Carbon paper processing method, device, electronic device and storage medium

Abstract

The present disclosure relates to a carbon paper processing method, device, electronic device, and storage medium. The method may include: obtaining an image to be detected and multiple preset carbon paper index information according to the technical solution provided by the present disclosure; inputting the image to be detected and the multiple preset carbon paper index information into a carbon paper performance prediction model, predicting the performance of the target carbon paper under the multiple preset carbon paper index information, and obtaining first performance prediction information corresponding to each of the target carbon paper under the multiple preset carbon paper index information; selecting target performance prediction information that meets preset conditions from the multiple first performance prediction information; and adjusting the gas diffusion layer of the target carbon paper based on the target performance prediction information. According to the technical solution provided by the present disclosure, the performance prediction of the gas diffusion layer is achieved, and the accuracy of predicting the performance of the gas diffusion layer is improved.

Description

Technical Field

[0001] The present disclosure relates to the field of fuel cells, and in particular to a carbon paper processing method, device, electronic device, and storage medium. Background Art

[0002] With the development of the hydrogen energy industry, hydrogen energy extraction technology and fuel cell manufacturing technology have also gradually developed. Among them, the gas diffusion layer is one of the most important components in the membrane electrode, which greatly affects the performance of the entire fuel cell.

[0003] One existing approach is to use a continuum model to model the gas diffusion layer (GDL) microstructure, treating it as a porous medium with a certain porosity for simulation. This approach is overly simplified, making it difficult to demonstrate the significant impact of microstructural deformation and fracture on GDL performance parameters.

[0004] Another approach considers the microstructure of the GDL, constructing a realistic geometric model of the GDL. Simulating this model using the explicit dynamics finite element method (EDFEM) yields the compressed geometry, and then using a pore-scale model to simulate and calculate the performance parameters of the compressed GDL. This approach involves a complex geometric model and a lengthy geometric processing process. Furthermore, as it involves a multi-scale problem, the computational cost is high and time-consuming. Summary of the Invention

[0005] The present disclosure provides a carbon paper processing method, device, electronic device, and storage medium to at least address the problem of predicting the performance of a gas diffusion layer in related technologies to improve the accuracy of the gas diffusion layer performance. The technical solutions of the present disclosure are as follows:

[0006] According to a first aspect of an embodiment of the present disclosure, a carbon paper processing method is provided, comprising:

[0007] In a possible implementation, an image to be detected and a plurality of preset carbon paper index information are obtained; the image to be detected is obtained by scanning a target carbon paper;

[0008] Inputting the image to be detected and the plurality of preset carbon paper index information into a carbon paper performance prediction model, performing prediction processing on the performance of the target carbon paper under the plurality of preset carbon paper index information, and obtaining first performance prediction information corresponding to each of the plurality of preset carbon paper index information; the first performance prediction information represents the predicted performance of the gas diffusion layer of the target carbon paper under the plurality of preset carbon paper index information;

[0009] Selecting target performance prediction information that meets a preset condition from the plurality of first performance prediction information;

[0010] Based on the target performance prediction information, the gas diffusion layer of the target carbon paper is adjusted.

[0011] According to a second aspect of an embodiment of the present disclosure, there is provided a carbon paper processing device, comprising:

[0012] An acquisition module is used to acquire an image to be detected and a plurality of preset carbon paper index information; the image to be detected is obtained by scanning the target carbon paper;

[0013] a first performance prediction information acquisition module, configured to input the image to be detected and the plurality of preset carbon paper index information into a carbon paper performance prediction model, perform prediction processing on the performance of the target carbon paper under the plurality of preset carbon paper index information, and obtain first performance prediction information corresponding to each of the plurality of preset carbon paper index information; the first performance prediction information representing the predicted performance of the gas diffusion layer of the target carbon paper under the plurality of preset carbon paper index information;

[0014] a target performance prediction information acquisition module, configured to select target performance prediction information that meets a preset condition from a plurality of said first performance prediction information;

[0015] An adjustment module is configured to adjust the gas diffusion layer of the target carbon paper based on the target performance prediction information.

[0016] According to a third aspect of an embodiment of the present disclosure, an electronic device is provided, comprising: a processor; and a memory for storing instructions executable by the processor; wherein the processor is configured to execute the instructions to implement a method as described in any one of the first aspects above.

[0017] According to a fourth aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided. When instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is capable of executing any method described in the first aspect of the embodiment of the present disclosure.

[0018] According to a fifth aspect of an embodiment of the present disclosure, a computer program product is provided, comprising computer instructions, which, when executed by a processor, enable a computer to execute any one of the methods according to the first aspect of the embodiment of the present disclosure.

[0019] The technical solutions provided by the embodiments of the present disclosure bring at least the following beneficial effects:

[0020] Based on the image to be detected and multiple preset carbon paper indicator information, the performance of the gas diffusion layer can be predicted; by selecting target performance prediction information that meets preset conditions from multiple first performance prediction information, the gas diffusion layer of the target carbon paper is adjusted, thereby improving the accuracy of the gas diffusion layer performance.

[0021] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The accompanying drawings herein are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description are used to explain the principles of the present disclosure, and do not constitute an improper limitation of the present disclosure.

[0023] Figure 1 The figure is a flow chart showing a carbon paper processing method according to an exemplary embodiment.

[0024] Figure 2 The flowchart of obtaining first performance prediction information corresponding to a target carbon paper under a plurality of preset carbon paper index information according to an exemplary embodiment is shown.

[0025] Figure 3 is a flow chart showing another carbon paper processing method according to an exemplary embodiment.

[0026] Figure 4 is a flow chart showing another carbon paper processing method according to an exemplary embodiment.

[0027] Figure 5 FIG. 4 is a schematic diagram showing loss information during the training of an initial performance prediction model according to an exemplary embodiment.

[0028] Figure 6 The flowchart of determining loss information according to carbon paper performance prediction information and label information is shown according to an exemplary embodiment.

[0029] Figure 7 The flowchart of the process of predicting the performance of a sample carbon paper under multiple preset carbon paper index information is shown according to an exemplary embodiment.

[0030] Figure 8 The figure is a block diagram of a carbon paper processing device according to an exemplary embodiment.

[0031] Figure 9 is a block diagram showing an electronic device for carbon paper processing according to an exemplary embodiment. DETAILED DESCRIPTION

[0032] In order to enable ordinary persons in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings.

[0033] It should be noted that the terms "first," "second," and the like in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or precedence. It should be understood that the numbers used in this manner are interchangeable where appropriate so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure as detailed in the appended claims.

[0034] Artificial Intelligence (AI) is the theory, methods, techniques, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to achieve optimal results. AI software technologies primarily encompass computer vision, speech processing, natural language processing, and machine learning / deep learning.

[0035] In recent years, with the research and progress of artificial intelligence technology, artificial intelligence technology has been widely used in many fields. The solutions provided in the embodiments of the present disclosure involve technologies such as machine learning / deep learning, which are specifically illustrated by the following embodiments.

[0036] It should be noted that the following diagrams illustrate a possible sequence of steps and are not intended to be strictly followed. Some steps can be performed in parallel without relying on each other.

[0037] Figure 1 This is a flow chart of a carbon paper processing method according to an exemplary embodiment, which can be executed by any terminal capable of running a training method for a carbon paper performance prediction model; specifically, any terminal can be a server terminal or a mobile terminal. Figure 1 As shown, the following steps may be included.

[0038] In step S101, an image to be detected and a plurality of preset carbon paper index information are obtained; the image to be detected is obtained by scanning the target carbon paper.

[0039] In an embodiment of the present disclosure, the image to be detected may represent a structural image of the target carbon paper. In one example, the image to be detected may be obtained by scanning the target carbon paper. For example, the image to be detected may be a scanning electron microscope (SEM) image of the target carbon paper, or a three-dimensional slice image of the target carbon paper, although this disclosure is not limited thereto.

[0040] The plurality of preset carbon paper index information may represent information measuring the size of the target carbon paper. In one example, the plurality of preset carbon paper index information may include thickness information of the target carbon paper and compression information of the target carbon paper. The compression information refers to the compression ratio of the target carbon paper.

[0041] For example, the thickness information may be 0.1 cm, 0.2 cm, 0.3 cm, etc., and the compression information may be 1%, 10%, 20%, etc., which is not limited in the present disclosure.

[0042] In the embodiment of the present disclosure, an image to be detected and a plurality of preset carbon paper index information can be obtained. In one example, the image to be detected can be obtained by scanning the target carbon paper, and a plurality of thickness information can be preset.

[0043] For example, a scanning electron microscope image of the target carbon paper may be obtained by scanning the target carbon paper, and the thickness information may be set to 0.1 cm, 0.2 cm, 0.3 cm, and the like.

[0044] In another example, a target carbon paper can be simulated to obtain an image to be tested, and multiple compression information can be preset. For example, a three-dimensional slice image of the target carbon paper can be obtained by scanning the target carbon paper, and the compression information can be set to 1%, 10%, 20%, etc.

[0045] In step S103, the image to be detected and multiple preset carbon paper index information are input into the carbon paper performance prediction model to obtain the first performance prediction information corresponding to the target carbon paper under the multiple preset carbon paper index information; the first performance prediction information represents the predicted performance of the gas diffusion layer of the target carbon paper under the multiple preset carbon paper index information.

[0046] In an embodiment of the present disclosure, the first performance prediction information can represent the predicted performance of the gas diffusion layer of the target carbon paper under multiple preset carbon paper index information. In one example, the first performance prediction information can include at least one of porosity information, vertical resistivity information, in-plane resistivity information, gas conductivity information, in-plane thermal conductivity information, vertical thermal conductivity information, in-plane gas diffusivity information, and vertical gas diffusivity information, which is not limited in the present disclosure. For example, the in-plane gas diffusivity can be 8.92%, 8.63%, etc., and the in-plane thermal conductivity can be 13.12%, 14.26%, etc.

[0047] In the disclosed embodiments, a carbon paper performance prediction model is inputted with an image to be inspected and multiple preset carbon paper index information. The performance of the target carbon paper under these multiple preset carbon paper index information is predicted, resulting in first performance prediction information corresponding to each of the multiple preset carbon paper index information. In one example, a scanning electron microscope image of the target carbon paper and multiple thickness information are inputted into the carbon paper performance prediction model. The performance of the target carbon paper under these multiple thickness information is predicted, resulting in porosity information, vertical resistivity information, in-plane resistivity information, gas conductivity information, and other corresponding information for the target carbon paper under these multiple thickness information.

[0048] In another example, a three-dimensional slice image and multiple compression information of the target carbon paper are input into a carbon paper performance prediction model, and the performance of the target carbon paper under multiple compression information is predicted and processed, thereby obtaining the in-plane thermal conductivity information, vertical thermal conductivity information, in-plane gas diffusivity information, vertical gas diffusivity information, etc. corresponding to the target carbon paper under multiple compression information.

[0049] For example, Table 1 shows the compression information, actual performance information, predicted performance information, and error information for some images to be tested. As shown in Table 1, when the target carbon paper has a compression of 0%, the in-plane thermal conductivity is 13.12%, the perpendicular thermal conductivity is 1.03%, the in-plane gas diffusivity is 8.92%, and the perpendicular gas diffusivity is 7.39%. When the target carbon paper has a compression of 5%, the in-plane thermal conductivity is 15.01%, the perpendicular thermal conductivity is 2.16%, the in-plane gas diffusivity is 8.63%, and the perpendicular gas diffusivity is 7.2%.

[0050] Table 1

[0051]

[0052]

[0053]

[0054] In step S105 , target performance prediction information that meets preset conditions is selected from the plurality of first performance prediction information.

[0055] In an embodiment of the present disclosure, in one example, the preset conditions may be multiple preset performance information intervals of the target carbon paper, such as a gas diffusion rate of 7.5%-9% and a thermal conductivity greater than 13%, etc., which are not limited in the present disclosure.

[0056] In an embodiment of the present disclosure, target performance prediction information that meets a preset condition can be selected from multiple first performance prediction information. In one example, multiple first performance prediction information can be filtered according to the preset condition, and when the filtering result is only one, the filtering result is used as the target performance prediction information that meets the preset condition.

[0057] In another example, multiple pieces of first performance prediction information may be screened according to preset conditions. When there are multiple screening results, the multiple screening results are compared to obtain target performance prediction information.

[0058] In step S107 , the gas diffusion layer of the target carbon paper is adjusted based on the target performance prediction information.

[0059] In the embodiment of the present disclosure, the gas diffusion layer of the target carbon paper can be adjusted based on the target performance prediction information. In one example, various parameters of the gas diffusion layer of the target carbon paper can be adjusted based on the target performance prediction information.

[0060] For example, the target performance prediction information includes in-plane thermal conductivity information of 15.01%, vertical thermal conductivity information of 2.16%, in-plane gas diffusivity information of 8.63%, vertical gas diffusivity information of 7.2%, and compression information of 5%. The target carbon paper is compressed by 5%, and the composition of the gas diffusion layer of the target carbon paper is adjusted.

[0061] In the embodiment of the present disclosure, the performance prediction of the gas diffusion layer can be achieved based on the image to be detected and multiple preset carbon paper indicator information; by selecting target performance prediction information that meets preset conditions from multiple first performance prediction information, the gas diffusion layer of the target carbon paper is adjusted, thereby improving the accuracy of the gas diffusion layer performance.

[0062] Figure 2 FIG. 1 is a flow chart showing a method for obtaining first performance prediction information corresponding to a target carbon paper under a plurality of preset carbon paper index information according to an exemplary embodiment. Figure 2 As shown, the following steps may be included.

[0063] In step S201, the image to be detected and a plurality of preset carbon paper index information are input into an image segmentation module, and image segmentation processing is performed to obtain a plurality of segmented images.

[0064] In an embodiment of the present disclosure, an image to be detected and multiple preset carbon paper index information can be input into an image segmentation module for image segmentation processing to obtain multiple segmented images. As an example, the image to be detected and multiple preset carbon paper index information can be input into the image segmentation module, and a preset number of evenly distributed points in the image to be detected can be used to segment the image to obtain multiple segmented images. The preset number can be 9, 4, etc., and this disclosure does not impose any restrictions on this; any reasonable number is sufficient.

[0065] For example, the image to be detected and a plurality of preset carbon paper index information can be input into the image segmentation module, and nine evenly distributed points of the image to be detected are taken to perform image segmentation processing to obtain nine evenly segmented images.

[0066] As another example, the image to be detected and a plurality of preset carbon paper index information are input into the image segmentation module, and the image to be detected can be segmented according to the area of ​​the image to be detected to obtain a plurality of segmented images.

[0067] In step S203, the multiple segmented images are input into the performance prediction module for performance prediction processing to obtain second performance prediction information; the second performance prediction information represents the performance prediction information corresponding to the multiple segmented images under multiple preset carbon paper index information.

[0068] In an embodiment of the present disclosure, the second performance prediction information can represent the performance prediction information corresponding to each of the multiple segmented images under multiple preset carbon paper index information. In one example, the second performance prediction information may include porosity information, vertical resistivity information, in-plane resistivity information, gas conductivity information, in-plane thermal conductivity information, vertical thermal conductivity information, in-plane gas diffusivity information, vertical gas diffusivity information, etc., which is not limited in the present disclosure. For example, the in-plane gas diffusivity can be 8.92%, 8.63%, etc., and the in-plane thermal conductivity can be 13.1%, 14.26%, etc.

[0069] In the disclosed embodiment, multiple segmented images can be input into a performance prediction module for performance prediction processing to obtain second performance prediction information. As an example, multiple segmented images can be input into the performance prediction module for performance prediction processing to obtain performance prediction information corresponding to each segmented image under multiple preset carbon paper index information, and the performance prediction information is used as the second performance prediction information.

[0070] For example, the image to be detected is divided into 9 segmented images, and the 9 segmented images are input into the performance prediction module. Performance prediction processing is performed on the 9 segmented images respectively to obtain the in-plane thermal conductivity information, vertical thermal conductivity information, in-plane gas diffusivity information, and vertical gas diffusivity information corresponding to each of the 9 segmented images under multiple preset carbon paper index information. The in-plane thermal conductivity information, vertical thermal conductivity information, in-plane gas diffusivity information, and vertical gas diffusivity information are used as the second performance prediction information.

[0071] In step S205, the second performance prediction information is input into the prediction processing module, and prediction information averaging processing is performed to obtain the first performance prediction information.

[0072] In the disclosed embodiment, the second performance prediction information can be input into the prediction processing module, and prediction information averaging processing can be performed on the second performance prediction information to obtain the first performance prediction information. As an example, the second performance prediction information can be input into the prediction processing module, and prediction information averaging processing can be performed on the second performance prediction information, and the plurality of performance prediction information can be used as the first performance prediction information.

[0073] For example, the image to be detected is divided into 9 segmented images, and the predicted information averaging processing is performed on the in-plane thermal conductivity information, vertical thermal conductivity information, in-plane gas diffusivity information, and vertical gas diffusivity information corresponding to each of the 9 segmented images under multiple preset carbon paper index information. The average processing result of the in-plane thermal conductivity information corresponding to each of the 9 segmented images under multiple preset carbon paper index information is used as the in-plane thermal conductivity information of the image to be detected; the average processing result of the vertical thermal conductivity information corresponding to each of the 9 segmented images under multiple preset carbon paper index information is used as the vertical thermal conductivity information of the image to be detected.

[0074] In the embodiment of the present disclosure, by performing segmentation and re-prediction on the band prediction image, the noise and unevenness of the image can be reduced, thereby improving the accuracy of the first performance prediction information.

[0075] Figure 3 FIG. 1 is a flow chart showing another method for processing carbon paper according to an exemplary embodiment. Figure 3 As shown, the following steps may be included.

[0076] In step S301, actual performance information corresponding to a plurality of preset carbon paper index information is obtained.

[0077] In the embodiment of the present disclosure, actual performance information corresponding to a plurality of preset carbon paper index information can be obtained. As an example, experiments can be conducted on the preset carbon paper to obtain actual performance information corresponding to the plurality of preset carbon paper index information.

[0078] As another example, actual performance information corresponding to multiple preset carbon paper index information can be calculated through pore scale model simulation.

[0079] In step S303, error information is determined based on the first performance prediction information and the actual performance information; the error information is the ratio of the error information between the prediction information and the actual performance information of the plurality of performance indicators to the actual performance information.

[0080] In the disclosed embodiment, the error information may represent the accuracy of the first performance prediction information. As an example, the error information may be the error information ratio between the prediction information and the actual performance information of the plurality of performance indicators. For example, the error information may be 3.12%, 4.27%, etc.

[0081] In the disclosed embodiments, error information may be determined based on the first performance prediction information and the actual performance information. For example, the mean absolute error between the first performance prediction information and the actual performance information may be calculated. Furthermore, the error information may be determined based on the ratio of the mean absolute error to the actual performance information.

[0082] For example, as shown in Table 1, the predicted information for the in-plane gas diffusivity is 8.92%, the actual information for the in-plane gas diffusivity is 8.6501%, and the calculated mean absolute error is 0.2699%. The ratio of the calculated mean absolute error to the actual information is 3.12%, which is used as the error information for the in-plane gas diffusivity.

[0083] In step S305, the first performance prediction information is evaluated based on the error information to obtain a prediction evaluation result.

[0084] In the disclosed embodiment, the first performance prediction information can be evaluated based on the error information to obtain a prediction evaluation result. As an example, the error information can be classified into levels based on the error information. The first performance prediction information is evaluated based on the level corresponding to the error information to obtain a prediction evaluation result.

[0085] For example, error information less than 5% can be classified as level 1, error information between 5% and 15% can be classified as level 2, and error information greater than 15% can be classified as level 3. If the error information is 3.97%, the error information level is level 1, and the predicted evaluation result for thermal conductivity corresponding to this error information is excellent.

[0086] In step S307 , target performance prediction information that meets preset conditions is selected from the plurality of first performance prediction information according to the prediction evaluation result.

[0087] In the disclosed embodiments, target performance prediction information that satisfies a preset condition can be selected from multiple first performance prediction information based on the prediction evaluation results. For example, multiple first performance prediction information can be filtered based on the preset condition. If multiple screening results are obtained, the multiple screening results can be further filtered based on the prediction evaluation results to obtain the target performance prediction information.

[0088] For example, multiple first performance prediction information pieces are screened according to preset conditions to obtain first performance prediction information pieces with compression rates of 5% and 10%. Based on error information of the first performance prediction information pieces with compression rates of 5% and 10%, the first performance prediction information pieces with compression rates of 5% and 10% are further screened to obtain target performance prediction information.

[0089] In the embodiment of the present disclosure, by evaluating the first performance prediction information and then screening the first performance prediction information to obtain the target performance prediction information, the accuracy of the target performance prediction information can be improved.

[0090] Figure 4 FIG. 1 is a flow chart showing another method for processing carbon paper according to an exemplary embodiment. Figure 4 As shown, the following steps may be included.

[0091] In step S401, a plurality of sample images, a plurality of preset carbon paper index information and label information corresponding to each sample image are obtained. The sample image is obtained by scanning the sample carbon paper, and the label information represents the carbon paper performance information marked on the sample carbon paper.

[0092] In the disclosed embodiment, the sample image can be used as a training set for training the initial performance prediction model. As an example, the sample image can be obtained by directly scanning the sample carbon paper, or by selecting an 11 cm*11 cm area in the center of the sample carbon paper as the sample image.

[0093] The label information may represent the carbon paper performance information of the sample carbon paper. As an example, the label information may be porosity information, vertical resistivity information, or in-plane resistivity information, which is not limited in the present disclosure.

[0094] In the disclosed embodiments, multiple sample images, multiple preset carbon paper index information, and label information corresponding to each sample image can be obtained. As an example, an 11 cm by 11 cm area in the center of the target carbon paper can be selected as a sample image, and the sample image can be measured to obtain the carbon paper index information corresponding to the sample image. The carbon paper index information corresponding to the multiple sample images can then be used as the multiple preset carbon paper index information. Furthermore, label information corresponding to each sample image can also be obtained through experimental measurement.

[0095] In step S403, multiple sample images and multiple preset carbon paper index information are input into the initial performance prediction model, and the performance of the sample carbon paper under the multiple preset carbon paper index information is predicted to obtain the corresponding carbon paper performance prediction information of the sample carbon paper under the multiple preset carbon paper index information.

[0096] In an embodiment of the present disclosure, multiple sample images and multiple preset carbon paper index information are input into an initial performance prediction model, and the performance of the sample carbon paper under the multiple preset carbon paper index information is predicted. This can provide corresponding carbon paper performance prediction information for each of the sample carbon paper under the multiple preset carbon paper index information. As an example, multiple sample images and multiple preset carbon paper index information are input into the initial performance prediction model, and a correspondence between the multiple sample images and the multiple preset carbon paper index information is established. Based on this correspondence, corresponding carbon paper performance prediction information for each of the sample carbon paper under the multiple preset carbon paper index information can be obtained.

[0097] In step S405, loss information is determined based on the carbon paper performance prediction information and the label information.

[0098] In the disclosed embodiments, loss information can be determined based on the carbon paper performance prediction information and label information. For example, mean absolute error (MAE) processing can be performed based on the carbon paper performance prediction information and label information to obtain a MAE result. This MAE result is used as the loss information.

[0099] As another example, the weight of the carbon paper performance prediction information and the weight of the label information may be calculated, and then weighted processing may be performed based on the weight of the carbon paper performance prediction information and the weight of the label information to obtain loss information.

[0100] In step S407, the initial performance prediction model is trained based on the loss information until a training iteration condition is satisfied, and the initial performance prediction model corresponding to when the training iteration condition is satisfied is used as the carbon paper performance prediction model.

[0101] In the disclosed embodiments, the initial performance prediction model can be trained based on the loss information until a training iteration condition is satisfied. The initial performance prediction model corresponding to the time when the training iteration condition is satisfied is used as the carbon paper performance prediction model. As an example, the initial performance prediction model can be trained based on the loss information, and the initial performance prediction model corresponding to the time when the training iteration condition is satisfied is used as the carbon paper performance prediction model.

[0102] For example, Figure 5 FIG. 1 is a schematic diagram showing loss information during the initial performance prediction model training process according to an exemplary embodiment. Figure 5As shown in the figure, the horizontal axis represents the number of sample images, and the vertical axis represents the loss information. When the loss information image converges, the iteration condition is met and the initial performance prediction model training is completed. This initial performance prediction model is used as the carbon paper performance prediction model.

[0103] In an embodiment of the present disclosure, the initial performance prediction model is trained by taking multiple sample images and multiple preset carbon paper index information as inputs to predict the carbon paper performance prediction information corresponding to each sample carbon paper under multiple preset carbon paper index information. This can improve the accuracy of the initial performance prediction model and thus improve the accuracy of the carbon paper performance prediction information.

[0104] Figure 6 FIG. 1 is a flow chart showing a method for determining loss information based on carbon paper performance prediction information and label information according to an exemplary embodiment. Figure 6 As shown, the following steps may be included.

[0105] In step S601 , difference information between the carbon paper performance prediction information and the label information is calculated.

[0106] In the disclosed embodiments, the difference between the predicted carbon paper performance information and the label information can be calculated. For example, the difference between the predicted carbon paper performance information and the label information can be calculated and used as the difference information. For example, the absolute difference between the predicted porosity and the labeled porosity can be calculated and used as the difference information.

[0107] In step S603, the average absolute value of the difference information is used as loss information.

[0108] In the embodiment of the present disclosure, the average absolute information is the average of the absolute values ​​of the difference information.

[0109] In the disclosed embodiment, the average absolute value of the difference information can be used as the loss information. As an example, the absolute values ​​of the difference information of multiple sample images can be calculated and the average of the multiple absolute values ​​can be calculated to obtain the average absolute value. The average absolute value is used as the loss information.

[0110] In the embodiment of the present disclosure, the loss information is obtained by calculating the average absolute information of the difference information between the carbon paper performance prediction information and the label information, which can improve the accuracy of the loss information.

[0111] Figure 7 FIG. 1 is a flowchart showing a method for predicting the performance of a sample carbon paper under multiple preset carbon paper index information according to an exemplary embodiment. Figure 7 As shown, the following steps may be included.

[0112] In step S701, a plurality of sample images and a plurality of preset carbon paper index information are input into an initial performance prediction model, and features are extracted from the plurality of sample images to obtain attribute features.

[0113] In the disclosed embodiment, multiple sample images and multiple preset carbon paper index information are input into the initial performance prediction model, and features can be extracted from the multiple sample images to obtain attribute features. As an example, multiple sample images and multiple preset carbon paper index information are input into the initial performance prediction model, and features can be extracted from the multiple sample images using sliding window segmentation and a 3D-UNet model to obtain attribute features.

[0114] For example, a 3D slice image is processed through a sliding window to obtain multiple block images. The multiple block images are then passed through a 3D-UNet model to generate multiple compressed images of the block images. The multiple compressed images are then spliced ​​into a compressed image of the 3D slice image, thereby obtaining the attribute features of the 3D slice image.

[0115] As another example, multiple sample images and multiple preset carbon paper index information are input into the initial performance prediction model. The Densenet model can be used to extract features from the multiple sample images to obtain attribute features.

[0116] For example, the scanned sample image is input into the Densenet model to extract fibers, pores, binders, etc. as attribute features.

[0117] In step S703, features are extracted from a plurality of preset carbon paper index information to obtain vector features.

[0118] In the embodiment of the present disclosure, the vector feature may represent a feature obtained by performing a vectorization operation on a plurality of preset carbon paper index information.

[0119] In the disclosed embodiments, features can be extracted from multiple preset carbon paper indicator information to obtain vector features. As an example, features can be extracted from multiple preset carbon paper indicator information using a DeepAE model to obtain vector features. Specifically, after DeepAE model training is completed, the weights in DeepAE can be set unchanged, and features can be extracted from multiple preset carbon paper indicator information.

[0120] In step S705, feature fusion is performed based on the attribute features and the vector features to obtain fused features.

[0121] In the embodiment of the present disclosure, feature fusion can be performed based on the attribute features and the vector features to obtain a fused feature. As an example, the attribute features and the vector features can be connected to obtain a fused feature.

[0122] As another example, the attribute feature and the vector feature may be combined into a composite feature, and the composite feature is used as the fused feature.

[0123] In step S707, the performance of the sample carbon paper under multiple preset carbon paper index information is predicted based on the fusion features.

[0124] In an embodiment of the present disclosure, based on the fusion features, the performance of a sample carbon paper under multiple preset carbon paper index information can be predicted. As an example, based on the fusion features, the performance of a sample carbon paper under multiple preset carbon paper index information can be predicted using a fusion feature extraction model. The fusion feature extraction model can be a 1D-RESNET model, a nonlocal model, a VIT model, etc., which is not limited in this disclosure.

[0125] In the embodiment of the present disclosure, by fusing attribute features and vector features, the performance of the sample carbon paper under multiple preset carbon paper index information is predicted, which can integrate the characteristics between attribute features and vector features, thereby improving the performance of the initial performance prediction model.

[0126] Figure 8 FIG. 1 is a block diagram of a carbon paper processing device according to an exemplary embodiment. Figure 8 , the apparatus may include:

[0127] The acquisition module 801 is used to acquire an image to be detected and a plurality of preset carbon paper index information; the image to be detected is obtained by scanning the target carbon paper;

[0128] The first performance prediction information acquisition module 803 is configured to input the image to be detected and multiple preset carbon paper index information into the carbon paper performance prediction model to obtain first performance prediction information corresponding to each of the multiple preset carbon paper index information. The first performance prediction information represents the predicted performance of the gas diffusion layer of the target carbon paper under the multiple preset carbon paper index information.

[0129] The target performance prediction information acquisition module 805 is configured to select target performance prediction information that meets a preset condition from a plurality of first performance prediction information;

[0130] The adjustment module 807 is configured to adjust the gas diffusion layer of the target carbon paper based on the target performance prediction information.

[0131] Based on the image to be detected and multiple preset carbon paper indicator information, the performance of the gas diffusion layer can be predicted; by selecting target performance prediction information that meets preset conditions from multiple first performance prediction information, the gas diffusion layer of the target carbon paper is adjusted, thereby improving the accuracy of the gas diffusion layer performance.

[0132] In one possible implementation, the first performance prediction information acquisition module 803 may include:

[0133] An image segmentation unit is used to input the image to be detected and a plurality of preset carbon paper index information into the image segmentation module, perform image segmentation processing, and obtain a plurality of segmented images;

[0134] A second performance prediction information acquisition unit is configured to input the plurality of segmented images into a performance prediction module, perform performance prediction processing, and obtain second performance prediction information; the second performance prediction information represents the performance prediction information corresponding to each of the plurality of segmented images under a plurality of preset carbon paper index information;

[0135] The first performance prediction information is used to input the second performance prediction information into the prediction processing module, perform prediction information averaging processing, and obtain the first performance prediction information.

[0136] In one possible implementation, the above device may include:

[0137] The actual performance information acquisition module is used to obtain actual performance information corresponding to a plurality of preset carbon paper index information;

[0138] An error information acquisition module, configured to determine error information based on the first performance prediction information and the actual performance information; the error information being a difference information ratio between the prediction information and the actual performance information of the plurality of performance indicators;

[0139] A prediction evaluation result module, configured to evaluate the first performance prediction information based on the error information to obtain a prediction evaluation result;

[0140] The target performance prediction information acquisition module is used to select target performance prediction information that meets preset conditions from multiple first performance prediction information based on the prediction evaluation result.

[0141] In one possible implementation, the above device may include:

[0142] A sample information acquisition module is used to acquire multiple sample images, multiple preset carbon paper index information, and label information corresponding to each sample image. The sample image is obtained by scanning the sample carbon paper, and the label information represents the carbon paper performance information marked on the sample carbon paper.

[0143] a prediction processing module, configured to input a plurality of sample images and a plurality of preset carbon paper index information into an initial performance prediction model, perform prediction processing on the performance of the sample carbon paper under the plurality of preset carbon paper index information, and obtain corresponding carbon paper performance prediction information of the sample carbon paper under the plurality of preset carbon paper index information;

[0144] A loss information acquisition module is used to determine loss information based on carbon paper performance prediction information and label information;

[0145] The carbon paper performance prediction model acquisition module trains the initial performance prediction model based on the loss information until the training iteration conditions are met, and uses the initial performance prediction model corresponding to the training iteration conditions as the carbon paper performance prediction model.

[0146] In one possible implementation, the loss information acquisition module may include:

[0147] a difference information obtaining unit, used for calculating difference information between carbon paper performance prediction information and label information;

[0148] The loss information acquiring unit is configured to use the average absolute information of the difference information as the loss information.

[0149] In one possible implementation, the prediction processing module may include:

[0150] An attribute feature acquisition unit is used to input a plurality of sample images and a plurality of preset carbon paper index information into an initial performance prediction model, extract features from the plurality of sample images, and obtain attribute features;

[0151] A vector feature acquisition unit is used to extract features from a plurality of preset carbon paper index information to obtain vector features;

[0152] A fusion feature acquisition unit is used to perform feature fusion based on attribute features and vector features to obtain fusion features;

[0153] The prediction processing unit is used to predict the performance of the sample carbon paper under multiple preset carbon paper index information based on the fusion characteristics.

[0154] Regarding the apparatus in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0155] Figure 9 is a block diagram of an electronic device for carbon paper processing according to an exemplary embodiment. The electronic device may be a terminal, and its internal structure diagram may be as shown in FIG. Figure 9As shown. The electronic device includes a processor, a memory, a network interface, a display screen and an input device connected via a system bus. The processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the electronic device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a method for treating carbon paper is implemented. The display screen of the electronic device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the electronic device can be a touch layer covering the display screen, or a button, trackball or touchpad provided on the housing of the electronic device, or an external keyboard, touchpad or mouse.

[0156] Those skilled in the art will understand that Figure 9 The structure shown in the figure is only a block diagram of a part of the structure related to the scheme of the present disclosure, and does not constitute a limitation on the electronic device to which the scheme of the present disclosure is applied. The specific electronic device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0157] In an exemplary embodiment, an electronic device is further provided, comprising: a processor; and a memory for storing instructions executable by the processor; wherein the processor is configured to execute the instructions to implement the carbon paper processing method as in the embodiment of the present disclosure.

[0158] In an exemplary embodiment, a computer-readable storage medium is also provided. When the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to perform the carbon paper processing method of the disclosed embodiment. The computer-readable storage medium may be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, or the like.

[0159] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, which can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided by the present disclosure can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0160] Other embodiments of the present disclosure will readily occur to those skilled in the art after considering the specification and practicing the invention disclosed herein. This disclosure is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, with the true scope and spirit of the present disclosure being indicated by the following claims.

[0161] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes can be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A carbon paper processing method, characterized in that: include: Acquire an image to be detected and a plurality of preset carbon paper index information; the image to be detected is obtained by scanning the target carbon paper; The carbon paper performance prediction model includes an image segmentation module, a performance prediction module, and a prediction processing module; the image to be detected and the plurality of preset carbon paper index information are input into the image segmentation module, image segmentation processing is performed, and a plurality of segmented images are obtained; the plurality of segmented images are input into the performance prediction module, performance prediction processing is performed, and second performance prediction information is obtained; The second performance prediction information represents performance prediction information corresponding to each of the plurality of segmented images under a plurality of preset carbon paper index information; Inputting the second performance prediction information into the prediction processing module, performing prediction information averaging processing, and obtaining first performance prediction information corresponding to each of the target carbon paper under the plurality of preset carbon paper index information; The first performance prediction information represents the predicted performance of the gas diffusion layer of the target carbon paper under the plurality of preset carbon paper index information; Selecting target performance prediction information that meets a preset condition from the plurality of first performance prediction information; Based on the target performance prediction information, the gas diffusion layer of the target carbon paper is adjusted.

2. The method according to claim 1, characterized in that The method further comprises: Acquire actual performance information corresponding to the plurality of preset carbon paper index information; Determining error information based on the first performance prediction information and the actual performance information; the error information is an error ratio between the prediction information of the plurality of performance indicators and the actual performance information; Evaluate the first performance prediction information according to the error information to obtain a prediction evaluation result; The selecting target performance prediction information that meets a preset condition from the plurality of first performance prediction information includes: According to the prediction evaluation result, the target performance prediction information that meets the preset conditions is selected from the plurality of first performance prediction information.

3. The method according to claim 1, characterized in that The method further comprises: Acquire multiple sample images, the multiple preset carbon paper index information, and label information corresponding to each sample image, wherein the sample images are obtained by scanning the sample carbon paper, and the label information represents the carbon paper performance information marked on the sample carbon paper; Inputting the plurality of sample images and the plurality of preset carbon paper index information into an initial performance prediction model, performing prediction processing on the performance of the sample carbon paper under the plurality of preset carbon paper index information, and obtaining corresponding carbon paper performance prediction information of the sample carbon paper under the plurality of preset carbon paper index information; determining loss information based on the carbon paper performance prediction information and the label information; The initial performance prediction model is trained based on the loss information until a training iteration condition is satisfied, and the initial performance prediction model corresponding to when the training iteration condition is satisfied is used as the carbon paper performance prediction model.

4. The method according to claim 3, characterized in that The label information includes at least one of porosity information, in-plane resistivity information, vertical resistivity information, and gas conductivity information.

5. The method according to claim 3, characterized in that The determining of loss information according to the carbon paper performance prediction information and the label information includes: Calculating difference information between the carbon paper performance prediction information and the label information; The average absolute information of the difference information is used as loss information.

6. The method according to claim 3, characterized in that Inputting the plurality of sample images and the plurality of preset carbon paper index information into an initial performance prediction model, and performing a performance prediction process on the sample carbon paper under the plurality of preset carbon paper index information includes: Inputting the plurality of sample images and the plurality of preset carbon paper index information into an initial performance prediction model, extracting features from the plurality of sample images to obtain attribute features; Extracting features from the plurality of preset carbon paper index information to obtain vector features; Performing feature fusion according to the attribute feature and the vector feature to obtain a fused feature; According to the fusion characteristics, the performance of the sample carbon paper under the plurality of preset carbon paper index information is predicted.

7. A carbon paper processing device, characterized in that: include: An acquisition module is used to acquire an image to be detected and a plurality of preset carbon paper index information; the image to be detected is obtained by scanning the target carbon paper; The first performance prediction information acquisition module is used for the carbon paper performance prediction model including an image segmentation module, a performance prediction module and a prediction processing module; Inputting the image to be detected and the plurality of preset carbon paper index information into the image segmentation module, performing image segmentation processing to obtain a plurality of segmented images; inputting the plurality of segmented images into the performance prediction module, performing performance prediction processing to obtain second performance prediction information; The second performance prediction information represents performance prediction information corresponding to each of the plurality of segmented images under a plurality of preset carbon paper index information; Inputting the second performance prediction information into the prediction processing module, performing prediction information averaging processing, and obtaining first performance prediction information corresponding to each of the target carbon paper under the plurality of preset carbon paper index information; The first performance prediction information represents the predicted performance of the gas diffusion layer of the target carbon paper under the plurality of preset carbon paper index information; a target performance prediction information acquisition module, configured to select target performance prediction information that meets a preset condition from a plurality of said first performance prediction information; An adjustment module is configured to adjust the gas diffusion layer of the target carbon paper based on the target performance prediction information.

8. An electronic device, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that When the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to perform the method according to any one of claims 1 to 6.

Images