A method and system for testing electromagnetic shielding performance of carbon fiber prepreg

Abstract

The present invention relates to a method and system for testing the electromagnetic shielding performance of carbon fiber prepreg, wherein the method includes: statistically analyzing the concentrated value of the uniformity of medium distribution and the concentrated value of the density of medium distribution; generating the first interlayer stability and the second interlayer stability; generating the electromagnetic shielding effectiveness prediction value through the electromagnetic shielding effectiveness simulator bound to the electromagnetic shielding carbon fiber prepreg formula; and calculating the weighted mean of the electromagnetic shielding performance test value in combination with the deviation weight set to obtain the electromagnetic shielding effectiveness correction value. The present application solves the technical problem in the prior art that when the electromagnetic shielding performance test of carbon fiber prepreg is carried out, it is susceptible to environmental interference and hardware precision deviation, resulting in low accuracy of the electromagnetic shielding performance test results, and achieves the technical effect of improving the accuracy of the electromagnetic shielding performance test results by combining multi-stage analysis with electromagnetic shielding effectiveness prediction.

Description

Technical Field

[0001] The invention relates to the field of electromagnetic shielding performance testing, and in particular to a method and system for testing the electromagnetic shielding performance of a carbon fiber prepreg. Background Art

[0002] The electromagnetic shielding performance test of carbon fiber prepreg is an important indicator for evaluating its application value. At present, the commonly used electromagnetic shielding performance test methods mainly include coaxial cable method, waveguide method, shielding room method, etc. In the actual test process, the electromagnetic shielding performance test results are often affected by many factors, such as environmental temperature and humidity, electromagnetic interference, test equipment accuracy deviation, etc. These factors will have a significant impact on the accuracy and reliability of the test results.

[0003] However, the current test methods for the electromagnetic shielding performance of carbon fiber prepregs are too dependent on the results of actual hardware tests, and there is no effective way to verify the stability of the test results. Testers usually directly use the data given by the test equipment as the final result, lacking further analysis and verification of the data. This approach is acceptable in scenarios where electromagnetic shielding effectiveness requirements are not high, but in high-end application scenarios with strict requirements, the credibility of the test results is significantly reduced, limiting the application scope of carbon fiber prepregs. On the other hand, existing test methods often ignore the influence of parameters in the material preparation process. The preparation of carbon fiber prepregs involves multiple stages, including film preparation, prepreg preparation, and composite board preparation. The wheel pressure parameters in each stage will affect the final electromagnetic shielding performance. Therefore, the prior art has the defect of low accuracy of the electromagnetic shielding performance test results of carbon fiber prepregs. Summary of the invention

[0004] The present invention aims to solve the technical problem in the prior art that when conducting electromagnetic shielding performance tests on carbon fiber prepregs, the carbon fiber prepregs are easily affected by environmental interference and hardware precision deviation, resulting in low accuracy of electromagnetic shielding performance test results. A method and system for testing the electromagnetic shielding performance of carbon fiber prepregs are provided to solve the problem.

[0005] The technical solution of the present invention to solve the above technical problems is as follows:

[0006] In a first aspect, the present invention provides a method for testing the electromagnetic shielding performance of a carbon fiber prepreg, comprising: counting the concentrated value of the uniformity of medium distribution and the concentrated value of the density of medium distribution of a first sample set that meets the adhesive film formula and rolling parameters; counting the proportion of stable samples of a second sample set that meets the carbon fiber prepreg formula and hot pressing-cooling parameters to generate a first interlayer stability, where a stable sample refers to a sample that has not been separated from a preset service time; counting the proportion of stable samples of a third sample set that meets the electromagnetic shielding carbon fiber prepreg formula and hot pressing composite parameters to generate a second interlayer stability; through an electromagnetic shielding effectiveness simulator bound to the electromagnetic shielding carbon fiber prepreg formula, processing the concentrated value of the uniformity of medium distribution, the concentrated value of the density of medium distribution, the first interlayer stability, the second interlayer stability and the test electromagnetic wave frequency band to generate an electromagnetic shielding effectiveness prediction value; according to the electromagnetic shielding effectiveness prediction value, performing deviation weight distribution on the electromagnetic shielding performance test value to obtain a deviation weight set, and performing weighted mean calculation on the electromagnetic shielding performance test value in combination with the deviation weight set to obtain an electromagnetic shielding effectiveness correction value.

[0007] Optionally, statistics of a first sample set satisfying the film formula and rolling parameters include: constructing a query constraint statement with the film formula and the rolling parameters as constraints; retrieving a local sample set satisfying the query constraint statement from a local history preparation log; when the number of the local sample set is less than or equal to a convergence sample number threshold, uploading the query constraint statement to a cloud server to obtain networked feedback information, wherein the networked feedback information includes a networked sample medium distribution uniformity concentration value and a networked sample medium distribution density concentration value; adding the networked sample medium distribution uniformity concentration value and the networked sample medium distribution density concentration value, together with the local sample set, into the first sample set.

[0008] Optionally, a query constraint statement is constructed with the film formula and the rolling parameters as constraints, including: the film formula includes parts by weight of raw materials and parts by weight of electromagnetic shielding media; the rolling parameters include rolling speed and roller gap; the parts by weight of raw materials and the parts by weight of electromagnetic shielding media are weighted for electromagnetic shielding correlation to obtain a first weight distribution; the rolling speed and the roller gap are weighted for medium distribution correlation to obtain a second weight distribution; a first distance function is constructed based on the film formula according to the first weight distribution, and a film formula query constraint statement is constructed in combination with a first distance threshold; a second distance function is constructed based on the rolling parameters according to the second weight distribution, and a rolling parameter query constraint statement is constructed in combination with a second distance threshold; the film formula query constraint statement and the rolling parameter query constraint statement are logically ANDed to obtain the query constraint statement.

[0009] Optionally, when the number of the local sample set is less than or equal to the convergence sample number threshold, the query constraint statement is uploaded to the cloud server to obtain networked feedback information, wherein the networked feedback information includes a networked sample medium distribution uniformity concentration value and a networked sample medium distribution density concentration value, including: when the cloud server receives the query constraint statement, the cloud server sends a query task and a key to the distributed node to obtain distributed encrypted feedback information, wherein any encrypted feedback information of the distributed encrypted feedback information includes: a sample film formula, a sample rolling parameter, a sample medium record distribution density, and a sample medium record distribution uniformity parameter; through the cloud server, based on the sample film formula and the sample rolling parameter, a sample medium record distribution uniformity parameter set and a sample medium record distribution density set that meet the query constraint statement are selected from the distributed encrypted feedback information; through the cloud server, the sample medium record distribution uniformity parameter set and the sample medium record distribution density set are respectively evaluated for concentration values ​​to obtain the networked sample medium distribution uniformity concentration value and the networked sample medium distribution density concentration value.

[0010] Optionally, an electromagnetic shielding effectiveness simulator bound to the electromagnetic shielding carbon fiber prepreg formula is used to process the medium distribution uniformity concentration value, the medium distribution density concentration value, the first interlayer stability, the second interlayer stability and the test electromagnetic wave frequency band to generate an electromagnetic shielding effectiveness prediction value, including: building a medium distribution feature extraction channel based on at least two layers of fully connected neural networks, building an interface stability feature extraction channel based on at least three layers of fully connected neural networks, putting the medium distribution uniformity concentration value, the medium distribution density concentration value, the first interlayer stability, the second interlayer stability, and the test frequency band into the input layer, and putting the electromagnetic shielding effectiveness identification value into the output layer to obtain the electromagnetic shielding effectiveness simulator architecture; collecting the electromagnetic shielding carbon fiber Historical preparation data of prepreg formula; collecting a concentrated interval of electromagnetic shielding effectiveness historical test values ​​that meet the historical preparation data, and setting it as an electromagnetic shielding effectiveness historical identification interval; processing the historical preparation data to obtain a dielectric distribution uniformity historical record value, a dielectric distribution density historical record value, a first interlayer stability historical record value, a second interlayer stability historical record value, and a historical test frequency band; training the electromagnetic shielding effectiveness simulator architecture based on the dielectric distribution uniformity historical record value, the dielectric distribution density historical record value, the first interlayer stability historical record value, the second interlayer stability historical record value, and the historical test frequency band in combination with the electromagnetic shielding effectiveness historical identification interval to generate the electromagnetic shielding effectiveness simulator.

[0011] Optionally, according to the electromagnetic shielding effectiveness prediction value, the electromagnetic shielding performance test value is subjected to deviation weight distribution to obtain a deviation weight set, and the electromagnetic shielding performance test value is weighted mean calculated in combination with the deviation weight set to obtain an electromagnetic shielding effectiveness correction value, including: the electromagnetic shielding effectiveness prediction value is an electromagnetic shielding effectiveness prediction interval; when the electromagnetic shielding performance test value is distributed in the electromagnetic shielding effectiveness prediction interval, a credible electromagnetic shielding performance test value is added; when the electromagnetic shielding performance test value does not belong to the electromagnetic shielding effectiveness prediction interval, an electromagnetic shielding performance test value to be weighted is added; according to the boundary distance between the electromagnetic shielding performance test value to be weighted and the electromagnetic shielding effectiveness prediction interval, a deviation weight distribution is performed to obtain a deviation weight set; according to the deviation weight set, a weighted mean calculation is performed on the electromagnetic shielding performance test value to be weighted to obtain a fitting value of the electromagnetic shielding performance test value to be weighted; and the mean of the fitting value of the electromagnetic shielding performance test value to be weighted and the credible electromagnetic shielding performance test value is calculated to obtain the electromagnetic shielding effectiveness correction value.

[0012] Optionally, deviation weight distribution is performed according to the boundary distance between the electromagnetic shielding performance test value to be weighted and the electromagnetic shielding effectiveness prediction interval to obtain a deviation weight set, including: when the boundary distance between the electromagnetic shielding performance test value to be weighted and the electromagnetic shielding effectiveness prediction interval is greater than or equal to a boundary distance threshold, the electromagnetic shielding performance test value to be weighted is deleted.

[0013] In a second aspect, the present invention provides a system for testing the electromagnetic shielding performance of carbon fiber prepregs, including: a concentrated value statistics module, which is used to count the concentrated value of medium distribution uniformity and the concentrated value of medium distribution density of a first sample set that meets the film formula and rolling parameters; a first stability module, which is used to count the proportion of stable samples of a second sample set that meets the carbon fiber prepreg formula and hot pressing-cooling parameters, and generate a first interlayer stability, where a stable sample refers to a sample that has not been separated from a preset service time; a second stability module, which is used to count the proportion of stable samples of a third sample set that meets the electromagnetic shielding carbon fiber prepreg formula and hot pressing composite parameters, and generate a first interlayer stability. The invention relates to a method for generating an electromagnetic shielding effectiveness prediction value by processing the medium distribution uniformity concentration value, the medium distribution density concentration value, the first interlayer stability, the second interlayer stability and the test electromagnetic wave frequency band through an electromagnetic shielding effectiveness simulator bound to the electromagnetic shielding carbon fiber prepreg formula, so as to generate an electromagnetic shielding effectiveness prediction value; a correction value acquisition module is used to perform deviation weight distribution on the electromagnetic shielding performance test value according to the electromagnetic shielding effectiveness prediction value, obtain a deviation weight set, perform weighted mean calculation on the electromagnetic shielding performance test value in combination with the deviation weight set, and obtain an electromagnetic shielding effectiveness correction value.

[0014] The beneficial effects of the present invention are:

[0015] By counting the concentrated values ​​of the uniformity of the medium distribution and the concentrated values ​​of the medium distribution density of the first sample set that meets the film formula and rolling parameters, the key indicators of the film quality are reflected, providing basic data for the subsequent prediction of the electromagnetic shielding effectiveness; by counting the proportion of stable samples of the second sample set that meets the carbon fiber prepreg formula and hot pressing-cooling parameters, the first interlayer stability is generated, reflecting the stability of the prepreg interlayer structure, providing a basis for analyzing the electromagnetic shielding performance; by counting the proportion of stable samples of the third sample set that meets the electromagnetic shielding carbon fiber prepreg formula and hot pressing composite parameters, the second interlayer stability is generated, and the structural stability of the composite prepreg is evaluated, which provides a basis for analyzing the electromagnetic shielding performance; by using the electromagnetic shielding bound to the electromagnetic shielding carbon fiber prepreg formula The effectiveness simulator processes the concentrated value of medium distribution uniformity, the concentrated value of medium distribution density, the stability between the first layer and the second layer, and the test electromagnetic wave frequency band to generate an electromagnetic shielding effectiveness prediction value. This prediction method based on multi-stage parameters is in line with the preparation process of electromagnetic shielding carbon fiber prepreg, avoids the limitations of one-stop analysis, and improves the accuracy of the prediction results. According to the electromagnetic shielding effectiveness prediction value, the electromagnetic shielding performance test value is distributed with deviation weights to obtain a deviation weight set, and the electromagnetic shielding performance test value is weighted averaged with the deviation weight set to obtain the electromagnetic shielding effectiveness correction value, which effectively corrects the environmental interference and hardware precision deviation that may exist in the actual test, making the final electromagnetic shielding performance test result more reliable and accurate.

[0016] Through the above technical scheme, each stage of carbon fiber prepreg preparation is analyzed, and the electromagnetic shielding effectiveness is predicted based on these analysis results. Then, the test results are accurately corrected through deviation weight distribution, which conforms to the actual process of material preparation and improves the accuracy and reliability of the electromagnetic shielding performance test results. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A schematic flow chart of a method for testing electromagnetic shielding performance of a carbon fiber prepreg provided by the present invention;

[0018] Figure 2 A structural schematic diagram of an electromagnetic shielding performance testing system for carbon fiber prepreg provided by the present invention.

[0019] In the accompanying drawings, the components represented by the reference numerals are as follows:

[0020] A concentrated value statistics module 11 , a first stability module 12 , a second stability module 13 , a predicted value generation module 14 , and a corrected value acquisition module 15 . DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0022] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0023] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or explanation". Any embodiment described as "for example" in the present invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any technician in the field to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes will not be elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed in the present invention.

[0024] Embodiment 1, as Figure 1 As shown, an embodiment of the present invention provides a method for testing electromagnetic shielding performance of carbon fiber prepreg, comprising:

[0025] S100: Counting the concentrated value of the uniformity of medium distribution and the concentrated value of the medium distribution density of the first sample set that meets the film formula and rolling parameters.

[0026] Specifically, first, obtain the film formula and rolling parameters for preparing carbon fiber prepreg. For example, the film formula is: select a single-component thermolatent epoxy resin as the raw material, add a conductive agent or a magnetic conductive agent thereto, and then evenly coat it on the release paper under the precise control of a three-roll coating machine to form a layer of resin film to obtain a film. In a specific embodiment, the conductive agent is carbon powder, and the addition amount is 3% of the mass of the epoxy resin; the thermolatent epoxy resin is a compound of epoxy resin, including a resin matrix, a curing agent, a promoter, and an auxiliary agent, and the mass ratio is 100:30:15:5; the resin matrix is ​​a mixture of E-51 epoxy resin and phenolic epoxy resin in a mass ratio of 7:3, and the viscosity is 35000cps at 70°C; the curing agent is a micronized dicyandiamide with an average particle size of 10μm; the promoter is a micronized diuron with an average particle size of 8μm; the auxiliary agent includes a composite of a toughening agent, a wetting agent and a coupling agent. For the rolling parameters, the surface temperature of the rollers was set to 120°C and the cylinder pressure was set to 0.6 MPa.

[0027] Then, for the obtained film formula and rolling parameters, the first sample set that meets the conditions is counted. The so-called first sample set that meets the conditions refers to the sample set recorded in the historical preparation log and produced using the same or similar film formula and rolling parameters. Specifically, by querying the local historical preparation log, the sample data that meets the obtained film formula and rolling parameters are retrieved to form a local sample set. When the number of local sample sets is insufficient to support subsequent analysis, the query constraint statement can be uploaded to the cloud server to obtain more sample data that meet the conditions, thereby expanding the first sample set. Then, the first sample set is statistically analyzed to calculate the medium distribution uniformity concentration value and the medium distribution density concentration value. Among them, the medium distribution uniformity concentration value represents the uniformity of the distribution of the conductive agent or the magnetic conductive agent in the film, which can be obtained by calculating the mean, median or weighted average of the medium distribution uniformity parameters in the first sample set; the medium distribution density concentration value represents the distribution density of the conductive agent or the magnetic conductive agent in the film, which can also be obtained by counting the concentration trend of the medium distribution density parameters in the first sample set.

[0028] Obtaining the concentrated value of medium distribution uniformity and medium distribution density is of great significance for evaluating the electromagnetic shielding performance of the film, and provides basic data support for subsequent prediction and correction of electromagnetic shielding effectiveness.

[0029] S200: Count the proportion of stable samples of the second sample set that meet the carbon fiber prepreg formula and hot pressing-cooling parameters to generate the first interlayer stability. Stable samples refer to samples that have not been out of the preset service time.

[0030] Specifically, first, determine the carbon fiber prepreg formula and hot pressing-cooling parameters. In one embodiment, the preparation of the carbon fiber prepreg uses carbon fiber woven cloth as the main raw material, and the weaving is arranged between the upper and lower layers of adhesive film, and the carbon fiber prepreg is obtained through hot pressing and cooling processes under the control of the prepreg impregnation machine. In a specific embodiment, the weight ratio per unit area of ​​the upper and lower layers of adhesive film is upper layer adhesive film: lower layer adhesive film = 6:3; the carbon fiber woven cloth used is T300-3K plain woven cloth; the hot pressing parameters are that the surface temperature of the rollers is set to 100°C, and the cylinder pressure is set to 0.5MPa; the cooling parameters are that the surface temperature of the cold plate is set to 10°C.

[0031] Based on the determined carbon fiber prepreg formula and hot pressing-cooling parameters, the corresponding second sample set is collected, and the proportion of stable samples is counted. The so-called stable sample refers to a sample in which the interface between the upper and lower layers of the carbon fiber prepreg film remains stable and no detachment occurs within the preset service time. By calculating the ratio of the number of stable samples to the total number of samples in the second sample set, the first interlayer stability is obtained, which reflects the firmness of the bonding between the upper and lower layers of the carbon fiber prepreg film and the carbon fiber woven cloth during the preparation process.

[0032] The stability of the first interlayer is an important parameter for evaluating the quality of carbon fiber prepreg, which directly affects the stability and reliability of its electromagnetic shielding performance. A higher stability of the first interlayer indicates that the carbon fiber and the film are more firmly bonded, and it is less likely to have problems such as delamination and detachment during subsequent use, which is conducive to maintaining a good electromagnetic shielding effect.

[0033] S300: Count the proportion of stable samples in the third sample set that meet the electromagnetic shielding carbon fiber prepreg formula and hot pressing composite parameters, and generate the second interlayer stability.

[0034] Specifically, first, determine the electromagnetic shielding carbon fiber prepreg formula and hot pressing composite parameters. In one embodiment, the preparation of the electromagnetic shielding carbon fiber prepreg is based on the carbon fiber prepreg, peeling off the release paper of the upper layer of the carbon fiber prepreg film, and then hot pressing the shielding net and the outer protective film together with the carbon fiber prepreg. The resin of the upper layer of the film overflows from the mesh of the shielding net and adheres to the outer protective film; forming an electromagnetic shielding carbon fiber prepreg with a four-layer structure of outer protective film, shielding net, carbon fiber prepreg and release paper from top to bottom. In a specific embodiment, the shielding net uses a continuous copper mesh with a surface density of 200g / m², and the shielding effectiveness of the copper mesh at a frequency of 1GHz is 45dB; the outer protective film uses a PE film with a thickness of 0.05mm; for the hot pressing composite parameters, the surface temperature of the rollers is set to 60°C and the cylinder pressure is set to 0.4MPa.

[0035] After collecting the third sample set that meets the statistical requirements of the electromagnetic shielding carbon fiber prepreg formula and hot pressing composite parameters, the proportion of stable samples is counted, that is, the ratio of the number of samples that did not show the phenomenon of separation between the shielding net and the carbon fiber prepreg layer within the preset service time to the total number of samples in the third sample set, thereby generating the second interlayer stability. The second interlayer stability reflects the firmness of the interface bonding between the shielding net and the carbon fiber prepreg layer in the electromagnetic shielding carbon fiber prepreg. A higher second interlayer stability indicates that a stable interface structure is formed during the hot pressing composite process, which is conducive to maintaining the stability of the electromagnetic shielding performance of the product during use and avoiding the decline in shielding effectiveness due to interlayer separation.

[0036] S400: Through the electromagnetic shielding effectiveness simulator bound to the electromagnetic shielding carbon fiber prepreg formula, the medium distribution uniformity concentration value, medium distribution density concentration value, first layer stability, second layer stability and test electromagnetic wave frequency band are processed to generate electromagnetic shielding effectiveness prediction values.

[0037] Specifically, first, an electromagnetic shielding effectiveness simulator corresponding to a specific electromagnetic shielding carbon fiber prepreg formula is obtained. The electromagnetic shielding effectiveness simulator is constructed using artificial intelligence algorithms such as neural networks, and can predict the electromagnetic shielding effectiveness of the electromagnetic shielding carbon fiber prepreg based on input parameters.

[0038] After obtaining the electromagnetic shielding effectiveness simulator, the dielectric distribution uniformity concentration value, dielectric distribution density concentration value, first interlayer stability and second interlayer stability are input into the electromagnetic shielding effectiveness simulator in combination with the test electromagnetic wave frequency band. Among them, the dielectric distribution uniformity concentration value and dielectric distribution density concentration value reflect the distribution state of the conductive agent or magnetic conductive agent in the adhesive film; the first interlayer stability reflects the interface firmness of the upper and lower adhesive films and the carbon fiber layer of the carbon fiber prepreg; the second interlayer stability reflects the interface bonding state between the shielding net and the carbon fiber prepreg layer; the test electromagnetic wave frequency band indicates the specific frequency range of performance evaluation, such as 0.1MHz~18000MHz. Then, the electromagnetic shielding effectiveness simulator processes the dielectric distribution uniformity concentration value, dielectric distribution density concentration value, first interlayer stability and second interlayer stability, and combines the test electromagnetic wave frequency band to analyze the intrinsic relationship between these parameters and the electromagnetic shielding effectiveness, comprehensively evaluate the electromagnetic shielding capability of the current electromagnetic shielding carbon fiber prepreg under the specific test electromagnetic wave frequency band, and output the electromagnetic shielding effectiveness prediction value. The electromagnetic shielding effectiveness prediction value represents the electromagnetic shielding effectiveness level that the electromagnetic shielding carbon fiber prepreg is expected to achieve under given parameter conditions.

[0039] S500: performing deviation weight distribution on the electromagnetic shielding performance test value according to the electromagnetic shielding effectiveness prediction value to obtain a deviation weight set, performing weighted mean calculation on the electromagnetic shielding performance test value in combination with the deviation weight set to obtain an electromagnetic shielding effectiveness correction value.

[0040] Specifically, first, the actual electromagnetic shielding performance test value of the electromagnetic shielding carbon fiber prepreg is obtained. The actual electromagnetic shielding performance test value is usually measured by professional testing equipment within the test electromagnetic wave frequency band. Then, the actual electromagnetic shielding performance test value is compared and analyzed with the electromagnetic shielding effectiveness prediction value generated by the electromagnetic shielding effectiveness simulator. In the comparison process, the electromagnetic shielding effectiveness prediction value is used as a reference benchmark to conduct a credibility assessment on the actual electromagnetic shielding performance test value. When the actual electromagnetic shielding performance test value falls within the electromagnetic shielding effectiveness prediction value, the test value is considered to have a high credibility and is classified as a credible electromagnetic shielding performance test value; when the test value deviates from the prediction interval, it indicates that there may be a test error or anomaly, and it is classified as an electromagnetic shielding performance test value to be weighted.

[0041] For the electromagnetic shielding performance test value to be weighted, a deviation weight distribution model is established according to its distance relationship with the boundary of the prediction interval. Specifically, by calculating the distance between the test value to be weighted and the boundary of the prediction interval, and taking the ratio of a single distance to the total distance as the weight, a deviation weight set is formed. When the distance between a test value and the boundary of the prediction interval exceeds a preset threshold, it can be regarded as an outlier and removed to improve data quality. Then, combined with the obtained deviation weight set, the weighted electromagnetic shielding performance test value to be weighted is calculated by weighted mean to obtain the fitting value of the electromagnetic shielding performance test value to be weighted. After that, the fitting value is comprehensively calculated with the mean of the credible electromagnetic shielding performance test value to obtain the electromagnetic shielding effectiveness correction value.

[0042] By correcting the electromagnetic shielding performance test value according to the electromagnetic shielding effectiveness prediction value, the electromagnetic shielding effectiveness correction value is obtained as the electromagnetic performance test result of the electromagnetic shielding carbon fiber prepreg, which can effectively reduce the impact of test environment interference, hardware accuracy deviation and other factors on the test results, and improve the accuracy and reliability of the electromagnetic shielding performance test. Using the electromagnetic shielding effectiveness correction value as the electromagnetic performance test result of the carbon fiber prepreg can objectively and accurately reflect the actual shielding performance of the electromagnetic shielding carbon fiber prepreg, and provide more reliable data support for product quality evaluation and application scenario adaptation.

[0043] Furthermore, the first sample set satisfying the film formulation and rolling parameters is statistically analyzed, including:

[0044] S110: constructing a query constraint statement with the film formula and rolling parameters as constraints;

[0045] S120: preparing logs from local history and retrieving a local sample set that satisfies the query constraint statement;

[0046] S130: When the number of local sample sets is less than or equal to the converged sample number threshold, the query constraint statement is uploaded to the cloud server to obtain networking feedback information, wherein the networking feedback information includes a networked sample medium distribution uniformity concentration value and a networked sample medium distribution density concentration value;

[0047] S140: Add the networked sample medium distribution uniformity concentration value and the networked sample medium distribution density concentration value, and the local sample set into the first sample set.

[0048] In an optional implementation, first, a query constraint statement is constructed with the film formula and rolling parameters as constraints. Specifically, the obtained film formula information (such as resin matrix type, curing agent type, accelerator type, conductive agent or magnetic conductive agent, etc.) and rolling parameter information (such as rolling temperature, pressure, etc.) are used as constraints to construct a query constraint statement with a specific structure. The query constraint statement contains logical relationships and matching rules, which are used to retrieve sample data that meet the conditions from the local historical preparation log in the future. In addition, the process of constructing the query constraint statement can be weighted based on the electromagnetic shielding correlation of specific formulas and parameters, so that the query results more accurately reflect the samples related to the performance of the target material. Then, a local sample set that meets the query constraint statement is retrieved from the local historical preparation log. Specifically, the constructed query constraint statement is used to query the local historical preparation log, and the sample set that meets the constraint conditions is screened out as the local sample set. Among them, the local historical preparation log records the material performance data under various film formulas and rolling parameters in the previous production process, including key parameters such as medium distribution uniformity and medium distribution density.

[0049] When the number of samples in the retrieved local sample set is insufficient to support reliable statistical analysis (i.e., the number of samples is less than or equal to the preset convergence sample number threshold), the query constraint statement is uploaded to the cloud server to obtain more qualified sample data information and obtain network feedback information. It is worth noting that in order to protect the security of process data, the network information fed back by the cloud only contains result data such as the concentrated value of the medium distribution uniformity and the concentrated value of the medium distribution density of the networked samples, but does not contain specific process parameters and formula details. This design can not only make full use of distributed data resources, but also effectively avoid the leakage of sensitive process information. Subsequently, the network feedback information obtained from the cloud server is merged with the local sample set to form a complete first sample set. This fusion method of local data and cloud data makes the sample set more representative and statistically significant, providing a more reliable data basis for subsequent analysis. The merged first sample set contains sufficient medium distribution uniformity and medium distribution density information, which can more accurately reflect the distribution of material properties under specific film formulations and rolling parameters.

[0050] By counting the first sample set, a statistically significant sample set can be efficiently obtained, laying a solid data foundation for the subsequent evaluation and prediction of electromagnetic shielding performance, and at the same time, the needs of process confidentiality and data security are fully considered in the data acquisition process. The statistical method of the second sample set and the third sample set can also be carried out in the same way as the statistical method of the first sample set.

[0051] Furthermore, with the film formula and rolling parameters as constraints, query constraint statements are constructed, including:

[0052] S111: The film formula includes parts by weight of raw material and parts by weight of electromagnetic shielding medium;

[0053] S112: Rolling parameters include rolling speed and roller gap;

[0054] S113: performing electromagnetic shielding correlation weight configuration on the weight portions of the raw material material and the weight portions of the electromagnetic shielding medium to obtain a first weight distribution;

[0055] S114: performing medium distribution correlation weight configuration on the roller speed and the roller gap to obtain a second weight distribution;

[0056] S115: constructing a first distance function based on the film formula according to the first weight distribution, and constructing a film formula query constraint statement in combination with the first distance threshold;

[0057] S116: constructing a second distance function based on the rolling parameters according to the second weight distribution, and constructing a rolling parameter query constraint statement in combination with the second distance threshold;

[0058] S117: Perform logical AND relationship configuration on the film formula query constraint statement and the rolling parameter query constraint statement to obtain a query constraint statement.

[0059] Specifically, first, the components of the film formula are clarified, including the weight parts of raw materials and the weight parts of electromagnetic shielding media. Among them, the weight parts of raw materials refer to the ratio of various materials that constitute the basic structure of the film, such as the mass ratio of components such as resin matrix, curing agent, accelerator, and auxiliary agent; the weight parts of electromagnetic shielding medium specifically refer to the amount of conductive agents or magnetic conductive agents added to the film, such as carbon powder, copper powder, silver powder or magnetic conductive powder particles. These weight parts data are the basic parameters for constructing query constraint statements and directly affect the electromagnetic shielding performance of the film. At the same time, the key process parameters of the rolling parameters are clarified, including the rolling speed and roller gap. Among them, the rolling speed refers to the linear speed of the roller rotation in the three-roller coating machine, which affects the time and uniformity of the film forming; the roller gap refers to the distance between the adjacent roller surfaces, which affects the thickness and density of the film. These parameters have a significant impact on the distribution state of the conductive agent or magnetic conductive agent in the film, and thus affect the electromagnetic shielding effectiveness.

[0060] Then, the weight of the raw material and the electromagnetic shielding medium are weighted to obtain the first weight distribution. For example, the gray correlation analysis method is used to study the correlation between the weight of each raw material and the weight of the electromagnetic shielding medium and the electromagnetic shielding effectiveness. By calculating the ratio of the correlation of each substance to the total correlation, the weight distribution of the influence of each component on the electromagnetic shielding effectiveness is obtained. For example, if the correlation of the conductive agent is high, it occupies a larger proportion in the first weight distribution, indicating that it has a more significant impact on the electromagnetic shielding effectiveness. The first weight distribution is used to more accurately identify samples related to the target electromagnetic shielding performance in the subsequent query constraint construction. At the same time, the roll speed and roller gap are weighted to obtain the second weight distribution. Specifically, the correlation between the roll speed and the roller gap and the medium distribution state (uniformity and density) is analyzed, and the relative influence weight of each parameter is calculated. As key process parameters, the roll speed and the roller gap indirectly affect the electromagnetic shielding performance by affecting the distribution state of the medium in the film. The second weight distribution determined by correlation analysis can reflect the influence of different roller pressure parameters on medium distribution and provide a basis for the subsequent construction of parameter query constraints.

[0061] Then, using the obtained first weight distribution, a distance function is established to measure the similarity between the target film formula and the historical sample formula, that is, the first distance function. For example, the first distance function can be in the form of weighted Euclidean distance or weighted Manhattan distance, where the weight of each formula parameter is determined according to the first weight distribution. Then, combined with the preset first distance threshold, a film formula query constraint statement is formed to screen samples whose formula similarity meets the requirements. Based on the obtained second weight distribution, a distance function for evaluating the similarity of rolling parameters is established, and a rolling parameter query constraint statement is constructed in combination with the second distance threshold. The constraint statement can identify historical samples with similar rolling process conditions, ensuring that the selected samples are comparable with the target conditions in terms of process parameters. The obtained film formula query constraint statement and rolling parameter query constraint statement are combined through logical AND operations to form a query constraint statement. The query constraint statement means that the screened samples must meet the requirements of film formula similarity and rolling parameter similarity at the same time, ensuring that the samples are similar to the target conditions in both material formula and process parameters, thereby improving the pertinence and reference value of the sample set.

[0062] By constructing query constraint statements, historical samples that are close to the target conditions can be accurately identified, providing a reliable data basis for subsequent electromagnetic shielding performance analysis. At the same time, the weight configuration method based on correlation makes the query process pay more attention to parameters that have a significant impact on electromagnetic shielding performance, further improving the effectiveness of sample screening.

[0063] Further, when the number of local sample sets is less than or equal to the convergence sample number threshold, the query constraint statement is uploaded to the cloud server to obtain networking feedback information, wherein the networking feedback information includes the networked sample medium distribution uniformity concentration value and the networked sample medium distribution density concentration value, including:

[0064] S131: When the cloud server receives the query constraint statement, the cloud server sends the query task and the key to the distributed node to obtain the distributed encrypted feedback information, wherein any piece of the distributed encrypted feedback information includes: the sample film formula, the sample rolling parameters, the sample medium record distribution density and the sample medium record distribution uniformity parameter;

[0065] S132: selecting, through the cloud server, a set of sample medium record distribution uniformity parameters and a set of sample medium record distribution density that meet the query constraint statement from the distributed encrypted feedback information based on the sample film formula and the sample rolling parameters;

[0066] S133: Through the cloud server, the sample medium record distribution uniformity parameter set and the sample medium record distribution density set are respectively evaluated for concentration values ​​to obtain the networked sample medium distribution uniformity concentration value and the networked sample medium distribution density concentration value.

[0067] In an optional embodiment, when the cloud server receives the query constraint statement, the query task and key are sent to the distributed node through the cloud server. Among them, the cloud server, as a central coordination unit, distributes the query task and the corresponding key to multiple distributed nodes in the network; these distributed nodes can be data centers of different production bases, research institutions or cooperative enterprises, storing their own electromagnetic shielding carbon fiber prepreg preparation and test data. In order to protect data security, the distributed returned information is in encrypted form, that is, distributed encrypted feedback information. Each encrypted feedback information contains four key data items: sample film formula (including the proportion of components such as resin, curing agent, and accelerator), sample rolling parameters (including process parameters such as temperature, pressure, and speed), sample medium record distribution density (reflecting the distribution concentration of conductive agents or magnetic agents) and sample medium record distribution uniformity parameters (reflecting the uniformity of the distribution of conductive agents or magnetic agents). This encrypted data transmission method containing complete information not only ensures the security of data during transmission, but also provides sufficient reference for subsequent screening.

[0068] Then, the data is screened and sorted on the cloud server. The cloud server first decrypts the received distributed encrypted feedback information, and then evaluates the sample film formula and sample rolling parameters according to the conditions set in the query constraint statement (including the first distance function and the second distance function). For samples that meet the query constraint conditions, their medium record distribution uniformity parameters and medium record distribution density are extracted to form a sample medium record distribution uniformity parameter set and a sample medium record distribution density set, respectively. This screening process ensures that the data used in subsequent analysis has a high similarity with the target preparation conditions, and improves the pertinence and reliability of the analysis results. Then, the screened medium record distribution uniformity parameter set and sample medium record distribution density set are statistically analyzed to calculate representative central tendency values. For example, the mean, median, weighted average or other statistical methods are used to evaluate the centralization value of the sample medium record distribution uniformity parameter set and the sample medium record distribution density set, respectively, to obtain the networked sample medium distribution uniformity centralization value and the networked sample medium distribution density centralization value. Through the centralization value evaluation, the influence of the random fluctuation of a single sample on the analysis results is effectively reduced, and more generally representative feature values ​​are extracted. At the same time, data processing is completed in the cloud, and only the processing results rather than the original data are returned, which not only improves computing efficiency, but also further protects the sensitive process data of each node from being directly obtained, thereby improving data security.

[0069] Through cloud-based encryption processing, the concentrated value of the uniformity of the medium distribution of networked samples and the concentrated value of the medium distribution density of networked samples are obtained. Under the premise of ensuring data security, the rich sample resources in the distributed network environment are fully utilized to obtain more statistically significant reference data, providing more reliable data support for the performance evaluation of electromagnetic shielding carbon fiber prepreg.

[0070] Furthermore, through the electromagnetic shielding effectiveness simulator bound to the electromagnetic shielding carbon fiber prepreg formula, the medium distribution uniformity concentration value, the medium distribution density concentration value, the first interlayer stability, the second interlayer stability and the test electromagnetic wave frequency band are processed to generate the electromagnetic shielding effectiveness prediction value, including:

[0071] S410: Based on at least two layers of fully connected neural networks, a medium distribution feature extraction channel is constructed; based on at least three layers of fully connected neural networks, an interface stability feature extraction channel is constructed; the medium distribution uniformity concentration value, the medium distribution density concentration value, the first inter-layer stability, the second inter-layer stability, and the test frequency band are put into the input layer; the electromagnetic shielding effectiveness identification value is put into the output layer, and the electromagnetic shielding effectiveness simulator architecture is obtained;

[0072] S420: Collect historical preparation data of electromagnetic shielding carbon fiber prepreg formula;

[0073] S430: Collecting a concentrated interval of electromagnetic shielding effectiveness historical test values ​​that meets historical preparation data and setting it as an electromagnetic shielding effectiveness historical identification interval;

[0074] S440: Processing the historical preparation data to obtain a historical record value of medium distribution uniformity, a historical record value of medium distribution density, a historical record value of first interlayer stability, a historical record value of second interlayer stability, and a historical test frequency band;

[0075] S450: According to the historical record value of medium distribution uniformity, the historical record value of medium distribution density, the historical record value of first layer stability, the historical record value of second layer stability, the historical test frequency band, and the electromagnetic shielding effectiveness historical identification interval, the electromagnetic shielding effectiveness simulator architecture is trained to generate an electromagnetic shielding effectiveness simulator.

[0076] Specifically, an electromagnetic shielding effectiveness simulator architecture is constructed. The electromagnetic shielding effectiveness simulator architecture adopts deep learning technology and includes two parallel feature extraction channels, namely, a medium distribution feature extraction channel and an interface stability feature extraction channel. Among them, the medium distribution feature extraction channel is based on at least two layers of fully connected neural networks, which is specifically used to process parameters related to the distribution of conductive agents or magnetic conductive agents; the interface stability feature extraction channel is based on at least three layers of fully connected neural networks, which is used to process parameters related to the interlayer bonding strength. A deeper network hierarchy is used to process the interface stability feature because the interlayer interface behavior usually has more complex nonlinear characteristics and requires stronger network expression capabilities for modeling. The input layer of the electromagnetic shielding effectiveness simulator receives five parameters, namely, the medium distribution uniformity concentration value, the medium distribution density concentration value, the first interlayer stability, the second interlayer stability, and the test electromagnetic wave frequency band. Among them, the medium distribution uniformity concentration value and the medium distribution density concentration value are processed by the medium distribution feature extraction channel, the first interlayer stability and the second interlayer stability are processed by the interface stability feature extraction channel, and the test frequency band parameters are input into both channels at the same time. The output layer is designed as an electromagnetic shielding effectiveness identification value, which can be a single value or an interval range, indicating the expected electromagnetic shielding performance under specific conditions. This dual-channel parallel structure design can more specifically extract and process the impact of different types of features on electromagnetic shielding effectiveness.

[0077] Then, the historical preparation data of the electromagnetic shielding carbon fiber prepreg formula is collected, including but not limited to: film formula (such as resin type, curing agent, accelerator and conductive / magnetic medium component ratio), rolling parameters (such as rolling temperature, pressure, hot pressing composite parameters, etc.) and performance test data of intermediate products and final products. At the same time, for the collected historical preparation data, the corresponding electromagnetic shielding effectiveness test results are collected and their concentrated distribution range is determined. These test values ​​reflect the electromagnetic wave attenuation ability of the material in different frequency bands. By statistically analyzing the distribution characteristics of these historical test values, a representative electromagnetic shielding effectiveness historical test value concentration interval is determined, which is set as the electromagnetic shielding effectiveness historical identification interval as the target output of simulator training. This interval setting takes into account the concentration trend of the test data while retaining a reasonable fluctuation range, which can more objectively reflect the performance distribution characteristics in actual production.

[0078] Next, the collected historical preparation data is preprocessed and feature extracted to convert the raw data into standardized input features acceptable to the simulator. Specifically, the historical record values ​​of dielectric distribution uniformity, dielectric distribution density, first interlayer stability, second interlayer stability, and historical test frequency bands are extracted from the historical preparation data. Among them, the historical record values ​​of dielectric distribution uniformity reflect the uniformity of the distribution of the conductive / magnetic medium in the film, the historical record values ​​of dielectric distribution density reflect the density distribution of the conductive / magnetic medium in the film, the historical record values ​​of first interlayer stability reflect the interlayer bonding strength of the carbon fiber prepreg, and the historical record values ​​of second interlayer stability reflect the interlayer bonding strength of the shielding net and the carbon fiber prepreg; the historical test frequency band records the electromagnetic wave frequency range used in the electromagnetic shielding performance test. Subsequently, the historical record values ​​of dielectric distribution uniformity, dielectric distribution density, first interlayer stability, second interlayer stability, historical test frequency bands, and electromagnetic shielding effectiveness historical identification intervals are prepared as training data to train the constructed electromagnetic shielding effectiveness simulator architecture. The training process adopts the supervised learning method, takes the historical records of dielectric distribution uniformity, dielectric distribution density, first inter-layer stability, second inter-layer stability and historical test frequency band as input, and the determined electromagnetic shielding effectiveness historical identification interval as the target output, and optimizes the weight parameters of the neural network through the back propagation algorithm. Appropriate learning rate adjustment strategy, regularization method and early stopping mechanism are used in the training process to avoid overfitting problems and improve the generalization ability of the model. At the same time, the cross-validation method can be used to evaluate the performance of the model to ensure that the trained simulator also has good prediction ability on unseen data. After the training is completed, the optimized network parameters are saved to form an electromagnetic shielding effectiveness simulator that can be directly applied.

[0079] By building and training an electromagnetic shielding effectiveness simulator, the electromagnetic shielding performance of electromagnetic shielding carbon fiber prepreg under specific conditions can be accurately predicted based on material properties and process parameter characteristics. Compared with traditional empirical formulas or simple regression models, this performance prediction method based on deep learning can more effectively capture the complex nonlinear relationship between parameters, provide more accurate prediction results, and provide support for electromagnetic shielding performance testing and evaluation.

[0080] Furthermore, according to the electromagnetic shielding effectiveness prediction value, the electromagnetic shielding performance test value is distributed with a deviation weight to obtain a deviation weight set, and the electromagnetic shielding performance test value is weighted averaged by combining the deviation weight set to obtain an electromagnetic shielding effectiveness correction value, including:

[0081] S510: The electromagnetic shielding effectiveness prediction value is the electromagnetic shielding effectiveness prediction interval;

[0082] S520: When the electromagnetic shielding performance test value is distributed within the electromagnetic shielding effectiveness prediction interval, a credible electromagnetic shielding performance test value is added;

[0083] S530: When the electromagnetic shielding performance test value does not belong to the electromagnetic shielding effectiveness prediction interval, add the electromagnetic shielding performance test value to be weighted;

[0084] S540: performing deviation weight distribution according to the boundary distance between the electromagnetic shielding performance test value to be weighted and the electromagnetic shielding effectiveness prediction interval to obtain a deviation weight set;

[0085] S550: performing weighted mean calculation on the weighted electromagnetic shielding performance test value to be weighted according to the deviation weight set to obtain a fitting value of the weighted electromagnetic shielding performance test value to be weighted;

[0086] S560: Calculate the average of the fitting value of the electromagnetic shielding performance test value to be weighted and the credible electromagnetic shielding performance test value to obtain the electromagnetic shielding effectiveness correction value.

[0087] Specifically, the electromagnetic shielding effectiveness prediction value is the electromagnetic shielding effectiveness prediction interval. For example, a decibel range from a minimum value to a maximum value, where the minimum value represents the lower limit of the interval and the maximum value represents the upper limit of the interval. The interval representation can reflect the inherent fluctuation range of electromagnetic shielding performance caused by factors such as the randomness of the material microstructure and fluctuations in the test environment.

[0088] After obtaining the electromagnetic shielding effectiveness prediction value, the actual measured electromagnetic shielding performance test values ​​are preliminarily screened. When the electromagnetic shielding performance test value falls within the electromagnetic shielding effectiveness prediction interval, that is, the conditions that the lower limit of the prediction interval is less than or equal to the test value and the test value is less than or equal to the upper limit of the prediction interval are met, indicating that the test result is consistent with the theoretical expectation and has a high degree of credibility, so it is classified as a credible electromagnetic shielding performance test value. These test values ​​within the prediction interval are considered to be less affected by random factors and systematic errors, and can more accurately reflect the actual electromagnetic shielding performance of the material. When the electromagnetic shielding performance test value does not fall within the electromagnetic shielding effectiveness prediction interval, that is, the conditions that the lower limit of the prediction interval is less than or equal to the test value and the test value is less than or equal to the upper limit of the prediction interval are not met, indicating that these test values ​​deviate from theoretical expectations and may be affected by test environment interference, improper operation or other abnormal factors, resulting in deviations in the test results. However, these test values ​​still contain valuable information and should not be discarded directly, but should be classified as electromagnetic shielding performance test values ​​to be weighted to prepare for subsequent reliability evaluation and correction processing.

[0089] Then, for the electromagnetic shielding performance test values ​​to be weighted, a weight allocation mechanism based on boundary distance is established. For each electromagnetic shielding performance test value to be weighted, first calculate its distance from the boundary of the electromagnetic shielding effectiveness prediction interval. If the electromagnetic shielding performance test value to be weighted is less than the lower limit of the prediction interval, calculate the distance from the test value to the lower limit; if the electromagnetic shielding performance test value to be weighted is greater than the upper limit of the prediction interval, calculate the distance from the test value to the upper limit. Next, calculate the sum of the boundary distances of all electromagnetic shielding performance test values ​​to be weighted, and divide the boundary distance of each electromagnetic shielding performance test value to be weighted by the sum to obtain the deviation weight corresponding to the test value. The deviation weights of all test values ​​constitute a deviation weight set. Then, using the obtained deviation weight set, calculate the weighted mean of all electromagnetic shielding performance test values ​​to be weighted. The specific method is to multiply each electromagnetic shielding performance test value to be weighted by its corresponding deviation weight, then add all the products, and finally divide by the sum of the deviation weights to obtain the fitting value of the electromagnetic shielding performance test value to be weighted. This weighted processing method can make reasonable use of all test data, while reducing the impact of test values ​​that deviate far from the prediction interval on the final result, making the fitting value more representative and reliable. Afterwards, the obtained fitting value of the electromagnetic shielding performance test value to be weighted is fused with the collected credible electromagnetic shielding performance test value. First, the arithmetic mean of all credible electromagnetic shielding performance test values ​​is calculated, and then this average is added to the fitting value of the electromagnetic shielding performance test value to be weighted and divided by two to obtain the final electromagnetic shielding effectiveness correction value. This fusion processing method balances the test data inside and outside the prediction interval, retains the benchmark role of high-credibility test values, and absorbs the effective information that may be contained in the deviation value, so that the final electromagnetic shielding effectiveness correction value more comprehensively and accurately reflects the actual shielding performance of the material.

[0090] By calculating the electromagnetic shielding effectiveness correction value, the problem of electromagnetic shielding performance testing being affected by environmental interference and hardware accuracy deviation is effectively solved, the reliability and stability of the test results are improved, and support is provided for the performance evaluation of electromagnetic shielding carbon fiber prepreg in high-demand application scenarios.

[0091] Furthermore, the deviation weight distribution is performed according to the boundary distance between the electromagnetic shielding performance test value to be weighted and the electromagnetic shielding effectiveness prediction interval to obtain a deviation weight set, including:

[0092] S541: When the boundary distance between the electromagnetic shielding performance test value to be weighted and the electromagnetic shielding effectiveness prediction interval is greater than or equal to the boundary distance threshold, the electromagnetic shielding performance test value to be weighted is deleted.

[0093] In a feasible implementation, in the process of obtaining a deviation weight set by distributing deviation weights according to the boundary distance between the weighted electromagnetic shielding performance test value and the electromagnetic shielding effectiveness prediction interval, an outlier screening mechanism is introduced. Specifically, before calculating the deviation weight distribution, the weighted electromagnetic shielding performance test value is first pre-screened to eliminate obviously abnormal test data. For each weighted electromagnetic shielding performance test value, calculate its distance to the nearest boundary of the electromagnetic shielding effectiveness prediction interval. If the test value is less than the lower limit of the prediction interval, the distance is the lower limit of the prediction interval minus the test value; if the test value is greater than the upper limit of the prediction interval, the distance is the test value minus the upper limit of the prediction interval.

[0094] Then, the calculated boundary distance is compared with the pre-set boundary distance threshold. The threshold can be determined based on the statistical analysis results of historical test data, or it can be dynamically adjusted according to the accuracy requirements of specific application scenarios. When the distance between a certain electromagnetic shielding performance test value to be weighted and the boundary of the prediction interval is greater than or equal to the boundary distance threshold, it indicates that the test value deviates too much from expectations, which may be abnormal data caused by factors such as test instrument failure, strong environmental interference or human operation error. It is not suitable for subsequent analysis, so it is deleted from the set of electromagnetic shielding performance test values ​​to be weighted.

[0095] Through the outlier deletion mechanism based on the distance threshold, extreme test results that are not of reference value can be effectively filtered out, ensuring that subsequent weight allocation and weighted calculation are only based on data within a reasonable deviation range, thereby improving the accuracy and reliability of the electromagnetic shielding effectiveness correction value. At the same time, this mechanism also avoids the distorted influence of outliers on the deviation weight distribution, making the weight allocation more reasonable and better reflecting the actual credibility of each test value, laying the foundation for obtaining accurate and reliable electromagnetic shielding effectiveness correction values ​​in the future.

[0096] Embodiment 2, as Figure 2 As shown, based on the same inventive concept as the method for testing electromagnetic shielding performance of a carbon fiber prepreg provided in Example 1, an embodiment of the present invention further provides a system for testing electromagnetic shielding performance of a carbon fiber prepreg, comprising:

[0097] A concentrated value statistics module 11 is used to count the concentrated value of medium distribution uniformity and the concentrated value of medium distribution density of the first sample set that meets the film formula and rolling parameters;

[0098] The first stability module 12 is used to count the proportion of stable samples of the second sample set that meet the carbon fiber prepreg formula and hot pressing-cooling parameters, and generate the first interlayer stability. The stable samples refer to samples that have not been separated from the preset service time;

[0099] The second stability module 13 is used to count the proportion of stable samples of the third sample set that meet the electromagnetic shielding carbon fiber prepreg formula and hot pressing composite parameters, and generate a second interlayer stability;

[0100] A prediction value generating module 14 is used to process the medium distribution uniformity concentration value, the medium distribution density concentration value, the first interlayer stability, the second interlayer stability and the test electromagnetic wave frequency band through an electromagnetic shielding effectiveness simulator bound to the electromagnetic shielding carbon fiber prepreg formula to generate an electromagnetic shielding effectiveness prediction value;

[0101] The correction value acquisition module 15 is used to perform deviation weight distribution on the electromagnetic shielding performance test value according to the electromagnetic shielding effectiveness prediction value to obtain a deviation weight set, and perform weighted mean calculation on the electromagnetic shielding performance test value in combination with the deviation weight set to obtain an electromagnetic shielding effectiveness correction value.

[0102] Furthermore, the concentrated value statistics module 11 includes the following execution steps:

[0103] Taking the film formula and the rolling parameters as constraints, constructing a query constraint statement;

[0104] Prepare logs from local history and retrieve a local sample set that satisfies the query constraint statement;

[0105] When the number of the local sample set is less than or equal to the converged sample number threshold, uploading the query constraint statement to the cloud server to obtain networking feedback information, wherein the networking feedback information includes a networked sample medium distribution uniformity concentration value and a networked sample medium distribution density concentration value;

[0106] The networked sample medium distribution uniformity concentration value and the networked sample medium distribution density concentration value, together with the local sample set, are added to the first sample set.

[0107] Furthermore, the concentrated value statistics module 11 further includes the following execution steps:

[0108] The film formula includes parts by weight of raw material and parts by weight of electromagnetic shielding medium;

[0109] The rolling parameters include rolling speed and roller gap;

[0110] Performing electromagnetic shielding correlation weight configuration on the weight portions of the raw material material and the weight portions of the electromagnetic shielding medium to obtain a first weight distribution;

[0111] Performing medium distribution correlation weight configuration on the roller pressing speed and the roller gap to obtain a second weight distribution;

[0112] Building a first distance function based on the film formula according to the first weight distribution, and building a film formula query constraint statement in combination with a first distance threshold;

[0113] Building a second distance function based on the rolling parameters according to the second weight distribution, and building a rolling parameter query constraint statement in combination with a second distance threshold;

[0114] The film formula query constraint statement and the rolling parameter query constraint statement are logically ANDed to obtain the query constraint statement.

[0115] Furthermore, the concentrated value statistics module 11 further includes the following execution steps:

[0116] When the cloud server receives the query constraint statement, the cloud server sends the query task and the key to the distributed node to obtain distributed encrypted feedback information, wherein any piece of encrypted feedback information of the distributed encrypted feedback information includes: sample film formula, sample rolling parameters, sample medium record distribution density and sample medium record distribution uniformity parameter;

[0117] By means of the cloud server, based on the sample film formula and the sample rolling parameters, a set of sample medium record distribution uniformity parameters and a set of sample medium record distribution density that satisfy the query constraint statement are selected from the distributed encrypted feedback information;

[0118] Through the cloud server, the sample medium record distribution uniformity parameter set and the sample medium record distribution density set are respectively evaluated for concentration values ​​to obtain the networked sample medium distribution uniformity concentration value and the networked sample medium distribution density concentration value.

[0119] Furthermore, the prediction value generating module 14 includes the following execution steps:

[0120] Based on at least two layers of fully connected neural networks, a medium distribution feature extraction channel is built. Based on at least three layers of fully connected neural networks, an interface stability feature extraction channel is built. The medium distribution uniformity concentration value, the medium distribution density concentration value, the first inter-layer stability, the second inter-layer stability, and the test frequency band are put into the input layer, and the electromagnetic shielding effectiveness identification value is put into the output layer to obtain the electromagnetic shielding effectiveness simulator architecture;

[0121] Collecting historical preparation data of the electromagnetic shielding carbon fiber prepreg formula;

[0122] Collecting a concentrated interval of electromagnetic shielding effectiveness historical test values ​​that meets the historical preparation data and setting it as an electromagnetic shielding effectiveness historical identification interval;

[0123] Processing the historical preparation data to obtain a historical record value of medium distribution uniformity, a historical record value of medium distribution density, a historical record value of first interlayer stability, a historical record value of second interlayer stability, and a historical test frequency band;

[0124] According to the historical record value of the medium distribution uniformity, the historical record value of the medium distribution density, the historical record value of the first interlayer stability, the historical record value of the second interlayer stability, the historical test frequency band, and the electromagnetic shielding effectiveness historical identification interval, the electromagnetic shielding effectiveness simulator architecture is trained to generate the electromagnetic shielding effectiveness simulator.

[0125] Furthermore, the correction value acquisition module 15 includes the following execution steps:

[0126] The electromagnetic shielding effectiveness prediction value is an electromagnetic shielding effectiveness prediction interval;

[0127] When the electromagnetic shielding performance test value is distributed in the electromagnetic shielding effectiveness prediction interval, adding a credible electromagnetic shielding performance test value;

[0128] When the electromagnetic shielding performance test value does not belong to the electromagnetic shielding effectiveness prediction interval, adding the electromagnetic shielding performance test value to be weighted;

[0129] Perform deviation weight distribution according to the boundary distance between the electromagnetic shielding performance test value to be weighted and the electromagnetic shielding effectiveness prediction interval to obtain a deviation weight set;

[0130] Performing weighted mean calculation on the electromagnetic shielding performance test value to be weighted according to the deviation weight set to obtain a fitting value of the electromagnetic shielding performance test value to be weighted;

[0131] The mean of the fitting value of the electromagnetic shielding performance test value to be weighted and the credible electromagnetic shielding performance test value is calculated to obtain the electromagnetic shielding effectiveness correction value.

[0132] Furthermore, the correction value acquisition module 15 also includes the following execution steps:

[0133] When the boundary distance between the electromagnetic shielding performance test value to be weighted and the electromagnetic shielding effectiveness prediction interval is greater than or equal to a boundary distance threshold, the electromagnetic shielding performance test value to be weighted is deleted.

[0134] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and for parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0135] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may 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.

[0136] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0137] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0138] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0139] Although preferred embodiments of the present invention have been described, additional changes and modifications may occur to these embodiments once those skilled in the art understand the basic inventive concepts.

[0140] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention belong to the scope of the present invention and its equivalent technologies, the present invention is also intended to include these changes and variations.

Claims

1. A method for testing the electromagnetic shielding performance of carbon fiber prepreg, characterized in that: include: Counting the concentrated value of the uniformity of medium distribution and the concentrated value of the density of medium distribution of the first sample set that meets the film formula and rolling parameters; The proportion of stable samples in the second sample set that meet the carbon fiber prepreg formula and hot pressing-cooling parameters is counted to generate the first interlayer stability. Stable samples refer to samples that have not deviated from the preset service time. Counting the proportion of stable samples in the third sample set that meet the electromagnetic shielding carbon fiber prepreg formula and hot pressing composite parameters, and generating the second interlayer stability; The electromagnetic shielding effectiveness simulator bound to the electromagnetic shielding carbon fiber prepreg formula processes the medium distribution uniformity concentration value, the medium distribution density concentration value, the first interlayer stability, the second interlayer stability and the test electromagnetic wave frequency band to generate an electromagnetic shielding effectiveness prediction value; According to the electromagnetic shielding effectiveness prediction value, the electromagnetic shielding performance test value is distributed with deviation weights to obtain a deviation weight set, and the electromagnetic shielding performance test value is weighted averaged in combination with the deviation weight set to obtain an electromagnetic shielding effectiveness correction value.

2. The method according to claim 1, characterized in that The first sample set that meets the film formulation and rolling parameters is statistically analyzed, including: Taking the film formula and the rolling parameters as constraints, constructing a query constraint statement; Prepare logs from local history and retrieve a local sample set that satisfies the query constraint statement; When the number of the local sample set is less than or equal to the converged sample number threshold, uploading the query constraint statement to the cloud server to obtain networking feedback information, wherein the networking feedback information includes a networked sample medium distribution uniformity concentration value and a networked sample medium distribution density concentration value; The networked sample medium distribution uniformity concentration value and the networked sample medium distribution density concentration value, together with the local sample set, are added to the first sample set.

3. The method according to claim 2, characterized in that Taking the film formula and the rolling parameters as constraints, a query constraint statement is constructed, including: The film formula includes parts by weight of raw material and parts by weight of electromagnetic shielding medium; The rolling parameters include rolling speed and roller gap; Performing electromagnetic shielding correlation weight configuration on the weight portions of the raw material material and the weight portions of the electromagnetic shielding medium to obtain a first weight distribution; Performing medium distribution correlation weight configuration on the roller speed and the roller gap to obtain a second weight distribution; Building a first distance function based on the film formula according to the first weight distribution, and building a film formula query constraint statement in combination with a first distance threshold; Building a second distance function based on the rolling parameters according to the second weight distribution, and building a rolling parameter query constraint statement in combination with a second distance threshold; The film formula query constraint statement and the rolling parameter query constraint statement are logically ANDed to obtain the query constraint statement.

4. The method according to claim 2, characterized in that When the number of the local sample set is less than or equal to the converged sample number threshold, the query constraint statement is uploaded to the cloud server to obtain networking feedback information, wherein the networking feedback information includes a networked sample medium distribution uniformity concentration value and a networked sample medium distribution density concentration value, including: When the cloud server receives the query constraint statement, the cloud server sends the query task and the key to the distributed node to obtain distributed encrypted feedback information, wherein any piece of encrypted feedback information of the distributed encrypted feedback information includes: sample film formula, sample rolling parameters, sample medium record distribution density and sample medium record distribution uniformity parameter; By means of the cloud server, based on the sample film formula and the sample rolling parameters, a set of sample medium record distribution uniformity parameters and a set of sample medium record distribution density that satisfy the query constraint statement are selected from the distributed encrypted feedback information; Through the cloud server, the sample medium record distribution uniformity parameter set and the sample medium record distribution density set are respectively evaluated for concentration values ​​to obtain the networked sample medium distribution uniformity concentration value and the networked sample medium distribution density concentration value.

5. The method according to claim 1, characterized in that The electromagnetic shielding effectiveness simulator bound to the electromagnetic shielding carbon fiber prepreg formula processes the medium distribution uniformity concentration value, the medium distribution density concentration value, the first interlayer stability, the second interlayer stability and the test electromagnetic wave frequency band to generate an electromagnetic shielding effectiveness prediction value, including: Based on at least two layers of fully connected neural networks, a medium distribution feature extraction channel is built. Based on at least three layers of fully connected neural networks, an interface stability feature extraction channel is built. The medium distribution uniformity concentration value, the medium distribution density concentration value, the first inter-layer stability, the second inter-layer stability, and the test electromagnetic wave frequency band are placed in the input layer, and the electromagnetic shielding effectiveness identification value is placed in the output layer to obtain the electromagnetic shielding effectiveness simulator architecture. Collecting historical preparation data of the electromagnetic shielding carbon fiber prepreg formula; Collecting a concentrated interval of electromagnetic shielding effectiveness historical test values ​​that meets the historical preparation data and setting it as an electromagnetic shielding effectiveness historical identification interval; Processing the historical preparation data to obtain a historical record value of medium distribution uniformity, a historical record value of medium distribution density, a historical record value of first interlayer stability, a historical record value of second interlayer stability, and a historical test electromagnetic wave frequency band; According to the historical record value of the medium distribution uniformity, the historical record value of the medium distribution density, the historical record value of the first interlayer stability, the historical record value of the second interlayer stability, and the historical test electromagnetic wave frequency band, the electromagnetic shielding effectiveness simulator architecture is trained in combination with the electromagnetic shielding effectiveness historical identification interval to generate the electromagnetic shielding effectiveness simulator.

6. The method according to claim 1, characterized in that According to the electromagnetic shielding effectiveness prediction value, performing deviation weight distribution on the electromagnetic shielding performance test value to obtain a deviation weight set, and performing weighted mean calculation on the electromagnetic shielding performance test value in combination with the deviation weight set to obtain an electromagnetic shielding effectiveness correction value, including: The electromagnetic shielding effectiveness prediction value is an electromagnetic shielding effectiveness prediction interval; When the electromagnetic shielding performance test value is distributed in the electromagnetic shielding effectiveness prediction interval, adding a credible electromagnetic shielding performance test value; When the electromagnetic shielding performance test value does not belong to the electromagnetic shielding effectiveness prediction interval, adding the electromagnetic shielding performance test value to be weighted; Perform deviation weight distribution according to the boundary distance between the electromagnetic shielding performance test value to be weighted and the electromagnetic shielding effectiveness prediction interval to obtain a deviation weight set; Performing weighted mean calculation on the electromagnetic shielding performance test value to be weighted according to the deviation weight set to obtain a fitting value of the electromagnetic shielding performance test value to be weighted; The mean of the fitting value of the electromagnetic shielding performance test value to be weighted and the credible electromagnetic shielding performance test value is calculated to obtain the electromagnetic shielding effectiveness correction value.

7. The method according to claim 6, characterized in that The deviation weight distribution is performed according to the boundary distance between the electromagnetic shielding performance test value to be weighted and the electromagnetic shielding effectiveness prediction interval to obtain a deviation weight set, including: When the boundary distance between the electromagnetic shielding performance test value to be weighted and the electromagnetic shielding effectiveness prediction interval is greater than or equal to a boundary distance threshold, the electromagnetic shielding performance test value to be weighted is deleted.

8. A carbon fiber prepreg electromagnetic shielding performance testing system, characterized in that: A method for testing electromagnetic shielding performance of a carbon fiber prepreg according to any one of claims 1 to 7, the system comprising: A concentrated value statistics module, used to count the concentrated value of medium distribution uniformity and the concentrated value of medium distribution density of the first sample set that meets the film formula and rolling parameters; The first stability module is used to count the proportion of stable samples of the second sample set that meet the carbon fiber prepreg formula and hot pressing-cooling parameters to generate the first interlayer stability. The stable samples refer to samples that have not been separated from the preset service time. The second stability module is used to count the proportion of stable samples of the third sample set that meet the electromagnetic shielding carbon fiber prepreg formula and hot pressing composite parameters, and generate the second interlayer stability; A prediction value generation module is used to process the medium distribution uniformity concentration value, the medium distribution density concentration value, the first interlayer stability, the second interlayer stability and the test electromagnetic wave frequency band through an electromagnetic shielding effectiveness simulator bound to the electromagnetic shielding carbon fiber prepreg formula to generate an electromagnetic shielding effectiveness prediction value; The correction value acquisition module is used to perform deviation weight distribution on the electromagnetic shielding performance test value according to the electromagnetic shielding effectiveness prediction value to obtain a deviation weight set, and perform weighted mean calculation on the electromagnetic shielding performance test value in combination with the deviation weight set to obtain an electromagnetic shielding effectiveness correction value.

Images