A method and system for detecting oxygen-free copper oxygen content combination

Abstract

The present application relates to the technical field of material detection, and particularly relates to a combined detection method and system for oxygen content of oxygen-free copper, which comprises the following steps: manufacturing oxygen-free copper castings and oxygen-containing copper castings, and obtaining a plurality of oxygen-free copper samples and oxygen-containing copper samples with the same mass; preparing a plurality of sample groups, mixing the oxygen-free copper samples and the oxygen-containing copper samples in the sample groups according to gradient proportions, and recording the mass fractions of the oxygen-free copper samples and the oxygen-containing copper samples in the sample groups; detecting the oxygen content of each sample group; constructing a design matrix, constructing a measurement value vector according to the oxygen content of each sample group; using a least square algorithm to solve a parameter estimation vector, and calculating the real oxygen content of the oxygen-free copper through the parameter estimation vector; verifying the credibility of the sample and the model, and if the verification is passed, calculating the uncertainty and outputting in combination with the real oxygen content of the oxygen-free copper. The present application can more accurately measure the oxygen content of the oxygen-free copper.

Description

Technical Field

[0001] This invention relates to the field of materials testing technology, and in particular to a method and system for combined detection of oxygen content in oxygen-free copper. Background Technology

[0002] Oxygen-free copper is a high-purity copper with extremely low levels of impurities, especially oxygen. This results in exceptionally high electrical and thermal conductivity, making it widely used in cutting-edge technology fields such as electronic devices and the semiconductor industry, where material performance requirements are extremely stringent. To ensure the performance of oxygen-free copper, its oxygen content needs to be tested. Current testing methods typically involve directly testing the oxygen-free copper product itself. However, due to the extremely low oxygen content, this places very high demands on the sensitivity of the testing environment and equipment. Furthermore, the test results are prone to significant errors due to instrument system bias and background signal interference. Therefore, a more accurate method and system for detecting the oxygen content of oxygen-free copper is needed. Summary of the Invention

[0003] This invention provides a method and system for combined detection of oxygen content in oxygen-free copper, which can effectively solve the problems in the background art.

[0004] This invention provides a method for combined detection of oxygen content in oxygen-free copper, comprising the following steps:

[0005] To produce oxygen-free copper castings and oxygen-containing copper castings, several oxygen-free copper samples and oxygen-containing copper samples of the same mass are obtained from the oxygen-free copper castings and oxygen-containing copper castings.

[0006] Multiple sample groups were prepared, and oxygen-free copper samples and oxygen-containing copper samples were mixed in a gradient ratio in the multiple sample groups. The mass fraction of oxygen-free copper samples and oxygen-containing copper samples in each sample group was recorded.

[0007] The oxygen content of each sample group was measured;

[0008] A design matrix is ​​constructed based on the mass fractions of oxygen-free and oxygen-containing copper samples in various sample groups, and a measurement value vector is constructed based on the oxygen content corresponding to each sample group.

[0009] The least squares algorithm is used to solve for the parameter estimation vector, and the true oxygen content of oxygen-free copper is calculated using the parameter estimation vector.

[0010] Perform credibility verification on the sample and model. If the verification passes, calculate the uncertainty and output it in conjunction with the actual oxygen content of oxygen-free copper.

[0011] Furthermore, the specific steps for preparing multiple sample groups include:

[0012] A total of N+1 sample groups were prepared. In the j-th sample group, there were N+1-j oxygen-free copper samples and j-1 oxygen-containing copper samples.

[0013] Furthermore, the specific construction method for the design matrix and measurement value vector is as follows:

[0014] Calculate the mass fraction of oxygen-free copper in the j-th sample group as αj=(N+1-j) / N, and the mass fraction of oxygen-containing copper as βj=(j-1) / N;

[0015] Design matrix X as an (N+1) row 3 column matrix, where the elements in the j-th row are [1, αj, βj];

[0016] Measurement vector O = [o1, o2, ..., o(N+1)] T , where oj is the oxygen content detected in the j-th sample group.

[0017] Furthermore, the specific steps for calculating the true oxygen content of oxygen-free copper include:

[0018] Constructing a modified equation for the least squares algorithm:

[0019] (X T ·W·X+λ·I)·θ=X T ·W·O;

[0020] Where W is the set weight matrix;

[0021] λ is the regularization coefficient;

[0022] I is the identity matrix;

[0023] θ is the parameter estimation vector, denoted as θ = [γ, U, V]. T ;

[0024] Solve the least squares algorithm transformed equations to obtain the specific values ​​of γ, U, and V;

[0025] The formula for calculating the true oxygen content (ou) of oxygen-free copper is:

[0026] ou = U·o(N+1) / V.

[0027] Furthermore, at least three samples of each type are prepared. When detecting oxygen content, the average value of the oxygen content detected in all samples corresponding to each type of sample group is calculated, and this average value is taken as the final oxygen content of that type of sample group.

[0028] The variance of oxygen content among multiple sample groups of the j-th sample group is calculated as σj²; the element value of the j-th row and j-th column of the weight matrix W is 1 / σj².

[0029] Furthermore, the specific steps for verifying sample credibility include:

[0030] Calculate the first-order difference Δoj = oj - o(j-1);

[0031] Calculate the second-order difference ΔDoj = Δoj - Δo(j-1);

[0032] Calculate the mean Δoavg and standard deviation σΔo of all Δoj. If σΔo / Δoavg < 0.05, then the first-order validation passes.

[0033] Calculate the mean △Doavg and standard deviation σ△Do for all △Doj. If |△Doavg| < 2σ△Do, then the second-order verification passes.

[0034] The sample credibility verification is successful only if both the first-order and second-order verifications pass.

[0035] Furthermore, the specific steps for model credibility verification include:

[0036] Calculate the judgment coefficient R = 1 - SR / ST;

[0037] Where, SR = (OX·θ) T ·W·(OX·θ);

[0038] ST=(O-omavg) T ·W·(O-omavg);

[0039] omavg is the weighted average of all measurements o;

[0040] If the value of R is less than the set threshold, the verification fails.

[0041] Furthermore, the specific algorithm for uncertainty is as follows:

[0042] Calculate the residual variance σθ 2 =(OX·θ) T ·W·(OX·θ) / (N+1-3);

[0043] Calculate the covariance matrix Σθ;

[0044] Σθ=σθ 2 ·(X T ·W·X+λ·I) -1 ·X T ·W·X·(X T ·W·X+λ·I) -1 ;

[0045] Extract the element from the second row and second column of matrix Σθ and denote it as uA(U);

[0046] Extract the element in the 3rd row and 3rd column of matrix Σθ and denote it as uA(V);

[0047] Extract the element in the 2nd row and 3rd column of matrix Σθ and denote it as cov(U,V);

[0048] The formula for calculating the uncertainty uc is:

[0049] ;

[0050] The final output value of the true oxygen content of oxygen-free copper is ou±uc.

[0051] Furthermore, the specific steps for detecting the oxygen content of each sample group include:

[0052] The sample group to be tested is placed into a graphite measuring chamber;

[0053] The sample group is heated by pulsed current to melt it, and then high-purity nitrogen gas is introduced into the measurement chamber, and the gas flowing out of the measurement chamber is introduced into the laser cell.

[0054] After storing a certain amount of gas in the laser pool, the degree of attenuation of the laser energy in the laser pool is detected, which is then converted into oxygen content.

[0055] The present invention also provides a combined detection system for oxygen-free copper and oxygen content, including a storage device and a processor. The storage device is used to store one or more program instructions; the processor is used to run one or more program instructions to perform the steps of the combined detection method for oxygen-free copper and oxygen content as described above.

[0056] The technical solution of this invention can achieve the following technical effects:

[0057] This method transforms the problem of measuring the minute oxygen content in oxygen-free copper into the problem of using multiple sample groups prepared by mixing oxygen-free copper and oxygen-containing copper in a gradient ratio. This allows for the quantitative analysis of the combination relationship between oxygen-free copper and oxygen-containing copper by using the increase in oxygen content. This enables high-precision analysis of the oxygen content in oxygen-free copper. 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 flowchart of the combined detection method for oxygen-free copper oxygen content in this invention. Detailed Implementation

[0060] The basic principles and main features of the technical solution of the present invention will be described below with reference to the accompanying drawings of the embodiments of the present invention. The following description will use one or more embodiments for a more intuitive understanding. These embodiments are merely some, not all, of the embodiments of the present invention.

[0061] In the description of this invention, the terms indicating orientation or positional relationship (such as up, down, left, right, etc.) are based on the orientation shown in the drawings or some conventional positional relationships, and are only for the convenience of describing this invention and simplifying the description, and are not intended to indicate or imply that the features referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as limiting this invention.

[0062] A combined detection method for oxygen-free copper content, such as Figure 1 As shown, the main steps include sample preparation, oxygen content detection, data analysis, and result output. The specific details of each step are as follows:

[0063] Sample preparation: First, prepare the oxygen-free copper casting and oxygen-containing copper casting required for testing. The oxygen content of the oxygen-containing copper casting should be relatively high. Cut the oxygen-free copper casting and oxygen-containing copper casting into several oxygen-free copper sample blocks and oxygen-containing copper sample blocks of the same mass.

[0064] Multiple sample groups were prepared, each with a different mixing ratio of oxygen-free copper and oxygen-containing copper samples. The oxygen-free and oxygen-containing copper samples in the multiple sample groups were mixed in a gradient ratio to create a series of sample groups with gradually increasing oxygen content. Since the oxygen content of castings produced in the same furnace is usually consistent, the comparability of oxygen-free and oxygen-containing copper samples in different sample groups was ensured by obtaining all oxygen-free copper samples from the same oxygen-free copper casting and all oxygen-containing copper samples from the same oxygen-containing copper casting.

[0065] The specific steps for preparing multiple sample groups are as follows:

[0066] Suppose that N+1 sample groups need to be prepared in total;

[0067] The first sample group includes N oxygen-free copper samples and 0 oxygen-containing copper samples;

[0068] The second sample group includes N-1 oxygen-free copper samples and 1 oxygen-containing copper sample.

[0069] ...

[0070] The j-th sample group includes N+1-j oxygen-free copper samples and j-1 oxygen-containing copper samples;

[0071] ...

[0072] The N+1th sample group includes 0 oxygen-free copper samples and N oxygen-containing copper samples.

[0073] When preparing the sample groups, the mass fractions of oxygen-free copper and oxygen-containing copper samples in each sample group were recorded simultaneously. The mass fractions mainly reflect the proportions of oxygen-free copper and oxygen-containing copper samples in the sample group. Since the mass of oxygen-free copper and oxygen-containing copper samples is the same, the proportion between oxygen-free copper and oxygen-containing copper is mainly related to the number of samples. That is, the mass fraction of oxygen-free copper is αj=(N+1-j) / N, and the mass fraction of oxygen-containing copper is βj=(j-1) / N.

[0074] Instruments for detecting oxygen content usually have an optimal range that is smaller than their standard range. Within this optimal range, the instrument's detection results are more accurate. Therefore, when designing N+1 sample groups, it is important to ensure that the estimated oxygen content of the sample group containing oxygen-containing copper blocks is within the instrument's optimal range.

[0075] Oxygen content detection: The oxygen content of each sample group is detected. Before detection, oxygen-free copper samples and oxygen-containing copper samples should be acid-washed (e.g., soaked in dilute hydrochloric acid or nitric acid solution) to remove the surface oxide layer, then rinsed with deionized water and dried to eliminate the interference of surface oxides on oxygen content detection. If the sample is contaminated with grease or organic matter, an organic solvent (e.g., acetone) cleaning step can be added, followed by rinsing with deionized water and drying to remove these oxygen-containing contaminants.

[0076] Data analysis and output: After obtaining the oxygen content of various sample groups, a design matrix is ​​constructed based on the mass fraction of oxygen-free copper and oxygen-containing copper samples in each sample group. The design matrix is ​​used to establish the relationship between the background noise of the detection equipment, the mass fraction of oxygen-free copper, and the mass fraction of oxygen-containing copper. A measurement value vector is constructed based on the oxygen content of each sample group.

[0077] The least squares algorithm is used to solve for the parameter estimation vector. The least squares algorithm is a fitting algorithm that can infer the relationship between multiple parameters based on the results. In this method, the principle of the least squares algorithm is utilized. However, the relationship between background noise, oxygen-free copper mass fraction and oxygenated copper mass fraction has been determined in this method. Therefore, this method forms a hypothetical parameter estimation vector based on the characteristics of the true oxygen content in this sample group. This hypothetical parameter estimation vector is used in the least squares algorithm to obtain a series of simultaneous corresponding equations. The variables in the equations are the elements in the parameter estimation vector. Through these simultaneous corresponding equations, the elements in the parameter estimation vector can be solved, and then the true oxygen content of oxygen-free copper can be restored.

[0078] Before outputting the calculation results, the reliability of the sample and model is verified using the calculated true oxygen content of oxygen-free copper. If the verification is successful, the uncertainty is calculated and output in conjunction with the true oxygen content of oxygen-free copper.

[0079] This method transforms the problem of measuring minute oxygen content in oxygen-free copper into a problem of using the incremental oxygen content to infer the quantitative combination relationship between oxygen-free and oxygen-containing copper by mixing and detecting oxygen-free copper and oxygen-containing copper in a gradient ratio. This allows for high-precision analysis of oxygen content in oxygen-free copper. Furthermore, this analysis can eliminate the influence of oxygen-containing copper, and the oxygen content of oxygen-containing copper itself can also be measured by this method. Therefore, it does not require overly stringent requirements on the actual oxygen content of oxygen-containing copper, making it easier to prepare oxygen-containing copper samples and effectively saving detection costs.

[0080] The preferred method for constructing the design matrix and measurement value vector is as follows:

[0081] Calculate the mass fraction of oxygen-free copper in the j-th sample group as αj=(N+1-j) / N, and the mass fraction of oxygen-containing copper as βj=(j-1) / N;

[0082] The design matrix X is an (N+1) row and 3 column matrix, where the elements in the j-th row are [1, αj, βj]. Each row of elements takes into account the background signal (first column) which is assumed to remain unchanged, the proportion of oxygen-free copper in the sample group (second column), and the proportion of oxygen-containing copper in the sample group (third column).

[0083] Measurement vector O = [o1, o2, ..., o(N+1)] T , where oj is the oxygen content detected in the j-th sample group.

[0084] Preferably, the specific steps for calculating the true oxygen content of oxygen-free copper include:

[0085] Let θ be the parameter estimation vector, denoted as θ = [γ, U, V] T Where γ represents the instrument’s systematic bias, including background signal, zero drift and other systematic influencing factors; U represents the degree of influence of oxygen-free copper on the measured value; V represents the degree of influence of oxygenated copper on the measured value; under the current procedure, the above three values ​​are all assumed values.

[0086] Based on the principle of the least squares algorithm, by introducing the measurement vector O and the weight matrix W, the equation of the least squares algorithm is transformed to obtain the following equation:

[0087] (X T ·W·X+λ·I)·θ=X T ·W·O;

[0088] Where W is the set weight matrix, which is an (N+1)×(N+1) matrix. Each element on the diagonal of the matrix is ​​a set weight value, which corresponds to the importance of each sample group. All other elements of the matrix are 0.

[0089] λ is a regularization coefficient set by the user. A suitable value of λ can minimize the prediction error of the model.

[0090] I is a 3×3 identity matrix, meaning that all elements on the diagonal of the matrix are 1, and all other elements of the matrix are 0.

[0091] In the above equation, the left side represents the theoretical oxygen content calculated under the influence of the background signal of the measuring instrument, oxygen-free copper, and oxygen-containing copper. The right side represents the actual oxygen content detected. By correlating the theoretical and actual results, the parameter estimation vector θ=[γ,U,V] can be obtained. T The specific values ​​of each element in the text.

[0092] Then, the true oxygen content ou of oxygen-free copper is calculated using the formula: ou=U·o(N+1) / V; where o(N+1) is the oxygen content in the N+1th sample group. Since there is no oxygen-free copper in this group, this value can be directly used as the true oxygen content of oxygenated copper if the sample verification is passed.

[0093] Preferably, at least three samples of each type are prepared. When detecting oxygen content, the average value of the oxygen content detected by all samples corresponding to each type of sample group is calculated. This average value is used as the final oxygen content of that type of sample group, thereby making the measured oxygen content of each sample group closer to the accurate value and minimizing the error. Furthermore, the weight matrix W can also be automatically generated and updated in this way. In the weight matrix W, the variance of the oxygen content of multiple sample groups of the j-th type of sample group is calculated as σj², and the element value of the j-th row and j-th column of the weight matrix W is 1 / σj². Thus, if the variance of the oxygen content of multiple sample groups within a sample group is smaller, it indicates that the measurement value of this sample group is more stable, and its corresponding weight value will increase, thereby increasing its influence in the less squares algorithm transformation equation and improving the accuracy of the calculation.

[0094] Preferably, the specific steps for sample credibility verification include:

[0095] Calculate the first-order difference Δoj = oj - o(j-1), where j ranges from 2 to N+1. Δoj represents the gradient of oxygen content changes detected in adjacent sample groups. Since the changes in the amount of oxygen-free and oxygen-containing copper in the sample group are regular, the gradient should be relatively uniform. Calculate the average value Δoavg and standard deviation σΔo of all Δoj. If σΔo / Δoavg < 0.05, it indicates a uniform gradient, and the first-order verification of the sample group passes. If the gradient is not uniform, it is highly likely that the oxygen content of the oxygen-free or oxygen-containing copper samples in the sample group is inconsistent, or that the instrument detection error is too large, requiring corresponding adjustments and re-detection.

[0096] Calculate the second-order difference ΔDoj = Δoj - Δo(j-1), where j ranges from 3 to N+1. ΔDoj is mainly used to determine the difference between the changing gradients, thereby determining whether the instrument has excessive instability during detection. Calculate the average ΔDoavg and standard deviation σΔDo of all ΔDoj. If |ΔDoavg| < 2σΔDo, it indicates that the difference between the changing gradients has not changed significantly, and the second-order verification is passed. However, if the difference between the changing gradients has changed significantly, it indicates that the instrument's detection results for the sample group are consistently too large or too small, or the offset is too large, requiring instrument adjustment and re-detection.

[0097] Only when both the first-order and second-order verifications pass can the sample credibility verification be passed, thus confirming that there are no obvious problems with the sample group and the instrument. Only then will subsequent model credibility verification be of practical significance.

[0098] Preferably, the specific steps for model credibility verification include:

[0099] Calculate the judgment coefficient R = 1 - SR / ST;

[0100] Where, SR = (OX·θ) T ·W·(OX·θ) represents the sum of squared residuals of the model, which is used to quantify the degree of discrepancy between the theoretical values ​​calculated in the model and the measured values.

[0101] ST=(O-omavg) T ·W·(O-omavg) represents the total sum of squares of the model, used to quantify the degree of discrepancy between the weighted average of the measured values ​​in the model and the total measured values;

[0102] omavg is the weighted average of all measured values ​​o. The specific algorithm for omavg is as follows:

[0103] ;

[0104] Where wj is the element value on the diagonal of the j-th row and j-th column in the weight matrix W, and oj is the measured value of the j-th sample group.

[0105] The physical meaning of the coefficient of determination R is the percentage of the total variation of the data that the model can explain. The more perfect the result calculated by the model, the closer the degree of non-correspondence will be to 0 (that is, the value of SR will be closer to 0), and the closer the coefficient of determination R will be to 1. If the value of R is less than the set threshold, it means that the result calculated by the model is too bad, and the verification cannot be passed. The values ​​in the model (such as λ) need to be adjusted and recalculated.

[0106] The preferred algorithm for uncertainty is as follows:

[0107] Calculate the residual variance σθ 2 =(OX·θ) T ·W·(OX·θ) / (N+1-3);

[0108] Calculate the covariance matrix Σθ;

[0109] Σθ=σθ 2 ·(X T ·W·X+λ·I) -1 ·X T ·W·X·(X T ·W·X+λ·I) -1 If we consider θ as an assumed value, then the final elements of Σθ will be expressions with γ, U, and V as variables.

[0110] Extract the element from the second row and second column of matrix Σθ and denote it as uA(U);

[0111] Extract the element in the 3rd row and 3rd column of matrix Σθ and denote it as uA(V);

[0112] Extract the element in the 2nd row and 3rd column of matrix Σθ and denote it as cov(U,V);

[0113] In the aforementioned calculation process, we have now obtained θ = [γ, U, V]. T Given the specific values ​​of each element, the uncertainty uc can be calculated.

[0114] The formula for calculating the uncertainty uc is:

[0115] ;

[0116] The final output value of the true oxygen content of oxygen-free copper is ou±uc.

[0117] Preferably, the specific steps for detecting the oxygen content of each sample group include:

[0118] The sample group to be tested is placed into a graphite measuring chamber;

[0119] The sample group is heated by pulsed current to melt it. During the melting process, oxygen combines with carbon in the graphite measurement cavity to form CO gas, which is then released. High-purity nitrogen gas is then introduced into the measurement cavity, and the nitrogen gas carries the CO gas out of the measurement cavity. The outflowing gas is then introduced into the laser cell.

[0120] After storing a certain amount of gas in the laser cell, the CO gas absorbs and blocks the laser energy. By detecting the degree of laser energy attenuation in the laser cell, the analog signal is converted into a digital signal for integration and various corrections to indirectly measure the O content in the laser cell, thereby obtaining the oxygen content. Before measuring each sample group, the measurement chamber is first flushed with high-purity nitrogen to eliminate the influence of residual gas.

[0121] The present invention also relates to a combined detection system for oxygen-free copper and oxygen content, comprising a storage device and a processor. The storage device is used to store one or more program instructions; the processor is used to run one or more program instructions to perform the steps of the above-described combined detection method for oxygen-free copper and oxygen content.

[0122] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present 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 the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for detecting oxygen and copper content in a combination, characterized by the steps of The application relates to a method for calculating the oxygen content of oxygen-free copper. The method comprises the following steps: manufacturing oxygen-free copper castings and oxygen-containing copper castings, and obtaining a plurality of oxygen-free copper samples and oxygen-containing copper samples with the same mass from the oxygen-free copper castings and the oxygen-containing copper castings; preparing a plurality of sample groups, mixing the oxygen-free copper samples and the oxygen-containing copper samples in the sample groups according to a gradient ratio, and recording the mass fractions of the oxygen-free copper samples and the oxygen-containing copper samples in the sample groups; a total of N+1 sample groups are prepared, and the jth sample group comprises N+1-j oxygen-free copper samples and j-1 oxygen-containing copper samples; detecting the oxygen content of each sample group; constructing a design matrix according to the mass fractions of the oxygen-free copper samples and the oxygen-containing copper samples in the sample groups, and constructing a measured value vector according to the oxygen contents of the sample groups; solving a parameter estimation vector by using a least square algorithm, and calculating the real oxygen content of the oxygen-free copper by using the parameter estimation vector; performing credibility verification of the sample and the model, calculating the uncertainty if the verification is passed, and outputting the real oxygen content of the oxygen-free copper in combination with the real oxygen content of the oxygen-free copper; wherein the specific construction method of the design matrix and the measured value vector is as follows: calculating the mass fraction of the oxygen-free copper in the jth sample group as alpha j = (N+1-j) / N, and the mass fraction of the oxygen-containing copper as beta j = (j-1) / N; The measured value vector O = [o1, o2... o(N+1)] T where oj is the oxygen content detected by the jth sample group; the design matrix X is a (N+1) * 3 matrix, and the elements of the jth row are [1, alpha j, beta j]; the specific steps for calculating the real oxygen content of the oxygen-free copper include: (X T • W · X + λ · I) · θ = X T • W · O; constructing a deformation equation of the least square algorithm: wherein W is a set weight matrix; lambda is a regularization coefficient; θ is a parameter estimation vector, denoted as θ = [γ, U, V] T ; I is a unit matrix; gamma is a system deviation of an instrument; U represents the influence degree of the oxygen-free copper on the measured value; and V represents the influence degree of the oxygen-containing copper on the measured value; solving the deformation equation of the least square algorithm to obtain the specific values of gamma, U and V; the calculation formula of the real oxygen content ou of the oxygen-free copper is: ou = U * o (N+1) / V; Compute residual variance σθ 2 = (O - X · θ) T · W · (O - X · θ) / (N + 1 - 3); the specific algorithm of the uncertainty is as follows: ∑θ = σθ 2 • (X T • W • X + λ • I) -1 • X T • W • X • (X T • W • X + λ • I) -1 ; calculating a covariance matrix Sigma theta; extracting the element of the second row and the second column of the matrix Sigma theta and marking the element as uA (U); extracting the element of the third row and the third column of the matrix Sigma theta and marking the element as uA (V); extracting the element of the second row and the third column of the matrix Sigma theta and marking the element as cov (U, V); ； the calculation formula of the uncertainty uc is:

2. The method of claim 1, wherein the method is performed without oxygen. the output value of the real oxygen content of the oxygen-free copper is ou+uc. At least three sample groups are prepared for each sample group, the oxygen contents of all the sample groups corresponding to each sample group are detected, and the average value of the oxygen contents is taken as the final oxygen content of the sample group; the oxygen content variance of the jth sample group is calculated as sigma j 2; 3. The method of claim 1, wherein the method is performed without oxygen. the element value of the jth row and the jth column of the weight matrix W is 1 / sigma j 2. The specific steps of the sample credibility verification include: calculating a first-order difference delta oj = oj-o (j-1); calculating a second-order difference delta Doj = delta oj-delta o (j-1); calculating the average value delta oavg and the standard deviation sigma delta o of all delta ojs, and if sigma delta o / delta oavg < 0.05, the first-order verification is passed; calculating the average value delta Doavg and the standard deviation sigma delta Do of all delta Dojs, and if |delta Doavg| < 2 * sigma delta Do, the second-order verification is passed; only when the first-order and the second-order verifications are both passed, the sample credibility verification is passed.

4. The method of claim 2, wherein the method is performed without oxygen. The model credibility verification specific steps include: Calculate the determination coefficient R=1-SR / ST; where SR= (O - X θ) T W (O - X θ); ST = (O - omavg) T • W • (O - omavg); omavg is the weighted average of all measured values o; If the value of R is less than the set threshold, the verification fails.

5. The method of claim 1, wherein the method is performed without oxygen. The specific steps of detecting the oxygen content of each sample group include: Put the sample group to be detected into a graphite measuring cavity; Use pulse current to heat the sample group to melt it, then introduce high-purity nitrogen into the measuring cavity, and pass the gas flowing out of the measuring cavity into a laser cell; After a certain amount of gas is stored in the laser cell, the degree of attenuation of the laser energy in the laser cell at this time is detected, which is further converted into oxygen content.

6. A system for detecting oxygen-free copper and oxygen content, characterized by, The device comprises a storage and a processor, the storage is used for storing one or more program instructions; the processor is used for running one or more program instructions, and is used for executing the steps of the oxygen-free copper oxygen content combined detection method in any one of claims 1-5.

Images