Glass production line product quality real-time monitoring system and method

Abstract

The invention provides a glass production line product quality real-time monitoring system and method, and relates to the technical field of data monitoring, a plurality of production stages are divided, product quality difference analysis is carried out, abnormal type positioning point information is obtained, stage glass acquisition abnormal images and comprehensive abnormal glass images are obtained, and a real-time monitoring result is obtained. Acquiring a category positioning anomaly coefficient of each abnormal category positioning point, calculating a category positioning anomaly difference value, and further acquiring category positioning anomaly sequence information; the method comprises the following steps of: calculating a comprehensive abnormal coefficient, further obtaining an abnormal difference coefficient, carrying out abnormal difference state analysis, obtaining abnormal difference state judgment information, and obtaining combination influence judgment information according to the abnormal difference state judgment information. And the quality monitoring accuracy and efficiency are improved.

Description

TECHNICAL FIELD

[0001] The application provides a glass production line product quality real-time monitoring system and method, and relates to the technical field of data monitoring, in particular to the technical field of glass production line product quality real-time monitoring. BACKGROUND

[0002] In the field of glass production, product quality control is crucial. Traditional production quality monitoring methods gradually expose limitations, such as complex glass production process, production quality monitoring relying on manual sampling inspection or single index detection, which not only is inefficient, but also easily misses some potential abnormalities. It is difficult to efficiently and accurately determine the source of serious faults caused by multiple types of cumulative minor abnormalities, and it is difficult to accurately assess the impact of abnormalities on product quality. SUMMARY

[0003] The application provides a glass production line product quality real-time monitoring system and method to solve the above problems:

[0004] The application provides a glass production line product quality real-time monitoring system and method, the method comprising:

[0005] S1, dividing the glass production line into multiple production stages, performing product quality difference analysis on each production stage, obtaining abnormal type positioning point information of stage collected glass data of multiple abnormal types, and further obtaining stage glass collection abnormal image;

[0006] S2, obtaining comprehensive abnormal glass image according to stage glass collection abnormal image, obtaining type positioning abnormal coefficient of each abnormal type positioning point, calculating type positioning abnormal difference value, and further obtaining type positioning abnormal sequence information;

[0007] S3, calculating comprehensive abnormal coefficient, further obtaining abnormal difference coefficient, performing abnormal difference state analysis, obtaining abnormal difference state judgment information, and obtaining combined influence judgment information according to abnormal difference state judgment information.

[0008] Further, the system comprises: a stage abnormality analysis module for dividing the glass production line into multiple production stages, performing product quality difference analysis on each production stage, obtaining abnormal type positioning point information of stage collected glass data of multiple abnormal types, and further obtaining stage glass collection abnormal image;

[0009] A type abnormality analysis module is configured to obtain comprehensive abnormal glass image according to stage glass collection abnormal image, obtain type positioning abnormal coefficient of each abnormal type positioning point, calculate type positioning abnormal difference value, and further obtain type positioning abnormal sequence information;

[0010] The comprehensive influence determination module is used for calculating a comprehensive abnormality coefficient, further obtaining an abnormality difference coefficient, performing abnormality difference state analysis, obtaining abnormality difference state determination information, and obtaining combination influence determination information according to the abnormality difference state determination information.

[0011] The present application has the advantages that: by dividing the production line into multiple stages and analyzing, the abnormality type information in each production stage can be accurately located, the problem source in the production process can be quickly found, and the problem solving efficiency is improved.

[0012] The calculation of the type positioning abnormality coefficient and the comprehensive abnormality coefficient realizes quantitative evaluation of abnormality, and the severity of abnormality can be more intuitively and accurately understood.

[0013] By calculating the abnormality difference coefficient and performing combination influence determination, the interaction and synergistic effect between different abnormality types can be analyzed in depth, the situation of only focusing on a single abnormality and ignoring the overall influence is avoided, and the influence of abnormality on glass quality can be more comprehensively evaluated.

[0014] Real-time monitoring of glass production line product quality can be realized, abnormality can be found and warned in time, measures can be taken in time, quality problems and production losses caused by abnormality can be reduced, and production efficiency and product quality stability can be improved.

[0015] By analyzing and mining abnormal data, weak links and potential problems in the production process can be found. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 FIG. 1 is a schematic diagram of a glass production line product quality real-time monitoring method. DETAILED DESCRIPTION

[0017] The preferred embodiments of the present application will be described below with reference to the accompanying drawings, and it should be understood that the preferred embodiments described herein are only used to illustrate and explain the present application, and are not used to limit the present application.

[0018] In an embodiment of the present application, a glass production line product quality real-time monitoring system and method are provided, and the method comprises:

[0019] S1, the glass production line is divided into multiple production stages, product quality difference analysis is performed on each production stage, abnormality type positioning point information of stage collected glass data of multiple abnormality types is obtained, and stage glass collection abnormality images are further obtained;

[0020] S2, comprehensive abnormality glass images are obtained according to the stage glass collection abnormality images, type positioning abnormality coefficients of each abnormality type positioning point are obtained, type positioning abnormality difference values are calculated, and type positioning abnormality sequence information is further obtained;

[0021] S3, calculate a comprehensive abnormality coefficient, further obtain an abnormality difference coefficient, perform abnormality difference state analysis, obtain abnormality difference state judgment information, and obtain combination influence judgment information according to the abnormality difference state judgment information.

[0022] The working principle and technical effects of the above technical solution are as follows: the glass production line is subdivided into multiple production stages, and product quality difference analysis is carried out for each stage. By comparing the data characteristics of normal and abnormal products, the positioning point information of multiple abnormal types in each production stage is determined, and the stage glass collection abnormal image is obtained accordingly, so that the accurate positioning and visual presentation of abnormal conditions in different stages are realized.

[0023] Based on the stage glass collection abnormal image, the type positioning abnormality coefficient of each abnormal type positioning point is extracted, the type positioning abnormality difference value is calculated, and the abnormal types are sorted according to the difference value to generate type positioning abnormality sequence information.

[0024] The comprehensive abnormality coefficient is calculated in combination with the abnormal information of each stage, which reflects the comprehensive abnormality of the abnormal conditions on the entire production line. The abnormality difference coefficient is further calculated, the abnormality difference state is analyzed, and the combination influence judgment information is obtained according to the judgment result. This process quantitatively analyzes and excavates the mutual relationship and combination influence between different abnormal types, and comprehensively evaluates the influence of abnormality on glass product quality.

[0025] By dividing the production line into multiple stages and analyzing, the abnormal type information in each production stage can be accurately positioned, and the problem source in the production process can be quickly found, improving the efficiency of problem solving.

[0026] The calculation of the type positioning abnormality coefficient and the comprehensive abnormality coefficient realizes the quantitative evaluation of abnormal conditions, and more intuitively and accurately understands the severity of abnormality.

[0027] By calculating the abnormality difference coefficient and performing combination influence judgment, the interaction and synergistic influence between different abnormal types can be deeply analyzed, avoiding the situation of only focusing on a single abnormality and ignoring the overall influence, and the influence of abnormality on glass quality can be more comprehensively evaluated.

[0028] Real-time monitoring of glass production line product quality can be realized, abnormal conditions can be found in time and warned, measures can be taken in time, quality problems and production losses caused by abnormality can be reduced, and production efficiency and product quality stability can be improved.

[0029] Through analysis and excavation of abnormal data, weak links and potential problems in the production process can be found.

[0030] In an embodiment of the present application, the S1 comprises:

[0031] The glass production line is divided into multiple stages to obtain multiple production stages;

[0032] Stage standard glass data of each stage of the glass production line is obtained; glass data indicators include stage product images and other environmental quality data, etc.

[0033] Product data of the glass production line is collected in stages to obtain stage collected glass data;

[0034] Product quality difference information of the stage standard glass data and the stage collected glass data is obtained;

[0035] According to the product quality difference information, the stage collected glass data is marked with difference points to obtain stage glass difference data with difference point marking information;

[0036] According to the stage glass difference data, abnormal type information of the stage collected glass data is obtained, and abnormal type positioning point information of the stage collected glass data is determined according to the abnormal type information.

[0037] The working principle and technical effect of the above technical solution are: the glass production line is divided into multiple stages according to the process flow, etc.

[0038] For each production stage, stage standard glass data is determined, which covers stage product images and environmental quality data indicators, etc.

[0039] In each stage of the glass production line, stage collected glass data is collected in real time, the data types are consistent with the standard data, and the feasibility of comparative analysis is ensured. Mainly including image information with quality characteristics;

[0040] The stage standard glass data and the stage collected glass data are compared to find the product quality difference information between the two, and the deviation between the actual production and the standard is determined.

[0041] According to the product quality difference information, the stage collected glass data is marked with difference points to generate stage glass difference data with difference point marking information, so that the difference information is visualized.

[0042] Based on the stage glass difference data, abnormal type information of the stage collected glass data is analyzed, and abnormal type positioning point information is further determined to accurately locate the problem.

[0043] Through the stage-by-stage data collection, comparison and analysis, the specific stage of the abnormal glass production process can be accurately located, the root cause of the problem can be quickly found, and the troubleshooting time is reduced.

[0044] The product quality difference and abnormality can be found in time, the production process and parameters can be adjusted in time, and the targeted improvement measures can be taken, so that the overall quality of the glass product is improved.

[0045] The unreasonable link in the production process can be found by analyzing the production data of each stage, and the production efficiency and resource utilization rate are improved.

[0046] With the aid of data acquisition, analysis and labeling technical means, intelligent monitoring and management of the glass production line are realized, manual intervention is reduced, and the scientificity and accuracy of management are improved.

[0047] In an embodiment of the application, the abnormal type information of the stage glass data is obtained according to the stage glass difference data, and the abnormal type positioning point information of the stage glass data is obtained according to the abnormal type information, comprising:

[0048] According to the preset abnormal type information, the stage glass abnormal data is classified to obtain stage type glass abnormal data;

[0049] According to the stage type glass abnormal data, the stage glass data is divided into data regions according to the abnormal type to obtain stage type abnormal regions;

[0050] The abnormal type positioning point of the stage type abnormal region is obtained as the abnormal type positioning point of the stage type abnormal region; the positioning point only represents symbolic meaning and does not have actual position meaning.

[0051] The abnormal type positioning points of the multiple stage type abnormal regions of the stage glass abnormal type are obtained, the stage glass data is labeled according to the abnormal type positioning points to obtain stage glass abnormal image information with abnormal type positioning point labeling.

[0052] The working principle and technical effect of the above technical solution are as follows: according to the preset abnormal type information, the stage glass abnormal data is classified, the data with the same or similar abnormal characteristics is classified into a category, and the stage type glass abnormal data is obtained. The scope of different abnormal types is determined. The abnormal types include scratches, cracks, stripes, bending and edge defects;

[0053] Based on the stage type glass abnormal data, the stage glass data is divided into data regions according to the abnormal type, and the region corresponding to each abnormal type in the glass data, i.e. the stage type abnormal region, is determined. The specific distribution range of the abnormality in the glass data can be accurately positioned.

[0054] In each stage category abnormal area, a particle is selected as the abnormal category positioning point of the area. The particle is usually a key point that can represent the characteristics of the abnormal area. By determining the particle, the abnormal position can be marked more concisely (without actual position reference significance).

[0055] The abnormal category positioning points of the multiple stage category abnormal areas of the stage glass abnormal category are collected, the positioning points are marked on the stage collected glass data, and finally the stage collected glass abnormal image information with the abnormal category positioning point marking is obtained. Abstract abnormal position information is converted into intuitive image marking.

[0056] Through classification, area division and positioning point determination of abnormal data, the abnormal situation information occurring in the glass production process can be accurately positioned, and the problem solving efficiency is greatly improved.

[0057] The abnormal category positioning points are marked on the stage collected glass abnormal image information, so that the abnormal situation is presented in the form of intuitive image, which facilitates quick understanding of the distribution and characteristics of the abnormality, and key information can be obtained without complex analysis.

[0058] Classifying and marking abnormal data according to categories can establish a systematic abnormal data management system, which facilitates statistics, analysis and traceability of different types of abnormalities.

[0059] Timely discovery and information abnormality can timely adjust the production process and parameters, take targeted measures to solve abnormal problems, and thus improve the production quality and stability of glass products.

[0060] In an embodiment of the present application, S2 comprises:

[0061] Obtain all stage collected glass abnormal images, integrate the abnormal category positioning points of the stage collected glass abnormal images in the final stage collected glass abnormal image, and obtain a comprehensive abnormal glass image;

[0062] Obtain the stage category glass abnormal data corresponding to each abnormal category positioning point information in the comprehensive abnormal glass image;

[0063] Calculate the category positioning abnormality coefficient of each abnormal category positioning point according to the stage category glass abnormal data;

[0064] The calculation formula of the category positioning abnormality coefficient is:

[0065]

[0066] Wherein, loys is the category positioning abnormality coefficient, S si is the real-time collected data of the i-th category, S ipQi is the average value of the stage category glass abnormal data of the i-th category i Qi is the preset weight data of the i-th category

[0067] Obtain the maximum value of the category positioning abnormality coefficient of the multiple abnormal category positioning points, and obtain the category abnormality maximum value

[0068] Obtain the difference value between the category abnormality maximum value and each category positioning abnormality coefficient, and obtain the category positioning abnormality difference value

[0069] Sort the category positioning abnormality difference values of the multiple abnormal category positioning points from small to large to obtain the category positioning abnormality sequence information.

[0070] The working principle and technical effects of the above technical solution are as follows: collect all the glass abnormal images collected in all stages, integrate the abnormal category positioning points in each image into the final stage glass abnormal image to form a comprehensive abnormal glass image. This process integrates the abnormal information of different production stages into the same image, which is convenient for overall observation and analysis.

[0071] For each abnormal category positioning point in the comprehensive abnormal glass image, find its corresponding stage category glass abnormal data. By establishing the association between the positioning point and the original abnormal data.

[0072] According to the stage category glass abnormal data, calculate the category positioning abnormality coefficient of each abnormal category positioning point. Find the maximum value from the multiple category positioning abnormality coefficients, which is the category abnormality maximum value. This maximum value represents the most serious abnormal situation in the entire production process.

[0073] Calculate the difference value between the category abnormality maximum value and each category positioning abnormality coefficient to obtain the category positioning abnormality difference value. The difference value reflects the gap between each abnormal point and the most serious abnormality.

[0074] Sort the category positioning abnormality difference values of the multiple abnormal category positioning points from small to large to form the category positioning abnormality sequence information. The sequence intuitively shows the severity ordering of each abnormal point relative to the most serious abnormality.

[0075] By integrating the positioning points of each stage abnormal image into the comprehensive image, the distribution and situation of all abnormalities in the entire glass production process can be comprehensively and intuitively understood, and key information can be avoided.

[0076] Calculate the category positioning abnormality coefficient to realize the quantitative evaluation of the severity of the abnormality, which can more accurately judge the influence degree of the abnormality on the product quality and the production process.

[0077] Find the category abnormality maximum value to quickly determine the most serious abnormal problem in the production process, which is convenient for priority processing and improves the efficiency of problem solving.

[0078] By calculating the abnormality difference of the category and generating sequence information, the priority order of each abnormal point is determined.

[0079] The abnormal glass image and the related abnormal analysis data are comprehensively analyzed, so as to provide rich information for continuous improvement of the production process.

[0080] In an embodiment of the present application, the S3 comprises:

[0081] The comprehensive abnormality coefficient is calculated by comprehensively analyzing the abnormal glass image information.

[0082] The formula of the comprehensive abnormality coefficient is:

[0083]

[0084] Wherein, zys is the comprehensive abnormality difference coefficient, N is the total number of abnormal categories, loys is the category positioning abnormality coefficient of the i-th category, λ is the correction coefficient, the value range is 0.1-0.3, e is the amount of additional monitored abnormal data, and ep is the existing average amount of abnormal data. i

[0085] The abnormality difference coefficient is calculated according to the category abnormality positioning coefficient and the comprehensive abnormality coefficient.

[0086] The abnormality difference coefficient is the difference between the comprehensive abnormality coefficient and the comprehensive abnormality coefficient of all category abnormality coefficients.

[0087] The abnormality difference coefficient is compared with the preset abnormality difference threshold value to obtain an abnormality difference comparison result.

[0088] According to the abnormality difference comparison result, the abnormality difference state is determined to obtain abnormality difference state determination information.

[0089] According to the abnormality difference state determination information, the combined influence determination information is obtained.

[0090] The working principle and technical effects of the above technical solution are as follows: the comprehensive abnormality coefficient is calculated based on the comprehensive abnormality glass image information. The coefficient is a quantitative expression of the comprehensive severity of the abnormal situation in the entire glass production process.

[0091] The category abnormality positioning coefficient and the comprehensive abnormality coefficient are combined to calculate the abnormality difference coefficient. The same data is obtained through two analysis methods, and the difference between the unified data obtained by the two analysis methods is determined.

[0092] ​The calculated abnormal difference coefficient is compared with a pre-set abnormal difference threshold value. The threshold value is determined according to production experience and other factors. Through comparison, an abnormal difference comparison result is obtained.

[0093] According to the abnormal difference comparison result, the abnormal difference state is determined. If the abnormal difference coefficient is greater than the threshold value, it may be determined that the abnormal difference is large, and it is determined whether further analysis is needed; if it is less than the threshold value, it is determined that the abnormal difference is small, and the production process is relatively stable.

[0094] According to the abnormal difference state determination information, the combined influence of each abnormal point is further analyzed. For example, when multiple abnormal difference states are small abnormal points that appear simultaneously in a specific area of the glass, they may have a synergistic effect on the overall performance of the glass, leading to more serious quality problems, and the comprehensive influence of these abnormal combinations on the glass quality is evaluated.

[0095] The calculation of the abnormal coefficient and the abnormal difference coefficient realizes the accurate quantification of the abnormal situation in the glass production process, accurately understands the severity of the abnormality and the difference, and avoids the uncertainty of subjective judgment.

[0096] By comparing the abnormal difference coefficient with the pre-set threshold value and determining the state, the situation of large abnormal difference in the production process can be found in time, and potential quality risks can be warned in advance.

[0097] According to the combined influence determination information, the interaction of multiple abnormal points can be considered comprehensively, and more reasonable production adjustment schemes can be developed, such as adjusting process parameters and strengthening quality detection, to improve production efficiency and product quality.

[0098] Through continuous monitoring and analysis of abnormal situations, abnormal differences can be found and handled in time, which can maintain the stability of the production process, reduce production interruptions and quality fluctuations caused by abnormal situations, and improve the continuity and reliability of production.

[0099] In an embodiment of the present application, the combined influence determination information is obtained according to the abnormal difference state determination information, comprising:

[0100] When the abnormal difference state determination information is the difference abnormal state, the comprehensive abnormal coefficient is divided into multiple category coefficients according to the category positioning abnormal sequence, and multiple category division coefficients are obtained;

[0101] The ratio of each category division coefficient to the corresponding category positioning abnormal coefficient is calculated to obtain the same category difference ratio;

[0102] The difference between each two same category difference ratios is calculated to obtain the category influence difference;

[0103] The category influence difference is compared with a preset influence difference threshold to obtain an influence difference comparison result;

[0104] An abnormal influence determination is made on each category according to the influence difference comparison result to obtain abnormal influence determination information, and a combined influence analysis is triggered according to the abnormal influence determination information;

[0105] Combined influence determination information is obtained according to the combined influence analysis data.

[0106] The working principle and technical effects of the above technical solution are as follows: when the abnormal difference state determination information shows a difference abnormal state, the comprehensive abnormal coefficient is segmented according to the previously generated category positioning abnormal sequence. The category positioning abnormal sequence reflects the priority or severity order of different abnormal category positioning points, and the comprehensive abnormal coefficient is divided into multiple category segmentation coefficients according to the sequence, each segmentation coefficient corresponding to an abnormal type.

[0107] For each category segmentation coefficient, the ratio of the category segmentation coefficient to the corresponding category positioning abnormal coefficient is calculated to obtain a same-category difference ratio. This ratio reflects the relative contribution degree or difference degree of the category abnormality in the comprehensive abnormality. The category with an abnormality is processed.

[0108] The difference between each two same-category difference ratios, i.e., the category influence difference, is calculated. By comparing the difference ratios of different category abnormalities, the relative influence difference between them in the comprehensive abnormality can be understood.

[0109] The category influence difference is compared with a preset influence difference threshold, and an abnormal influence determination is made on each category according to the comparison result. If the category influence difference exceeds the threshold, it means that the influence difference of the category abnormality and other category abnormalities in the comprehensive abnormality is large, and the corresponding category can be directly processed.

[0110] According to the abnormal influence determination information, a combined influence analysis is triggered, the interaction and influence between multiple abnormal categories are comprehensively considered, and finally combined influence determination information is obtained.

[0111] Through coefficient segmentation and same-category difference ratio calculation, the abnormal categories that contribute more or have obvious differences in the comprehensive abnormality can be accurately identified, and these key abnormalities can be focused on and processed, improving the efficiency of problem solving.

[0112] The category influence difference is calculated and compared with the threshold, realizing the quantitative analysis of the influence difference of different category abnormalities.

[0113] Triggering the combined influence analysis and obtaining the combined influence determination information can deeply explore the interaction and synergistic influence between multiple abnormal categories, avoiding the situation of only focusing on a single abnormality and ignoring the overall influence, and can more comprehensively evaluate the influence of the abnormality on the glass quality.

[0114] Based on the combined effect determination information, the adverse effects of abnormal combinations on glass quality can be reduced, and the production efficiency and product quality can be improved.

[0115] In one embodiment of the present application, the abnormal effect determination of each category according to the difference comparison result of the influence is obtained, and the combined effect analysis is triggered according to the abnormal effect determination information, which comprises:

[0116] When the category influence difference is less than or equal to the preset influence difference threshold, it is determined that the category has low influence;

[0117] When the category influence difference is greater than the preset influence difference threshold, it is determined that the category has high influence;

[0118] When the abnormal effect determination information is low influence of the category, the combined effect analysis is triggered;

[0119] When the abnormal effect determination information is high influence of the category, the abnormal category determination and adjustment of the corresponding category are performed.

[0120] When the category influence is low, it can be understood that the abnormal factor is not caused by a single category, but by the combination of multiple low-influence categories (two are taken as an example).

[0121] The working principle and technical effects of the above technical solution are as follows: based on the comparison result of the category influence difference and the preset influence difference threshold, the influence of each abnormal category is determined. If the category influence difference is less than or equal to the preset threshold, it is determined that the category has low influence; if the category influence difference is greater than the preset threshold, it is determined that the category has high influence. This rule clearly defines the relative influence degree of different abnormal categories in the comprehensive abnormality.

[0122] When the abnormal effect determination information shows low influence of the category, the combined effect analysis is triggered. This is because when the influence of a single category is low, it may mean that the abnormality is not dominated by a single factor, but caused by the combination and synergistic effect of multiple low-influence categories, so further analysis of the combined effect of these low-influence categories is needed.

[0123] When the abnormal effect determination information is high influence of the category, the abnormal category determination and adjustment of the corresponding category are directly performed. This indicates that the category abnormality has a large contribution to the comprehensive abnormality and is a key factor leading to the abnormal problem, which needs to be processed first to reduce its adverse effects on glass quality.

[0124] When it is determined that the category influence is low, it is assumed that the abnormal factor is caused by the combination of multiple low-influence categories (two are taken as an example). By analyzing the interaction and superposition effect between these low-influence categories, the causes of the comprehensive abnormality are explained.

[0125] By setting the influence difference threshold and performing abnormal influence determination, high influence categories that have greater impact on the comprehensive abnormality can be quickly identified, improving the efficiency of problem solving.

[0126] Triggering the combined influence analysis when determining that the category is low influence can find potential problems that have small individual influence but may have greater impact on glass quality when combined, avoiding the omission of quality problems caused by ignoring these low influence categories.

[0127] For high influence categories, directly perform abnormal category determination and adjustment to take targeted measures to solve problems in time; for low influence categories, develop more reasonable handling schemes through combined influence analysis to avoid blind handling and improve the accuracy and effectiveness of abnormal handling.

[0128] By classifying and processing abnormal categories with different influence degrees, abnormal problems in the production process can be more comprehensively solved, the disturbance of abnormalities to the production process is reduced, and the stability and continuity of production are improved.

[0129] In an embodiment of the present application, the combined influence analysis data is used to obtain combined influence determination information, which includes:

[0130] The sum of the category division coefficients of each two categories is obtained;

[0131] The ratio of the sum of the category division coefficients to the sum of the category positioning abnormality coefficients is calculated to obtain a combined category difference ratio.

[0132] The difference of each two combined category difference ratios is calculated to obtain a combined influence difference.

[0133] The combined influence difference is compared with a preset combined difference threshold to obtain a combined difference comparison result, and each combined category is subjected to combined influence determination according to the combined difference comparison result to obtain combined influence determination information.

[0134] Abnormal combined categories are determined and adjusted according to the combined influence determination information.

[0135] The working principle and technical effects of the above technical solution are: the sum of the category division coefficients of each two categories is obtained, and the contributions of different abnormal categories in the comprehensive abnormality are quantified.

[0136] The ratio of the sum of the category division coefficients to the sum of the corresponding category positioning abnormality coefficients is calculated to obtain a combined category difference ratio. This ratio reflects the relative difference degree of different abnormal category combinations in the comprehensive abnormality, and comprehensively considers the contributions of each abnormal category in the combination and their characteristics in the original abnormal positioning.

[0137] Calculate the difference between the difference ratios of each two combinations to obtain the difference in combined impact. By comparing the difference ratios between different combinations, we can understand the relative impact differences between different abnormal combinations and determine which combinations may have a more significant impact on glass quality.

[0138] The combination impact difference is compared with the preset combination difference threshold, and the combination impact of each combination type is determined based on the comparison result. If the combination impact difference exceeds the threshold, it means that there is a significant difference in impact between the combination and other combinations, and there may be a special abnormal combination situation.

[0139] Based on the combination impact judgment information, the types of abnormal combinations are determined, and it is clear which combinations need to be paid special attention to and handled. Corresponding adjustment measures are taken, such as adjusting process parameters, optimizing production processes, etc., to reduce the adverse effects of abnormal combinations on glass quality.

[0140] By calculating the combination type difference ratio and combination impact difference, we can comprehensively evaluate the comprehensive impact of different abnormal type combinations on glass quality, avoiding the problem of only focusing on a single abnormal type and ignoring the combination effect.

[0141] By comparing the combined impact difference with the preset threshold, we can accurately identify the key abnormal combinations that have a greater impact on glass quality and improve the efficiency of problem solving.

[0142] By determining and adjusting the types of abnormal combinations based on the combination impact judgment information, it is possible to formulate more scientific and reasonable abnormality handling plans, take targeted measures for different abnormal combinations, improve the accuracy and effectiveness of abnormality handling, and reduce unnecessary interference with the production process.

[0143] Timely detection and handling of key abnormal combinations can reduce quality defects of glass products and improve product consistency and stability.

[0144] In one embodiment of the present invention, comparing the combination impact difference with a preset combination difference threshold to obtain a combination difference comparison result, and performing a combination impact determination on each combination type based on the combination difference comparison result to obtain combination impact determination information includes:

[0145] When the combination impact difference is greater than the preset combination difference threshold, the corresponding combination type is judged as having a high combination impact;

[0146] When the combination impact difference is less than or equal to the preset combination difference threshold, a combination low impact determination is performed on the corresponding combination type.

[0147] The working principle and technical effects of the above technical solution are: by comparing the calculated combination influence difference with the preset combination difference threshold value, the influence degree of each abnormal combination type is determined according to the comparison result. The preset combination difference threshold value is determined based on production experience and other factors, and is used to distinguish the high and low limits of the influence degree of different abnormal combinations on glass quality. When the combination influence difference exceeds the threshold value, the combination type is determined as combination high influence; when the combination influence difference does not exceed the threshold value (i.e. less than or equal to), the combination type is determined as combination low influence.

[0148] The key combination (combination high influence) that may have a significant impact on glass quality can be quickly screened out from a large number of abnormal combinations, improving the efficiency of problem solving and avoiding the expansion of quality problems due to untimely processing.

[0149] For abnormal combinations determined as combination low influence, relatively conventional or simplified processing measures can be taken, while for abnormal combinations determined as combination high influence, more resources and efforts are invested for in-depth analysis and processing, realizing reasonable allocation of resources and improving the fine degree of production management.

[0150] For abnormal combinations of different influence degrees, more targeted processing schemes are developed to improve the accuracy and effectiveness of abnormal processing and reduce unnecessary interference to the normal production process.

[0151] Through accurate determination and timely processing of abnormal combinations, the quality fluctuation of glass products caused by abnormal combinations can be reduced, the consistency and stability of products can be improved, and the requirements of customers for product quality can be met.

[0152] In an embodiment of the present application, the system comprises: a stage abnormality analysis module for dividing the glass production line into multiple production stages, performing product quality difference analysis on each production stage, obtaining abnormality type positioning point information of the stage collected glass data of multiple abnormality types, and further obtaining a stage glass collection abnormality image;

[0153] A type abnormality analysis module is used to obtain a comprehensive abnormal glass image according to the stage glass collection abnormality image, obtain a type positioning abnormality coefficient of each abnormality type positioning point, calculate a type positioning abnormality difference value, and further obtain type positioning abnormality sequence information.

[0154] A comprehensive influence determination module is used to calculate a comprehensive abnormality coefficient, further obtain an abnormality difference coefficient, perform abnormality difference state analysis, obtain abnormality difference state determination information, and obtain combination influence determination information according to the abnormality difference state determination information.

[0155] The working principle and technical effects of the above technical solution are: the glass production line is subdivided into multiple production stages, and product quality difference analysis is carried out for each stage. By comparing the data characteristics of normal and abnormal products, the positioning point information of multiple abnormal types in each production stage is determined, and the stage glass collection abnormal image is obtained accordingly, realizing accurate positioning and visual presentation of abnormal conditions in different stages.

[0156] Based on the stage glass collection abnormal image, the type positioning abnormality coefficient of each abnormal type positioning point is extracted, the type positioning abnormality difference value is calculated, and the abnormal types are sorted according to the difference value, and the type positioning abnormality sequence information is generated.

[0157] The comprehensive abnormality coefficient is calculated combined with the abnormal information of each stage, which reflects the comprehensive abnormality of the entire production line. Further calculation of abnormal difference coefficient and analysis of abnormal difference state, according to the judgment result to obtain combination influence judgment information. This process quantitatively analyzes the relationship between different abnormal types and the combination influence, and comprehensively evaluates the influence of abnormality on glass product quality.

[0158] By dividing the production line into multiple stages and analyzing, the abnormal type information in each production stage can be accurately positioned, the problem source in the production process can be quickly found, and the problem solving efficiency can be improved.

[0159] The calculation of type positioning abnormality coefficient and comprehensive abnormality coefficient realizes the quantitative evaluation of abnormality, and more intuitively and accurately understands the severity of abnormality.

[0160] By calculating the abnormal difference coefficient and making combination influence judgment, the interaction and synergistic effect between different abnormal types can be deeply analyzed, avoiding the situation of only focusing on single abnormality and ignoring the overall influence, and the influence of abnormality on glass quality can be more comprehensively evaluated.

[0161] Real-time monitoring of glass production line product quality can be realized, abnormal conditions can be found and warned in time, measures can be taken in time, quality problems and production losses caused by abnormality can be reduced, and production efficiency and product quality stability can be improved.

[0162] Through analysis and mining of abnormal data, weak links and potential problems in the production process can be found.

[0163] Obviously, those skilled in the art can make various modifications and changes to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and changes of the present application belong to the scope of the claims of the present application and their equivalent technologies, the present application also intends to include these modifications and changes.

Claims

1. A method for real-time monitoring of product quality of a glass production line, characterized in that: The method comprises: S1. Divide the glass production line into multiple production stages, analyze the product quality differences in each production stage, obtain the abnormal type location point information of the glass data collected in the stage for multiple abnormal types, and then obtain the abnormal image of the glass collected in the stage; S2. Obtain a comprehensive abnormal glass image based on the abnormal image collected from the staged glass, obtain the type positioning abnormality coefficient of each abnormal type positioning point, calculate the type positioning abnormality difference, and then obtain type positioning abnormality sequence information; S3. Calculate the comprehensive abnormality coefficient, and then obtain the abnormal difference coefficient, perform abnormal difference state analysis, obtain abnormal difference state determination information, and obtain combined impact determination information based on the abnormal difference state determination information.

2. The method for real-time monitoring of product quality of a glass production line according to claim 1, characterized in that: Said S1 comprises: Divide the glass production line into multiple stages to obtain multiple production stages; Obtain stage standard glass data for each stage of the glass production line; Carry out product data collection of glass production lines in stages to obtain stage-collected glass data; Obtain product quality difference information between stage standard glass data and stage collected glass data; Marking the difference points of the glass data collected at the stage according to the product quality difference information to obtain the stage glass difference data with the difference point marking information; According to the stage glass difference data, abnormality type information of the glass data collected in the stage is acquired, and according to the abnormality type information, abnormality type positioning point information of the glass data collected in the determination stage is acquired.

3. The method for real-time monitoring of product quality of a glass production line according to claim 2, characterized in that: The step of acquiring abnormality type information of the glass data collected in the stage according to the glass difference data in the stage, and acquiring abnormality type positioning point information of the glass data collected at the node determined according to the abnormality type information, includes: Classify the glass abnormality data of the stage according to the preset abnormality type information to obtain the glass abnormality data of the stage type; According to the stage type glass abnormal data, the stage collected glass data is divided into data areas of abnormal types to obtain stage type abnormal areas; Obtaining a particle point in the stage type abnormal region as an abnormal type positioning point in the stage type abnormal region; Acquire abnormal type positioning points of multiple stage type abnormal areas of stage glass abnormal types, mark the abnormal type positioning points of stage collected glass data according to the abnormal type positioning points, and obtain stage collected glass abnormal image information with abnormal type positioning point markings.

4. The method for real-time monitoring of product quality of a glass production line according to claim 1, characterized in that: The S2 includes: Acquire abnormal glass images collected at all stages, integrate abnormal type positioning points of the abnormal glass images collected at the stages into the abnormal glass images collected at the final stage, and obtain a comprehensive abnormal glass image; Obtaining stage type glass abnormality data corresponding to each abnormality type positioning point information in the comprehensive abnormal glass image; Calculate the type location anomaly coefficient of each abnormal type location point according to the glass abnormality data of the stage type; Obtain the maximum value of the type positioning anomaly coefficients of multiple abnormal type positioning points to obtain the maximum value of the type anomaly; Obtain the difference between the maximum value of the category anomaly and the location anomaly coefficient of each category to obtain the category location anomaly difference; The type location anomaly differences of multiple abnormal type location points are sorted from small to large to obtain type location anomaly sequence information.

5. The method for real-time monitoring of product quality of a glass production line according to claim 1, characterized in that: The S3 includes: Calculate the comprehensive anomaly coefficient by integrating the abnormal glass image information; The anomaly difference coefficient is calculated based on the type anomaly location coefficient combined with the comprehensive anomaly coefficient; Compare the abnormal difference coefficient with the preset abnormal difference threshold to obtain an abnormal difference comparison result; Determine the abnormal difference state according to the abnormal difference comparison result to obtain abnormal difference state determination information; The combined impact determination information is obtained based on the abnormal difference state determination information.

6. The method for real-time monitoring of product quality of a glass production line according to claim 5, characterized in that: The acquiring of combined impact determination information according to abnormal difference state determination information includes: When the abnormal difference state determination information is a difference abnormal state, the comprehensive abnormality coefficient is divided into multiple categories of coefficients according to the category positioning abnormality sequence to obtain multiple category division coefficients; The ratio of the segmentation coefficient of each species to the positioning anomaly coefficient of the corresponding species was calculated to obtain the difference ratio of the same species; Calculate the difference between the difference ratios of each two species to obtain the species impact difference; Comparing the category impact difference with a preset impact difference threshold to obtain an impact difference comparison result; Performing abnormal impact determination on each category according to the impact difference comparison result to obtain abnormal impact determination information, and triggering combined impact analysis according to the abnormal impact determination information; The combined impact determination information is obtained based on the combined impact analysis data.

7. The method for real-time monitoring of product quality of a glass production line according to claim 6, characterized in that: The abnormal impact determination is performed on each category according to the impact difference comparison result, and abnormal impact determination information is obtained, and combined impact analysis is triggered according to the abnormal impact determination information, including: When the category impact difference is less than or equal to the preset impact difference threshold, the category is determined to have low impact; When the category impact difference is greater than the preset impact difference threshold, the category is determined to have high impact; When the abnormal impact judgment information is low impact, the combined impact analysis is triggered; When the abnormal impact determination information indicates that the category has a high impact, abnormal category determination and adjustment are performed on the corresponding category.

8. The method for real-time monitoring of product quality of a glass production line according to claim 6, characterized in that: The obtaining of combined impact determination information according to the combined impact analysis data includes: Get the sum of the category splitting coefficients of every two categories; Calculating the ratio of the sum of the category segmentation coefficients to the sum of the category positioning anomaly coefficients to obtain a combined category difference ratio; Calculate the difference between the type difference ratios of each two combinations to obtain the combination impact difference; Comparing the combination impact difference with a preset combination difference threshold to obtain a combination difference comparison result, and performing a combination impact determination on each combination type according to the combination difference comparison result to obtain combination impact determination information; Determine and adjust the abnormal combination type based on the combination impact judgment information.

9. The method for real-time monitoring of product quality of a glass production line according to claim 8, characterized in that: The step of comparing the combination impact difference with a preset combination difference threshold to obtain a combination difference comparison result, and performing a combination impact determination on each combination type according to the combination difference comparison result to obtain combination impact determination information includes: When the combination impact difference is greater than the preset combination difference threshold, the corresponding combination type is judged as having a high combination impact; When the combination impact difference is less than or equal to the preset combination difference threshold, a combination low impact determination is performed on the corresponding combination type.

10. A real-time monitoring system for product quality of a glass production line, characterized in that: The system includes: a stage abnormality analysis module for dividing the glass production line into multiple production stages, performing product quality difference analysis on each production stage, obtaining abnormality type location point information of multiple abnormal types of stage-collected glass data, and then obtaining stage glass collection abnormality images; The type anomaly analysis module is used to obtain a comprehensive abnormal glass image based on the abnormal image collected by the stage glass, obtain the type positioning anomaly coefficient of each abnormal type positioning point, calculate the type positioning anomaly difference, and then obtain the type positioning anomaly sequence information; The comprehensive impact determination module is used to calculate the comprehensive abnormality coefficient, thereby obtaining the abnormal difference coefficient, performing abnormal difference state analysis, obtaining abnormal difference state determination information, and obtaining combined impact determination information based on the abnormal difference state determination information.

Images