of the conductivity anomaly data, an equivalent matrix defect model characterizing the location and depth of matrix defects is generated; S3, Energy Parameter Optimization Processing: Based on the defect depth data in the equivalent matrix defect model, the output power and focal position of the laser cutting machine are adjusted in real time to perform adaptive processing.

2. The ship plate identification processing method according to claim 1, characterized in that, S1 includes: S11, Dual-frequency excitation signal generation: The dual-frequency eddy current excitation method is adopted to control the output of sinusoidal excitation current of fixed frequency by the low-frequency and high-frequency coils respectively, and the current amplitude ratio is set. S12, Spiral Path Scanning: Drives the dual-frequency eddy current detection probe to scan the surface of the plate along a spiral trajectory, with the scanning interval set to be less than one-third of the minimum coil diameter; S13, Impedance signal acquisition: During the scanning process of the dual-frequency eddy current detection probe, the complex impedance signals of the low-frequency and high-frequency coils are acquired in real time, and their resistance and reactance components are measured. S14, Conductivity decoupling calculation: Extract the reactance increment and introduce the coating coupling coefficient for decoupling calculation to obtain the relative conductivity deviation value reflecting the degree of substrate defects; S15, Abnormal Data Extraction: Based on the decoupling calculation results, set an empirical threshold to determine defects. When the relative conductivity deviation value exceeds the set empirical threshold, record the corresponding location and its relative conductivity deviation value.

3. The method for identifying and processing ship plate materials according to claim 2, characterized in that, S13 includes: S131, Impedance Response Measurement: During the scanning process of the dual-frequency eddy current detection probe, the complex impedance of the low-frequency coil is measured in real time, including the resistive component and the reactance component. S132, High-frequency characteristic acquisition: Synchronously acquire the complex impedance information of the high-frequency coil and record the reactance changes.

4. The method for identifying and processing ship plate materials according to claim 2, characterized in that, S14 includes: S141, Reactance increment calculation: Based on the currently measured reactance value and the baseline value of the defect-free area, calculate the increments of high-frequency reactance and low-frequency reactance respectively, and extract abnormal change characteristics; S142, Coating Interference Compensation: Introduce a coating thickness coupling coefficient and use low-frequency reactance variation as a compensation term; S143, Deviation value calculation: Combine the probe calibration coefficient to correct the decoupling result, and finally calculate the relative conductivity deviation value.

5. The method for identifying and processing ship plate materials according to claim 4, characterized in that, S2 includes: S21, Spatial Interpolation Reconstruction: Weighted interpolation is performed on the conductivity deviation data obtained from the scan to construct a global continuous distribution function; S22, Depth Inversion Calculation: Based on the magnitude of the conductivity deviation function obtained by interpolation, the depth distribution of defects in the thickness direction of the plate is further estimated; S23, Equivalent crack width calculation: Combine the spatial change rate of deviation and depth information to calculate the equivalent crack width in the transverse direction; S24, Voxel Model Generation: Divide the plate into small voxel meshes, assign values ​​based on depth and width information, and construct a three-dimensional crack morphology model.

6. The method for identifying and processing ship plate materials according to claim 5, characterized in that, S22 includes: S221, Deviation Amplitude Analysis: Based on the interpolated conductivity deviation function, obtain the absolute value of the deviation at each location point; S222, Nonlinear Depth Mapping: Combining the maximum detection depth and the sensitivity coefficient determined by calibration tests, the deviation value is converted into the defect depth through an exponential mapping relationship.

7. The method for identifying and processing ship plate materials according to claim 5, characterized in that, S23 includes: S231, Spatial gradient extraction: Calculate the rate of change of the conductivity deviation function in space and obtain the gradient value at each location; S232, Crack width estimation: Combine the gradient value with the defect depth at the corresponding location and introduce the material property coefficient to estimate the equivalent crack width at that location.

8. A method for identifying and processing ship plate materials according to claim 5, characterized in that, S24 includes: S241, Spatial Mesh Generation: Divide the area of ​​the ship plate to be inspected into a uniform voxel mesh in three-dimensional space; S242, Voxel assignment calculation: Based on the position of each voxel in the thickness direction, compare it with the defect depth and width at the corresponding position, and assign a value to each voxel according to the set rules; S243, Crack Region Labeling: Determine whether the value of each voxel is greater than the set threshold, mark the voxels that meet the conditions as crack regions, and construct a three-dimensional crack morphology model.

9. A method for identifying and processing ship plate materials according to claim 8, characterized in that, S3 includes: S31, Dynamic Power Compensation: Calculates the laser power compensation value based on the defect depth at the current processing point; S32, Focus position shifted downward: Adjust the focus position according to the defect depth and crack width, and calculate the downward shift amount; S33, Scanning speed optimization: Adjust the laser scanning speed according to crack gradient information; S34, Real-time processing execution: Based on the above calculation results, the processing parameters of the laser cutting machine are controlled in real time; S35, Online Quality Monitoring and Self-Learning Adjustment: Based on infrared thermal imaging monitoring of the cutting seam temperature, the power compensation coefficient is updated when the conditions are met.

10. A ship plate identification and processing system, used to implement the ship plate identification and processing method as described in any one of claims 1-9, characterized in that, Includes the following modules: Substrate feature data acquisition module: Scans the surface of ship plate with anti-corrosion coating using a dual-frequency eddy current detection probe, collects abnormal conductivity data of the substrate material under the coating, and outputs a dataset for subsequent modeling. Equivalent defect model generation module: Based on the spatial distribution and amplitude variation of the conductivity anomaly data, construct an equivalent matrix defect model characterizing the location and depth of matrix defects, and output the spatial location parameters and depth feature parameters of the defects; Energy parameter optimization processing module: Based on the defect depth data in the equivalent matrix defect model, dynamically calculate the output power and focal position of the laser cutting equipment to achieve adaptive processing for different defect areas.

Images