Method for determining mechanical property of cable insulation layer based on near infrared spectrum data

Abstract

The invention discloses a method and a system for determining the mechanical property of a cable insulation layer based on near infrared spectrum data. The method comprises the following steps: acquiring near infrared spectrum data and mechanical property data of a cable sample; wherein the mechanical property data comprises the elongation at break, the tensile strength and the elastic modulus; preprocessing the near infrared spectrum data to obtain near infrared spectrum processing data; classifying the near infrared spectrum data and the mechanical performance data to determine a sample set and a verification set; performing wavelength selection based on the sample set, and determining a target wavelength point and a weight corresponding to the target wavelength point; and based on the target wavelength point and the weight corresponding to the target wavelength point, determining a quantitative analysis model of the mechanical properties of the cable insulation layer, so as to determine the mechanical properties of the cable insulation layer based on the quantitative analysis model. According to the method, the elongation at break, the tensile strength and the elastic modulus of the cable insulation layer can be rapidly detected at the same time, and the method is suitable for on-site full-coverage testing of the mechanical performance of the cable.

Description

Technical Field

[0001] The present invention relates to the technical field of non-destructive testing of power cables by near-infrared spectroscopy, and more particularly, to a method and system for determining the mechanical properties of a cable insulation layer based on near-infrared spectroscopy data. Background Art

[0002] The mechanical property indexes of XLPE in cables include: tensile strength, elongation at break, and elastic modulus. The tests for these indexes can refer to ASTM methods such as ASTM D638 (dumbbell-shaped specimens), ASTM D882 (plastic films and thin sheets), ASTM D3039 (fiber-reinforced composites), ASTM D2637 (plastic materials at high temperatures), and ASTM D4895 (continuous and discontinuous fiber-reinforced materials). At present, the current test method for XLPE mechanical property indexes in China is the general test method in GB / T2951.1, which corresponds to ASTM D638 and is also a general test method for the mechanical properties of polymer materials. It has the following deficiencies: 1) Using a mechanical tensile test, steps such as sampling - making specimens - pressing dumbbell shapes - testing are required for cable segments. The experimental steps are complicated, involving various instruments and taking a long time; 2) The sampling amount is large, and at least 500 mm of cable is used for testing a single sample. Obviously, this method cannot be used for on-site testing scenarios of cable insulation mechanical properties, nor can it be used to achieve the goal of quickly judging and evaluating the performance of in-service cables in the construction of smart grids.

[0003] Therefore, researching a method for quickly and simultaneously testing cable insulation mechanical property indexes is of great significance for the quality control of power grid cables. A method for determining the mechanical properties of a cable insulation layer based on near-infrared spectroscopy data is needed. Summary of the Invention

[0004] The present invention provides a method and system for determining the mechanical properties of a cable insulation layer based on near-infrared spectroscopy data to solve the problem of how to efficiently detect the performance of cable insulation materials.

[0005] To solve the above problems, according to one aspect of the present invention, a method for determining the mechanical properties of a cable insulation layer based on near-infrared spectroscopy data is provided. The method includes:

[0006] Obtaining near-infrared spectroscopy data and mechanical property data of a cable sample; wherein, the mechanical property data includes: elongation at break, tensile strength, and elastic modulus;

[0007] Preprocessing the near-infrared spectroscopy data to obtain processed near-infrared spectroscopy data;

[0008] Classifying the processed near-infrared spectroscopy data and mechanical property data to determine a sample set and a validation set;

[0009] Perform wavelength selection based on the sample set to determine the target wavelength points and the weights corresponding to the target wavelength points;

[0010] Based on the target wavelength points and the weights corresponding to the target wavelength points, determine a quantitative analysis model for the mechanical properties of the cable insulation layer, and determine the mechanical properties of the cable insulation layer based on the quantitative analysis model.

[0011] Preferably, preprocess the near-infrared spectral data to obtain processed near-infrared spectral data, including:

[0012] Use first-order differentiation to eliminate spectral baseline drift, and use Savitzky-Golay convolution to smooth the near-infrared spectral data to obtain processed near-infrared spectral data.

[0013] Preferably, classify the processed near-infrared spectral data and mechanical property data to determine the sample set and the validation set, including:

[0014] Adopt the Kennard-Stone method, and uniformly select samples in the feature space based on the Euclidean distance of the near-infrared spectrum for sample classification to determine the sample set and the validation set.

[0015] Preferably, perform wavelength selection based on the sample set to determine the target wavelength points and the weights corresponding to the target wavelength points, including:

[0016] For any kind of mechanical property, determine the target wavelength points and the weights corresponding to the target wavelength points by the following method, including:

[0017] Decompose the performance parameter matrix y composed of any kind of mechanical property data into two mutually orthogonal parts, that is, y = P x y + A x y; where, P x = X(X T X) -1 X T is the orthogonal projection matrix, P x y projects y onto the column space of X, denoted as A x y projects y onto the orthogonal complement of the column space of X, where A x =(I - P x ), and I is the identity matrix;

[0018] Perform orthogonal decomposition on the spectral matrix X composed of processed infrared spectral data, that is, Use the principal component analysis method to extract the first n principal components T A =(t1, t2, t3,..., t A ) to calculate the loading vector And solve the equation Xw i =t i , to obtain the weight vector w i ;

[0019] Calibrate the spectral matrix To remove more information unrelated to y.

[0020] Sort |w i | from largest to smallest, and denote the sorted result as |w (j) |, |w (1) | is the largest, |w (2) | is the next, and so on;

[0021] For each j = 0, 1, 2,... i, select wavelengths And based on the selected wavelengths, establish a PLS model and calculate where k is the initial number of wavelength points, S is the step size; y i is the measured value; is the predicted value of the PLS model; n is the number of calibration set samples;

[0022] Select the model j0 with the minimum RMSECV, then determine the target wavelength point corresponding to any one of the mechanical properties as And determine the weight corresponding to the target wavelength point.

[0023] Preferably, based on the target wavelength point and the weight corresponding to the target wavelength point, a quantitative analysis model for determining the mechanical properties of the cable insulation layer includes:

[0024] For any one of the mechanical properties, the quantitative analysis model is: where a0 is the calibration coefficient, w i is the weight corresponding to the wavelength point, x i is the absorbance at the corresponding wavelength after pretreatment, and n is the modeling wavelength coefficient.

[0025] Preferably, the method further includes:

[0026] Based on the validation set for validation to determine the accuracy of the quantitative analysis model.

[0027] According to another aspect of the present invention, there is provided a system for determining the mechanical properties of a cable insulation layer based on near-infrared spectral data, the system includes:

[0028] A data acquisition unit for acquiring near-infrared spectral data and mechanical property data of a cable sample; wherein, the mechanical property data includes: elongation at break, tensile strength, and elastic modulus;

[0029] A preprocessing unit for preprocessing the near-infrared spectral data to obtain preprocessed near-infrared spectral data;

[0030] A classification unit for classifying the preprocessed near-infrared spectral data and mechanical property data to determine a sample set and a validation set;

[0031] A wavelength selection unit for performing wavelength selection based on the sample set to determine target wavelength points and weights corresponding to the target wavelength points;

[0032] A quantitative analysis model determination unit for determining a quantitative analysis model of the mechanical properties of the cable insulation layer based on the target wavelength points and the weights corresponding to the target wavelength points, so as to determine the mechanical properties of the cable insulation layer based on the quantitative analysis model.

[0033] Preferably, the preprocessing unit preprocesses the near-infrared spectral data to obtain preprocessed near-infrared spectral data, including:

[0034] Using first-order differentiation to eliminate spectral baseline drift and using Savitzky-Golay convolution to smooth the near-infrared spectral data to obtain preprocessed near-infrared spectral data.

[0035] Preferably, the classification unit classifies the preprocessed near-infrared spectral data and mechanical property data to determine a sample set and a validation set, including:

[0036] Using the Kennard-Stone system to uniformly select samples in the feature space based on the Euclidean distance of the near-infrared spectrum for sample classification to determine a sample set and a validation set.

[0037] Preferably, the wavelength selection unit performs wavelength selection based on the sample set to determine target wavelength points and weights corresponding to the target wavelength points, including:

[0038] For any kind of mechanical property, the following method is used to determine the target wavelength points and the weights corresponding to the target wavelength points, including:

[0039] Decompose the performance parameter matrix y composed of any kind of mechanical property data into two mutually orthogonal parts, that is, y = P x y + A x y; where P x = X(X T X) -1 X T is an orthogonal projection matrix, P x y projects y onto the column space of X, denoted as A x y then projects y onto the orthogonal complement of the column space of X, where A x = (I - P x), where I is the identity matrix;

[0040] Perform orthogonal decomposition on the spectral matrix X composed of infrared spectrum processed data, that is Extract the first n principal components T using the principal component analysis method A =(t1, t2, t3,..., t A ) to calculate the loading vector and solve the equation Xw i =t i , to obtain the weight vector w i ;

[0041] Correct the spectral matrix to remove more information unrelated to y.

[0042] Sort |w i | from largest to smallest, and denote it as |w (j) | after sorting, |w (1) | is the largest, |w (2) | is the second largest, and so on;

[0043] For each j = 0, 1, 2,... i, select the wavelength and based on the selected wavelength, establish a PLS model and calculate where k is the initial number of wavelength points, S is the step size; y i is the measured value; is the predicted value of the PLS model; n is the number of samples in the calibration set;

[0044] Select the model j0 with the minimum RMSECV, then determine the target wavelength point corresponding to any one of the mechanical properties as and determine the weight corresponding to the target wavelength point.

[0045] Preferably, the quantitative analysis model determination unit, based on the target wavelength point and the weight corresponding to the target wavelength point, determines a quantitative analysis model for the mechanical properties of the cable insulation layer, including:

[0046] For any one of the mechanical properties, the quantitative analysis model is: where a0 is the calibration coefficient, w i is the weight corresponding to the wavelength point, x i is the absorbance at the corresponding wavelength after pretreatment, and n is the wavelength coefficient for modeling.

[0047] Preferably, the system further includes:

[0048] A verification unit for verifying based on the verification set to determine the accuracy of the quantitative analysis model.

[0049] The present invention provides a method and system for determining the mechanical properties of a cable insulation layer based on near-infrared spectral data, including: obtaining the near-infrared spectral data and mechanical property data of a cable sample; wherein, the mechanical property data includes: elongation at break, tensile strength, and elastic modulus; preprocessing the near-infrared spectral data to obtain processed near-infrared spectral data; classifying the processed near-infrared spectral data and mechanical property data to determine a sample set and a validation set; performing wavelength selection based on the sample set to determine target wavelength points and weights corresponding to the target wavelength points; and determining a quantitative analysis model for the mechanical properties of the cable insulation layer based on the target wavelength points and the weights corresponding to the target wavelength points, so as to determine the mechanical properties of the cable insulation layer based on the quantitative analysis model. The method of the present invention can simultaneously and rapidly detect the elongation at break, tensile strength, and elastic modulus of the cable insulation layer, realizing the rapid detection of multiple mechanical property indexes of the insulation layer, and is suitable for on-site full-coverage testing of the mechanical properties of cables. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The exemplary embodiments of the present invention can be more fully understood by referring to the following drawings:

[0051] Figure 1 FIG. 100 is a flowchart of a method for determining the mechanical properties of a cable insulation layer based on near-infrared spectral data according to an embodiment of the present invention;

[0052] Figure 2 FIG. 200 is a schematic structural diagram of a system for determining the mechanical properties of a cable insulation layer based on near-infrared spectral data according to an embodiment of the present invention. DETAILED DESCRIPTION

[0053] The exemplary embodiments of the present invention will now be described with reference to the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. These embodiments are provided to disclose the present invention in detail and completely, and to fully convey the scope of the present invention to those skilled in the art. The terms used in the exemplary embodiments shown in the drawings are not intended to limit the present invention. In the drawings, the same unit / element is denoted by the same reference numeral.

[0054] Unless otherwise specified, the terms (including scientific and technical terms) used herein have the ordinary meaning understood by those skilled in the art. In addition, it can be understood that the terms defined in a commonly used dictionary should be understood as having a meaning consistent with the context of their related fields, and should not be understood as having an idealized or overly formal meaning.

[0055] The cable insulation material is composed of polyolefins, containing -CH3, -CH2, and -CH groups, all of which have certain overtone band assignments in the near-infrared spectrum. The mechanical properties of the material are closely related to crystallinity, lamellar thickness, long helical structure, crosslinking degree, etc. The content of the long helical structure corresponding to crystallinity and lamellar thickness, the content of terminal methyl groups corresponding to molecular weight, and the number of tertiary and quaternary carbon atoms corresponding to crosslinking degree and other structural parameters are reflected in the near-infrared spectrum. Therefore, the near-infrared spectrum can be used to judge the mechanical properties of the material.

[0056] Figure 1 FIG. is a flowchart of a method 100 for determining the mechanical properties of a cable insulation layer based on near-infrared spectral data according to an embodiment of the present invention. As Figure 1 shown, the method for determining the mechanical properties of a cable insulation layer based on near-infrared spectral data provided by the embodiment of the present invention can simultaneously and quickly detect the elongation at break, tensile strength, and elastic modulus of the cable insulation layer, realizing the rapid detection of multiple mechanical property indexes of the insulation layer, and is suitable for on-site full-coverage testing of the mechanical properties of cables. The method 100 for determining the mechanical properties of a cable insulation layer based on near-infrared spectral data provided by the embodiment of the present invention starts at step 101. At step 101, near-infrared spectral data and mechanical property data of a cable sample are obtained; wherein, the mechanical property data includes: elongation at break, tensile strength, and elastic modulus.

[0057] At step 102, the near-infrared spectral data is preprocessed to obtain near-infrared spectral processing data. Preferably, preprocessing the near-infrared spectral data to obtain near-infrared spectral processing data includes:

[0058] Using first-order differentiation to eliminate spectral baseline drift, and using Savitzky-Golay convolution to smooth the near-infrared spectral data to obtain near-infrared spectral processing data.

[0059] In the present invention, at room temperature, after the spectrometer is turned on, it is preheated for 30 min, placed in a fiber optic probe holder, the probe is fixed using a bracket, and left to stand for 30 s. Due to the light transmission characteristics of the cable insulation layer, the near-infrared spectrum of the insulation at the cable cross-section is collected in a diffuse reflectance mode. Spectral range: 1100 - 2200 nm, resolution 0.4 nm, number of scans 25. Three spectra are collected at multiple points, and their average spectrum is taken as the sample spectrum. During preprocessing, first-order differentiation is used to eliminate spectral baseline drift, and Savitzky-Golay convolution is used to smooth the spectrum (window width is 5, using least squares fitting).

[0060] In the present invention, mechanical property data is obtained by testing the mechanical properties of a cable segment. Among them, the mechanical property indexes include: elongation at break, elastic modulus, and tensile strength. Samples are prepared according to GB / T 2951.1, and then multiple measurements are respectively carried out according to the test methods of GB / T 2951.1 to obtain mechanical property data.

[0061] In step 103, the near-infrared spectrum processed data and the mechanical property data are classified to determine a sample set and a validation set.

[0062] Preferably, classifying the near-infrared spectrum processed data and the mechanical property data to determine a sample set and a validation set includes:

[0063] The Kennard-Stone method is adopted to uniformly select samples in the feature space based on the Euclidean distance of the near-infrared spectrum for sample classification to determine the sample set and the validation set.

[0064] In the present invention, 4 / 5 of the original data is used for the modeling set, and 1 / 5 is used as the validation set. The Kennard-Stone (K-S) method is used to select samples for the modeling set. Based on the Euclidean distance of the near-infrared spectrum of the samples, samples are uniformly selected in the feature space, so that the samples in the modeling set contain more information on the insulating mechanical properties, improving the effectiveness and accuracy of the model.

[0065] In step 104, wavelength selection is performed based on the sample set to determine the target wavelength points and the weights corresponding to the target wavelength points.

[0066] Preferably, wavelength selection is performed based on the sample set to determine the target wavelength points and the weights corresponding to the target wavelength points, including:

[0067] For any one of the mechanical properties, the target wavelength points and the weights corresponding to the target wavelength points are determined by the following method, including:

[0068] The performance parameter matrix y composed of the mechanical property data of any one kind is decomposed into two mutually orthogonal parts, that is, y = P x y + A x y; where P x = X(X T X) -1 X T is the orthogonal projection matrix, P x y projects y onto the column space of X, denoted as A x y projects y onto the orthogonal complement of the column space of X, where A x =(I - P x ), and I is the identity matrix;

[0069] Perform orthogonal decomposition on the spectral matrix X composed of infrared spectrum processed data, that is Extract the first n principal components T using the principal component analysis method A =(t1, t2, t3,..., t A ) Calculate the loading vector and solve the equation Xw i =t i , to obtain the weight vector w i ;

[0070] Perform calibration on the spectral matrix to remove more information unrelated to y.

[0071] Sort |w i | from largest to smallest, and denote the sorted result as |w (j) |, |w (1) | is the largest, |w (2) | is the second largest, and so on;

[0072] For each j = 0, 1, 2,... i, select the wavelength and based on the selected wavelength, establish a PLS model and calculate where k is the initial number of wavelength points, S is the step size; y i is the measured value; is the predicted value of the PLS model; n is the number of calibration set samples;

[0073] Select the model j0 with the minimum RMSECV, then determine the target wavelength point corresponding to any one of the mechanical properties as and determine the weight corresponding to the target wavelength point.

[0074] In step 105, based on the target wavelength point and the weight corresponding to the target wavelength point, determine a quantitative analysis model for the mechanical properties of the cable insulation layer, so as to determine the mechanical properties of the cable insulation layer based on the quantitative analysis model.

[0075] Preferably, where the determining of the quantitative analysis model for the mechanical properties of the cable insulation layer based on the target wavelength point and the weight corresponding to the target wavelength point includes: for any one of the mechanical properties, the quantitative analysis model is: where a0 is the calibration coefficient, w i is the weight corresponding to the wavelength point, x i is the wavelength absorbance corresponding to after pretreatment, and n is the modeling wavelength coefficient.

[0076] In the present invention, for any one of the mechanical properties, the steps of wavelength selection include:

[0077] ①Decompose the performance parameter matrix y composed of any mechanical property data into two mutually orthogonal parts, i.e., y = P x y + A x y; where, P x = X(X T X) -1 X T is an orthogonal projection matrix, and P x y projects y onto the column space of X, denoted as A x y projects y onto the orthogonal complement of the column space of X, where A x = (I - P x ), and I is the identity matrix.

[0078] ②Perform orthogonal decomposition on the spectral matrix X, i.e., extract the first n principal components T using the principal component analysis method A = (t1, t2, t3,..., t A ) to calculate the loading vector and solve the equation Xw i = t i , obtaining the weight vector w i .

[0079] ③Correct the spectral matrix to remove more information unrelated to y.

[0080] ④Sort |w i | from largest to smallest, denoted as |w (j) | after sorting, |w (1) | is the largest,, |w (2) | is the next, and so on;.

[0081] ⑤For each j = 0, 1, 2,... N, select the wavelengths {i||w i | > |w (k+j×s) |}. Where k is the initial number of wavelength points, S is the step size, and at most k + N × s wavelength points are considered.. And based on the selected wavelengths, establish a PLS model and calculate where, y i is the measured value; is the predicted value of the PLS model; n is the number of calibration set samples.

[0082] ⑥Select the model j0 with the minimum RMSECV, then the finally selected target wavelength points are and determine the corresponding weights according to the target wavelength points.

[0083] Finally, the quantitative analysis model for the insulation mechanical properties obtained according to the above steps is: where, a0 is the calibration coefficient, wi is the weight corresponding to the wavelength point, x i is the wavelength absorbance corresponding after preprocessing, and n is the modeling wavelength coefficient.

[0084] Preferably, the method further includes:

[0085] Verify based on the validation set to determine the accuracy of the quantitative analysis model.

[0086] The method of the present invention can realize the performance detection of the insulating material of the cable outside the national standard range and improve the access quality of the power cable.

[0087] The following specifically illustrates the implementation manner of the present invention

[0088] A method for quickly measuring the elongation at break, tensile strength and elastic modulus of the cable insulation layer based on near-infrared spectral data. The specific process is as follows:

[0089] 1) Use a tool to remove the outer sheath, armor layer and waterproof layer of the cable, take out the single-core cable, and then cut it into 20 cm and 50 cm, which are respectively used for the sample preparation of spectral acquisition and mechanical property index detection.

[0090] 2) Collect the near-infrared spectra of 40 cable segments: At room temperature, after the spectrometer is turned on, preheat for 30 min, put it into the fiber optic probe holder, fix the probe with a bracket, and let it stand for 30 s. Due to the light transmission characteristics of the cable insulation layer, the diffuse reflection mode is used to collect the near-infrared spectra of the insulation at the cable cross-section. Spectral range: 1100 - 2200 nm, resolution 0.4 nm, number of scans 25. Collect 3 spectra at multiple points and take their average spectrum as the sample spectrum.

[0091] 3) Test the mechanical properties of 40 cable segments. The mechanical property indexes include elongation at break, elastic modulus, and tensile strength. Make specimens according to GB / T2951.1 to obtain five dumbbell-shaped specimens, and respectively measure 5 values for each index according to the test method of GB / T2951.1, take their average value as the index value, and retain the results of these 5 values at the same time.

[0092] 4) Spectral preprocessing: First, use the first-order derivative to eliminate the spectral baseline drift, and use the Savitzky-Golay convolution to smooth the spectrum (window width is 5, using the least squares fitting).

[0093] 5) Sample set classification: Four-fifths of the original data, i.e., the spectral and performance data of 32 cables, are used for the modeling set, and one-fifth, i.e., the spectral and performance data of 8 cables, are used as the validation set. The Kennard-Stone (K-S) method is used to select the samples in the modeling set. Based on the Euclidean distance of the near-infrared spectra of the samples, samples are uniformly selected in the feature space to make the samples in the modeling set contain more information on the mechanical properties of the insulation and improve the effectiveness and accuracy of the model.

[0094] 6) Wavelength selection: ① First, decompose a certain parameter performance parameter matrix y into two mutually orthogonal parts, i.e., y = P x y + A x y; where P x = X(X T X) -1 X T is the orthogonal projection matrix, and P x y projects y onto the column space of X, denoted as A x y projects y onto the orthogonal complement of the column space of X, where A x = (I - P x ), and I is the identity matrix.

[0095] ② Perform orthogonal decomposition on the spectral matrix X, i.e., Use the principal component analysis method to extract the first n principal components T A = (t1, t2, t3,..., t A ) to calculate the loading vector and solve the equation Xw i = t i , obtaining the weight vector w i .

[0096] ③ Calibrate the spectral matrix to remove more information unrelated to y.

[0097] ④ Sort |w i | from largest to smallest, and denote the sorted result as |w (j) |, |w (1) | is the largest,, |w (2) | is the second, and so on;

[0098] ⑤ For each j = 0, 1, 2,... N, select the wavelengths {i||w i | > |w (k+j×s) |}. Where k is the initial number of wavelength points, S is the step size, and at most k + N × s wavelength points are considered. And based on the selected wavelengths, establish a PLS model and calculate where, y i is the measured value; is the predicted value of the PLS model; n is the number of calibration set samples.

[0099] ⑥ Select the model j0 with the smallest RMSECV, then the finally selected wavelength points

[0100] In this case, the number of initial wavelength points is k = 10, the step size s = 1, N = 100, and 10 characteristic wavelength points are obtained: 1784 nm, 1684 nm, 1859 nm, 1783 nm, 1767 nm, 1795 nm, 1695 nm, 1845 nm, 1749 nm, 1766 nm. The 10 wavelength points are concentrated between 1680 - 1780 nm (near the two characteristic peaks at 1730 and 1764 nm).

[0101] 10 characteristic wavelength points are obtained: 1784 nm, 1684 nm, 1859 nm, 1783 nm, 1767 nm, 1795 nm, 1695 nm, 1845 nm, 1749 nm, 1766 nm. The 10 wavelength points are concentrated between 1680 - 1780 nm (near the two characteristic peaks at 1730 and 1764 nm).

[0102] 5) Quantitative analysis: According to the above steps, the quantitative analysis model for the mechanical properties of the insulation is: where a0 is the calibration coefficient, w i is the weight corresponding to the wavelength point, x i is the absorbance at the corresponding wavelength after pretreatment, and n is the modeling wavelength coefficient..

[0103] Table 1 is a comparison table of the mechanical property test values of the validation set obtained through experimental tests and the predicted values of the mechanical energy performance of the validation set obtained through spectroscopy. As can be seen from Table 1, the relative deviation between the predicted values of the elongation at break, tensile strength, and elastic modulus obtained by using the method of the present invention and the experimental values obtained by tensile testing is within 10%, which proves that the prediction method for the mechanical properties of the cable insulation layer proposed based on the present invention has high accuracy.

[0104] Data comparison table of Table 1

[0105]

[0106] Figure 2 is a schematic structural diagram of a system 200 for determining the mechanical properties of a cable insulation layer based on near-infrared spectral data according to an embodiment of the present invention. As Figure 2 shown, the system 200 for determining the mechanical properties of a cable insulation layer based on near-infrared spectral data provided by the embodiment of the present invention includes: a data acquisition unit 201, a preprocessing unit 202, a classification unit 203, a wavelength selection unit 204, and a quantitative analysis model determination unit 205.

[0107] Preferably, the data acquisition unit 201 is configured to acquire near-infrared spectral data and mechanical property data of a cable sample. Wherein, the mechanical property data includes: elongation at break, tensile strength, and elastic modulus.

[0108] Preferably, the preprocessing unit 202 is configured to preprocess the near-infrared spectral data to obtain processed near-infrared spectral data.

[0109] Preferably, the preprocessing unit 202 preprocesses the near-infrared spectral data to obtain processed near-infrared spectral data, including:

[0110] Using first-order differentiation to eliminate spectral baseline drift, and using Savitzky-Golay convolution to smooth the near-infrared spectral data to obtain processed near-infrared spectral data.

[0111] Preferably, the classification unit 203 is configured to classify the processed near-infrared spectral data and mechanical property data to determine a sample set and a validation set.

[0112] Preferably, the classification unit 203 classifies the processed near-infrared spectral data and mechanical property data to determine a sample set and a validation set, including:

[0113] Using the Kennard-Stone system, based on the Euclidean distance of the near-infrared spectrum, uniformly selecting samples in the feature space for sample classification to determine a sample set and a validation set.

[0114] Preferably, the wavelength selection unit 204 is configured to perform wavelength selection based on the sample set to determine target wavelength points and weights corresponding to the target wavelength points.

[0115] Preferably, the wavelength selection unit 204 performs wavelength selection based on the sample set to determine target wavelength points and weights corresponding to the target wavelength points, including:

[0116] For any one of the mechanical properties, the target wavelength points and weights corresponding to the target wavelength points are determined by the following method, including:

[0117] Decompose the performance parameter matrix y composed of any one of the mechanical property data into two mutually orthogonal parts, that is, y = P x y + A x y; where, P x = X(X T X) -1 X T is an orthogonal projection matrix, P x y projects y onto the column space of X, denoted as A x y projects y onto the orthogonal complement of the column space of X, where Ax =(I - P x ), where I is the identity matrix;

[0118] Perform orthogonal decomposition on the spectral matrix X composed of infrared spectrum processed data, that is Extract the first n principal components T using the principal component analysis method A =(t1, t2, t3,..., t A ) to calculate the loading vector and solve the equation Xw i =t i , to obtain the weight vector w i ;

[0119] Correct the spectral matrix to remove more information unrelated to y.

[0120] Sort |w i | from largest to smallest, and denote the sorted one as |w (j) |, |w (1) | is the largest, |w (2) | is the second, and so on;

[0121] For each j = 0, 1, 2,... i, select the wavelength and based on the selected wavelength, establish a PLS model and calculate where k is the initial number of wavelength points, S is the step size; y i is the measured value; is the predicted value of the PLS model; n is the number of calibration set samples;

[0122] Select the model j0 with the smallest RMSECV, then determine the target wavelength point corresponding to any one of the mechanical properties as and determine the weight corresponding to the target wavelength point.

[0123] Preferably, the quantitative analysis model determination unit 205 is configured to determine a quantitative analysis model of the mechanical properties of the cable insulation layer based on the target wavelength point and the weight corresponding to the target wavelength point, so as to determine the mechanical properties of the cable insulation layer based on the quantitative analysis model.

[0124] Preferably, the quantitative analysis model determination unit 205 determines a quantitative analysis model of the mechanical properties of the cable insulation layer based on the target wavelength point and the weight corresponding to the target wavelength point, including:

[0125] For any one of the mechanical properties, the quantitative analysis model is: where a0 is the calibration coefficient, w i is the weight corresponding to the wavelength point, x iis the absorbance corresponding to the wavelength after preprocessing, and n is the modeling wavelength coefficient.

[0126] Preferably, the system further includes:

[0127] a verification unit for verifying based on the verification set to determine the accuracy of the quantitative analysis model.

[0128] The system 200 for determining the mechanical properties of a cable insulating layer based on near-infrared spectral data in an embodiment of the present invention corresponds to the method 100 for determining the mechanical properties of a cable insulating layer based on near-infrared spectral data in another embodiment of the present invention, and will not be elaborated here.

[0129] The present invention has been described by referring to a few embodiments. However, as is well known to those skilled in the art, other embodiments equivalent to those disclosed above in the present invention equally fall within the scope of the present invention.

[0130] Generally, all terms used in the present invention are interpreted according to their ordinary meanings in the technical field, unless otherwise clearly defined therein. All references to "a / the [device, component, etc.]" are to be interpreted openly as at least one instance of the device, component, etc., unless otherwise clearly stated. The steps of any method disclosed herein need not be run in the exact order disclosed, unless clearly stated.

[0131] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0132] The present application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in Figure 1 one or more flows and / or Figure 1 blocks or multiple blocks.

[0133] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to work in a particular manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction device that implements the functions specified in one or more of the processes and / or blocks Figure 1 in one or more of the processes and / or blocks Figure 1 specified in the block or blocks.

[0134] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, so that the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one or more of the processes and / or blocks Figure 1 in one or more of the processes and / or blocks Figure 1 specified in the block or blocks.

[0135] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications or equivalent replacements can still be made to the specific embodiments of the present invention. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention shall be covered by the protection scope of the present invention.

Claims

1. A method for determining the mechanical properties of a cable insulation layer based on near-infrared spectral data, characterized in that The method includes: Obtaining near-infrared spectral data and mechanical property data of a cable sample; wherein, the mechanical property data includes: elongation at break, tensile strength, and elastic modulus; Preprocessing the near-infrared spectral data to obtain processed near-infrared spectral data; Classifying the processed near-infrared spectral data and mechanical property data to determine a sample set and a validation set; Performing wavelength selection based on the sample set to determine target wavelength points and weights corresponding to the target wavelength points; Determining a quantitative analysis model for the mechanical properties of the cable insulation layer based on the target wavelength points and weights corresponding to the target wavelength points, so as to determine the mechanical properties of the cable insulation layer based on the quantitative analysis model.

2. The method according to claim 1, wherein Preprocessing the near-infrared spectral data to obtain processed near-infrared spectral data, including: Using first-order differentiation to eliminate spectral baseline drift and using Savitzky-Golay convolution integral to smooth the near-infrared spectral data to obtain processed near-infrared spectral data.

3. The method according to claim 1, wherein Classifying the processed near-infrared spectral data and mechanical property data to determine a sample set and a validation set, including: Using the Kennard-Stone method to uniformly select samples in the feature space based on the Euclidean distance of the near-infrared spectrum for sample classification to determine a sample set and a validation set.

4. The method according to claim 1, characterized in that, Performing wavelength selection based on the sample set to determine target wavelength points and weights corresponding to the target wavelength points, including: For any one of the mechanical properties, determining the target wavelength points and weights corresponding to the target wavelength points by the following method, including: Decompose the performance parameter matrix y composed of any kind of mechanical property data into two mutually orthogonal parts, i.e., y = P x y + A x y; where, P x = X(X T X) -1 X T is an orthogonal projection matrix, and P x y projects y onto the column space of X, denoted as A x y then projects y onto the orthogonal complement of the column space of X, where A x = (I - P x ), and I is the identity matrix; Perform orthogonal decomposition on the spectral matrix X composed of data processed by infrared spectroscopy, that is Extract the first n principal components T using the principal component analysis method A =(t1, t2, t3,..., t A ) to calculate the loading vector and solve the equation Xw i =t i to obtain the weight vector w i ; Calibrate the spectral matrix To remove more information unrelated to y. Sort |w i in descending order, and denote the sorted result as |w (j) |, |w (1) | is the largest, |w (2) | is the second largest, and so on; For each j = 0, 1, 2,... i, select a wavelength and based on the selected wavelength, establish a PLS model and calculate where k is the initial number of wavelength points, S is the step size; y i is the measured value; is the predicted value of the PLS model; n is the number of calibration set samples; Select the model j0 with the smallest RMSECV, then determine that the target wavelength point corresponding to any one of the mechanical properties is and determine the weight corresponding to the target wavelength point.

5. The method according to claim 1, characterized in that, Determining a quantitative analysis model for the mechanical properties of the cable insulation layer based on the target wavelength points and weights corresponding to the target wavelength points, including: For any kind of mechanical property, the quantitative analysis model is as follows: where a0 is the calibration coefficient, w i is the weight corresponding to the wavelength point, x i is the absorbance at the wavelength corresponding to the preprocessing, and n is the modeling wavelength coefficient.

6. The method according to claim 1, wherein The method further includes: Performing validation based on the validation set to determine the accuracy of the quantitative analysis model.

7. A system for determining the mechanical properties of a cable insulation layer based on near-infrared spectral data, characterized in that, The system includes: A data acquisition unit for obtaining near-infrared spectral data and mechanical property data of a cable sample; wherein, the mechanical property data includes: elongation at break, tensile strength, and elastic modulus; A preprocessing unit for preprocessing the near-infrared spectral data to obtain processed near-infrared spectral data; A classification unit for classifying the processed near-infrared spectral data and mechanical property data to determine a sample set and a validation set; A wavelength selection unit for performing wavelength selection based on the sample set to determine target wavelength points and weights corresponding to the target wavelength points; A quantitative analysis model determination unit for determining a quantitative analysis model for the mechanical properties of the cable insulation layer based on the target wavelength points and weights corresponding to the target wavelength points, so as to determine the mechanical properties of the cable insulation layer based on the quantitative analysis model.

8. The system according to claim 7, wherein The preprocessing unit preprocesses the near-infrared spectral data to obtain processed near-infrared spectral data, including: Using first-order differentiation to eliminate spectral baseline drift and using Savitzky-Golay convolution integral to smooth the near-infrared spectral data to obtain processed near-infrared spectral data.

9. The system according to claim 7, wherein The classification unit classifies the processed near-infrared spectral data and mechanical property data to determine a sample set and a validation set, including: The Kennard-Stone system is adopted to uniformly select samples in the feature space based on the Euclidean distance of near-infrared spectra for sample classification, and determine the sample set and the validation set.

10. The system according to claim 7, wherein The wavelength selection unit performs wavelength selection based on the sample set to determine the target wavelength points and the weights corresponding to the target wavelength points, including: For any mechanical property, the target wavelength points and the weights corresponding to the target wavelength points are determined by the following method, including: Decompose the performance parameter matrix y composed of any mechanical property data into two mutually orthogonal parts, i.e., y = P x y + A x y; where P x = X(X T X) -1 X T is an orthogonal projection matrix, and P x y projects y onto the column space of X, denoted as A x y then projects y onto the orthogonal complement of the column space of X, where A x = (I - P x ), and I is the identity matrix; Perform orthogonal decomposition on the spectral matrix X composed of data processed by infrared spectroscopy, that is Extract the first n principal components T using the principal component analysis method A =(t1, t2, t3,..., t A ) to calculate the loading vector and solve the equation Xw i =t i to obtain the weight vector w i ; Calibrate the spectral matrix To remove more information unrelated to y. Sort |w i in descending order, and denote the sorted result as |w (j) ; |w (1) | is the largest, |w (2) | is the second largest, and so on; For each j = 0, 1, 2,... i, select a wavelength and based on the selected wavelength, establish a PLS model and calculate where k is the initial number of wavelength points, S is the step size; y i is the measured value; is the predicted value of the PLS model; n is the number of calibration set samples; Select the model j0 with the smallest RMSECV, then determine that the target wavelength point corresponding to any one of the mechanical properties is and determine the weight corresponding to the target wavelength point.

11. The system according to claim 7, wherein The quantitative analysis model determination unit determines the quantitative analysis model of the mechanical properties of the cable insulation layer based on the target wavelength points and the weights corresponding to the target wavelength points, including: For any mechanical property, the quantitative analysis model is as follows: where a0 is the calibration coefficient, w i is the weight corresponding to the wavelength point, x i is the wavelength absorbance corresponding to after pretreatment, and n is the modeling wavelength coefficient.

12. The system according to claim 7, characterized in that The system further includes: A verification unit for verifying based on the validation set to determine the accuracy of the quantitative analysis model.

Images