Tempering film full-process antibacterial performance monitoring method and system

Abstract

This application belongs to the field of intelligent detection technology, specifically relating to a method and system for monitoring the antibacterial performance of tempered glass film throughout its entire process. The method includes: acquiring color-changing image sequences of tempered glass film samples with various coating formulations, extracting the color-changing degree sequence, brightness distribution, and color shift values; determining the performance fluctuation range by combining the antibacterial agent release rate and environmental variables; generating preliminary mapping results when the performance fluctuation range exceeds a fluctuation threshold; screening samples, grouping them according to the release process, and performing consistency screening to obtain grouped data; constructing and adjusting the correlation matrix to obtain a stable correlation matrix; fitting a quantization function based on the stable correlation matrix, outputting antibacterial performance indicators, and obtaining the antibacterial stability range through time-series prediction. This application can improve the continuity and stability of tempered glass film antibacterial performance evaluation.

Description

Technical Field

[0001] This application belongs to the field of intelligent detection technology, specifically relating to a method and system for monitoring the antibacterial performance of tempered glass film throughout the entire process. Background Technology

[0002] Tempered glass screen protectors, as a crucial protective material for electronic display terminals, have seen their surface functional coatings expand from simple scratch and explosion protection to include antibacterial, antifouling, and weather resistance. For manufacturers, the stability of antibacterial coating performance throughout the preparation, curing, aging, testing, and factory evaluation stages directly impacts product quality grading, batch consistency, and subsequent quality traceability. Existing testing methods typically rely on sampling antibacterial experiments or surface color observation, which are time-consuming and struggle to continuously reflect the effects of coating formulation differences, antibacterial agent release states, and environmental changes on antibacterial performance.

[0003] With the application of smart sensors in the field of materials testing, image, environmental and release process data can be continuously collected. However, existing methods mostly remain at the level of judging single data items and lack a processing mechanism to establish a stable correspondence between color change response, release process and antibacterial test results.

[0004] In actual production and testing, the same type of tempered glass film may exhibit similar color changes but different antibacterial properties under different temperature, humidity, light and contact conditions, or the antibacterial effect may be similar but the surface color changes may be significantly different. This leads to unstable test results and is not conducive to coating formulation screening, batch release and long-term performance evaluation. Summary of the Invention

[0005] To address the above issues, this application provides a method and system for monitoring the antibacterial performance of tempered glass throughout the entire process, which aims to at least solve the problem of how to accurately establish a stable correlation between the color change response and antibacterial performance of tempered glass under different coating formulations and complex environmental influences.

[0006] To achieve the above objectives, the technical solution adopted in this application is as follows: In a first aspect, this application provides a method for monitoring the antibacterial performance of tempered glass film throughout the entire process, the method comprising: Obtain the color change image sequence of tempered film samples with each coating formulation, and extract the color change degree sequence, brightness distribution and color shift value; Key areas are determined based on brightness distribution and color shift values, and the performance fluctuation range is determined by combining antimicrobial agent release rate and environmental variables. When the performance fluctuation range exceeds the fluctuation threshold, a preliminary mapping result is generated based on the color deviation characteristics and antibacterial performance labeling. The tempered glass film samples in the preliminary mapping results were screened, grouped by release process, and screened for consistency to obtain grouped data; A stable correlation matrix was obtained by constructing an association matrix based on the grouped data, antimicrobial agent release rate, and environmental variables and adjusting the weights. Based on the stable correlation matrix, a quantization function is fitted, and the antibacterial performance index is output when the prediction deviation is lower than the deviation threshold. The antibacterial stability range is obtained through time-series prediction.

[0007] In one possible implementation, the process involves acquiring color-changing image sequences of tempered film samples with various coating formulations, and extracting color-changing degree sequences, brightness distributions, and color shift values. This includes: continuously acquiring images of the tempered film samples under various simulated environments using an optical sensor to obtain color-changing image sequences; performing denoising and standardization processing on the color-changing image sequences to obtain standardized image sequences; and determining the color-changing degree sequences, brightness distributions, and color shift values ​​based on the brightness values ​​at each sampling time and preset reference color values ​​in the standardized image sequences.

[0008] In one possible implementation, key regions are determined based on brightness distribution and color shift values, and the performance fluctuation range is determined by combining antimicrobial agent release rate and environmental variables. This includes: identifying color change regions based on brightness distribution and color shift values, and defining the color change regions as key regions; corresponding the brightness distribution, color shift values, antimicrobial agent release rate, and environmental variables within the key regions according to the sampling times of the color-changing image sequence to obtain regional fluctuation characteristics; and determining the performance fluctuation range based on the regional fluctuation characteristics.

[0009] In one possible implementation, when the performance fluctuation range exceeds the fluctuation threshold, a preliminary mapping result is generated based on the color deviation features and antibacterial performance annotations. This includes: extracting color deviation features from key areas when the performance fluctuation range exceeds the fluctuation threshold, and constructing a feature vector based on the color deviation features; matching the feature vector with the antibacterial performance annotations, and determining the correspondence between the degree of color change and the antibacterial effect through a classification boundary optimization method; and generating a preliminary mapping result based on the correspondence. The antibacterial performance annotations are annotation information determined based on the antibacterial effect test records of the tempered glass film samples.

[0010] In one possible implementation, the tempered glass film samples in the preliminary mapping results are screened, grouped by release process, and screened for consistency to obtain grouped data. This includes: selecting relevant samples from the tempered glass film samples based on the degree of matching between discoloration and antibacterial effect in the preliminary mapping results; grouping the relevant samples by release process based on differences in coating formulation and changes in antibacterial agent release rate; calculating the intra-cluster consistency index of relevant samples within the same group, and retaining groups with intra-cluster consistency indices not lower than the consistency threshold to obtain grouped data; wherein, the intra-cluster consistency index is used to indicate the degree of consistency of relevant samples within the same group in terms of discoloration, antibacterial agent release rate, and antibacterial effect.

[0011] In one possible implementation, a stable correlation matrix is ​​obtained by constructing a correlation matrix based on grouped data, antimicrobial agent release rate, and environmental variables, and adjusting the weights. This includes: constructing an initial correlation matrix based on the degree of discoloration, antimicrobial effect, antimicrobial agent release rate, and environmental variables corresponding to the grouped data; iteratively adjusting the matrix weights in the initial correlation matrix; and determining the correlation matrix obtained in the current iteration as the stable correlation matrix when the change in matrix weights obtained in two adjacent iterations is less than a stability threshold. The initial correlation matrix is ​​used to represent the correspondence between the degree of discoloration, antimicrobial agent release rate, environmental variables, and antimicrobial effect.

[0012] In one possible implementation, a quantization function is fitted based on a stable correlation matrix, and an antibacterial performance index is output when the prediction deviation is below a deviation threshold. This includes: determining the mapping sample from the degree of discoloration to the antibacterial performance based on the stable correlation matrix; performing error analysis and function fitting on the mapping sample to obtain the quantization function; determining the prediction deviation based on the predicted antibacterial performance value output by the quantization function and the measured antibacterial performance value in the antibacterial effect test record; and outputting the antibacterial performance index based on the quantization function when the prediction deviation is below the deviation threshold.

[0013] In one possible implementation, the method further includes: generating antibacterial effect evaluation results based on antibacterial performance indicators; cross-checking the antibacterial effect evaluation results using historical evaluation data and additional environmental simulation data; and adjusting the parameters of the quantization function based on the additional environmental simulation data when the cross-check results meet preset consistency conditions to obtain optimized antibacterial performance indicators.

[0014] In one possible implementation, the antimicrobial stability range is obtained through time-series prediction, including: constructing a performance trend model based on the discoloration degree sequence, historical evaluation data, and optimized antimicrobial performance indicators; obtaining the future discoloration trend through a time-series prediction algorithm based on the performance trend model; and determining the antimicrobial stability range based on the correspondence between the future discoloration trend and the optimized antimicrobial performance indicators.

[0015] Secondly, this application provides a tempered glass film antibacterial performance monitoring system for the entire process, used to implement a method for monitoring the antibacterial performance of tempered glass film throughout the entire process. The system includes: The image processing module is used to acquire the color change image sequence of tempered film samples with various coating formulations, and extract the color change degree sequence, brightness distribution and color shift value; The fluctuation determination module is used to identify key areas based on brightness distribution and color offset values, and to determine the performance fluctuation range by combining antimicrobial agent release rate and environmental variables. The mapping generation module is used to generate preliminary mapping results based on color deviation characteristics and antibacterial performance labels when the performance fluctuation range exceeds the fluctuation threshold. The grouping and filtering module is used to filter tempered glass samples in the preliminary mapping results, group them according to the release process, and perform consistency screening to obtain grouped data. The matrix construction module is used to construct an association matrix based on grouped data, antimicrobial agent release rate, and environmental variables, and adjust the weights to obtain a stable association matrix. The performance prediction module is used to fit a quantization function based on a stable correlation matrix and output antibacterial performance indicators when the prediction deviation is lower than the deviation threshold. The antibacterial stability range is obtained through time-series prediction.

[0016] Compared with existing technologies, the advantages and beneficial effects of this application are as follows: By acquiring color-changing image sequences of tempered film samples with various coating formulations and extracting color-changing degree sequences, brightness distributions, and color shift values, continuous quantification of surface optical response was achieved, avoiding judgment bias caused by relying solely on single visual observation or single-point detection.

[0017] By combining the antibacterial agent release rate and environmental variables to determine the performance fluctuation range, simultaneous analysis of color change response, release state and external environmental influences was achieved, enabling the test results to more closely reflect the real changes in the antibacterial performance of the coating.

[0018] By generating preliminary mapping results when the performance fluctuation range exceeds the fluctuation threshold, targeted modeling of abnormal fluctuation samples is achieved, reducing the interference of invalid samples on performance judgment.

[0019] By screening samples in the preliminary mapping results, grouping release processes, and performing consistency screening, stratified processing of different coating formulations and different release processes was achieved, reducing mapping instability caused by formulation differences.

[0020] By constructing an association matrix and adjusting the weights, the relationship between the degree of color change, the release rate of antibacterial agents, environmental variables, and the antibacterial effect was solidified.

[0021] By using quantification functions and time-series predictions to output antimicrobial performance indicators and antimicrobial stability ranges, continuous evaluation of current performance and long-term stability is achieved, providing an actionable basis for quality control, formulation adjustment, and batch traceability. Attached Figure Description

[0022] Figure 1 This is a flowchart illustrating the method described in this application; Figure 2 This is a block diagram of the module combination of the system in this application. Detailed Implementation

[0023] To enable those skilled in the art to better understand the technical solution, the present application will be described in detail below with reference to the embodiments. The description in this section is only exemplary and explanatory, and should not be used to limit the scope of protection of the present application in any way.

[0024] Antimicrobial performance monitoring typically refers to the collection, analysis, and determination of observable characteristics related to the antimicrobial effect, focusing on the formation, maintenance, and attenuation of the antimicrobial ability of a material surface. For tempered glass films, antimicrobial performance is not solely reflected in the final antimicrobial test results; it is also influenced by the coating formulation, the antimicrobial agent release process, surface optical changes, and environmental conditions. By using intelligent sensors to acquire color-changing images, release states, and environmental parameters, dispersed detection data can be converted into continuous performance characterization information, shifting antimicrobial performance monitoring from single-point result judgment to a comprehensive evaluation that is process-oriented, correlated, and traceable. Based on this approach, this application jointly processes the color-changing response, antimicrobial agent release rate, and environmental variables of tempered glass film samples to establish a quantitative relationship between the degree of color change and the antimicrobial effect, and further determines the antimicrobial performance indicators and the range of antimicrobial stability.

[0025] In this application, the antibacterial effect is the measured result obtained from the antibacterial effect test record, the antibacterial performance index is the evaluation result output based on the stable correlation matrix and quantization function, the antibacterial effect is used for training, verification and correction of mapping relationship, and the antibacterial performance index is used to form the performance evaluation of the current sample and subsequent stability prediction.

[0026] like Figure 1 As shown, a method for monitoring the antibacterial performance of tempered glass film throughout the entire process can be used for sample monitoring during the preparation, curing, aging, testing, and factory evaluation stages of the tempered glass film coating. The method links the data collected at each stage through sample numbers. The method includes: Obtain the color change image sequence of tempered film samples with each coating formulation, and extract the color change degree sequence, brightness distribution and color shift value; During implementation, the system assigns a sample number and a coating formula number to each tempered glass film sample, and binds the sample number to the acquisition time, simulated environmental parameters, and image acquisition parameters. An optical sensor continuously acquires images of the tempered glass film samples under constant illumination conditions. The simulated environmental parameters include one or more of temperature, humidity, light intensity, and contact time. After image acquisition, the data processing unit performs denoising and standardization on the color-changing image sequence, forming a standardized image sequence with consistent size, brightness range, and color space. The standardized image sequence extracts brightness and color values ​​according to the sampling time and compares them with preset reference color values ​​to obtain the color-changing degree sequence, brightness distribution, and color shift value. These results, along with the sample number, are stored as the data basis for subsequent key area determination and performance fluctuation range calculation.

[0027] Obtain color-changing image sequences of tempered glass film samples with various coating formulations, and extract the color-changing degree sequence, brightness distribution, and color shift value. This includes: continuously acquiring images of tempered glass film samples under various simulated environments using an optical sensor to obtain color-changing image sequences; performing noise reduction and standardization processing on the color-changing image sequences to obtain standardized image sequences; and determining the color-changing degree sequence, brightness distribution, and color shift value based on the brightness value and preset reference color value at each sampling time in the standardized image sequences.

[0028] In one embodiment, a unified data caliber is set for the acquisition, preprocessing, and feature extraction of color-changing image sequences. Tempered glass film samples are placed in a simulated environment chamber with constant temperature, humidity, and light regulation functions. Optical sensors are fixed at a position perpendicular or nearly perpendicular to the tempered glass film surface, and the acquisition distance remains constant throughout a single detection process. A standard white board or standard color chart is provided in the simulated environment chamber. The data processing unit acquires images of the standard white board or standard color chart before each round of acquisition to correct for lighting fluctuations and sensor response differences. The sampling period is set according to the color-changing response speed of the antibacterial coating; faster-changing materials can use shorter sampling periods, and slower-changing materials can use longer sampling periods. Samples from the same batch use the same sampling period to avoid incomparability of color-changing degree sequences due to different sampling intervals. When local reflections, dust obstruction, or edge shadows exist in the acquired original images, the data processing unit checks the image quality based on areas of brightness abrupt changes, the proportion of saturated pixels, and image sharpness. When the saturation pixel ratio exceeds the set ratio, or the image clarity is lower than the device calibration lower limit, the corresponding sampled image will be marked as an abnormal image. Abnormal images will not participate in the color offset value calculation, and the abnormal mark and sampling time will be retained for supplementary collection or subsequent traceability.

[0029] Noise reduction employs one of median filtering, mean filtering, or bilateral filtering, with the filter window size determined based on image resolution and the fineness of the tempered glass surface texture. Normalization includes size standardization, color space conversion, and luminance normalization. The CIELAB color space can be used, as it can separately represent luminance and chrominance components, making it suitable for quantifying subtle color variations on the tempered glass surface. For the first... At each sampling time, the color offset value can be determined by the following formula:

[0030] In the formula, For the first Color offset value at each sampling time, For the first The brightness component at each sampling time. For the first The red-green color components at each sampling time. For the first The yellow and blue color components at each sampling time. The luminance component in the preset reference color value. The red and green color components in the preset base color value. The preset reference color value represents the yellow and blue color components. This preset reference color value can be determined from the same sample image before entering the simulation environment, or from the average color value of qualified samples with the same coating formulation. When using the average color value, the number of samples and their batch origins should be recorded in the testing configuration. The color change degree sequence is formed by arranging the color offset values ​​at each sampling time in chronological order. Alternatively, the color offset values ​​can be normalized according to the maximum permissible color offset value set in the product testing specification to obtain color change degree values ​​that facilitate comparison of samples from different coating formulations. The brightness distribution is formed by the brightness components of each pixel in the standardized image or the average brightness of each detection window. The data processing unit writes the brightness distribution, color offset values, sampling time, and sample number into the same testing record, enabling subsequent processing to be called according to the same timeline.

[0031] Key areas are determined based on brightness distribution and color shift values, and the performance fluctuation range is determined by combining antimicrobial agent release rate and environmental variables. The data processing unit reads the brightness distribution, color shift values, and color-changing image sequence generated in the previous stage, and establishes a detection record according to the sample number and sampling time. For each standardized image frame, the data processing unit identifies color-changing regions based on the spatial distribution of local brightness changes and color shift values ​​in the brightness distribution. After excluding image edge shadows, fixed reflective points, and invalid background areas, the color-changing regions are designated as key regions. The data processing unit maps the brightness distribution, color shift values, antibacterial agent release rate, and environmental variables within the key regions to the same sampling time, forming regional fluctuation characteristics. These regional fluctuation characteristics describe the color-changing response and release state of the same location under different environmental conditions, and are used to calculate the performance fluctuation range, serving as the data basis for subsequent determination of whether to proceed to the initial mapping process.

[0032] The key regions are determined based on brightness distribution and color shift values, and the performance fluctuation range is determined by combining antimicrobial agent release rate and environmental variables. This includes: identifying color change regions based on brightness distribution and color shift values, and defining the color change regions as key regions; matching the brightness distribution, color shift values, antimicrobial agent release rate, and environmental variables within the key regions according to the sampling times of the color-changing image sequence to obtain regional fluctuation characteristics; and determining the performance fluctuation range based on the regional fluctuation characteristics.

[0033] In one embodiment, the determination of the key region is based on the effective detection region in the color-changing image sequence, and spatial and temporal continuity constraints are set for the color-changing region. The effective detection region is determined by the contour boundary of the tempered glass film sample in the image. The contour boundary can be obtained through sample template matching, edge detection, or calibration point localization at a fixed shooting position. The data processing unit divides the effective detection region into multiple detection windows, each corresponding to a fixed image position and window number. The size of the detection window is set according to the minimum identifiable area of ​​the color-changing patch on the tempered glass film surface and the resolution of the optical sensor. If the size is too small, it is easily affected by noise; if the size is too large, it will mask the local color-changing differences. Therefore, it can be set as a fixed pixel block or an equal-area grid according to the sample surface area and image resolution.

[0034] The data processing unit calculates the average brightness, brightness dispersion, and average color shift value of each detection window at the same sampling time. If the difference between the average brightness and the average brightness of surrounding windows exceeds a brightness variation threshold, or if the average color shift value exceeds a color shift threshold, the detection window is marked as a candidate color change region. The brightness variation threshold is determined by the upper limit of brightness fluctuation of qualified samples of the same coating formulation under stable conditions, and the color shift threshold is determined by the reference color value and the product's allowable color change range. Candidate color change regions must appear continuously in adjacent sampling times or form a connected region with adjacent windows to be confirmed as color change regions. Candidate regions that appear only in a single sampling time and do not have spatial connectivity are marked as isolated abnormal regions and are not included in the performance fluctuation range calculation.

[0035] The antimicrobial agent release rate is derived from release detection records of the same sample or batch of samples near the same sampling time, and can be obtained by extraction solution detection, electrochemical sensor detection, or other quantifiable release detection methods. Environmental variables include one or more of temperature, humidity, light intensity, and exposure time, which are recorded by simulated environment equipment or environmental sensors. To ensure stable data correspondence, the data processing unit uses the sampling time of the color-changing image sequence as a reference to map the antimicrobial agent release rate and environmental variables to the same time axis. If the release detection record and the image sampling time are not completely consistent, an interpolation of adjacent time points or the matching of the nearest time point is used to establish a correspondence. If the matching time difference exceeds the set allowable time range, the sampling point is not included in the calculation of regional fluctuation characteristics.

[0036] Regional fluctuation characteristics can be represented by performance fluctuation values:

[0037] In the formula, For the first Performance fluctuation value at each sampling time, For the first Color offset values ​​within key regions at each sampling time. For the first Brightness changes in key areas at each sampling time. For the first Antimicrobial agent release rate at each sampling time point For the first Environmental impact values ​​at each sampling time, , , and The weighting coefficients are used to determine the environmental impact values, which are determined by the degree of deviation of temperature, humidity, light intensity, and contact time from the corresponding baseline conditions. The weighting coefficients can be set based on the correlation between various characteristics and antibacterial effect test records in historical testing data, or they can be pre-configured by the process testing specifications. The same set of weighting coefficients is used for samples in the same batch to avoid variations in judgment criteria within the batch. The performance fluctuation range is determined by the maximum, minimum, and fluctuation span of the same sample's performance fluctuation values ​​within the testing period. The data processing unit saves the performance fluctuation range along with the sample number, key area location, sampling time range, and environmental variable records for subsequent use in judging cases exceeding the fluctuation threshold.

[0038] When the performance fluctuation range exceeds the fluctuation threshold, a preliminary mapping result is generated based on the color deviation characteristics and antibacterial performance labeling. The data processing unit reads the performance fluctuation range formed in the previous stage and compares it with the fluctuation threshold. The fluctuation threshold is set based on the historical fluctuation range of qualified samples of the same coating formulation, the allowable error of antibacterial effect testing, and the stable range of the simulated environment. When the performance fluctuation range exceeds the fluctuation threshold, the data processing unit extracts color deviation features from key areas. The color deviation features include color shift amplitude, color shift duration, color change area ratio, and color shift rate. The data processing unit organizes the color deviation features into feature vectors and matches the feature vectors with the antibacterial performance labels using sample numbers. The antibacterial performance labels are derived from the antibacterial effect test records of tempered glass samples. After matching, the correspondence between the degree of color change and the antibacterial effect is determined using a classification boundary optimization method, generating preliminary mapping results for subsequent screening, grouping, and association matrix construction.

[0039] When the performance fluctuation range exceeds the fluctuation threshold, a preliminary mapping result is generated based on the color deviation features and antibacterial performance annotations. This includes: extracting color deviation features from key areas and constructing feature vectors based on the color deviation features; matching the feature vectors with the antibacterial performance annotations and determining the correspondence between the degree of color change and the antibacterial effect through a classification boundary optimization method; and generating a preliminary mapping result based on the correspondence. The antibacterial performance annotations are annotation information determined based on the antibacterial effect test records of the tempered glass film samples.

[0040] In one embodiment, the triggering judgment of performance fluctuation range adopts a stratified threshold at the sample formulation level to avoid false triggering caused by differences in color change sensitivity between different coating formulations. The data processing unit reads the corresponding fluctuation threshold according to the coating formulation number. The fluctuation threshold is jointly determined by the performance fluctuation range of the coating formulation in qualified sample testing, the repeatability error of antibacterial effect testing, and the allowable fluctuation of simulated environment. For new coating formulations without historical testing records, the system uses the fluctuation threshold of the same antibacterial agent content, the same coating structure, or similar production batches as the initial threshold, and updates it after the subsequent testing records accumulate to a set number. During the update, the original threshold and the updated threshold are retained to avoid the inability to trace back due to different standards used for historical samples and new samples.

[0041] Sampling intervals where performance fluctuations exceed the fluctuation threshold are marked as unmapped intervals. Color deviation features are extracted from the key regions corresponding to these unmapped intervals. Color offset amplitude characterizes the degree of offset of the key region relative to the baseline color value; color offset duration characterizes the duration of continuous sampling exceeding the color offset threshold; color change area ratio characterizes the proportion of the key region where color change occurs; and color offset rate characterizes the trend of color offset values ​​at adjacent sampling times. These features are standardized in terms of dimensions before constructing the feature vector. This standardization can be achieved using max-min normalization or standardization based on the statistical range of historical valid samples. When missing sampling points exist, the data processing unit prioritizes using adjacent valid sampling points to fill in the gaps. When the number of consecutive missing points exceeds a set upper limit, a missing label is generated for the unmapped interval, which is not used in classification boundary training but is retained as an anomaly record.

[0042] Antimicrobial performance labeling is determined by antimicrobial effect test records. These records include sample number, test batch, test time, measured antimicrobial performance value, and test conditions. The measured antimicrobial performance value can be expressed as inhibition rate, bactericidal rate, or antimicrobial activity value; the specific method used depends on product testing specifications and laboratory records. The data processing unit matches feature vectors with antimicrobial performance labels according to sample number and test batch. When multiple antimicrobial effect test records exist for the same sample, the record closest to the acquisition time of the color-changing image is used first, or the average test result under the same test conditions is used. Antimicrobial performance labels can be classified into qualified, borderline, and unqualified categories according to testing specifications, or multi-level categories can be formed based on the range of measured antimicrobial performance values. The classification boundary optimization method can be implemented using a support vector machine classifier. The training data consists of matched feature vectors and antimicrobial performance labels, and the classification boundary parameters are adjusted based on classification errors in the training and validation sets. If the validation set classification results fail to reach the set accuracy or class differentiation stability, the data processing unit re-examines the outlier fields in the feature vectors and performs a consistency check on missing markers, outlier image markers, and antibacterial effect test conditions. After the classification boundary stabilizes, the system outputs the correspondence between the degree of color change and the antibacterial effect, and writes the correspondence, classification boundary parameters, sample numbers participating in training, and annotation sources into the preliminary mapping result. This preliminary mapping result is not directly used as the final judgment result, but rather serves as the data foundation for subsequent high-relevance sample screening, release process grouping, and correlation matrix construction.

[0043] The tempered glass film samples in the preliminary mapping results were screened, grouped by release process, and screened for consistency to obtain grouped data; The data processing unit reads the preliminary mapping results and obtains the tempered glass film samples participating in the preliminary mapping, their discoloration degree, antibacterial effect, and matching degree. Tempered glass film samples with a matching degree reaching the correlation threshold are identified as relevant samples. The correlation threshold is set based on the mapping error of historical samples, the repeatability of antibacterial effect testing, and the stability of test batches. The data processing unit groups the relevant samples according to the differences in coating formulation and the characteristics of antibacterial agent release rate changes, so that relevant samples within the same group have similar release processes. After grouping, the data processing unit calculates the intra-cluster consistency index of relevant samples within the same group and compares the intra-cluster consistency index with the consistency threshold. Groups with intra-cluster consistency indices not lower than the consistency threshold are retained and form grouped data, which enters the subsequent correlation matrix construction process.

[0044] The tempered glass film samples in the preliminary mapping results are screened, grouped by release process, and screened for consistency to obtain grouped data. This includes: selecting relevant samples from the tempered glass film samples based on the matching degree between discoloration degree and antibacterial effect in the preliminary mapping results; grouping the relevant samples by release process based on the differences in coating formulation and the variation characteristics of antibacterial agent release rate; calculating the intra-cluster consistency index of relevant samples within the same group, and retaining groups with intra-cluster consistency index not lower than the consistency threshold to obtain grouped data; wherein, the intra-cluster consistency index is used to indicate the degree of consistency of relevant samples within the same group in terms of discoloration degree, antibacterial agent release rate, and antibacterial effect.

[0045] In one embodiment, the grouping of relevant samples during screening and release is constrained by sample number, coating formulation number, and test batch number to prevent samples from different batches from being incorrectly grouped into the same group due to different environmental conditions. The data processing unit reads the degree of discoloration, antibacterial effect, and matching degree of each tempered film sample from the preliminary mapping results. The matching degree can be determined by the category confidence, mapping error, or validation set matching results output by the classification boundary optimization method. When using category confidence, the correlation threshold can be set based on the lower confidence limit of correctly classified samples in the validation set; when using mapping error, the correlation threshold can be set based on the allowable error of the antibacterial effect test. Samples whose matching degree does not reach the correlation threshold are retained in the original test record but do not participate in the release process grouping to avoid low-confidence samples affecting the subsequent weight adjustment of the correlation matrix.

[0046] The release process grouping is primarily based on differences in coating formulations and the characteristics of antimicrobial agent release rate variations. Coating formulation differences include one or more of the following: antimicrobial agent type, antimicrobial agent content, coating thickness, curing conditions, and surface treatment method. Antimicrobial agent release rate variation characteristics include the initial release rate, peak release rate, time to reach the peak release rate, and the release rate decline trend. The data processing unit compares the antimicrobial agent release rate variation characteristics within the same coating formulation category, grouping related samples with similar release processes into the same candidate group. For related samples with the same coating formulation number but significantly different antimicrobial agent release rate variations, candidate groups are re-divided according to the peak release rate range or release duration. For related samples with different coating formulation numbers but highly similar release processes, they can be recorded as adjacent candidate groups without direct merging, facilitating subsequent tracing of the impact of formulation differences on antimicrobial efficacy.

[0047] After the same candidate group is formed, the data processing unit calculates the intra-cluster consistency index. The intra-cluster consistency index can be determined by the following formula:

[0048] In the formula, For the first Intra-cluster consistency index of candidate groups For the first Discrete values ​​of the degree of color change within each candidate group For the first Discrete values ​​of antimicrobial agent release rate within each candidate group For the first Discrete values ​​of antibacterial efficacy within each candidate group , and These are the weighting coefficients. The discrete values ​​of discoloration degree, antimicrobial agent release rate, and antimicrobial effect can all be expressed using standard deviation, range, or normalized deviation, and the same calculation method should be used within the same test batch. The weighting coefficients are set based on the influence of the three types of discrete values ​​on the mapping error in historical test data, or they can be pre-configured according to the process testing specifications. When historical data is lacking, the three weighting coefficients can be set to the same value and updated after data accumulation. The closer the intra-cluster consistency index is to one, the higher the degree of consistency among related samples within the candidate group in terms of discoloration degree, antimicrobial agent release rate, and antimicrobial effect.

[0049] The consistency threshold is determined based on the statistical lower limit of intra-cluster consistency index for qualified batch samples, the allowable range of detection error, and the sample stability required for correlation matrix modeling. When the intra-cluster consistency index of the same candidate group is not lower than the consistency threshold, the data processing unit retains the candidate group as a valid group and generates a group number, a sample number list, a coating formulation record, release process characteristics, and consistency screening results. When the intra-cluster consistency index is lower than the consistency threshold, the data processing unit checks whether there are abnormal image markers, missing release rate records, or samples with inconsistent antibacterial effect test conditions within the candidate group; if abnormal samples are found, they are removed and the intra-cluster consistency index is recalculated; if no abnormal samples can be removed, the candidate group is marked as a low-consistency group. Low-consistency groups do not enter the correlation matrix construction process, but the group record and failure reason are retained for subsequent supplementary collection or manual review.

[0050] A stable correlation matrix was obtained by constructing an association matrix based on the grouped data, antimicrobial agent release rate, and environmental variables and adjusting the weights. The data processing unit reads the grouped data and obtains the degree of discoloration, antibacterial effect, antibacterial agent release rate, and environmental variables within the same valid group according to the group number. Each data point is matched with its sample number and sampling time. Missing sampling records, records of out-of-bounds environmental variables, and records of abnormal antibacterial agent release rates are marked before matrix construction. The data processing unit uses the degree of discoloration, antibacterial agent release rate, and environmental variables as input dimensions of the correlation matrix, and the antibacterial effect as the mapping target to construct an initial correlation matrix. The matrix weights in the initial correlation matrix are initialized according to the sample distribution of the grouped data and historical detection errors, and adjusted in each iteration based on the prediction deviation of the antibacterial effect. When the change in matrix weights between two consecutive iterations is less than a stability threshold, the correlation matrix obtained in the current iteration is determined as a stable correlation matrix and enters the quantization function fitting process.

[0051] A stable correlation matrix is ​​obtained by constructing a correlation matrix based on grouped data, antimicrobial agent release rate, and environmental variables, and adjusting the weights. This includes: constructing an initial correlation matrix based on the degree of discoloration, antimicrobial effect, antimicrobial agent release rate, and environmental variables corresponding to the grouped data; iteratively adjusting the matrix weights in the initial correlation matrix; and determining the correlation matrix obtained in the current iteration as the stable correlation matrix when the change in matrix weights obtained in two adjacent iterations is less than a stability threshold. The initial correlation matrix is ​​used to represent the correspondence between the degree of discoloration, antimicrobial agent release rate, environmental variables, and antimicrobial effect.

[0052] In one embodiment, the association matrix construction process limits the data integrity and dimensional consistency of the grouped data. After reading the valid groups, the data processing unit checks whether each sample within the same group simultaneously possesses the degree of color change, antibacterial effect, antibacterial agent release rate, and environmental variables. The antibacterial effect comes from antibacterial effect test records, the antibacterial agent release rate comes from release detection records of the same sample or batch of samples, and the environmental variables come from simulated environmental equipment or environmental sensors. When a single data item is missing, the data processing unit fills it in based on adjacent records at the same sampling time; if the length of consecutive missing data exceeds the set allowable length, the corresponding sample is removed from the current matrix construction task, and the reason for removal is recorded. The set allowable length is determined based on the sampling period and the rate of change of the antibacterial agent release process; a shorter allowable length is used when the release process changes rapidly, and a longer allowable length is used when the release process changes slowly.

[0053] Before matrix construction, the data processing unit standardized the dimensions of discoloration degree, antimicrobial agent release rate, environmental variables, and antimicrobial effect. Discoloration degree was represented by the normalized result of color offset value; antimicrobial agent release rate was normalized according to the release rate range within the same testing period; environmental variables were represented by the degree of deviation from the baseline environmental conditions; and antimicrobial effect was represented by the measured value or grade numerical result from the antimicrobial effect test record. For the first... The input features and the first The matrix elements between the antibacterial effect categories can be determined by the following formula:

[0054] In the formula, The first in the initial correlation matrix The input feature and the first Initial correlation values ​​between antibacterial effect categories This represents the number of samples participating in matrix construction within the current valid group. For the first The first sample Each input feature can take a value. For the first The nth sample pair The annotation values ​​for each antibacterial effect category are specified. Input features include the degree of color change, antibacterial agent release rate, and specific features from environmental variables. Antibacterial effect categories are determined based on antibacterial effect test records; when using measured values, the antibacterial effect category can be derived from the measured value range. This relationship is used to convert sample features and antibacterial effect records within the same group into iteratively adjustable matrix elements.

[0055] Matrix weight adjustments are made based on the prediction deviation of antibacterial effect. In each iteration, the data processing unit uses the current association matrix to calculate the predicted antibacterial effect for each sample and compares it with the measured results in the antibacterial effect test records to obtain the current iteration error. When the error exceeds the set error range, the data processing unit adjusts the matrix weight of the corresponding input feature upwards or downwards according to the error direction. The stability threshold is jointly set by the upper limit of the matrix weight change of historical qualified samples in a stable state, the repeatability error of the antibacterial effect test, and the allowed number of iterations. The matrix weight change between two adjacent iterations can be determined by the following formula:

[0056] In the formula, For the first Round iteration and the first The change in matrix weights between iterations For the first After the first iteration The input feature and the first Correlation values ​​between antibacterial effect categories This represents the correlation value corresponding to the previous iteration. When the change in matrix weights is less than the stability threshold, it indicates that the current correlation has reached a stable state within the valid group. If the change in matrix weights has not reached the stability condition and the number of iterations has reached the upper limit, the data processing unit marks the current group as an unstable group and returns to the group data to check for abnormal samples, environmental variable mutation records, and abnormal antimicrobial agent release rate records. Groups that still cannot reach the stability condition after review are not included in the quantization function fitting process, and the group number, the cause of the anomaly, and the current matrix state are saved in the detection record.

[0057] Based on the stable correlation matrix, a quantization function is fitted, and the antibacterial performance index is output when the prediction deviation is lower than the deviation threshold. The antibacterial stability range is obtained through time-series prediction.

[0058] During implementation, the data processing unit reads the stable correlation matrix and extracts the mapping samples between discoloration degree and antibacterial performance from it. The mapping samples include sample number, discoloration degree, antibacterial agent release rate, environmental variables, matrix weights, and measured antibacterial performance values. The data processing unit performs error analysis on the mapping samples, eliminating samples with inconsistent test conditions or significantly abnormal prediction deviations. Then, it performs function fitting on the remaining mapping samples to obtain a quantization function. The quantization function outputs a predicted antibacterial performance value, which is compared with the measured antibacterial performance values ​​in the antibacterial effect test records. When the prediction deviation is below the deviation threshold, the data processing unit outputs an antibacterial performance index. After verification with historical data and additional environmental simulation data, the antibacterial performance index is entered into the performance trend model. A time-series prediction algorithm calculates the antibacterial stability range, and the antibacterial performance index, stability range, and corresponding sample records are written into the database.

[0059] Based on a stable correlation matrix, a quantization function is fitted, and the antibacterial performance index is output when the prediction deviation is lower than the deviation threshold. This process includes: determining the mapping sample from the degree of discoloration to the antibacterial performance based on the stable correlation matrix; performing error analysis and function fitting on the mapping sample to obtain the quantization function; determining the prediction deviation based on the predicted antibacterial performance value output by the quantization function and the measured antibacterial performance value in the antibacterial effect test record; and outputting the antibacterial performance index based on the quantization function when the prediction deviation is lower than the deviation threshold.

[0060] In one embodiment, the fitting process of the quantization function is based on a stable correlation matrix as a constraint, and a consistency requirement is set for the source of the mapped samples. The data processing unit reads the matrix weights from the stable correlation matrix of each valid group, and combines the degree of color change, antimicrobial agent release rate, and environmental variables in the group data to form a mapped sample for fitting. The measured antimicrobial performance value of the mapped sample comes from the antimicrobial effect test record. When there are multiple test results for the same sample, the test result that is closest to the acquisition time of the color change image sequence, the simulated environmental conditions, and the detection time of the antimicrobial agent release rate is preferred. When multiple test results are under the same test conditions, the average value can be used as the measured antimicrobial performance value. If the sample test conditions are inconsistent, the sample number cannot be matched, the antimicrobial agent release rate is missing, or the environmental variable record is out of bounds, the sample will not be included in the quantization function fitting, but the reason for exclusion will be retained.

[0061] The quantification function can be one of linear regression, piecewise linear regression, or polynomial regression. The function form is determined based on the relationship between the degree of discoloration and the measured antibacterial performance in historical samples. When the relationship is close to monotonic linear, linear regression is used; when different discoloration intervals correspond to different release states, piecewise linear regression is used; when the degree of discoloration and antibacterial performance show a smooth, non-linear change, low-order polynomial regression is used. To avoid overly complex models, a low-order function form that meets the deviation threshold requirement is preferred. The quantification function can be expressed as follows:

[0062] In the formula, For the first The predicted antibacterial performance of each mapped sample. For the first The degree of color change of each mapped sample For the first Antimicrobial release rate of each mapped sample For the first The influence value of environmental variables on each mapped sample. To stabilize the correlation matrix, These are the parameters for the quantization function. The quantization function uses the degree of color change, antimicrobial agent release rate, environmental variable influence values, and stable correlation matrix to generate predicted antimicrobial performance values. The quantization function parameters are obtained by fitting the measured antimicrobial performance values ​​in the mapping sample.

[0063] Prediction bias can be determined by the following formula:

[0064] In the formula, To predict bias, The number of mapped samples used in the bias calculation. For the first The measured antibacterial performance values ​​of each mapped sample. The deviation threshold is set based on the repeatability error of the antibacterial effect test, the allowable error for product quality judgment, and the historical fitting error. When the predicted deviation is lower than the deviation threshold, the quantization function is used to output the antibacterial performance index; when the predicted deviation is not lower than the deviation threshold, the data processing unit checks abnormal mapped samples, antibacterial effect test records, and the status of the stable correlation matrix, and returns to the matrix weight repetition kernel or the mapped sample screening stage if necessary. The antibacterial performance index, along with the quantization function parameters, deviation calculation records, and sample numbers, are stored together for evaluation and traceability.

[0065] The method further includes: generating antibacterial effect evaluation results based on antibacterial performance indicators; cross-checking the antibacterial effect evaluation results using historical evaluation data and additional environmental simulation data; and adjusting the parameters of the quantization function based on the additional environmental simulation data when the cross-check results meet the preset consistency conditions to obtain optimized antibacterial performance indicators.

[0066] In one embodiment, the antimicrobial efficacy evaluation result is generated after the antimicrobial performance index is output, and historical evaluation data and additional environmental simulation data are incorporated for cross-checking. The data processing unit establishes a correspondence between the antimicrobial performance index and sample number, coating formulation number, discoloration degree sequence, antimicrobial agent release rate, environmental variables, and quantification function parameters to form the antimicrobial efficacy evaluation result. The antimicrobial efficacy evaluation result may include the antimicrobial performance index, prediction bias, number of samples involved in the fitting, corresponding stable correlation matrix number, and test batch. The antimicrobial performance index in the evaluation result can be classified into qualified, critical, or unqualified levels according to product testing specifications, and the level boundaries are determined by the antimicrobial efficacy testing standard, enterprise quality control documents, or the statistical range of historical qualified batches. The level classification is only used for output and traceability and does not change the antimicrobial performance index already formed by the quantification function.

[0067] Cross-checking is conducted using a combination of historical evaluation data and additional environmental simulation data. Historical evaluation data includes antimicrobial performance indicators, antimicrobial effect test records, and discoloration sequences of samples with the same or similar coating formulations in previous tests. Additional environmental simulation data is generated by simulation equipment using extended combinations of temperature, humidity, light intensity, and contact time, ensuring these combinations do not deviate from the actual application range of the product. The data processing unit compares the current antimicrobial effect evaluation results with historical evaluation data to check whether the differences in indicators for the same coating formulation under similar environmental conditions are within preset consistency conditions. These preset consistency conditions are determined by the fluctuation range of indicators from historically qualified batches, the repeatability error of antimicrobial effect tests, and environmental control errors. When the differences in indicators exceed the preset consistency conditions, the data processing unit checks for abnormal images, missing release rate records, and abrupt changes in environmental variables in the current test records, creating cross-check anomaly records.

[0068] Additional environmental simulation data is used to test the stability of the quantification function under adjacent environmental conditions. The data processing unit inputs the additional environmental simulation data into the quantification function to obtain the predicted antimicrobial performance under the extended environment, and compares it with the results of similar conditions in historical evaluation data. When the cross-check results meet the preset consistency conditions, the data processing unit fine-tunes the quantification function parameters based on the additional environmental simulation data. The parameter adjustment is limited to the range of influence of the recorded environmental variables and does not change the sample grouping and matrix weight structure corresponding to the stable correlation matrix. After the parameter adjustment is completed, the data processing unit recalculates the antimicrobial performance index to obtain the optimized antimicrobial performance index. When the cross-check results do not meet the preset consistency conditions, the current antimicrobial effect evaluation result is marked as a result to be reviewed. The system retains the original quantification function parameters and the current antimicrobial performance index, and does not use additional environmental simulation data to overwrite the original result. The result to be reviewed enters the manual review or supplementary data collection process to avoid inconsistent data directly affecting subsequent stability predictions. The optimized antimicrobial performance index is stored simultaneously with the antimicrobial performance index before adjustment, and the source of parameter adjustment, the range of additional environmental simulation data, and the cross-check results are recorded.

[0069] The antimicrobial stability range is obtained through time-series prediction, including: constructing a performance trend model based on the discoloration degree sequence, historical evaluation data, and optimized antimicrobial performance indicators; obtaining the future discoloration trend through a time-series prediction algorithm based on the performance trend model; and determining the antimicrobial stability range based on the correspondence between the future discoloration trend and the optimized antimicrobial performance indicators.

[0070] In one embodiment, the antimicrobial stability range is determined by a performance trend model, which describes the relationship between the discoloration trend and antimicrobial performance changes of the tempered glass sample within the testing period and under extended use conditions. The data processing unit reads the discoloration degree sequence, historical evaluation data, and optimized antimicrobial performance indicators, and establishes a time-series record based on the sampling time. The time-series record includes the sampling time, discoloration degree, antimicrobial agent release rate, environmental variables, antimicrobial performance indicators, and the corresponding coating formulation number. Historical evaluation data provides a reference for the changes of the same or similar coating formulations over a longer testing period, and the optimized antimicrobial performance indicators determine the performance status of the current sample at the prediction starting point. When a sampling point is missing from the time-series record, the data processing unit fills in the missing points according to the sampling period and adjacent valid records; when the length of consecutive missing points exceeds a set upper limit, stability prediction is not performed, and a data insufficiency record is generated.

[0071] The performance trend model can employ a moving average model, an exponential smoothing model, or an autoregressive model. A moving average model is suitable when the data volume is small and the changes are stable; an exponential smoothing model is suitable when the color change is more sensitive to recent environmental changes; and an autoregressive model is suitable when historical evaluation data is sufficient and shows significant time correlation. The model type is determined by the number of valid time points in the test batch, the stationarity of the color change sequence, and the length of the historical evaluation data. Before constructing the performance trend model, the data processing unit checks for abrupt changes in the color change sequence. If an abrupt change corresponds to abrupt changes in environmental variables, abnormal image markers, or abnormal antibacterial agent release rates, it is treated as an anomaly; if abrupt changes are continuous and consistent with antibacterial effect test records, they are retained as true trend changes.

[0072] The time-series prediction algorithm outputs future color change trends based on a performance trend model. These trends include the direction and magnitude of color change within the predicted time period, as well as the time interval for reaching the set color change level. The data processing unit inputs these future color change trends into a validated quantization function or the antimicrobial performance mapping relationship within the performance trend model to obtain the antimicrobial performance change range within the predicted time period. The antimicrobial stability range is determined by the upper and lower limits of the antimicrobial performance indicators and the time range within which the performance continuously meets the qualification conditions within the predicted time period. The duration of the antimicrobial stability range is set based on the product usage scenario, the simulated environment cycle, and the testing specifications, and must not exceed the prediction span supported by historical evaluation data. When the prediction span exceeds the range supported by historical data, the system outputs a low-confidence prediction flag and requests supplementary aging test data or an extended simulated environment testing cycle.

[0073] The data processing unit writes the antimicrobial stability range, future discoloration trend, performance trend model parameters, historical evaluation data sources, and optimized antimicrobial performance indicators into the testing database. If the prediction results show that the antimicrobial performance indicators are below the acceptable boundary within the target service life, the system generates a stability deficiency record and associates it with the corresponding coating formulation number and key environmental variables. If the antimicrobial performance indicators remain above the acceptable boundary within the target service life, the system generates a stability acceptable record and retains the prediction interval and prediction confidence information. The prediction confidence information is determined by the number of valid time points, the coverage of historical evaluation data, and cross-check results. When subsequent samples of the same formulation enter the testing process, the system can call the stored performance trend model as a reference model, but each batch of samples still regenerates prediction results based on the discoloration degree sequence, antimicrobial agent release rate, and environmental variables of this batch.

[0074] like Figure 2 As shown, a tempered glass film antibacterial performance monitoring system is used to implement a method for monitoring the antibacterial performance of tempered glass film throughout the entire process. The system includes: The image processing module is used to acquire color-changing image sequences of tempered film samples with various coating formulations, and extract color-changing degree sequences, brightness distributions, and color shift values. The image processing module consists of an optical sensor, a standard light source, an image acquisition card, and an image processor. It is used to continuously photograph tempered film samples and to perform noise reduction, standardization, and color feature extraction on the acquired images.

[0075] The fluctuation determination module is used to determine key areas based on brightness distribution and color offset values, and to determine the performance fluctuation range by combining antimicrobial agent release rate and environmental variables. The fluctuation determination module consists of an environmental sensor, an antimicrobial agent release rate detection unit, an area recognition processor and a data synchronization circuit. It is used to collect environmental variables and antimicrobial agent release rate, and to correlate them with brightness distribution and color offset values ​​in a time sequence.

[0076] The mapping generation module is used to generate preliminary mapping results based on color deviation features and antibacterial performance labels when the performance fluctuation range exceeds the fluctuation threshold. The mapping generation module consists of a feature calculation processor, a label data storage, and a classification operation processor. It is used to extract color deviation features when the performance fluctuation range exceeds the fluctuation threshold and generate preliminary mapping results based on antibacterial performance labels.

[0077] The grouping and screening module is used to screen the tempered glass film samples in the preliminary mapping results, group them during the release process, and perform consistency screening to obtain grouped data. The grouping and screening module consists of a sample data storage, a grouping operation processor, and a consistency judgment processor. It is used to read the sample data in the preliminary mapping results and complete the relevant sample screening, release process grouping, and consistency screening.

[0078] The matrix construction module is used to construct an association matrix based on grouped data, antimicrobial agent release rate, and environmental variables, and adjust the weights to obtain a stable association matrix. The matrix construction module consists of a matrix operation processor, a weight update circuit, and a parameter storage. It is used to construct an association matrix based on grouped data, antimicrobial agent release rate, and environmental variables, and to iteratively adjust the matrix weights.

[0079] The performance prediction module is used to fit a quantization function based on the stable correlation matrix, and outputs an antibacterial performance index when the prediction deviation is below a deviation threshold. The antibacterial stability range is obtained through time-series prediction. The performance prediction module consists of a function fitting processor, a time-series prediction processor, and a result output interface. It is used to fit a quantization function based on the stable correlation matrix and output the antibacterial performance index and the antibacterial stability range.

[0080] It should be noted that, in this document, the terms "comprising," "including," and any other variations are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Specific examples have been used in this document to illustrate the principles and implementation methods of the technical solutions of this application. The above examples are only for the purpose of helping to understand the methods and core ideas of this application. The above descriptions are merely preferred embodiments of this application. It should be pointed out that, due to the limitations of written expression and the objective existence of infinite specific structures, those skilled in the art can make several improvements, modifications, or changes without departing from the principles of this application, and can also combine the above technical features in an appropriate manner; these improvements, modifications, changes, or combinations, or the direct application of the concept and technical solutions of this application to other situations without modification, should all be considered within the scope of protection of this application.

Claims

1. A method for monitoring the antibacterial performance of a full process of a toughening film, characterized in that, The method includes: Obtain the color change image sequence of tempered film samples with each coating formulation, and extract the color change degree sequence, brightness distribution and color shift value; Key areas are determined based on the brightness distribution and color offset values, and the performance fluctuation range is determined by combining the antibacterial agent release rate and environmental variables. When the performance fluctuation range exceeds the fluctuation threshold, a preliminary mapping result is generated based on the color deviation characteristics and antibacterial performance labeling. The tempered glass film samples in the preliminary mapping results are screened, grouped by release process, and screened for consistency to obtain grouped data; A stable correlation matrix is ​​obtained by constructing an association matrix and adjusting the weights based on the grouped data, the antimicrobial agent release rate, and the environmental variables. Based on the stable correlation matrix, a quantization function is fitted, and an antibacterial performance index is output when the prediction deviation is lower than the deviation threshold. The antibacterial stability range is obtained through time-series prediction.

2. The method of claim 1, wherein, The process of acquiring the color-changing image sequence of each coating formulation tempered film sample, and extracting the color-changing degree sequence, brightness distribution, and color shift value includes: The tempered film sample was continuously photographed under various simulated environments using an optical sensor to obtain the color-changing image sequence. The color-changing image sequence is subjected to denoising and standardization processing to obtain a standardized image sequence; Based on the brightness value and preset reference color value at each sampling time in the standardized image sequence, the color change degree sequence, the brightness distribution, and the color offset value are determined.

3. The method according to claim 1, characterized in that, The process of determining key areas based on the brightness distribution and color shift values, and determining the performance fluctuation range in conjunction with the antibacterial agent release rate and environmental variables, includes: Identify color change regions based on the brightness distribution and the color offset value, and determine the color change regions as the key regions; The brightness distribution, color shift value, antibacterial agent release rate, and environmental variables within the key area are correlated with the sampling time of the color-changing image sequence to obtain the regional fluctuation characteristics; The performance fluctuation range is determined based on the regional fluctuation characteristics.

4. The method according to claim 1, characterized in that, When the performance fluctuation range exceeds the fluctuation threshold, a preliminary mapping result is generated based on the color deviation characteristics and antibacterial performance labeling, including: When the performance fluctuation range exceeds the fluctuation threshold, the color deviation feature is extracted from the key area, and a feature vector is constructed based on the color deviation feature; The feature vector is matched with the antibacterial performance label, and the correspondence between the degree of color change and the antibacterial effect is determined by the classification boundary optimization method. The preliminary mapping result is generated based on the correspondence; The antibacterial performance label is determined based on the antibacterial effect test records of the tempered glass film sample.

5. The method according to claim 4, characterized in that, The process of screening, grouping, and consistency screening of the tempered glass samples in the preliminary mapping results yields grouped data, including: Based on the degree of matching between the degree of color change and the antibacterial effect in the preliminary mapping results, relevant samples are selected from the tempered glass film samples; Based on the differences in coating formulations and the variation characteristics of the antibacterial agent release rate, the relevant samples were grouped according to their release process. Calculate the intra-cluster consistency index of the related samples within the same group, and retain the groups whose intra-cluster consistency index is not lower than the consistency threshold to obtain the group data; The intra-cluster consistency index is used to indicate the degree of consistency among related samples within the same group in terms of discoloration degree, antimicrobial agent release rate, and antimicrobial effect.

6. The method according to claim 5, characterized in that, The step of constructing an association matrix and adjusting the weights based on the grouped data, the antimicrobial agent release rate, and the environmental variables to obtain a stable association matrix includes: An initial correlation matrix is ​​constructed based on the degree of color change, antibacterial effect, antibacterial agent release rate, and environmental variables corresponding to the grouped data. The matrix weights in the initial correlation matrix are iteratively adjusted; When the change in matrix weights obtained in two consecutive iterations is less than the stability threshold, the correlation matrix obtained in the current iteration is determined as the stable correlation matrix. The initial correlation matrix is ​​used to represent the correspondence between the degree of discoloration, the release rate of the antibacterial agent, the environmental variables, and the antibacterial effect.

7. The method according to claim 6, characterized in that, The step of fitting a quantization function based on the stable correlation matrix and outputting an antibacterial performance index when the prediction deviation is below a deviation threshold includes: The mapping samples from the degree of discoloration to antibacterial performance are determined based on the stable correlation matrix. Error analysis and function fitting are performed on the mapped samples to obtain the quantization function; The prediction deviation is determined based on the predicted antibacterial performance value output by the quantification function and the measured antibacterial performance value in the antibacterial effect test record. When the prediction deviation is lower than the deviation threshold, the antibacterial performance index is output according to the quantization function.

8. The method according to claim 7, characterized in that, The method further includes: Based on the aforementioned antibacterial performance indicators, an antibacterial effect evaluation result is generated; The antibacterial efficacy evaluation results were cross-checked using historical evaluation data and additional environmental simulation data. When the cross-check results meet the preset consistency conditions, the parameters of the quantization function are adjusted according to the additional environmental simulation data to obtain the optimized antibacterial performance index.

9. The method according to claim 8, characterized in that, The time-predicted antimicrobial stability range includes: A performance trend model is constructed based on the discoloration degree sequence, the historical evaluation data, and the optimized antibacterial performance index. Based on the performance trend model, the future color change trend is obtained through a time-series prediction algorithm; The antibacterial stability range is determined based on the correspondence between the future color change trend and the optimized antibacterial performance indicators.

10. A tempered glass film antibacterial performance monitoring system for the entire process, used to implement the tempered glass film antibacterial performance monitoring method according to any one of claims 1 to 9, characterized in that, The system includes: The image processing module is used to acquire the color change image sequence of tempered film samples with various coating formulations, and extract the color change degree sequence, brightness distribution and color shift value; The fluctuation determination module is used to determine key areas based on the brightness distribution and the color offset value, and to determine the performance fluctuation range by combining the antibacterial agent release rate and environmental variables. The mapping generation module is used to generate preliminary mapping results based on color deviation characteristics and antibacterial performance labels when the performance fluctuation range exceeds the fluctuation threshold. The grouping and filtering module is used to filter the tempered film samples in the preliminary mapping results, group them according to the release process, and perform consistency screening to obtain grouped data. The matrix construction module is used to construct an association matrix and adjust the weights based on the grouped data, the antibacterial agent release rate and the environmental variables to obtain a stable association matrix. The performance prediction module is used to fit a quantization function based on the stable correlation matrix, output antibacterial performance index when the prediction deviation is lower than the deviation threshold, and obtain the antibacterial stability range through time-series prediction.

Images