A method and system for detecting surface oxide layers on copper rods

Abstract

The present application relates to copper rod oxidation layer detection technical field, especially in kind of copper rod surface oxidation layer detection method and system, method includes: establishing calibration database, detecting the surface oxidation layer of the copper rod to be measured, collecting spectral image data and laser scattering data; Based on the spectral image data, get the composition parameter and thickness parameter, based on the laser scattering data analysis scattering light intensity, get scattering parameter; Refer to the calibration database, convert the composition parameter, the thickness parameter and the scattering parameter into composition index, thickness index and compactness index; Construct a judgment rule containing multiple judgment channels, judge the composition index, thickness index and compactness index through the judgment channel respectively, and each judgment channel outputs an initial detection signal; Integrate all the initial detection signals to generate a detection result. The present application effectively detects the multi-dimensional characteristics of copper rod oxidation layer, improves the detection accuracy, and obtains comprehensive and reliable detection results.

Description

Technical Field

[0001] This invention relates to the field of copper rod oxide layer detection technology, and in particular to a method and system for detecting the oxide layer on the surface of copper rods. Background Technology

[0002] Copper rods are essential raw materials for industries such as power, electronics, and communications. Their surface quality directly affects the success of subsequent processing steps such as wire drawing and tin plating, as well as the performance of the final product. During production and storage, the surface of copper rods inevitably reacts with oxygen in the air to form an oxide layer. This oxide layer mainly consists of cuprous oxide and copper oxide, significantly increasing the contact resistance in subsequent processing, leading to abnormal wear or even wire breakage of the wire drawing die, and affecting the adhesion of the plating layer. Rapid and accurate detection and evaluation of the oxide layer on the surface of copper rods is a crucial step in quality control during the production process.

[0003] Currently, in industrial production, the detection of oxide layers on the surface of copper rods typically employs machine vision or single-spectral analysis. Machine vision methods primarily respond to color changes, but color is easily affected by lighting conditions, camera performance, and surface cleanliness. Furthermore, it cannot effectively distinguish between cuprous oxide and copper oxide, which have similar colors but different properties, and it cannot obtain information on the thickness and density of the oxide layer. While single-spectral analysis can provide some information on composition and thickness, it is insensitive to the microscopic physical structure of the oxide layer. However, the microscopic structure is a key factor affecting its conductivity and processing performance. Moreover, relying on a single data source for analysis makes it susceptible to environmental interference, leading to misjudgments, and it cannot provide a comprehensive test report on composition, thickness, and density.

[0004] Therefore, there is a need for a method and system for detecting the oxide layer on the surface of copper rods that can overcome the above-mentioned shortcomings. Summary of the Invention

[0005] In view of at least one of the above technical problems, the present invention provides a method and system for detecting oxide layers on the surface of copper rods, which adopts a multi-source data detection method combining spectral feature extraction and laser scattering analysis to achieve integrated quantification and stable determination of oxide layer composition, thickness and density.

[0006] This invention provides a method for detecting the oxide layer on the surface of a copper rod, comprising the following steps:

[0007] S1: Establish a calibration database, detect the oxide layer on the surface of the copper rod to be tested, and collect spectral image data and laser scattering data;

[0008] S2: Based on the spectral image data, extract the first spectral feature located in the first preset wavelength range and the second spectral feature located in the second preset wavelength range, and obtain the composition parameter and thickness parameter according to the comparison relationship between the first spectral feature and the second spectral feature. Analyze the scattered light intensity based on the laser scattering data to obtain the scattering parameter.

[0009] S3: Referring to the calibration database, convert the composition parameters, the thickness parameters, and the scattering parameters into composition indices, thickness indices, and density indices;

[0010] S4: Construct a judgment rule containing multiple judgment channels, and use the judgment channels to judge the composition index, thickness index and density index respectively, and output an initial detection signal for each judgment channel;

[0011] S5: Integrate all the initial detection signals to generate the detection results.

[0012] In some embodiments of the present invention, extracting a first spectral feature located in a first preset wavelength range and a second spectral feature located in a second preset wavelength range specifically involves:

[0013] The first preset wavelength range is used to extract the first spectral features corresponding to the characteristic absorption peak of cuprous oxide;

[0014] The second preset wavelength range is used to extract the second spectral feature corresponding to the characteristic absorption peak of copper oxide.

[0015] In some embodiments of the present invention, the composition parameters and thickness parameters are obtained based on the comparison relationship between the first spectral feature and the second spectral feature, including:

[0016] The first spectral feature and the second spectral feature are respectively normalized in terms of spectral intensity. The ratio of the difference in absorption peaks between the two to the baseline is calculated. Based on the ratio of the difference in absorption peaks to the baseline, the relative contents of copper oxide and cuprous oxide are determined to obtain the component parameters.

[0017] Construct a thickness absorption curve and obtain the thickness parameter based on the difference in absorption peaks.

[0018] In some embodiments of the present invention, a judgment rule comprising multiple judgment channels is constructed, including:

[0019] The judgment rules include a composition judgment channel, a thickness judgment channel, and a density judgment channel;

[0020] The component determination channel is used to compare the component index with a preset component threshold to determine the main component category of the oxide layer;

[0021] The thickness determination channel is used to compare the thickness index with a preset thickness threshold range to determine the oxide layer thickness.

[0022] The density determination channel is used to compare the density index with a preset density threshold to determine the uniformity and density of the oxide layer structure.

[0023] Each of the aforementioned determination channels outputs a copy of the initial detection signal.

[0024] In some embodiments of the present invention, the component determination channel is used to compare the component index with a preset component threshold to determine the main component category of the oxide layer, including:

[0025] Establish a copper oxide standard information group and a cuprous oxide standard information group;

[0026] The difference between the first spectral feature and the second spectral feature is calculated on the wavelength axis to plot the spectral difference curve of the component index;

[0027] The wavelength position, amplitude, and adjacent spacing of multiple extreme points of the spectral difference curve are extracted to form a group of information to be measured;

[0028] The test information group is matched point by point with the copper oxide standard information group and the cuprous oxide standard information group to determine the optimal matching path, and the total cost of the corresponding optimal matching path is calculated through the test information group.

[0029] The main category of the oxide layer is determined based on the total cost.

[0030] The copper oxide standard information group and the cuprous oxide standard information group contain the wavelength position, amplitude and adjacent spacing that are matched point by point with the information group to be tested.

[0031] In some embodiments of the present invention, calculating the total cost corresponding to the optimal matching path using the group of information to be tested includes:

[0032] Based on the wavelength position, the wavelength difference D is calculated for each matching point. λ (i),

[0033] The wavelength difference ;

[0034] Based on the amplitude, the amplitude difference D is calculated for each matching point. A (i),

[0035] The amplitude difference ;

[0036] Based on the adjacent spacing, the spacing difference D is calculated for each matching point. Δ (i),

[0037] The distance difference ;

[0038] The total cost is obtained by adding the wavelength difference, amplitude difference, and spacing difference of all matching points.

[0039] Where i is the extreme point in the information group to be tested, and λ i For the wavelength position of this extreme point, A i The amplitude of the extreme point; j is the extreme point matched in the copper oxide standard information group or the cuprous oxide standard information group, λ j For the wavelength position of this extreme point, A j The amplitude at that extreme point; Δλ i Δλ is the adjacent distance between adjacent extreme points in the group of information to be measured. j The adjacent spacing of the extreme point positions matched in the standard information group; δ λ The tolerance parameter for the preset wavelength position is a constant; δ A The preset tolerance parameter for the amplitude is a constant; W λ (i) represents the wavelength weight, W A (i) represents the magnitude weight, W Δ (i) represents the spacing weights, all of which are constants, and W λ (i)+W A (i)+W Δ (i)=1.

[0040] In some embodiments of the present invention, the total cost further includes a global penalty:

[0041] For each extreme point in the information group to be tested that fails to match the standard information group, the penalty value for the missing point is increased by one.

[0042] For each extreme point in the standard information group that fails to match the information group to be tested, the penalty value of the insertion point is increased by one.

[0043] When a match crosses adjacent extreme points, causing the match to be discontinuous, add a penalty value for inconsistent trends.

[0044] The sum of the penalty values ​​is included in the total cost as the global penalty.

[0045] In some embodiments of the present invention, the thickness determination channel is used to compare the thickness index with a preset thickness threshold range, including:

[0046] The thickness index is compared with the boundary values ​​of each preset thickness threshold range in turn. When the thickness index falls into a certain range, the thickness level signal corresponding to that range is output.

[0047] When the thickness index is near the interval boundary, it is further compared with the difference of the center value of the adjacent interval, and the thickness level signal corresponding to the one with smaller difference is output.

[0048] In some embodiments of the present invention, step S6 is further included to perform auxiliary determination of the oxide layer:

[0049] A micro-region excitation signal is applied to the surface of the copper rod under test, and transient response data of the region is collected.

[0050] Based on the transient response data, the physical properties of the oxide layer are analyzed to obtain the corresponding auxiliary judgment signal;

[0051] Based on the auxiliary judgment signal and the detection result, a comprehensive detection result is obtained.

[0052] The present invention also provides a copper rod surface oxide layer detection system, comprising:

[0053] The data acquisition and storage module establishes a calibration database, detects the oxide layer on the surface of the copper rod under test, and acquires spectral image data and laser scattering data.

[0054] The parameter extraction module extracts a first spectral feature located in a first preset wavelength range and a second spectral feature located in a second preset wavelength range based on the spectral image data. It then obtains composition parameters and thickness parameters based on the comparison relationship between the first spectral feature and the second spectral feature. Finally, it analyzes the scattered light intensity based on the laser scattering data to obtain scattering parameters.

[0055] The index conversion module converts the composition parameters, thickness parameters, and scattering parameters into composition indices, thickness indices, and density indices by referring to the calibration database;

[0056] The judgment and detection module constructs judgment rules containing multiple judgment channels. The composition index, thickness index and density index are judged through the judgment channels. Each judgment channel outputs an initial detection signal. All the initial detection signals are integrated to generate a detection result.

[0057] The beneficial effects of this invention are as follows: By simultaneously acquiring spectral image data and laser scattering data within the detection process and establishing three judgment channels for composition, thickness, and density, this invention overcomes the shortcomings of traditional single detection methods in terms of information dimension fragmentation, and significantly improves the comprehensiveness and reliability of the detection results. By utilizing spectral difference feature matching, thickness absorption curves, and scattering parameters for joint judgment, the invention effectively reduces the interference effects of surface roughness, light disturbance, and copper oxide and cuprous oxide mixed phase conditions, thereby maintaining high detection stability even in complex production line environments. Attached Figure Description

[0058] 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 recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0059] Figure 1 This is a schematic diagram illustrating the steps of a method for detecting the oxide layer on the surface of a copper rod in an embodiment of the present invention;

[0060] Figure 2 This is a flowchart illustrating how compositional parameters and thickness parameters are obtained based on the comparison between first and second spectral features in an embodiment of the present invention.

[0061] Figure 3 This is a flowchart illustrating the steps of the component determination channel in this embodiment of the invention to compare component indicators with preset component thresholds to determine the main component categories of the oxide layer.

[0062] Figure 4 This is a diagram illustrating the steps of comparing the thickness index with a preset thickness threshold range in the thickness determination channel according to an embodiment of the present invention.

[0063] Figure 5 This is a schematic diagram illustrating the steps for auxiliary determination of the oxide layer in an embodiment of the present invention. Detailed Implementation

[0064] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0065] It should be noted that when an element is referred to as being "fixed to" another element, it can be directly attached to the other element or there may be an intervening element. When an element is referred to as being "connected to" another element, it can be directly connected to the other element or there may be an intervening element. The terms "vertical," "horizontal," "left," "right," and similar expressions used herein are for illustrative purposes only and do not represent the only possible implementation.

[0066] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0067] This invention provides a method such as Figures 1 to 5 The method for detecting the oxide layer on the surface of a copper rod, as shown, includes the following steps:

[0068] S1: Establish a calibration database, detect the oxide layer on the surface of the copper rod under test, and collect spectral image data and laser scattering data. A hyperspectral camera is used as the acquisition tool. Under uniform illumination from a specific broadband light source, the hyperspectral camera continuously acquires hyperspectral image cubes of the copper rod surface. This data volume contains two-dimensional spatial information and one-dimensional spectral information, serving as spectral image data. A semiconductor laser is used as the light source. Its beam is expanded and collimated before being incident on the copper rod surface at a certain angle, such as 45°. A CCD detector is then used to receive the scattered light spot image of the copper rod surface under test. The intensity distribution of the scattered light is directly related to the roughness and grain size of the surface microstructure. The calibration database can be established in a laboratory environment by preparing a series of standard copper rod oxide samples with known composition, known thickness, and known density determined by ellipsometer. The same equipment described above is used to collect spectral image data and laser scattering data of these standard samples, and feature parameters are extracted and correlated with known indicators to establish a calibration database for subsequent parameter conversion.

[0069] S2: Based on spectral image data, extract the first spectral feature located in the first preset wavelength range and the second spectral feature located in the second preset wavelength range. Obtain the composition parameter and thickness parameter according to the comparison relationship between the first spectral feature and the second spectral feature. Analyze the scattered light intensity based on laser scattering data to obtain scattering parameters. After performing preprocessing such as grayscale conversion and noise reduction on the acquired laser scattering image, calculate the grayscale standard deviation and average grayscale value of the entire spot image. Input these two feature values ​​into the model trained based on the calibration database to output scattering parameters that reflect the density and uniformity of the oxide layer. The stronger the scattering and the more diffuse the spot, the higher the scattering parameter value, indicating that the surface is rougher and the oxide layer structure may be more porous.

[0070] S3: Referring to the calibration database, the composition parameters, thickness parameters, and scattering parameters are converted into composition indices, thickness indices, and density indices. After extracting key features from the spectral data, they are input into the corresponding mapping relationship in the database to obtain a standardized value. This value reflects the ratio of copper oxide to cuprous oxide in the oxide layer. When the value is close to a certain standard sample, it indicates that the composition of the copper rod to be tested is similar to that of the standard sample. Based on the overall absorption characteristics of the spectrum and combined with the thickness information of existing samples in the database, a thickness index is obtained through interpolation or fitting methods. This index can reflect whether the oxide layer to be tested is in the thin, medium, or thick range. The statistical characteristics of the scattering image are input into the database to obtain a density index. This index can distinguish between dense layers and porous layers, thereby reflecting the compactness of the oxide layer structure.

[0071] S4: Construct a judgment rule containing multiple judgment channels. The judgment channels are used to judge the composition index, thickness index and density index respectively. Each judgment channel outputs an initial detection signal. The main component category, thickness level and density level of the oxide layer are judged, which ensures that the results of different dimensions do not interfere with each other and can remain stable under complex conditions.

[0072] S5: Integrate all initial detection signals to generate detection results; the integration method can be simple weighting or fusion based on logical rules. The final result is output in the form of a comprehensive conclusion, for example, the oxide layer is mainly copper oxide, with medium thickness and poor density; it can intuitively reflect the overall state of the oxide layer, making it easy for operators to quickly judge whether the production process is reasonable, or whether annealing, cooling and other steps need to be adjusted.

[0073] This embodiment achieves unified quantification and comprehensive judgment of three key indicators: oxide layer composition, thickness, and density by fusing multi-source data of detected spectral features and laser scattering features, supplemented by a calibration database and multi-channel judgment rules.

[0074] In some embodiments of the present invention, such as Figure 2As shown, the extraction of first spectral features located in the first preset wavelength range and second spectral features located in the second preset wavelength range is specifically as follows:

[0075] The first preset wavelength range is used to extract the first spectral feature corresponding to the characteristic absorption peak of cuprous oxide;

[0076] The second preset wavelength range is used to extract the second spectral features corresponding to the characteristic absorption peak of copper oxide.

[0077] The first preset wavelength range is selected to cover the position of the characteristic absorption peak of cuprous oxide, and the first spectral feature is extracted from this range; the second preset wavelength range covers the position of the characteristic absorption peak of copper oxide, and the second spectral feature is extracted from this range. By directly capturing the difference signals exhibited by different oxides in their spectral responses, the problem of inaccurate judgment caused by redundant spectral data or interfering components can be effectively avoided. Comparing and calculating the two spectral features can not only improve the pertinence and accuracy of component analysis, but also provide a reliable reference for the derivation of thickness parameters. Using the characteristic peaks corresponding to different ranges as the basis for component judgment can significantly enhance the sensitivity and discrimination of the detection results to the true composition of the oxide layer, making the final output component index and thickness index more consistent with the actual situation, and improving the accuracy and stability of the overall detection method.

[0078] Based on the above embodiments, such as Figure 2 As shown, the composition parameters and thickness parameters are obtained based on the comparison between the first spectral features and the second spectral features, including:

[0079] The first and second spectral features were normalized by spectral intensity, and the ratio of the difference between their absorption peaks to the baseline was calculated. Based on the ratio of the difference between their absorption peaks to the baseline, the relative contents of copper oxide and cuprous oxide were determined, and the composition parameters were obtained.

[0080] Construct thickness absorption curves and obtain thickness parameters based on the differences in absorption peaks.

[0081] The first and second spectral features are normalized to eliminate the influence of light source intensity, surface reflection conditions, or changes in the detection environment on the spectral intensity. On the normalized spectral curve, the ratio of the absorption difference between the two characteristic peaks to their corresponding baseline is calculated, thereby obtaining stable characteristic quantities related to the composition of the oxide layer on the surface of the copper rod under test. By comprehensively analyzing the absorption peak difference and the baseline ratio, the relative content relationship between copper oxide and cuprous oxide can be effectively derived, thus obtaining the composition parameters. Based on the correspondence between spectral absorption intensity and thickness, a thickness absorption curve is constructed, and the thickness parameters of the oxide layer are determined by combining the position and amplitude of the absorption peak difference. This method maintains the accuracy of composition analysis in complex detection environments and avoids the bias that may be caused by the single peak method, making the final obtained composition and thickness parameters more representative and stable, thus providing reliable input data for subsequent index conversion and judgment.

[0082] Thickness absorption curves can be modeled based on the Lambert-Beer law to establish the relationship between absorption intensity and material thickness. By calculating or simulating the absorption changes of the spectrum under different thickness conditions, the trend curve of the difference of spectral characteristic peaks with thickness can be obtained. Alternatively, a series of standard samples of copper rod oxide layers with different thicknesses can be prepared, and the actual thickness can be accurately measured by means of microscopic cross-sectional observation or film thickness gauge. Then, combined with the spectral test results, the corresponding curve of absorption peak difference and thickness can be directly plotted to ensure that the thickness absorption curve is highly consistent with the actual material state.

[0083] In some embodiments of the present invention, a judgment rule comprising multiple judgment channels is constructed, including:

[0084] The judgment rules include the composition judgment channel, the thickness judgment channel, and the density judgment channel;

[0085] The component determination channel is used to compare component indicators with preset component thresholds to determine the main component categories of the oxide layer;

[0086] The thickness determination channel is used to compare the thickness index with a preset thickness threshold range to determine the oxide layer thickness.

[0087] The density determination channel is used to compare the density index with a preset density threshold to determine the uniformity and density of the oxide layer structure.

[0088] Each decision channel outputs an initial detection signal.

[0089] The composition index of the copper rod to be tested is input into the composition determination channel. This channel compares the index with a preset composition threshold to determine the main composition category of the oxide layer. For example, it can determine whether the oxide layer is mainly composed of copper oxide or cuprous oxide. Subsequently, the thickness index is input into the thickness determination channel. This channel compares the thickness with a preset thickness threshold range and outputs a signal indicating whether the oxide layer thickness is thin, medium, or thick. The density index is also input into the density determination channel. This channel compares the density with a density threshold to determine whether the oxide layer structure is uniform and dense, and outputs the corresponding level information.

[0090] The component threshold can be established by performing spectral tests on a series of standard samples with known component ratios, and establishing a correspondence between the actual analytical results of copper oxide and cuprous oxide content and spectral characteristic indicators. When the indicator value is in the transition region between copper oxide and cuprous oxide, a critical value is set as the component threshold to distinguish between cuprous oxide-dominated and copper oxide-dominated cases. The thickness threshold range can be obtained by preparing oxide layer samples of different thicknesses and performing spectral tests, and mapping their spectral absorption differences to the actual thickness to plot thickness absorption curves. Subsequently, according to the actual needs of production and quality control, the thickness range can be divided into several intervals, such as thin layer, medium layer, and thick layer, with the corresponding points on the curve as interval boundaries, reflecting the physical correspondence between spectral characteristics and thickness, and meeting the needs of rapid determination in engineering testing. The density threshold can usually be obtained through comparative experiments. Compact and loose control samples are selected, their scattering data are collected, and the corresponding density indicators are extracted. Based on this, the range of the density indicator is determined, and the boundary value between the two types of samples is used as the density threshold. Alternatively, the threshold can be further corrected by statistical methods in combination with actual testing needs to ensure stability and consistency in batch testing.

[0091] Each judgment channel operates independently and outputs an initial detection signal. These signals can be in the form of a level, category, or numerical range, all of which support subsequent comprehensive judgment. This multi-channel parallel judgment method avoids the excessive influence of a single indicator on the overall detection results, making the final detection conclusion more comprehensive, objective, and stable.

[0092] Based on the above embodiments, such as Figure 3 As shown, the component determination channel is used to compare component indicators with preset component thresholds to determine the main component categories of the oxide layer, including:

[0093] Establish a copper oxide standard information group and a cuprous oxide standard information group;

[0094] The difference between the first spectral feature and the second spectral feature is calculated on the wavelength axis, and the spectral difference curve of the component index is plotted.

[0095] The wavelength position, amplitude, and adjacent spacing of multiple extreme points of the spectral difference curve are extracted to form a group of information to be measured;

[0096] The test information group is matched point by point with the copper oxide standard information group and the cuprous oxide standard information group to determine the optimal matching path, and the total cost of the corresponding optimal matching path is calculated through the test information group.

[0097] The main category of the oxide layer is determined based on the total cost.

[0098] Among them, the copper oxide standard information group and the cuprous oxide standard information group contain the wavelength position, amplitude and adjacent spacing that are matched point by point with the information group to be tested.

[0099] A standard information group for copper oxide and a standard information group for cuprous oxide were established. Both information groups contain the location, amplitude, and distance between adjacent extreme points on the spectral difference curve extracted during the spectral testing of typical samples. In actual testing, the first and second spectral features of the copper rod under test are calculated on the wavelength axis to obtain the spectral difference curve of the component index. The wavelength position, amplitude value, and adjacent spacing of multiple extreme points are extracted from this curve to form the information group to be tested. This allows for the capture of multi-dimensional feature information on a single difference curve, rather than relying solely on a single peak, thereby effectively enhancing the robustness of the judgment.

[0100] The test information group is matched point-by-point with the copper oxide standard information group and the cuprous oxide standard information group, respectively. During the matching process, differences in wavelength position, amplitude, and adjacent spacing are considered to find the optimal matching path with the standard information group. After obtaining the optimal matching path, the corresponding total cost is calculated based on the test information group. The total cost is a quantitative index obtained by comprehensively calculating the differences of each matching point. The smaller the value, the higher the similarity between the test copper rod and a certain standard information group. By comparing the total cost of the test information group with the copper oxide standard information group and the cuprous oxide standard information group, the main component category of the oxide layer can be determined. For example, when the total cost with the copper oxide standard information group is significantly less than the total cost with the cuprous oxide standard information group, it can be determined that the oxide layer is mainly composed of copper oxide; otherwise, it is determined to be mainly composed of cuprous oxide.

[0101] Based on the above embodiments, the total cost of the corresponding optimal matching path is calculated using the group of information to be tested, including:

[0102] The wavelength difference D is calculated for each matching point based on the wavelength location. λ (i),

[0103] Wavelength difference ;

[0104] The peak positions of the spectral curves are directly related to the chemical bond energies of the materials. The peak positions of cuprous oxide and copper oxide often exhibit slight shifts. These shifts can be addressed by calculating the wavelength difference D. λ (i) can directly reflect the difference in chemical composition between the test information group and the standard information group; wavelength difference D λ (i) is calculated using absolute value normalization to ensure that the difference is always positive regardless of the offset direction. Physically, this corresponds to the degree of offset in the energy level structure. The closer the two extreme points are on the wavelength axis, the smaller their wavelength difference; conversely, the greater the difference, the greater the difference. This difference is then weighted by wavelength W. λ (i) Adjustments can be made to adjust the contribution of the wavelength position to the total cost based on the importance of the wavelength position in the actual detection.

[0105] The amplitude difference D is calculated for each matching point based on the amplitude. A (i),

[0106] Amplitude difference ;

[0107] Spectral amplitude is related to the absorption intensity of the material, reflecting differences in the concentration and thickness of the oxide layer; however, amplitude data is easily affected by noise such as light source intensity and detector sensitivity; amplitude difference D A (i) By highlighting the sensitivity of small differences through the squared term, slight changes in composition can be captured; and by suppressing the influence of large differences through the exponential decay factor, the occasional noise generated by the detector saturation point can be avoided from dominating the judgment. This combination can both amplify small differences and filter large anomalies, which is equivalent to the combination of signal enhancement and noise suppression.

[0108] Calculate the distance difference D for each matching point based on the adjacent distance. Δ (i),

[0109] Spacing difference ;

[0110] The spacing between adjacent extreme points reflects the overall spectral structure. Influenced by the material's lattice and electronic energy level structure, the relative spacing is often more stable than that of a single peak and less susceptible to systematic errors; the spacing difference D Δ (i) Using relative ratios is because the spectral curve may drift as a whole, but the relative spacing between adjacent peaks is stable. Using ratios can capture the inherent energy level structure characteristics of the material and has stronger discriminative power.

[0111] The total cost is obtained by adding up the wavelength difference, amplitude difference, and spacing difference of all matching points;

[0112] Where i represents the extreme point in the information group to be measured, and λ i Let A be the wavelength position of this extreme point. iλ represents the amplitude of the extreme point; j represents the extreme point matched in the copper oxide standard information group or the cuprous oxide standard information group; λ represents the amplitude of the extreme point. j Let A be the wavelength position of this extreme point. j Δλ represents the magnitude of the extreme point. i Δλ represents the distance between adjacent extreme points in the data set to be measured. j The distance between adjacent extreme points matched in the standard information group; δ λ The tolerance parameter for the preset wavelength position is a constant; δ A The tolerance parameter for the preset amplitude is a constant; W λ (i) represents the wavelength weight, W A (i) represents the magnitude weight, W Δ (i) represents the spacing weights, all of which are constants, and W λ (i)+W A (i)+W Δ (i)=1, the weights can be optimized according to the detection requirements, so that the method is suitable for both component classification and thickness and density analysis.

[0113] In spectral analysis, the differences between the test data set and the standard data set are mainly reflected in three aspects: wavelength shift (different chemical compositions or energy level structures can cause changes in peak positions); amplitude variation (absorption intensity is affected by oxide layer thickness and component concentration, and amplitude differences can reflect differences in substance content); and differences in adjacent spacing (different materials have different energy level spacing, which will change the interval between peaks). Therefore, the difference method is used to measure the deviation between the two spectra in calculations. This is the most direct and physically realistic comparison method, which can clearly quantify the degree of deviation between the test extreme point and the standard extreme point, avoiding fuzzy judgments.

[0114] Tolerance parameter δ λ and δ A The introduction of this parameter allows minute offsets to be normalized within a reasonable range, avoiding amplification and misjudgment, reducing the impact of errors caused by equipment resolution and environmental fluctuations during measurement, and enabling comparisons of differences across different dimensions on the same scale; the tolerance parameter δ λ and δ A This can be achieved in several ways. For example, a set of copper rod samples with known oxide layer types and thicknesses can be selected as calibration objects. The peak positions and amplitude data of the samples can be obtained through spectral measurements. The fluctuation range of these data under different experimental conditions (such as temperature, detection angle, and instrument resolution) can be analyzed, and the statistical values ​​of these fluctuation ranges can be used as tolerance parameters. Alternatively, the tolerance parameter, δ, can be set directly based on the instrument resolution. λ Not less than the instrument resolution, δ A The noise level should not be less than the system noise level to ensure that the judgment process remains stable within the normal measurement error range.

[0115] The total cost represents the overall similarity between the information group to be tested and the standard information group. It takes into account both local features (wavelength, amplitude) and overall features (spacing), is sensitive to small differences, and avoids the influence of noise. It provides a unified evaluation index, and the optimal matching path can be selected by comparing the magnitude of the total cost. When the total cost is the minimum, it means that the similarity between the information group to be tested and a certain standard information group is the highest, and the corresponding standard information group is the main component category of the oxide layer to be determined.

[0116] In addition to the above embodiments, the total cost also includes a global penalty:

[0117] For each extreme point in the test information group that fails to match in the standard information group, a penalty value of one point is added for the missing point. If some extreme points exist in the test curve but not in the standard curve, it indicates that the curve contains additional spectral features, and this difference may mean that the material composition is different.

[0118] For each extreme point in the standard information group that cannot be matched by the test information group, the penalty value of the insertion point is increased by one point. If the standard curve has an extreme point of a characteristic peak, but the test curve does not have an extreme point of a corresponding peak, it means that the test material lacks the key component of the standard substance. The insertion point penalty ensures that this missing feature is included in the difference.

[0119] When a match skips adjacent extreme points, causing a discontinuity in the match, a penalty value for inconsistency in trend is added. Spectral extreme points are usually arranged in order of energy level structure. If a match skips or is discontinuous, it indicates that the curve to be tested and the standard curve have significant differences in overall shape. The penalty for inconsistency in trend is used to prevent the algorithm from pursuing only local optima while ignoring overall consistency.

[0120] The sum of the penalty values ​​is included in the total cost as a global penalty, which ensures that the matching path can accurately reflect local differences while maintaining global continuity and trend consistency, thereby improving the stability and reliability of component category determination.

[0121] In some embodiments of the present invention, such as Figure 4 As shown, the thickness determination channel is used to compare the thickness index with a preset thickness threshold range, including:

[0122] The thickness index is compared with the boundary values ​​of each preset thickness threshold range in turn. When the thickness index falls into a certain range, the thickness level signal corresponding to that range is output.

[0123] When the thickness index is near the boundary of the interval, it is further compared with the difference of the center value of the adjacent interval, and the thickness level signal corresponding to the one with smaller difference is output.

[0124] The thickness index is compared sequentially with the boundary values ​​of each preset thickness threshold interval. When the thickness index falls into a certain interval, the thickness level signal corresponding to that interval is output. Furthermore, when the thickness index is near the interval boundary, to avoid misjudgment and jump errors, this embodiment preferably introduces a secondary comparison mechanism. The difference between the thickness index and the center value of the adjacent interval is calculated separately, and the one with the smaller difference is selected as the final level determination basis, thereby outputting the corresponding thickness level signal. This can quickly give the thickness level through interval division, improving the efficiency of the determination. Introducing secondary comparison in the interval with ambiguous boundaries can effectively reduce misjudgment caused by critical point fluctuations, making the thickness determination smoother and more stable. The obtained thickness level signal is not only intuitive and easy to understand, but also provides a clear thickness grading reference for subsequent comprehensive detection results.

[0125] In some embodiments of the present invention, such as Figure 5 As shown, it also includes step S6, which assists in the determination of the oxide layer:

[0126] A micro-region excitation signal is applied to the surface of the copper rod under test, and transient response data of the region is collected.

[0127] Based on transient response data, the physical properties of the oxide layer are analyzed to obtain corresponding auxiliary judgment signals;

[0128] Based on the auxiliary judgment signal and the detection results, a comprehensive detection result is obtained.

[0129] The excitation signal can be a transient thermal excitation generated by a low-power laser pulse, or an electrochemical excitation generated by a small AC voltage. The intensity of the excitation signal is precisely controlled to ensure that it does not damage the surface of the copper rod. The transient response of the micro-region under the excitation signal is acquired using a high-speed infrared thermal imager or a high-speed electrochemical impedance spectroscopy device. When thermal excitation is used, the temperature decay curve during the cooling process is acquired; when electrochemical excitation is used, the corresponding impedance spectrum changes over time are acquired. The acquired transient response data is analyzed to extract parameters reflecting the characteristics of the oxide layer. Under thermal excitation, the thermal conductivity and thickness of the oxide layer are inferred from the slope and decay rate of the cooling curve; under electrochemical excitation... In this case, the resistance and capacitance characteristics of the oxide layer are obtained by fitting the equivalent circuit model of the impedance spectrum, thus reflecting its density and uniformity. A corresponding auxiliary judgment signal is output, reflecting the thickness level, density level, or correction opinion on the main component categories of the oxide layer. The auxiliary judgment signal is compared with the detection results based on spectroscopy and scattering. When the two are consistent, the detection result is confirmed; when there is a difference, the detection result is corrected according to the auxiliary judgment signal, thus obtaining the final comprehensive detection result. This method can adapt to various interference factors in complex production environments, achieving multi-angle verification of oxide layer thickness, density, and composition, avoiding the limitations of a single method, and making the final judgment closer to the true state. Of course, the above embodiments only list implementation methods of thermal or electrochemical excitation; auxiliary excitation can also be extended to different probes and other methods.

[0130] The present invention also provides a copper rod surface oxide layer detection system, comprising:

[0131] The data acquisition and storage module establishes a calibration database, detects the oxide layer on the surface of the copper rod under test, and acquires spectral image data and laser scattering data.

[0132] The parameter extraction module extracts first spectral features located in a first preset wavelength range and second spectral features located in a second preset wavelength range based on spectral image data. It obtains composition parameters and thickness parameters based on the comparison relationship between the first and second spectral features. It also obtains scattering parameters by analyzing the intensity of scattered light based on laser scattering data.

[0133] The index conversion module converts composition parameters, thickness parameters, and scattering parameters into composition indices, thickness indices, and density indices by referring to the calibration database.

[0134] The judgment and detection module constructs judgment rules containing multiple judgment channels. The judgment channels determine the composition index, thickness index, and density index. Each judgment channel outputs an initial detection signal. All initial detection signals are integrated to generate the detection result.

[0135] The detection system described above in this invention can effectively realize a method for detecting the oxide layer on the surface of a copper rod, and the technical effects it can achieve are as described in the above embodiments, and will not be repeated here.

[0136] Those skilled in the art should understand that this invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to this invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A method for detecting the oxide layer on the surface of a copper rod, characterized in that, Includes the following steps: S1: Establish a calibration database, detect the oxide layer on the surface of the copper rod to be tested, and collect spectral image data and laser scattering data; S2: Based on the spectral image data, extract the first spectral feature located in the first preset wavelength range and the second spectral feature located in the second preset wavelength range, and obtain the composition parameter and thickness parameter according to the comparison relationship between the first spectral feature and the second spectral feature. Analyze the scattered light intensity based on the laser scattering data to obtain the scattering parameter. S3: Referring to the calibration database, convert the composition parameters, the thickness parameters, and the scattering parameters into composition indices, thickness indices, and density indices; S4: Construct a judgment rule containing multiple judgment channels, and use the judgment channels to judge the composition index, thickness index and density index respectively, and output an initial detection signal for each judgment channel; The judgment rules include a composition judgment channel, a thickness judgment channel, and a density judgment channel; The component determination channel is used to compare the component index with a preset component threshold to determine the main component category of the oxide layer; The thickness determination channel is used to compare the thickness index with a preset thickness threshold range to determine the oxide layer thickness. The density determination channel is used to compare the density index with a preset density threshold to determine the uniformity and density of the oxide layer structure. S5: Integrate all the initial detection signals to generate the detection results.

2. The method for detecting the oxide layer on the surface of a copper rod according to claim 1, characterized in that, Extracting the first spectral feature located in the first preset wavelength range and the second spectral feature located in the second preset wavelength range, specifically: The first preset wavelength range is used to extract the first spectral features corresponding to the characteristic absorption peak of cuprous oxide; The second preset wavelength range is used to extract the second spectral feature corresponding to the characteristic absorption peak of copper oxide.

3. The method for detecting the oxide layer on the surface of a copper rod according to claim 2, characterized in that, The composition parameters and thickness parameters are obtained based on the comparison between the first spectral features and the second spectral features, including: The first spectral feature and the second spectral feature are respectively normalized in terms of spectral intensity. The ratio of the difference in absorption peaks between the two to the baseline is calculated. Based on the ratio of the difference in absorption peaks to the baseline, the relative contents of copper oxide and cuprous oxide are determined to obtain the component parameters. Construct a thickness absorption curve and obtain the thickness parameter based on the difference in absorption peaks.

4. The method for detecting the oxide layer on the surface of a copper rod according to claim 1, characterized in that, The component determination channel is used to compare the component index with a preset component threshold to determine the main component category of the oxide layer, including: Establish a copper oxide standard information group and a cuprous oxide standard information group; The difference between the first spectral feature and the second spectral feature is calculated on the wavelength axis to plot the spectral difference curve of the component index; The wavelength position, amplitude, and adjacent spacing of multiple extreme points of the spectral difference curve are extracted to form a group of information to be measured; The test information group is matched point by point with the copper oxide standard information group and the cuprous oxide standard information group to determine the optimal matching path, and the total cost of the corresponding optimal matching path is calculated through the test information group. The main category of the oxide layer is determined based on the total cost. The copper oxide standard information group and the cuprous oxide standard information group contain the wavelength position, amplitude and adjacent spacing that are matched point by point with the information group to be tested.

5. The method for detecting the oxide layer on the surface of a copper rod according to claim 4, characterized in that, The total cost corresponding to the optimal matching path is calculated using the group of information to be tested, including: Based on the wavelength position, the wavelength difference D is calculated for each matching point. λ (i), The wavelength difference ; Based on the amplitude, the amplitude difference D is calculated for each matching point. A (i), The amplitude difference ; Based on the adjacent spacing, the spacing difference D is calculated for each matching point. Δ (i), The distance difference ; The total cost is obtained by adding the wavelength difference, amplitude difference, and spacing difference of all matching points. Where i is the extreme point in the information group to be tested, and λ i For the wavelength position of this extreme point, A i The amplitude of the extreme point; j is the extreme point matched in the copper oxide standard information group or the cuprous oxide standard information group, λ j For the wavelength position of this extreme point, A j The amplitude at that extreme point; Δλ i Δλ is the adjacent distance between adjacent extreme points in the group of information to be measured. j The adjacent spacing of the extreme point positions matched in the standard information group; δ λ The tolerance parameter for the preset wavelength position is a constant; δ A The tolerance parameter for the preset amplitude is a constant; W λ (i) represents the wavelength weight, W A (i) represents the magnitude weight, W Δ (i) represents the spacing weights, all of which are constants, and W λ (i)+W A (i)+W Δ (i)=1.

6. The method for detecting the oxide layer on the surface of a copper rod according to claim 5, characterized in that, The total cost also includes a global penalty: For each extreme point in the information group to be tested that fails to match the standard information group, the penalty value for the missing point is increased by one. For each extreme point in the standard information group that fails to match the information group under test, the penalty value of the insertion point is increased by one. When a match crosses adjacent extreme points, causing the match to be discontinuous, add a penalty value for inconsistent trends. The sum of the penalty values ​​is included in the total cost as the global penalty.

7. The method for detecting the oxide layer on the surface of a copper rod according to claim 1, characterized in that, The thickness determination channel is used to compare the thickness index with a preset thickness threshold range, including: The thickness index is compared with the boundary values ​​of each preset thickness threshold range in turn. When the thickness index falls into a certain range, the thickness level signal corresponding to that range is output. When the thickness index is near the interval boundary, it is further compared with the difference of the center value of the adjacent interval, and the thickness level signal corresponding to the one with smaller difference is output.

8. The method for detecting the oxide layer on the surface of a copper rod according to claim 1, characterized in that, It also includes step S6, which assists in the determination of the oxide layer: A micro-region excitation signal is applied to the surface of the copper rod under test, and transient response data of the surface of the copper rod under test with the applied micro-region excitation signal are collected. Based on the transient response data, the physical properties of the oxide layer are analyzed to obtain the corresponding auxiliary judgment signal; Based on the auxiliary judgment signal and the detection result, a comprehensive detection result is obtained.

9. A copper rod surface oxide layer detection system, characterized in that, include: The data acquisition and storage module establishes a calibration database, detects the oxide layer on the surface of the copper rod under test, and acquires spectral image data and laser scattering data. The parameter extraction module extracts a first spectral feature located in a first preset wavelength range and a second spectral feature located in a second preset wavelength range based on the spectral image data. It then obtains composition parameters and thickness parameters based on the comparison relationship between the first spectral feature and the second spectral feature. Finally, it analyzes the scattered light intensity based on the laser scattering data to obtain scattering parameters. The index conversion module converts the composition parameters, thickness parameters, and scattering parameters into composition indices, thickness indices, and density indices by referring to the calibration database; The judgment and detection module constructs judgment rules containing multiple judgment channels. The judgment channels are used to judge the composition index, thickness index, and density index respectively. Each judgment channel outputs an initial detection signal. All the initial detection signals are integrated to generate a detection result. The judgment rules include composition judgment channels, thickness judgment channels, and density judgment channels. The component determination channel is used to compare the component index with a preset component threshold to determine the main component category of the oxide layer; the thickness determination channel is used to compare the thickness index with a preset thickness threshold range to determine the thickness of the oxide layer; the density determination channel is used to compare the density index with a preset density threshold to determine the uniformity and density of the oxide layer structure.

Images