A method for evaluating and analyzing the density distribution state of nanomaterials

Abstract

The present invention discloses a method for evaluating and analyzing the density distribution state of nanomaterials, which relates to the technical field of nanomaterials. The method ensures the uniformity and representativeness of the sample by preprocessing and regional segmenting the nanomaterial to be tested, and reduces the impact of sample heterogeneity on the evaluation results. The method selects n cluster centers and combines them with a density evaluation data set to determine the number Nks of nanoparticles in each cluster center. A particle distribution coefficient Kfxs is constructed based on the relative abundance Xdfd of the nanoparticles in each cluster center, effectively evaluating the distribution state of the nanoparticles. Feature extraction is used to obtain the effective surface area and porosity of the sample to be tested, and these two parameters are combined to construct a structural state coefficient Jzxs, thereby evaluating the structural characteristics and distribution state of the nanomaterial from multiple dimensions. The overall density distribution evaluation index Zmzs of the sample to be tested is calculated using a trained model to ensure that the model has high prediction accuracy and generalization ability.

Description

Technical Field

[0001] The present invention relates to the technical field of nanomaterials, and in particular to a method for evaluating and analyzing the density distribution state of nanomaterials. Background Art

[0002] In the fields of nanotechnology and materials science, nanomaterials, including nanoparticles, nanotubes, and nanofibers, are widely used in industries such as electronics, medicine, energy, and the environment due to their unique physical and chemical properties. These materials are typically composed of nanoscale particles, typically ranging in size from 1 to 100 nanometers. The structure and properties of nanomaterials are largely influenced by the size, morphology, dispersion, and aggregation of these nanoparticles. The distribution of these nanoparticles within the material directly affects the overall performance of the nanomaterial, such as mechanical strength, electrical conductivity, thermal conductivity, and catalytic activity.

[0003] In nanomaterials, the structure and distribution of nanoparticles within them significantly influence their density distribution. However, current approaches to assessing and analyzing the density distribution of nanomaterials face several limitations. Traditional methods may not be able to accurately and quickly assess the density distribution of nanomaterials, resulting in inaccurate results. Furthermore, these methods often rely on microscopic observations or simple surface analysis techniques, which are limited in resolution and data processing capabilities, making it difficult to comprehensively and accurately reflect the internal distribution of nanomaterials. Summary of the Invention

[0004] In view of the shortcomings of the prior art, the present invention provides a method for evaluating and analyzing the density distribution state of nanomaterials, which solves the problems in the above-mentioned background technology.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: A method for evaluating and analyzing the density distribution state of nanomaterials, comprising the following steps:

[0006] S1. Clean the nanomaterials to be tested in advance and segment the cleaned nanomaterials to be tested to obtain several groups of samples to be tested;

[0007] S2. Collect relevant structural data and relevant particle state data for several groups of samples to be tested to construct a density assessment data set, and preprocess the relevant data in the density assessment data set. Combined with dimensionless processing technology, the preprocessed relevant data are unified into units;

[0008] S3. Preselect n cluster centers and, in combination with the density assessment data set, determine the number of nanoparticles Nks in each cluster center. Based on the number of nanoparticles Nks in each cluster center, obtain the relative abundance Xdfd of the nanoparticles in each cluster center. Based on the relative abundance Xdfd of the nanoparticles in each cluster center and related particle state data, construct the particle distribution coefficient Kfxs.

[0009] S4. After feature extraction of the relevant structural data, the effective surface area Yxmd and the porosity Kxz of the sample to be tested are obtained. Based on the effective surface area Yxmd and the porosity Kxz, the structural state coefficient Jzxs is obtained. After dimensionless processing of the particle distribution coefficient Kfxs and the structural state coefficient Jzxs, the overall density distribution evaluation index Zmzs of the sample to be tested is calculated in combination with the trained distribution prediction model;

[0010] S5. Pre-set an evaluation threshold P, and compare and analyze the evaluation threshold P with the overall density distribution evaluation index Zmzs to evaluate the current density distribution state of the nanomaterial, and issue a corresponding early warning command according to the corresponding density distribution state.

[0011] Preferably, the specific steps of S1 include:

[0012] S11, pre-filling a washing tank with clean water, and placing the nanomaterial to be detected in the washing tank for soaking. At the same time, the clean water needs to be regularly replaced three times to repeat the washing step. After the washing is completed, the nanomaterial to be detected is removed from the washing tank and is low-temperature dried in a dryer;

[0013] S12, performing regional segmentation on the nanomaterial to be detected after the processing step of the nanomaterial to be detected in step S11, so as to evenly divide the nanomaterial to be detected into several groups of samples to be detected.

[0014] Preferably, the specific steps of S2 include:

[0015] S21, performing data detection on the plurality of groups of samples to be detected obtained in step S12 to respectively obtain relevant structural data and relevant particle state data in the samples to be detected, so as to construct a density assessment data set, wherein the relevant data in the density assessment data set includes relevant structural data and relevant particle state data;

[0016] S211 , the relevant particle state data includes the number Zks of nanoparticles in the centers of all clusters in the sample to be detected, and the size of each nanoparticle, wherein a nanoparticle size information unit is constructed according to the size of each nanoparticle.

[0017] S22. Detect and delete duplicate data of relevant data in the density assessment data set, identify missing data and outliers, and then use dimensionless processing technology to standardize the processed data to eliminate the influence of different feature dimensions.

[0018] Preferably, the specific steps of S3 include:

[0019] S31, pre-selecting n cluster centers, and preliminarily assigning each nanoparticle size in the nanoparticle size information unit to each cluster center based on the nanoparticle size information unit in step S212, so as to calculate the distance between each nanoparticle size and all cluster centers, wherein each cluster center is represented by the average size of the nanoparticles in each cluster, and the average size of the nanoparticles in each cluster is preliminarily randomly set;

[0020] S311. Use the Euclidean distance algorithm to calculate the distance from each nanoparticle size to the center of all clusters. Take the distance Jz(a, b) from the nanoparticle of size a to the center of the b-th cluster as an example. The distance Jz(a, b) is obtained by the following method:

[0021]

[0022] Preferably, S32, assigning each nanoparticle size to the nearest cluster center according to the calculated distance from each nanoparticle size to all cluster centers, to calculate the mean of all nanoparticle sizes in each cluster, and using the currently calculated mean of all nanoparticle sizes in each cluster as a new cluster center;

[0023] S33, repeating steps S31 and S32 until the pre-selected n cluster centers no longer change, stopping the repetitive operation, obtaining the final cluster center, and finally determining the number Nks of nanoparticles in each cluster center.

[0024] Preferably, S34, based on the number of nanoparticles Nks in each cluster center and in combination with the nanoparticle size information unit, the relative abundance Xdfd of the nanoparticles in each cluster center is calculated and obtained specifically in the following manner:

[0025]

[0026] In the formula, Xdfd i Expressed as the relative abundance of nanoparticles in the center of the i-th cluster, Nks i It is represented as the number of nanoparticles in the center of the i-th cluster, and Zks is represented as the number of nanoparticles in the centers of all clusters.

[0027] Preferably, in S35, a particle distribution coefficient Kfxs is constructed based on the relative abundance Xdfd of the nanoparticles in the center of each cluster and related particle state data. The particle distribution coefficient Kfxs is obtained by the following formula:

[0028]

[0029] The significance of this formula is: n represents the number of cluster centers, i = 1, 2, 3, ..., n, Kmd i It is expressed as the density of nanoparticles in the center of the i-th cluster, Xdfd i It is expressed as the relative abundance of nanoparticles in the center of the i-th cluster.

[0030] Preferably, the specific steps of S4 include:

[0031] S41. Based on the relevant structural data, the effective surface area Yxmd and the porosity Kxz of the sample to be tested are dimensionlessly processed to construct a structural state coefficient Jzxs. The structural state coefficient Jzxs is obtained by the following formula:

[0032]

[0033] Where α and β represent the weight values ​​of the effective surface area Yxmd and porosity Kxz of the sample to be tested, respectively, and C represents the first correction constant.

[0034] Preferably, S42, using convolutional neural network technology to construct an original model, and training and testing the original model with relevant structural data and relevant particle state data, and using the trained original model as a distribution state recognition model, respectively obtaining feature information in the distribution state recognition model, and training and testing the distribution state recognition model with the obtained feature information, and finally using the trained distribution state recognition model as a distribution prediction model, by performing dimensionless processing on the particle distribution coefficient Kfxs and the structural state coefficient Jzxs, obtaining the overall density distribution evaluation index Zmzs of the sample to be tested, specifically obtaining it in the following manner:

[0035]

[0036] Where F1 and F2 represent the weight values ​​of the particle distribution coefficient Kfxs and the structural state coefficient Jzxs, respectively, and V represents the second correction constant.

[0037] Preferably, the specific steps of S5 include:

[0038] S51, comparing and analyzing the overall density distribution evaluation index Zmzs with the evaluation threshold P to evaluate the current density distribution state of the nanomaterial. The specific evaluation content is as follows:

[0039] When the overall density distribution evaluation index Zmzs is greater than or equal to the evaluation threshold P, that is, when Zmzs≥P, it indicates that the density distribution of the current nanomaterial is in a normal state. At this time, three green warning commands will flash outwards;

[0040] When the overall density distribution evaluation index Zmzs is less than the evaluation threshold P, that is, when Zmzs<P, it indicates that the density distribution of the current nanomaterial is in an abnormal state. At this time, a red warning command will be continuously sent outwards until the system obtains the confirmation of the background operator;

[0041] S52. After receiving the corresponding warning command, generate a corresponding report, and the specific content is as follows:

[0042] If a green warning command is issued, the existing process and treatment flow will continue to be maintained at this time, and the process parameters and operating conditions of the current production will be recorded as part of the standard operating procedures;

[0043] If a red warning command is issued, the current process flow and parameters will be re-evaluated and optimized at this time, the factors affecting the density distribution will be found and improved, the quality control of raw materials and intermediate products will be monitored frequently, and small-scale tests will be carried out at the same time to verify the adjusted process.

[0044] The present invention provides an evaluation and analysis method for the density distribution state of nanomaterials, which has the following beneficial effects:

[0045] (1) By preprocessing and region segmentation of the to-be-detected nanomaterial, the uniformity and representativeness of the sample are ensured, the influence of sample non-uniformity on the evaluation result is reduced. By selecting n cluster centers and combining with the density evaluation data set, the number Nks of nanoparticles in each cluster center is determined, and the particle distribution coefficient Kfxs is constructed according to the relative abundance Xdfd of nanoparticles in each cluster center, effectively evaluating the distribution state of nanoparticles; By feature extraction, the effective surface area and porosity of the to-be-detected sample are obtained, and the structure state coefficient Jzxs is constructed by combining these two parameters, evaluating the structural characteristics and distribution state of nanomaterials from multiple dimensions. The distribution prediction model is constructed and trained by using convolutional neural network technology. The overall density distribution evaluation index Zmzs of the to-be-detected sample is calculated through the trained model, ensuring that the model has high prediction accuracy and generalization ability, and comparing and analyzing with the threshold to provide a clear evaluation standard and warning mechanism, facilitating the timely discovery and response to abnormal situations. By accurately evaluating the density distribution state of nanomaterials, problems in the production process can be found in time, the production process can be optimized, and the product quality and consistency can be improved. Combining with the warning command issuing mechanism, a complete quality control system is established to ensure the reliability and stability of nanomaterials in production and application.

[0046] (2) Based on the pre-selected n groups of cluster centers and the collected nanoparticle size information units, the nanoparticle sizes are comprehensively covered and preliminarily classified. The distance from each nanoparticle size to all cluster centers is calculated using the Euclidean distance algorithm to ensure that each nanoparticle size is accurately assigned to the closest cluster center, thereby establishing a preliminary nanoparticle distribution model. Based on the calculated distance from the nanoparticle size to all cluster centers, each nanoparticle size is assigned to the nearest cluster center. By calculating the mean of all nanoparticle sizes in each cluster, a new cluster center is determined, and the selection and distribution of cluster centers are further optimized to obtain the final cluster center and the number of nanoparticles in each cluster center. During this period, multiple iterations are performed to ensure the stability and final accuracy of the cluster center, thereby improving the accuracy and reliability of the nanoparticle distribution model. This process not only optimizes the classification and distribution of nanoparticles, but also provides accurate basic data for subsequent density evaluation and prediction models, making the analysis results more reliable. At the same time, by constructing the particle distribution coefficient Kfxs, the distribution of nanoparticles in different size ranges is further reflected. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 The figure is a flow chart of a method for evaluating and analyzing the density distribution state of nanomaterials according to the present invention. DETAILED DESCRIPTION

[0048] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0049] Example 1

[0050] See also Figure 1 The present invention provides a method for evaluating and analyzing the density distribution state of nanomaterials, comprising the following steps:

[0051] S1. Clean the nanomaterials to be tested in advance and segment the cleaned nanomaterials to be tested to obtain several groups of samples to be tested;

[0052] S2. Collect relevant structural data and relevant particle state data for several groups of samples to be tested to construct a density assessment data set, and preprocess the relevant data in the density assessment data set. Combined with dimensionless processing technology, the preprocessed relevant data are unified into units;

[0053] S3. Preselect n cluster centers and, in combination with the density assessment data set, determine the number of nanoparticles Nks in each cluster center. Based on the number of nanoparticles Nks in each cluster center, obtain the relative abundance Xdfd of the nanoparticles in each cluster center. Based on the relative abundance Xdfd of the nanoparticles in each cluster center and related particle state data, construct the particle distribution coefficient Kfxs.

[0054] S4. After feature extraction of the relevant structural data, the effective surface area Yxmd and porosity Kxz of the sample to be tested are obtained. Based on the effective surface area Yxmd and the porosity Kxz, the structural state coefficient Jzxs is obtained. After dimensionless processing of the particle distribution coefficient Kfxs and the structural state coefficient Jzxs, the overall density distribution evaluation index Zmzs of the sample to be tested is calculated in combination with the trained distribution prediction model.

[0055] S5. Pre-set an evaluation threshold P, and compare and analyze the evaluation threshold P with the overall density distribution evaluation index Zmzs to evaluate the current density distribution state of the nanomaterial, and issue a corresponding early warning command according to the corresponding density distribution state.

[0056] In this embodiment, the representativeness of the sample to be tested is further ensured through cleaning operations and regional segmentation; by collecting and preprocessing relevant data and combining dimensionless processing technology, the unit of data is unified, thereby improving the accuracy of density distribution assessment; a density assessment data set is constructed and preprocessed, and by selecting cluster centers and determining the number of nanoparticles Nks, the relative abundance Xdfd and particle distribution coefficient Kfxs are obtained. A systematic method is used to process the data, ensuring the versatility and accuracy of data processing; key parameters are obtained through feature extraction, and combined with a distribution prediction model, the overall density distribution assessment index Zmzs of the sample to be tested is calculated, further ensuring the scientificity and reliability of the assessment results. Evaluation thresholds are set and compared and analyzed, and warning commands are issued based on the density distribution status. Abnormal density distribution of nanomaterials is promptly identified and feedback is provided, which helps to improve the quality control and management level in the production and application of nanomaterials. In summary, this method further improves the accuracy and reliability of nanomaterial density distribution status assessment through standardized operating steps and scientific evaluation mechanisms, providing effective technical support for the research and application of nanomaterials.

[0057] Example 2

[0058] Please refer to Figure 1 , specifically: S1 specific steps include:

[0059] S11. Add clean water to the washing tank in advance. Make sure that there are no impurities or residues on the surface of the washing tank, and place the nanomaterial to be tested in the washing tank for soaking. Make sure that the sample is completely covered in water to ensure uniform cleaning. At the same time, you can also gently stir the nanomaterial to be tested to help remove dirt or residue attached to the surface. However, avoid excessive stirring during stirring to avoid introducing air or damaging the sample. If the nanomaterial to be tested is relatively dirty or requires particularly thorough cleaning, it is necessary to change the clean water three times at a time to repeat the cleaning steps to ensure the cleaning effect. After cleaning, remove the nanomaterial to be tested from the washing tank and use a dryer to dry the nanomaterial to be tested at low temperature to remove excess moisture to avoid the influence of moisture or other liquid residues on the evaluation results. It is best to use a hair dryer at a low temperature setting to avoid overheating the sample.

[0060] S12, performing regional segmentation on the nanomaterial to be detected after the processing step of the nanomaterial to be detected in step S11, so as to evenly divide the nanomaterial to be detected into several groups of samples to be detected.

[0061] In this embodiment, by adding clean water to the washing tank and ensuring that the washing tank surface is free of impurities or residue, the nanomaterials to be tested are protected from external contaminants during the cleaning process to the greatest extent possible. The immersion process allows the sample to be completely covered in water, ensuring that dirt and residue on the sample surface are evenly cleaned. Gentle stirring helps remove surface dirt or residue, while excessive stirring avoids the introduction of air or damage to the sample, thus protecting the integrity of the nanomaterial. For nanomaterials that are relatively dirty or require particularly thorough cleaning, repeated washing steps with three regular changes of clean water further ensure that the sample is well cleaned, effectively removing contaminants on the sample surface and reducing potential interference factors during the evaluation process. After cleaning, a dryer is used for low-temperature drying to ensure that excess water in the sample is removed, preventing the impact of moisture or other liquid residue on the evaluation results. Using a hair dryer with a low temperature setting further prevents damage to the sample due to overheating, thereby ensuring the quality and integrity of the sample. After cleaning and drying, the nanomaterial to be tested is segmented and evenly divided into several groups of samples to be tested, ensuring the consistency and representativeness of each group of samples and reducing the impact of sample differences on the evaluation results.

[0062] Uniformly segmenting the sample not only improves the accuracy of density distribution assessments but also lays a solid foundation for subsequent data collection and analysis, making each sample group comparable in subsequent analyses. In summary, the specific measures in the cleaning and pretreatment steps of this method effectively improve the accuracy and reliability of nanomaterial density distribution assessments by ensuring sample cleanliness and integrity, improving cleaning efficiency and sample consistency, and achieving uniform sample segmentation, providing an excellent prerequisite for subsequent evaluation and analysis.

[0063] Example 3

[0064] Please refer to Figure 1 , specifically: S2 specific steps include:

[0065] S21, performing data detection on the plurality of groups of samples to be detected obtained in step S12 to respectively obtain relevant structural data and relevant particle state data in the samples to be detected, so as to construct a density assessment data set, wherein the relevant data in the density assessment data set includes relevant structural data and relevant particle state data;

[0066] S211 , the relevant particle state data includes the number Zks of nanoparticles in the centers of all clusters in the sample to be detected, and the size of each nanoparticle, wherein a nanoparticle size information unit is constructed according to the size of each nanoparticle.

[0067] Among them, the size of each nanoparticle and the number of nanoparticles Zks in the center of all clusters can be monitored and obtained by transmission electron microscopy (TEM) and scanning electron microscopy (SEM). These microscopes can provide high-resolution images, so that the morphology and distribution of nanoparticles can be directly observed. Through image processing technology, the size of the particles can be accurately measured.

[0068] S22. Detect and delete duplicate data of relevant data in the density assessment data set to avoid bias in the analysis caused by data duplication. At the same time, identify missing data and outliers, and then use dimensionless processing technology to standardize the processed data to eliminate the influence of different feature dimensions, ensure that the data is comparable under different conditions, and improve the accuracy of data processing.

[0069] In this embodiment, in step S21, by performing data detection on several groups of samples to be detected, relevant structural data and relevant particle state data in the samples are obtained, and by constructing a nanoparticle size information unit, it is helpful to record and analyze the size distribution of each nanoparticle in detail, providing a rich and accurate data basis for subsequent evaluation and analysis; in step S22, duplicate data detection and deletion are performed on the relevant data in the density assessment data set, avoiding the influence of data repeatability on the analysis results, thereby improving the accuracy and reliability of the data. By identifying missing data and outliers and performing corresponding processing, noise and errors in the data are eliminated, and the integrity and quality of the data are further ensured. The processed data are standardized using dimensionless processing technology to eliminate the influence of different feature dimensions, so that the data are comparable on the same scale. Standardization helps to improve the accuracy of data analysis and modeling, making the relationship between different features clearer, and providing reliable data support for subsequent evaluation and prediction. Through the above-mentioned specific data detection and preprocessing steps, this method effectively solves the problems of data duplication, missing data, and anomalies, ensuring the accuracy, completeness, and comparability of the data, thereby improving the accuracy and reliability of the assessment of the density distribution state of nanomaterials and providing a solid data foundation for subsequent analysis and prediction.

[0070] Example 4

[0071] Please refer to Figure 1 , specifically: S3 specific steps include:

[0072] S31, pre-selecting n cluster centers, and preliminarily assigning each nanoparticle size in the nanoparticle size information unit to each cluster center based on the nanoparticle size information unit in step S212, so as to calculate the distance between each nanoparticle size and all cluster centers, wherein each cluster center is represented by the average size of the nanoparticles in each cluster, and the average size of the nanoparticles in each cluster is preliminarily randomly set;

[0073] S311. Use the Euclidean distance algorithm to calculate the distance from each nanoparticle size to the center of all clusters. Take the distance Jz(a, b) from the nanoparticle of size a to the center of the b-th cluster as an example. The distance Jz(a, b) is obtained by the following method:

[0074]

[0075] S32, assigning each nanoparticle size to the nearest cluster center according to the calculated distance from each nanoparticle size to all cluster centers, to calculate the mean of all nanoparticle sizes in each cluster, and using the currently calculated mean of all nanoparticle sizes in each cluster as a new cluster center;

[0076] S33, repeating steps S31 and S32 until the pre-selected n cluster centers no longer change, stopping the repetitive operation, obtaining the final cluster center, and finally determining the number Nks of nanoparticles in each cluster center.

[0077] The formula for the distance Jz(a, b) from a nanoparticle of size a to the center of the bth cluster can be simplified to Jz(a, b) = |ab|;

[0078] Example: Set 3 cluster centers and the average size of nanoparticles in the first cluster is 5, the average size of nanoparticles in the second cluster is 10, and the average size of nanoparticles in the third cluster is 20;

[0079] For a nanoparticle of size 1: the distance to the center of the first cluster is: |1-5|=4; the distance to the center of the second cluster is: 11-10|=9; the distance to the center of the third cluster is: |1-20|=19; repeat the above calculations for the distances to all cluster centers for other nanoparticle sizes.

[0080] In the present embodiment, by preselected n group cluster center, each nanoparticle size is tentatively assigned to each group cluster center, utilize Euclidean distance algorithm to calculate the distance of each nanoparticle size to all cluster centers, this step is understood the distribution of each nanoparticle size in space, for follow-up cluster center is determined to lay the foundation, according to the distance calculated, each nanoparticle size is assigned to nearest cluster center, calculate the mean of all nanoparticle sizes in each cluster, and use it as new cluster center, repeat this process, no longer change until preselected n group cluster center, namely reach convergence state, this iterative optimization has guaranteed the accuracy and stability of cluster center, effectively reflected the average size characteristics of nanoparticles in each cluster. Through repeated calculation and optimization, determined the nanoparticle number Nks in each cluster center, this step provides concrete data support for follow-up density assessment, makes assessment result more accurate and reliable. By the enforcement of above S3 concrete steps, this method effectively realizes accurate assessment and analysis to nanomaterial density distribution state, guaranteed the scientificity and validity of data processing, for further density distribution state assessment provides solid foundation and reliable data support.

[0081] Example 5

[0082] Please refer to Figure 1 Specifically: S34, based on the number of nanoparticles Nks in each cluster center and in combination with the nanoparticle size information unit, calculate and obtain the relative abundance Xdfd of the nanoparticles in the center of each cluster, specifically in the following way:

[0083]

[0084] In the formula, Xdfd i Expressed as the relative abundance of nanoparticles in the center of the i-th cluster, Nks i It is represented as the number of nanoparticles in the center of the i-th cluster, and Zks is represented as the number of nanoparticles in the centers of all clusters.

[0085] S35. Based on the relative abundance Xdfd of the nanoparticles in the center of each cluster and the relevant particle state data, a particle distribution coefficient Kfxs is constructed. The particle distribution coefficient Kfxs is obtained by the following formula:

[0086]

[0087] The significance of this formula is: n represents the number of cluster centers, i = 1, 2, 3, ..., n, Kmd i It is expressed as the density of nanoparticles in the center of the i-th cluster, Xdfd i It is expressed as the relative abundance of nanoparticles in the center of the i-th cluster.

[0088] Among them, the density of nanoparticles in the center of each cluster can be monitored and obtained through density gradient centrifugation, suspension method, X-ray diffraction method and nitrogen adsorption method. Among them, the principle of density gradient centrifugation is: place the nanoparticle suspension in a density gradient solution, and through centrifugation, the particles will be distributed in different density ranges according to their density. According to the position of the particles in the centrifuge tube, their density can be determined.

[0089] The principle of the suspension method is to suspend nanoparticles in a liquid with a known density and adjust the density of the liquid until the particles are suspended without settling. At this point, the density of the liquid is the density of the particles.

[0090] The principle of X-ray diffraction is: by measuring the lattice parameters and relative density of nanoparticles, the theoretical density of the particles can be calculated.

[0091] The principle of the nitrogen adsorption method is to measure the specific surface area (BET area) of the nanoparticles and combine it with the known volume of the particles to infer the density of the particles.

[0092] In this embodiment, based on the number of nanoparticles Nks in each cluster center, combined with the nanoparticle size information unit, the relative abundance Xdfd of the nanoparticles in each cluster center is calculated. This step, by counting the number of nanoparticles in each cluster center, reflects the distribution of nanoparticles of different sizes in detail. The relative abundance Xdfd of the nanoparticles in each cluster center and the related particle state data are used to construct the particle distribution coefficient Kfxs. This coefficient reflects the density distribution of nanoparticles in different cluster centers. The constructed particle distribution coefficient Kfxs provides a quantitative description of the nanoparticle distribution, providing a scientific basis and support for further research and application. Through these steps, this method effectively optimizes the evaluation and analysis process of the density distribution state of nanomaterials, making the analysis results more reliable and operational, and providing a powerful analytical tool and method for nanomaterial research and industrial application.

[0093] Example 6

[0094] Please refer to Figure 1 , specifically: S4 specific steps include:

[0095] S41. Based on the relevant structural data, the effective surface area Yxmd and the porosity Kxz of the sample to be tested are dimensionlessly processed to construct a structural state coefficient Jzxs. The structural state coefficient Jzxs is obtained by the following formula:

[0096]

[0097] Wherein, α and β represent the weight values ​​of the effective surface area Yxmd and porosity Kxz of the sample to be tested, respectively, and C represents the first correction constant, wherein 0<α≤1, 0<β≤1, and α+β=1.

[0098] The effective surface area Yxmd of the sample to be tested refers to the surface area of ​​the nanoparticles or nanostructures, which is usually related to their particle size and surface morphology. The effective surface area Yxmd can be obtained by the following methods: Gas adsorption method (specific surface area analysis): The specific surface area is measured by the adsorption and desorption process of gas (such as nitrogen) on the surface of the nanomaterial. This method uses the Brunner-Emmett-Traut-Thiele method (BET method) or other adsorption isotherm analysis methods to calculate the effective surface area;

[0099] Porosity Kxz refers to the ratio of the volume of the voids on the surface of the sample to be tested to the total volume. The porosity Kxz can be obtained by the following method:

[0100] Gas adsorption-desorption method: Similar to measuring specific surface area, gas adsorption-desorption method can measure the pore structure of nanomaterials, including pore size distribution and total volume.

[0101] Meanwhile, the cross-section of the nanomaterial can also be observed through an electron microscope (such as a scanning electron microscope) or an optical microscope to directly observe the pore structure, and the volume of the pores can be measured using image processing software.

[0102] S42. Use the convolutional neural network technology to construct the original model, and train and test the original model with relevant structure data and relevant particle state data. Then, use the trained original model as the distribution state recognition model, respectively obtain the feature information within the distribution state recognition model, and use the obtained feature information to train and test the distribution state recognition model. Finally, use the trained distribution state recognition model as the distribution prediction model. After dimensionless processing of the particle distribution coefficient Kfxs and the structure state coefficient Jzxs, obtain the overall density distribution evaluation index Zmzs of the sample to be detected, which is specifically obtained in the following manner:

[0103]

[0104] In the formula, F1 and F2 respectively represent the weight values of the particle distribution coefficient Kfxs and the structure state coefficient Jzxs, and V represents the second correction constant. Among them, 0 < F1 ≤ 1, 0 < F2 ≤ 1, and F1 + F2 = 1.

[0105] The specific steps of S5 include:

[0106] S51. Compare and analyze the overall density distribution evaluation index Zmzs with the evaluation threshold P to evaluate the current density distribution state of the nanomaterial. The specific evaluation content is as follows:

[0107] If the overall density distribution evaluation index Zmzs is greater than or equal to the evaluation threshold P, that is, Zmzs ≥ P, it indicates that the density distribution of the current nanomaterial is in a normal state, and at this time, three green warning commands are flashed outward;

[0108] If the overall density distribution evaluation index Zmzs is less than the evaluation threshold P, that is, Zmzs < P, it indicates that the density distribution of the current nanomaterial is in an abnormal state, and at this time, a continuous red warning command is issued outward until the system obtains the confirmation of the background operator;

[0109] S52. After receiving the corresponding warning command, generate the corresponding report. The specific content is as follows:

[0110] If a green warning command is issued, the existing process and treatment flow will continue to be maintained at this time, and the current production process parameters and operating conditions will be recorded as part of the standard operating procedures for reference in subsequent batch production;

[0111] If a red alert is issued, the current process flow and parameters will be re-evaluated and optimized to identify and improve factors affecting density distribution. The quality control of raw materials and intermediate products will be frequently monitored to ensure that each step of the operation meets the standard requirements. At the same time, small-scale tests will be carried out to verify the adjusted process to ensure that it can effectively improve the density distribution.

[0112] In this embodiment, the structural state coefficient Jzxs comprehensively considers the sample's surface area and porosity, providing an important structural feature foundation for subsequent density distribution assessment. A convolutional neural network (CNN) is used to construct an initial model. Through training and testing using relevant structural and particle state data, the trained distribution state recognition model is optimized into a distribution prediction model. This model effectively learns from the characteristic information and predicts the overall density distribution evaluation index Zmzs of the sample to be tested, further ensuring the accuracy and reliability of the assessment results. The overall density distribution evaluation index Zmzs is compared with a pre-set assessment threshold P, and the current density distribution state of the nanomaterial is determined based on the comparison results. Furthermore, corresponding production reports are generated based on the triggering of different warning commands. When a green warning is issued, the current process parameters and operating conditions are recorded for reference in standard operating procedures. When a red warning is issued, the process flow and parameters are reassessed and optimized to ensure continuous improvement in production quality and density distribution. Through these steps, this method effectively improves the accuracy and real-time performance of nanomaterial density distribution assessment, providing a reliable scientific basis and operational guidance for quality control and process optimization during production.

[0113] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A method for evaluating and analyzing the density distribution of nanomaterials, characterized by: The following steps are included: S1. Clean the nanomaterials to be tested in advance and segment the cleaned nanomaterials to be tested to obtain several groups of samples to be tested; S2. Collect relevant structural data and relevant particle state data for several groups of samples to be tested to construct a density assessment data set, and preprocess the relevant data in the density assessment data set. Combined with dimensionless processing technology, the preprocessed relevant data are unified into units; S3. Preselect n cluster centers and, in combination with the density assessment data set, determine the number of nanoparticles Nks in each cluster center. Based on the number of nanoparticles Nks in each cluster center, obtain the relative abundance Xdfd of the nanoparticles in each cluster center. Based on the relative abundance Xdfd of the nanoparticles in each cluster center and related particle state data, construct the particle distribution coefficient Kfxs. The S3 specific steps include: S31, pre-selecting n cluster centers, and preliminarily assigning each nanoparticle size in the nanoparticle size information unit to each cluster center based on the nanoparticle size information unit in step S212, so as to calculate the distance between each nanoparticle size and all cluster centers, wherein each cluster center is represented by the average size of the nanoparticles in each cluster, and the average size of the nanoparticles in each cluster is preliminarily randomly set; S311. Use the Euclidean distance algorithm to calculate the distance from each nanoparticle size to the center of all clusters. Take the distance Jz(a, b) from the nanoparticle of size a to the center of the b-th cluster as an example. The distance Jz(a, b) is obtained by the following method: S32, assigning each nanoparticle size to the nearest cluster center according to the calculated distance from each nanoparticle size to all cluster centers, to calculate the mean of all nanoparticle sizes in each cluster, and using the currently calculated mean of all nanoparticle sizes in each cluster as a new cluster center; S33, repeating steps S31 and S32 until the pre-selected n cluster centers no longer change, stopping the repetitive operation, obtaining the final cluster center, and finally determining the number Nks of nanoparticles in each cluster center; S4. After feature extraction of the relevant structural data, the effective surface area Yxmd and the porosity Kxz of the sample to be tested are obtained. Based on the effective surface area Yxmd and the porosity Kxz, the structural state coefficient Jzxs is obtained. After dimensionless processing of the particle distribution coefficient Kfxs and the structural state coefficient Jzxs, the overall density distribution evaluation index Zmzs of the sample to be tested is calculated in combination with the trained distribution prediction model; S5. Pre-set an evaluation threshold P, and compare and analyze the evaluation threshold P with the overall density distribution evaluation index Zmzs to evaluate the current density distribution state of the nanomaterial, and issue a corresponding early warning command according to the corresponding density distribution state.

2. The method for evaluating and analyzing the density distribution of nanomaterials according to claim 1, wherein: The specific steps of S1 include: S11, pre-filling a washing tank with clean water, and placing the nanomaterial to be detected in the washing tank for soaking. At the same time, the clean water needs to be regularly replaced three times to repeat the washing step. After the washing is completed, the nanomaterial to be detected is removed from the washing tank and is low-temperature dried in a dryer; S12, performing regional segmentation on the nanomaterial to be detected after the processing step of the nanomaterial to be detected in step S11, so as to evenly divide the nanomaterial to be detected into several groups of samples to be detected.

3. The method for evaluating and analyzing the density distribution of nanomaterials according to claim 1, wherein: The specific steps of S2 include: S21, performing data detection on the plurality of groups of samples to be detected obtained in step S12 to respectively obtain relevant structural data and relevant particle state data in the samples to be detected, so as to construct a density assessment data set, wherein the relevant data in the density assessment data set includes relevant structural data and relevant particle state data; S211, the relevant particle state data includes the number Zks of nanoparticles in the center of all clusters in the sample to be detected, and the size of each nanoparticle, wherein a nanoparticle size information unit is constructed according to the size of each nanoparticle; S22. Detect and delete duplicate data of relevant data in the density assessment data set, identify missing data and outliers, and then use dimensionless processing technology to standardize the processed data to eliminate the influence of different feature dimensions.

4. The method for evaluating and analyzing the density distribution of nanomaterials according to claim 3, wherein: S34. Based on the number of nanoparticles Nks in each cluster center and in combination with the nanoparticle size information unit, the relative abundance Xdfd of the nanoparticles in each cluster center is calculated and obtained in the following manner: In the formula, Xdfd i Expressed as the relative abundance of nanoparticles in the center of the i-th cluster, Nks i It is represented as the number of nanoparticles in the center of the i-th cluster, and Zks is represented as the number of nanoparticles in the centers of all clusters.

5. The method for evaluating and analyzing the density distribution of nanomaterials according to claim 1, wherein: S35. Based on the relative abundance Xdfd of the nanoparticles in the center of each cluster and the relevant particle state data, a particle distribution coefficient Kfxs is constructed. The particle distribution coefficient Kfxs is obtained by the following formula: The significance of this formula is: n represents the number of cluster centers, i = 1, 2, 3, ..., n, Kmd i It is expressed as the density of nanoparticles in the center of the i-th cluster, Xdfd i It is expressed as the relative abundance of nanoparticles in the center of the i-th cluster.

6. The method for evaluating and analyzing the density distribution of nanomaterials according to claim 1, wherein: The specific steps of S4 include: S41. Based on the relevant structural data, the effective surface area Yxmd and the porosity Kxz of the sample to be tested are dimensionlessly processed to construct a structural state coefficient Jzxs. The structural state coefficient Jzxs is obtained by the following formula: Where α and β represent the weight values ​​of the effective surface area Yxmd and porosity Kxz of the sample to be tested, respectively, and C represents the first correction constant.

7. The method for evaluating and analyzing the density distribution of nanomaterials according to claim 1, wherein: S42. Use convolutional neural network technology to construct an original model, and train and test the original model with relevant structural data and relevant particle state data, and use the trained original model as a distribution state recognition model, respectively obtain feature information within the distribution state recognition model, and train and test the distribution state recognition model with the obtained feature information, and finally use the trained distribution state recognition model as a distribution prediction model, which is specifically obtained in the following manner: Where F1 and F2 represent the weight values ​​of the particle distribution coefficient Kfxs and the structural state coefficient Jzxs, respectively, and V represents the second correction constant.

8. The method for evaluating and analyzing the density distribution of nanomaterials according to claim 1, wherein: The specific steps of S5 include: S51, comparing and analyzing the overall density distribution evaluation index Zmzs with the evaluation threshold P to evaluate the current density distribution state of the nanomaterial. The specific evaluation content is as follows: If the overall density distribution evaluation index Zmzs is greater than or equal to the evaluation threshold P, that is, Zmzs ≥ P, it indicates that the current density distribution of the nanomaterial is in a normal state, and a green warning command flashes outward three times; If the overall density distribution evaluation index Zmzs is less than the evaluation threshold P, that is, Zmzs<P, it means that the current density distribution of the nanomaterial is in an abnormal state. At this time, a red warning command is continuously issued to the outside until the system receives confirmation from the background operator; S52. After receiving the corresponding warning command, a corresponding report is generated. The specific contents are as follows: If a green warning command is triggered, the existing process and treatment process will continue to be maintained, and the current production process parameters and operating conditions will be recorded as part of the standard operating procedures; If a red alert is issued, the current process flow and parameters will be re-evaluated and optimized, factors affecting density distribution will be identified and improved, the quality control of raw materials and intermediate products will be frequently monitored, and small-scale tests will be conducted to verify the adjusted process.

Images