CN117872048A - Cable insulation material performance prediction method based on near infrared spectrum and related equipment - Google Patents
Cable insulation material performance prediction method based on near infrared spectrum and related equipment Download PDFInfo
- Publication number
- CN117872048A CN117872048A CN202311680616.3A CN202311680616A CN117872048A CN 117872048 A CN117872048 A CN 117872048A CN 202311680616 A CN202311680616 A CN 202311680616A CN 117872048 A CN117872048 A CN 117872048A
- Authority
- CN
- China
- Prior art keywords
- correction
- sample
- spectral data
- nearest
- samples
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 59
- 238000002329 infrared spectrum Methods 0.000 title claims abstract description 41
- 239000012774 insulation material Substances 0.000 title claims description 33
- 238000012937 correction Methods 0.000 claims abstract description 296
- 230000003595 spectral effect Effects 0.000 claims abstract description 178
- 238000001228 spectrum Methods 0.000 claims abstract description 75
- 230000006870 function Effects 0.000 claims description 44
- 238000004364 calculation method Methods 0.000 claims description 14
- 238000010276 construction Methods 0.000 claims description 14
- 238000004590 computer program Methods 0.000 claims description 11
- 238000009413 insulation Methods 0.000 claims description 11
- 239000011159 matrix material Substances 0.000 claims description 8
- 238000004422 calculation algorithm Methods 0.000 claims description 5
- 238000004497 NIR spectroscopy Methods 0.000 claims 1
- 230000008569 process Effects 0.000 abstract description 13
- 238000001514 detection method Methods 0.000 abstract description 6
- 230000009286 beneficial effect Effects 0.000 abstract description 5
- 239000011810 insulating material Substances 0.000 abstract description 5
- 238000013528 artificial neural network Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000012360 testing method Methods 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 230000001174 ascending effect Effects 0.000 description 2
- 229920003020 cross-linked polyethylene Polymers 0.000 description 2
- 239000004703 cross-linked polyethylene Substances 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 150000002430 hydrocarbons Chemical class 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000012706 support-vector machine Methods 0.000 description 2
- 229920000181 Ethylene propylene rubber Polymers 0.000 description 1
- 238000012569 chemometric method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000012821 model calculation Methods 0.000 description 1
- JRZJOMJEPLMPRA-UHFFFAOYSA-N olefin Natural products CCCCCCCC=C JRZJOMJEPLMPRA-UHFFFAOYSA-N 0.000 description 1
- 239000004800 polyvinyl chloride Substances 0.000 description 1
- 229920000915 polyvinyl chloride Polymers 0.000 description 1
- 238000000513 principal component analysis Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000009864 tensile test Methods 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention belongs to the technical field of power cable performance detection, and particularly relates to a cable insulating material performance prediction method based on near infrared spectrum and related equipment; acquiring spectrum data of a sample to be measured, and acquiring k correction samples nearest to the sample to be measured from a correction set; constructing a spectral data constraint condition and a weight function of the sample to be detected according to the spectral data of the sample to be detected and the spectral data of k correction samples; constructing a target correction model based on the spectral data of the sample to be detected, the weight function and the spectral data constraint condition, and calculating the weight coefficient corresponding to the spectral data of each correction sample in the k correction samples through the model; and determining target performance parameters of the sample to be tested based on the weight coefficients of the k correction samples and the corresponding performance data. The method has the advantages of simple model and small calculated amount, can better process the nonlinear relation, and is beneficial to improving the prediction accuracy of the target correction model.
Description
Technical Field
The invention belongs to the technical field of power cable performance detection, and particularly relates to a cable insulation material performance prediction method based on near infrared spectrum and related equipment.
Background
In the detection of the insulation performance of the power cable, the traditional method for checking the arrival goods by sampling by manpower has the defects of long test period, low efficiency, large loss and inconvenient operation, so that the high-efficiency performance judgment on the insulation material is particularly necessary. The cable insulation material is mainly hydrocarbon compounds containing various different C-H group information, such as: group information of methyl C-H (913 nm), olefin C-H (895 nm), methylene C-H (934 nm), and aromatic C-H (875 nm). The content of hydrocarbon compounds with different structures in the material can change in the near infrared spectrum, and the information of the composition change of a sample can be obtained by processing spectrum data through a chemometric method although the change is very fine. Therefore, in the prior art, it is proposed to establish a correlation between the near infrared spectrum of the insulating material and the composition data for analysis, and determine the material performance parameters.
However, the PCR (principle component regression, principal component analysis) and PLS (Partial Least Squares) methods widely used at present can only build a linear model, and in the spectrum acquisition process, nonlinear relations exist between spectra and component or property data due to the influence of various physical and chemical factors, so that the prediction accuracy is not high. Although nonlinear modeling methods such as an ANN (Artificial Neural Network ) and an SVM (support vector machines, support vector machine) can describe the nonlinear relationship more accurately, the ANN and the SVM have the problems of complex model, multiple parameters and large calculation amount, and the neighborhood of spectral data of a sample to be detected needs to be determined firstly, so that the calculation amount is increased, and the application of the method in cable insulation performance detection is severely restricted.
Disclosure of Invention
Aiming at the problems that in the prior art, the linear modeling prediction accuracy is low, the model calculation amount is large and the model is complex, the invention provides a cable insulating material performance prediction method and related equipment based on near infrared spectrum, the method is characterized in that the calculation amount is smaller compared with the existing mode of calculating the neighborhood construction constraint condition and model of the spectrum data of a sample to be measured by establishing the association relation between the spectrum data and the performance data of the cable insulating material and selecting k correction samples nearest to the sample to be measured from the acquired correction set as model parameters when constructing a target correction model; and the weight function is determined by combining the spectrum data of the k nearest correction samples, and the final target correction model is established based on the data rather than according to experience.
In a first aspect, the present invention provides a method for predicting performance of a cable insulation material based on near infrared spectrum, the method comprising the steps of:
Acquiring a correction set, wherein the correction set comprises spectrum data and performance data of a plurality of correction samples;
acquiring spectrum data of a sample to be detected, and acquiring k correction samples nearest to the sample to be detected from the correction set;
constructing a spectral data constraint condition and a weight function of the sample to be detected according to the spectral data of the sample to be detected and the spectral data of the nearest k correction samples;
constructing a target correction model based on the spectral data of the sample to be detected, the weight function and the spectral data constraint condition, and calculating a weight coefficient corresponding to the spectral data of each correction sample in the k nearest correction samples through the target correction model;
and determining target performance parameters of the sample to be tested according to the weight coefficient corresponding to the spectral data of each correction sample in the k nearest correction samples and the performance data of a plurality of correction samples in the correction set.
Further, the acquiring the correction set includes:
acquiring spectrum data and performance data of a plurality of correction samples, wherein the spectrum data of each correction sample is a point in a p-dimensional space, and p is more than or equal to 1;
constructing the correction set (x) based on the spectral data and the performance data of the plurality of correction samples i ,y i )，i＝1,2,3,...,n，x i Is the spectral data of the ith calibration sample, y i Is the performance data for each calibration sample at i.
Further, the acquiring spectral data of the sample to be measured, and acquiring k correction samples nearest to the sample to be measured from the correction set includes:
acquiring the sample to be detected, and acquiring spectrum data of the sample to be detected based on spectrum acquisition equipment;
calculating the difference value between the spectrum data of the sample to be detected and the spectrum data of each correction sample in the correction set based on a preset distance algorithm;
and constructing a difference sequence according to each difference value, and determining k correction samples corresponding to the smallest difference value in the difference sequence as k correction samples nearest to the sample to be detected.
Further, the spectral data constraint condition includes a nearest sample point constraint condition and a spectral data range constraint condition, and the constructing the spectral data constraint condition of the sample to be measured according to the spectral data of the sample to be measured and the spectral data of the nearest k correction samples includes:
constructing a nearest sample point constraint condition of the sample to be detected according to the spectral data of the sample to be detected and the spectral data of the nearest k correction samples, wherein the nearest sample point constraint condition is as follows:
Wherein x is u For the spectrum data of the sample to be tested, delta is equal to x u The range between the spectral data of the nearest k calibration samples,is x u A set of sample points within the spectral data of the nearest k calibration samples;
constructing spectral data range constraint conditions of correction samples in the k nearest correction samples based on the nearest sample point constraint conditions, wherein the spectral data range constraint conditions are as follows:
wherein x is j The spectral data of the j-th calibration sample among the k nearest calibration samples,for the correction set (x i ,y i ) Is of the category X u The nearest k adjacent sets of sample points of the spectral data of the corrected samples.
Further, the constructing a weight function of the sample to be measured according to the spectral data of the sample to be measured and the spectral data of the k nearest correction samples includes:
introducing weight coefficients corresponding to the spectral data of the k nearest correction samples, and constructing a weight function of the sample to be detected according to the spectral data of the sample to be detected, the spectral data of the k nearest correction samples and the weight coefficients corresponding to the spectral data of the k nearest correction samples, wherein the weight function is as follows:
the constraint conditions of the weight coefficients are as follows: Wherein w is uj Weighting coefficient of spectral data of j-th correction sample among k nearest correction samples, ε (W u ) As a weight function, W u The overall weight coefficient is calculated for the weight coefficients based on the nearest k correction samples.
Further, the constructing a target correction model based on the spectral data of the sample to be measured, the weight function and the spectral data constraint condition, calculating a weight coefficient corresponding to the spectral data of each correction sample of the k nearest correction samples by the target correction model, including:
constructing the target correction model based on the spectrum data of the sample to be detected, the weight function and the spectrum data constraint condition, wherein the target correction model is as follows:
L(W u ；θ)＝(x u -X T B u W u ) T (x u -X T B u W u )+θ(1-W u T )
wherein B is u Is a matrix of n X k, X is a correction set, X T B u Is X and X u The nearest k sets of spectral data of the correction samples, θ being the lagrangian, T representing the transpose of the matrix;
the overall weight coefficient is derived through the target correction model and is brought into the weight coefficient constraint condition to be calculated, and the Lagrange operator is determined;
calculating the integral weight coefficient corresponding to the spectral data of the nearest k correction samples based on the Lagrangian operator obtained by calculation;
And calculating the weight coefficient corresponding to the spectral data of each correction sample in the k nearest correction samples based on the integral weight coefficient.
Further, the determining the target performance parameter of the sample to be measured according to the weight coefficient corresponding to the spectral data of each correction sample in the k nearest correction samples and the performance data of a plurality of correction samples in the correction set includes: according to the weight coefficient corresponding to the spectral data of each correction sample in the k nearest correction samples and the performance data of the k nearest correction samples in the correction set, predicting the target performance parameter of the sample to be measured, wherein the prediction formula is as follows:
wherein,is the target performance parameter.
In a second aspect, the present invention also provides a device for predicting performance of a cable insulation material based on near infrared spectrum, the device comprising:
the data acquisition module is used for acquiring a correction set, wherein the correction set comprises spectrum data and performance data of a plurality of correction samples; the adjacent sample acquisition module is used for acquiring spectral data of a sample to be detected and acquiring k correction samples nearest to the sample to be detected from the correction set;
the first construction module is used for constructing spectral data constraint conditions and weight functions of the sample to be tested according to the spectral data of the sample to be tested and the spectral data of the k nearest correction samples;
The second construction module is used for constructing a target correction model based on the spectrum data of the sample to be detected, the weight function and the spectrum data constraint condition, and calculating the weight coefficient corresponding to the spectrum data of each correction sample in the k nearest correction samples through the target correction model;
and the performance prediction module is used for determining target performance parameters of the sample to be detected according to the weight coefficient corresponding to the spectral data of each correction sample in the k nearest correction samples and the performance data of a plurality of correction samples in the correction set.
In a third aspect, the present invention also provides an electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, the processor implementing the steps of the near infrared spectrum based cable insulation material performance prediction method as described in the embodiments when the computer program is executed.
In a fourth aspect, the present invention also provides a computer readable storage medium storing a computer program which, when executed by a processor, implements the steps of the near infrared spectrum based cable insulation performance prediction method as described in the embodiments.
The invention has the beneficial effects that: according to the invention, k correction samples nearest to the sample to be detected are selected from the obtained correction set to serve as model parameters, so that a spectrum data constraint condition and a model are built, and compared with the existing mode of building the constraint condition and the model by calculating the neighborhood of spectrum data of the sample to be detected, the calculation amount is smaller; and the weight function is determined by combining the spectrum data of the k nearest correction samples, and the final target correction model is established based on the data rather than according to experience.
The foregoing summary is merely an overview of the present invention, and is intended to be implemented in accordance with the teachings of the present invention in order that the same may be more fully understood, and in order that the same or additional objects, features and advantages of the present invention may be more fully understood.
Drawings
Other features, objects and advantages of the present invention will become more apparent upon reading of the detailed description of non-limiting embodiments made with reference to the following drawings. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the invention. Also, like reference numerals are used to designate like parts throughout the figures.
Fig. 1 is a flowchart of a method for predicting performance of a cable insulation material based on near infrared spectrum according to an embodiment of the present invention.
Fig. 2 is a schematic structural diagram of a cable insulation material performance prediction device based on near infrared spectrum according to an embodiment of the present invention.
Fig. 3 is a schematic structural diagram of a computer device according to an embodiment of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the accompanying drawings and examples, it being understood that the detailed description herein is merely a preferred embodiment of the present invention, which is intended to illustrate the present invention, and not to limit the scope of the invention, as all other embodiments obtained by those skilled in the art without making any inventive effort fall within the scope of the present invention.
Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although a flowchart depicts operations (or steps) as a sequential process, many of the operations (or steps) can be performed in parallel, concurrently, or at the same time. Furthermore, the order of the operations may be rearranged. The process may be terminated when its operations are completed, but may have additional steps not included in the figures; the processes may correspond to methods, functions, procedures, subroutines, and the like.
The terms "first," "second," "third," "fourth," and the like in the description of the invention and in the above figures, if any, are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the invention described herein may be implemented in sequences other than those illustrated or otherwise described herein. It should also be understood that, in various embodiments of the present invention, the sequence number of each process does not mean the order of execution, and the order of execution of each process should be determined by its functions and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.
It should be understood that in the present invention, "plurality" means two or more. "and/or" is merely a variable relationship describing an associated object, meaning that there may be three relationships, e.g., and/or B, may represent: a exists alone, A and B exist together, and B exists alone. The character "/" generally indicates that the context-dependent object is an "or" relationship. "comprising A, B and C", "comprising A, B, C" means that all three of A, B, C comprise, "comprising A, B or C" means that one of the three comprises A, B, C, and "comprising A, B and/or C" means that any 1 or any 2 or 3 of the three comprises A, B, C.
It should be understood that in the present invention, "B corresponding to a", "a corresponding to B", or "B corresponding to a" means that B is associated with a, from which B can be determined. Determining B from a does not mean determining B from a alone, but may also determine B from a and/or other information. The matching of A and B is that the similarity of A and B is larger than or equal to a preset threshold value.
Example 1
Referring to fig. 1, fig. 1 is a flowchart of a method for predicting performance of a cable insulation material based on near infrared spectrum according to an embodiment of the present invention, and the method for predicting performance of a cable insulation material based on near infrared spectrum specifically includes the following steps:
s1, acquiring a correction set, wherein the correction set comprises spectrum data and performance data of a plurality of correction samples.
The cable insulation material performance prediction method based on the near infrared spectrum provided by the embodiment of the invention is suitable for a scene of performance detection of wires and cables, and the electronic equipment running the cable insulation material performance prediction method based on the near infrared spectrum can communicate with other electronic equipment in a wireless connection/wired connection mode to realize data interaction, for example: and receiving data uploaded by other electronic equipment, or sending the prediction result to other electronic equipment and the like. It should be noted that the wireless connection includes, but is not limited to, a 4G/5G connection, a WiFi connection, a bluetooth connection, a WiMAX connection, a Zigbee connection, a UWB connection, and other wireless connection methods now known or developed in the future.
Wherein a correction set may be used to provide data support for model construction, and a plurality of correction samples of cable insulation material may be included in the correction set, each correction sample including corresponding spectral data and performance data. The cable insulation may include, but is not limited to, crosslinked polyethylene, polyvinyl chloride, ethylene propylene rubber, and the like.
Specifically, the step S1 includes:
s11, acquiring spectrum data and performance data of a plurality of correction samples, wherein the spectrum data of each correction sample is a point in a p-dimensional space, and p is more than or equal to 1;
s12, constructing the correction set (x) based on the spectrum data and the performance data of a plurality of correction samples i ,y i )，i＝1,2,3,...,n，x i Is the spectral data of the ith calibration sample, y i Is the performance data for each calibration sample at i.
The spectral data may be the spectral data of the calibration sample collected in real time or historically by the near infrared spectrophotometer, and the spectral data refers to near infrared spectral data. Wherein, can carry out tensile test to the correction sample of cable insulation material and obtain the performance data of correction sample, for example: the elongation at break, tensile strength, etc. were tested. The spectral data may be data in one-dimensional space, may be data in two-dimensional space, or may be data in three-dimensional or even higher-dimensional space. Obtaining spectral data and performance data of each correction sample Then, a correction set (x) including the spectrum and performance of each correction sample can be constructed based on the order i ,y i )。
S2, acquiring spectral data of a sample to be detected, and acquiring k correction samples nearest to the sample to be detected from the correction set.
Specifically, in order to predict performance parameters according to the near infrared spectrum of the insulating material, the spectrum data of the sample to be measured may be collected first after obtaining the sample to be measured of a certain component or property data. The nearest neighbor data tends to have more similar spectra and reflect more similar properties. For more convenient and accurate prediction of performance parameters, k correction samples nearest to the sample to be measured may be obtained from the correction set.
Specifically, the step S2 specifically includes:
s21, acquiring the sample to be detected, and acquiring spectrum data of the sample to be detected based on spectrum acquisition equipment;
s22, calculating the difference value between the spectrum data of the sample to be detected and the spectrum data of each correction sample in the correction set based on a preset distance algorithm;
s23, constructing a difference sequence according to each difference value, and determining k correction samples corresponding to the smallest difference value in the difference sequence as k correction samples nearest to the sample to be detected.
The spectrum data of the sample to be detected can be collected through the far infrared spectrophotometer, the spectrum data are points in space, and the distance relation between the point where the spectrum data of the sample to be detected are located and the point where the spectrum data of each correction sample in the correction set are located can be calculated based on the position relation between any two points.
Specifically, as a possible implementation manner, if the difference is a point in a one-dimensional space, the difference between two points may be directly calculated, for example: the point where the spectrum data of the sample to be measured is 8, the point where the spectrum data of the correction sample is 12, and the distance difference is 4. As another possible implementation, if the two-dimensional or higher dimensional space is a point, the difference between the two points may be calculated by means of the euclidean distance, the manhattan distance, the chebyshev distance, or the like, for example: in the two-dimensional space, the point where the spectrum data of the sample to be detected is (3, 4), the point where the spectrum data of the correction sample is (3, 7), and the distance difference value is obtained through Euclidean distance calculation.
More specifically, after the difference between the sample to be measured and each correction sample in the correction set is calculated based on the spectrum data, the difference sequence can be constructed by arranging in ascending order and descending order according to the magnitude of the difference. If the correction samples are arranged in ascending order, k correction samples which are the k correction samples at the back in the difference sequence are selected as k correction samples which are nearest to the sample to be detected; if the order is descending, the k correction samples in the difference sequence, which are the k correction samples closest to the sample to be measured, are selected. The k correction samples nearest to the sample to be measured are samples with the nearest spectral data and the nearest performance parameters.
S3, constructing a spectral data constraint condition and a weight function of the sample to be detected according to the spectral data of the sample to be detected and the spectral data of the nearest k correction samples.
The spectrum data constraint condition can be used for constraining the spectrum in the process of constructing the model according to the spectrum data of the sample to be detected and the spectrum data of the k nearest correction samples, and the spectrum data constraint condition is ensured to be in a constraint range. Meanwhile, the weight of the corresponding correction sample can be introduced, a weight function is established according to the spectral data of the sample to be detected and each correction sample in the spectral data of the k nearest correction samples, and the association relationship between the spectral data of the sample to be detected and the spectral data of the k nearest correction samples is established through the weight function.
Specifically, in the step S3, a spectral data constraint condition of the sample to be measured is constructed, including:
s31, constructing a constraint condition of nearest sample points of the sample to be detected according to the spectral data of the sample to be detected and the spectral data of the nearest k correction samples, wherein the constraint condition of the nearest sample points is shown in the following formula (1):
wherein x is u For the spectrum data of the sample to be tested, delta is equal to x u The range between the spectral data of the nearest k calibration samples,is x u A set of sample points within the spectral data of the nearest k calibration samples;
s32, constructing spectral data range constraint conditions of correction samples in the k nearest correction samples based on the nearest sample point constraint conditions, wherein the spectral data range constraint conditions are shown in the following formula (2):
wherein x is j The spectral data of the j-th calibration sample among the k nearest calibration samples,for the correction set (x i ,y i ) Is of the category X u The nearest k adjacent sets of sample points of the spectral data of the corrected samples.
Specifically, the spectral data constraint condition of the sample to be measured includes a nearest sample point constraint condition and a spectral data range constraint condition. In order to limit the range of occurrence of k correction samples within the correction set, it is necessary to construct the spectral data range constraint, as in equation (2) above, that is
Specifically, in the step S3, a weight function of the sample to be measured is constructed, including:
s33, introducing weight coefficients corresponding to the spectral data of the k nearest correction samples, and constructing a weight function of the sample to be detected according to the spectral data of the sample to be detected, the spectral data of the k nearest correction samples and the weight coefficients corresponding to the spectral data of the k nearest correction samples, wherein the weight function is shown in the following formula (3):
The constraint condition of the weight coefficient is shown as a formula (4):
wherein w is uj Weighting coefficient of spectral data of j-th correction sample among k nearest correction samples, ε (W u ) As a weight function, W u The overall weight coefficient is calculated for the weight coefficients based on the nearest k correction samples.
Considering that the weight function has direct influence on the accuracy of predicting the target correction model, the weight function with self-adaptive adjustment is constructed based on the selected data, rather than directly determining the weight coefficient through past experience, thereby being beneficial to constructing the target correction model with higher prediction accuracy.
More specifically, by introducing a weight coefficient, a weight coefficient w is configured for each of the spectral data of the nearest k correction samples uj The weight function shown in the formula (3) above can be constructed by combining the acquired spectral data of the sample to be detected, the spectral data of the k nearest correction samples and the weight coefficients corresponding to the spectral data of the k nearest correction samples, and the weight coefficients corresponding to the spectral data of the k nearest correction samples are constrained by the formula (4) above, so that the sum of the weight coefficients is ensured to be 1. The weight coefficient constraint condition ensures that the weight coefficient has translational invariance to the sample to be measured and the adjacent points thereof.
S4, constructing a target correction model based on the spectral data of the sample to be detected, the weight function and the spectral data constraint condition, and calculating a weight coefficient corresponding to the spectral data of each correction sample in the k nearest correction samples through the target correction model.
Specifically, step S4 includes:
s41, constructing the target correction model based on the spectral data of the sample to be detected, the weight function and the spectral data constraint condition, wherein the target correction model is shown in the following formula (5):
L(W u ；θ)＝(x u -X T B u W u ) T (x u -X T B u W u )+θ(1-W u T ) (5)
wherein B is u Is a matrix of n X k, X is a correction set, X T B u Is X and X u The nearest k sets of spectral data of the correction samples, θ being the lagrangian, T representing the transpose of the matrix;
s42, deriving the overall weight coefficient through the target correction model and carrying the overall weight coefficient into the weight coefficient constraint condition to calculate, and determining the Lagrange operator;
s43, calculating the integral weight coefficient corresponding to the spectral data of the nearest k correction samples based on the Lagrange operator obtained by calculation;
s44, calculating the weight coefficient corresponding to the spectral data of each correction sample in the k nearest correction samples based on the integral weight coefficient.
The target correction model constructed may be a model constructed based on Lagrange function, as shown in the above formula (5). First, solving the overall weight coefficient W u W is represented by the above formula (5) u The bias guide is calculated as shown in the following formula (6):
converting to W based on the above formula (6) u ＝(B u T XX T B u ) -1 (B u T Xx u - θ) is substituted into the above formula (4) to calculate a lagrangian θ, and after the lagrangian θ is calculated, the overall weight coefficient can be calculatedW u According to the overall weight coefficient W u Weight coefficient w corresponding to spectral data of each of the nearest k corrected samples uj The relation between the two can calculate the weight coefficient w corresponding to the spectrum data of each correction sample in the k nearest correction samples uj 。
In the process of constructing the target correction model and calculating, the nearest k correction samples are the only parameters involved in the algorithm, the model is simple, and the calculation efficiency is higher.
S5, determining target performance parameters of the sample to be tested according to the weight coefficient corresponding to the spectral data of each correction sample in the k nearest correction samples and the performance data of a plurality of correction samples in the correction set.
Specifically, according to the weight coefficient corresponding to the spectral data of each correction sample in the k nearest correction samples and the performance data of the k nearest correction samples in the correction set, predicting the target performance parameter of the sample to be measured, wherein the prediction formula is shown in the following formula (7):
Wherein,is the target performance parameter.
Specifically, after determining the weight coefficient corresponding to the spectral data of each of the k nearest correction samples, the performance data of the k nearest correction samples in the correction set may be combined, and the target performance parameter of the sample to be measured may be obtained by calculating according to the above formula (7). The method is characterized in that a nonlinear model is built based on the near infrared spectrum and the performance parameters of the power cable, so that the performance prediction of the sample to be tested is realized, the model is simple, the calculated amount is small, the nonlinear relation can be better processed, and the method is more beneficial to the application in cable insulation performance detection.
In the embodiment of the invention, k correction samples nearest to the sample to be detected are selected from the acquired correction set to be used as model parameters, so that a spectrum data constraint condition and a model are constructed, and compared with the existing mode of constructing the constraint condition and the model by calculating the neighborhood of the spectrum data of the sample to be detected, the calculation amount is smaller; and the weight function is determined by combining the spectrum data of the k nearest correction samples, and the final target correction model is established based on the data rather than according to experience.
For a better understanding of the present invention, the following description is made in connection with examples. The near infrared spectrum of the crosslinked polyethylene and its elongation at break, tensile strength test and prediction procedure are provided below.
The spectrum acquisition range measured by the experiment is 1100-2200 nm, the step length is 2nm, and the total wavelength is 550. The total samples included 50 samples of the correction set, 15 samples of the prediction set and the test set. The target correction model established by adopting PLS, ANN, SVM and the cable insulation material performance prediction method based on the near infrared spectrum provided by the invention is used for prediction, and the prediction errors of the 4 models are evaluated according to the result of a test set, as shown in the following table 1:
TABLE 1 comparison of prediction errors for different correction models
| Correction model | Elongation at break | Tensile Strength |
| PLS | 9.35％ | 8.67％ |
| ANN | 10.24％ | 9.64％ |
| SVM | 7.68％ | 7.12％ |
| The method provided by the invention | 5.68％ | 6.37％ |
From table 1 above, it can be seen that the prediction error of predicting the elongation at break and the tensile strength of the insulation material by the target correction model established by the cable insulation material performance prediction method based on the near infrared spectrum provided by the invention is minimum, and the prediction accuracy of the model is higher.
Example two
Referring to fig. 2, fig. 2 is a schematic structural diagram of a cable insulation material performance prediction device based on near infrared spectrum according to an embodiment of the present invention. A near infrared spectrum based cable insulation material performance prediction apparatus, the apparatus M20 comprising:
The data acquisition module M201 is used for acquiring a correction set, wherein the correction set comprises spectrum data and performance data of a plurality of correction samples;
the adjacent sample acquisition module M202 is used for acquiring spectral data of a sample to be detected, and k correction samples which are nearest to the sample to be detected are acquired from the correction set;
the first construction module M203 is used for constructing a spectral data constraint condition and a weight function of the sample to be detected according to the spectral data of the sample to be detected and the spectral data of the k nearest correction samples;
the second construction module M204 is configured to construct a target correction model based on the spectral data of the sample to be tested, the weight function and the spectral data constraint condition, and calculate a weight coefficient corresponding to the spectral data of each correction sample of the k nearest correction samples through the target correction model;
and the performance prediction module M205 is used for determining the target performance parameters of the sample to be detected according to the weight coefficient corresponding to the spectral data of each correction sample in the k nearest correction samples and the performance data of a plurality of correction samples in the correction set.
Optionally, the data acquisition module M201 includes:
the first data acquisition submodule M2011 is used for acquiring spectrum data and performance data of a plurality of correction samples, wherein the spectrum data of each correction sample is a point in a p-dimensional space, and p is more than or equal to 1;
A dataset construction sub-module M2012 for constructing the correction set (x i ，y i )，i＝1，2，3，...·，n，x i Is the spectral data of the ith calibration sample, y i Is the performance data for each calibration sample at i.
Optionally, adjacent to the sample acquisition module M202, comprising:
a second data acquisition sub-module M2021, configured to acquire the sample to be detected, and acquire spectral data of the sample to be detected based on a spectral acquisition device;
a first calculating submodule M2022 for calculating a difference value between the spectral data of the sample to be measured and the spectral data of each correction sample in the correction set based on a preset distance algorithm;
and the adjacent sample selection submodule M2023 is used for constructing a difference sequence according to each difference value, and determining k correction samples corresponding to the smallest difference value in the difference sequence as k correction samples nearest to the sample to be detected.
Optionally, the spectral data constraint includes a nearest neighbor sample point constraint and a spectral data range constraint, and the first building module M203 includes:
the first construction submodule M2031 is configured to construct a nearest sample point constraint condition of the sample to be measured according to the spectral data of the sample to be measured and the spectral data of the k nearest correction samples, where the nearest sample point constraint condition is:
Wherein x is u For the spectrum data of the sample to be tested, delta is equal to x u The range between the spectral data of the nearest k calibration samples,is x u A set of sample points within the spectral data of the nearest k calibration samples;
a second construction submodule M2033, configured to construct, based on the nearest sample point constraint condition, a spectral data range constraint condition of a correction sample among the nearest k correction samples, where the spectral data range constraint condition is:
wherein x is j The spectral data of the j-th calibration sample among the k nearest calibration samples,for the correction set (x i ,y i ) Is of the category X u The nearest k adjacent sets of sample points of the spectral data of the corrected samples.
Optionally, the first construction module M203 further includes:
a third construction submodule M2033, configured to introduce weight coefficients corresponding to the spectral data of the k nearest correction samples, and construct a weight function of the sample to be measured according to the spectral data of the sample to be measured, the spectral data of the k nearest correction samples, and the weight coefficients corresponding to the spectral data of the k nearest correction samples, where the weight function is:
the constraint conditions of the weight coefficients are as follows:
wherein w is uj Weighting coefficient of spectral data of j-th correction sample among k nearest correction samples, ε (W u ) As a weight function, W u The overall weight coefficient is calculated for the weight coefficients based on the nearest k correction samples.
Optionally, the second building module M204 includes:
a fourth construction submodule M2041, configured to construct the target correction model based on the spectral data of the sample to be measured, the weight function, and the spectral data constraint condition, where the target correction model is:
L(W u ；θ)＝(x u -X T B u W u ) T (x u -X T B u W u )+θ(1-W u T )
wherein B is u Is a matrix of n X k, X is a correction set, X T B u Is X and X u The nearest k sets of spectral data of the correction samples, θ being the lagrangian, T representing the transpose of the matrix;
the second calculation submodule M2042 is used for deriving the overall weight coefficient through the target correction model and carrying the overall weight coefficient into the weight coefficient constraint condition to calculate, so as to determine the Lagrangian operator;
a third calculation sub-module M2043, configured to calculate, based on the lagrangian obtained by calculation, the overall weight coefficients corresponding to the spectral data of the k nearest correction samples;
and a fourth calculating submodule M2044 for calculating a weight coefficient corresponding to the spectral data of each of the nearest k correction samples based on the overall weight coefficient.
Optionally, the performance prediction module M205 is specifically configured to predict, according to a weight coefficient corresponding to the spectral data of each of the k nearest correction samples and the performance data of the k nearest correction samples in the correction set, a target performance parameter of the sample to be measured, where a prediction formula is as follows:
wherein,is the target performance parameter.
The device for predicting the performance of the cable insulation material based on the near infrared spectrum provided by the embodiment of the invention can realize each process in the method for predicting the performance of the cable insulation material based on the near infrared spectrum, and can achieve the same beneficial effects. In order to avoid repetition, a description thereof is omitted.
An embodiment of the present invention further provides a computer device, please refer to fig. 3, fig. 3 is a schematic structural diagram of the computer device provided in the embodiment of the present invention, and the computer device D30 includes: a processor D301, a memory D302, and a computer program stored on the memory D302 and executable on the processor D301. The processor D301 invokes the computer program stored in the memory D302 to execute each step in the method for predicting the performance of the cable insulation material based on the near infrared spectrum provided in the embodiment of the present invention, and can achieve the same technical effects, which are not described herein again with reference to the description in the above embodiment.
It should be noted that, as will be understood by those skilled in the art, the electronic device in the embodiments of the present invention is a device capable of automatically performing numerical calculation and/or information processing according to a preset or stored instruction, and its hardware includes, but is not limited to, a microprocessor, an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), a programmable gate array (Field-Programmable Gate Array, FPGA), a digital processor (Digital Signal Processor, DSP), an embedded device, and the like. The electronic device may be a computing device such as a desktop computer, a notebook computer, a palm computer, a cloud server, and the like. The electronic equipment can perform man-machine interaction in a mode of a keyboard, a mouse, a remote controller, a touch pad or voice control equipment and the like.
The embodiment of the invention also provides a computer readable storage medium, on which a computer program is stored, which when executed by a processor, implements each process and step in the method for predicting performance of cable insulation material based on near infrared spectrum provided by the embodiment of the invention, and can implement the same technical effects, and in order to avoid repetition, the description is omitted here.
The readable storage medium includes flash memory, a hard disk, a multimedia card, a card memory (e.g., SD or DX memory, etc.), a Random Access Memory (RAM), a Static Random Access Memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, and the like. In some embodiments, the memory may be an internal storage unit of the electronic device, such as a hard disk or a memory of the electronic device. In other embodiments, the memory may also be an external storage device of the electronic device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash Card (Flash Card) or the like. Of course, the memory may also include both internal storage units of the electronic device and external storage devices. In this embodiment, the memory is generally used to store an operating device installed in an electronic apparatus and various kinds of application software, such as program codes of a cable insulation material performance prediction method based on near infrared spectrum. In addition, the memory can be used to temporarily store various types of data that have been output or are to be output.
Those skilled in the art will appreciate that the implementation of all or part of the above-described embodiment of the near infrared spectrum-based method for predicting the performance of a cable insulation material may be implemented by a computer program for instructing relevant hardware, and the program may be stored in a computer readable storage medium, and the program may include the above-described embodiment of the near infrared spectrum-based method for predicting the performance of a cable insulation material when executed. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM) or the like.
The above embodiments are preferred embodiments of the method for predicting performance of cable insulation material based on near infrared spectrum, and are not intended to limit the scope of the invention, which includes but is not limited to the embodiments, and equivalent changes of shape and structure according to the invention are all within the scope of the invention.
Claims (10)
1. The cable insulation material performance prediction method based on the near infrared spectrum is characterized by comprising the following steps of:
acquiring a correction set, wherein the correction set comprises spectrum data and performance data of a plurality of correction samples;
Acquiring spectrum data of a sample to be detected, and acquiring k correction samples nearest to the sample to be detected from the correction set;
constructing a spectral data constraint condition and a weight function of the sample to be detected according to the spectral data of the sample to be detected and the spectral data of the nearest k correction samples;
constructing a target correction model based on the spectral data of the sample to be detected, the weight function and the spectral data constraint condition, and calculating a weight coefficient corresponding to the spectral data of each correction sample in the k nearest correction samples through the target correction model;
and determining target performance parameters of the sample to be tested according to the weight coefficient corresponding to the spectral data of each correction sample in the k nearest correction samples and the performance data of a plurality of correction samples in the correction set.
2. The method for predicting performance of cable insulation based on near infrared spectrum according to claim 1, wherein the obtaining the correction set comprises:
acquiring spectrum data and performance data of a plurality of correction samples, wherein the spectrum data of each correction sample is a point in a p-dimensional space, and p is more than or equal to 1;
constructing the correction set (x) based on the spectral data and the performance data of the plurality of correction samples i ,y i )，i＝1,2,3,...,n，x i Is the spectral data of the ith calibration sample, y i Is the performance data for each calibration sample at i.
3. The method for predicting performance of cable insulation based on near infrared spectrum according to claim 1, wherein the obtaining spectral data of a sample to be measured, obtaining k correction samples nearest to the sample to be measured from the correction set, comprises:
acquiring the sample to be detected, and acquiring spectrum data of the sample to be detected based on spectrum acquisition equipment;
calculating the difference value between the spectrum data of the sample to be detected and the spectrum data of each correction sample in the correction set based on a preset distance algorithm;
and constructing a difference sequence according to each difference value, and determining k correction samples corresponding to the smallest difference value in the difference sequence as k correction samples nearest to the sample to be detected.
4. The method for predicting performance of a cable insulation material based on near infrared spectroscopy according to claim 2, wherein the spectral data constraint includes a nearest sample point constraint and a spectral data range constraint, and the constructing the spectral data constraint of the sample to be measured according to the spectral data of the sample to be measured and the spectral data of the nearest k correction samples includes:
Constructing a nearest sample point constraint condition of the sample to be detected according to the spectral data of the sample to be detected and the spectral data of the nearest k correction samples, wherein the nearest sample point constraint condition is as follows:
wherein x is u For the spectrum data of the sample to be tested, delta is equal to x u The range between the spectral data of the nearest k calibration samples,is x u A set of sample points within the spectral data of the nearest k calibration samples;
constructing spectral data range constraint conditions of correction samples in the k nearest correction samples based on the nearest sample point constraint conditions, wherein the spectral data range constraint conditions are as follows:
wherein x is j The spectral data of the j-th calibration sample among the k nearest calibration samples,for the correction set (x i ,y i ) Is of the category X u The nearest k adjacent sets of sample points of the spectral data of the corrected samples.
5. The method for predicting performance of cable insulation based on near infrared spectrum according to claim 4, wherein the constructing the weight function of the sample to be measured according to the spectral data of the sample to be measured and the spectral data of the k nearest correction samples comprises:
introducing weight coefficients corresponding to the spectral data of the k nearest correction samples, and constructing a weight function of the sample to be detected according to the spectral data of the sample to be detected, the spectral data of the k nearest correction samples and the weight coefficients corresponding to the spectral data of the k nearest correction samples, wherein the weight function is as follows:
The constraint conditions of the weight coefficients are as follows:
wherein w is uj Weighting coefficient of spectral data of j-th correction sample among k nearest correction samples, ε (W u ) As a weight function, W u The overall weight coefficient is calculated for the weight coefficients based on the nearest k correction samples.
6. The method for predicting performance of cable insulation material based on near infrared spectrum according to claim 5, wherein the constructing a target correction model based on the spectral data of the sample to be measured, the weight function and the spectral data constraint condition, calculating the weight coefficient corresponding to the spectral data of each of the k nearest correction samples by the target correction model comprises:
constructing the target correction model based on the spectrum data of the sample to be detected, the weight function and the spectrum data constraint condition, wherein the target correction model is as follows:
L(W u ；θ)＝(x u -X T B u W u ) T (x u -X T B u W u )+θ(1-W u T )
wherein B is u Is a matrix of n X k, X is a correction set, X T B u Is X and X u The nearest k sets of spectral data of the correction samples, θ being the lagrangian, T representing the transpose of the matrix;
the overall weight coefficient is derived through the target correction model and is brought into the weight coefficient constraint condition to be calculated, and the Lagrange operator is determined;
Calculating the integral weight coefficient corresponding to the spectral data of the nearest k correction samples based on the Lagrangian operator obtained by calculation;
and calculating the weight coefficient corresponding to the spectral data of each correction sample in the k nearest correction samples based on the integral weight coefficient.
7. The method for predicting performance of cable insulation based on near infrared spectrum according to claim 6, wherein determining the target performance parameter of the sample to be measured according to the weight coefficient corresponding to the spectral data of each of the k nearest correction samples and the performance data of a plurality of correction samples in the correction set comprises:
according to the weight coefficient corresponding to the spectral data of each correction sample in the k nearest correction samples and the performance data of the k nearest correction samples in the correction set, predicting the target performance parameter of the sample to be measured, wherein the prediction formula is as follows:
wherein,is the target performance parameter.
8. A near infrared spectrum-based cable insulation material performance prediction apparatus, characterized in that the apparatus comprises:
the data acquisition module is used for acquiring a correction set, wherein the correction set comprises spectrum data and performance data of a plurality of correction samples;
The adjacent sample acquisition module is used for acquiring spectral data of a sample to be detected and acquiring k correction samples nearest to the sample to be detected from the correction set;
the first construction module is used for constructing spectral data constraint conditions and weight functions of the sample to be tested according to the spectral data of the sample to be tested and the spectral data of the k nearest correction samples;
the second construction module is used for constructing a target correction model based on the spectrum data of the sample to be detected, the weight function and the spectrum data constraint condition, and calculating the weight coefficient corresponding to the spectrum data of each correction sample in the k nearest correction samples through the target correction model;
and the performance prediction module is used for determining target performance parameters of the sample to be detected according to the weight coefficient corresponding to the spectral data of each correction sample in the k nearest correction samples and the performance data of a plurality of correction samples in the correction set.
9. An electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor implements the steps of the near infrared spectrum based cable insulation performance prediction method according to any one of claims 1 to 7 when the computer program is executed.
10. A computer-readable storage medium storing a computer program, characterized in that the computer program, when executed by a processor, implements the steps of the near infrared spectrum based cable insulation performance prediction method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311680616.3A CN117872048A (en) | 2023-12-08 | 2023-12-08 | Cable insulation material performance prediction method based on near infrared spectrum and related equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311680616.3A CN117872048A (en) | 2023-12-08 | 2023-12-08 | Cable insulation material performance prediction method based on near infrared spectrum and related equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117872048A true CN117872048A (en) | 2024-04-12 |
Family
ID=90589153
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311680616.3A Pending CN117872048A (en) | 2023-12-08 | 2023-12-08 | Cable insulation material performance prediction method based on near infrared spectrum and related equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117872048A (en) |
-
2023
- 2023-12-08 CN CN202311680616.3A patent/CN117872048A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Raissi et al. | Statistical process optimization through multi-response surface methodology | |
| JP2017049674A (en) | Prediction performance curve estimation program, prediction performance curve estimation device and prediction performance curve estimation method | |
| US9523635B2 (en) | Apparatus and methods of spectral searching using wavelet transform coefficients | |
| CN108681697B (en) | Feature selection method and device | |
| US11550823B2 (en) | Preprocessing for a classification algorithm | |
| CN114026510A (en) | Method and system for optimizing process parameters in an additive manufacturing process | |
| WO2018162047A1 (en) | Tester and method for testing a device under test and tester and method for determining a single decision function | |
| CN107016416B (en) | Data classification prediction method based on neighborhood rough set and PCA fusion | |
| Mehtätalo et al. | The use of quantile trees in the prediction of the diameter distribution of a stand | |
| Bhushan et al. | New efficient logarithmic estimators using multi‐auxiliary information under ranked set sampling | |
| Portier et al. | Bootstrap testing of the rank of a matrix via least-squared constrained estimation | |
| US20200279148A1 (en) | Material structure analysis method and material structure analyzer | |
| CN113869642A (en) | Enterprise activity determination method and device, electronic equipment and storage medium | |
| CN117872048A (en) | Cable insulation material performance prediction method based on near infrared spectrum and related equipment | |
| US11187992B2 (en) | Predictive modeling of metrology in semiconductor processes | |
| CN106485286B (en) | Matrix classification model based on local sensitivity discrimination | |
| CN114398228A (en) | Method and device for predicting equipment resource use condition and electronic equipment | |
| Serrà et al. | On tuning consistent annealed sampling for denoising score matching | |
| CN109933579B (en) | Local K neighbor missing value interpolation system and method | |
| CN110795689B (en) | Matrix data processing method, device and storage medium | |
| CN117934019B (en) | Copper concentrate sample tracing method and system based on deep learning | |
| JP2004227279A (en) | Failure cause diagnosing method using mahalanobis distance and program | |
| CN113987909B (en) | Oilpaper insulation aging prediction method, device, computer equipment and storage medium | |
| JP7424509B2 (en) | Learning device, identification device, learning method, identification method, learning program, and identification program | |
| US20240011910A1 (en) | Methods and systems for raman spectra-based identification of chemical compounds |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |