A glass production line cold end stress detection system and method

Abstract

The application provides a glass production line cold end stress detection system and method. It belongs to the technical field of glass production detection. The method comprises the following steps: collecting stress-related spectral data of the cold end region of the glass production line through a multispectral stress sensing array, and preprocessing the collected spectral data to obtain a pretreated spectral data set; extracting spectral features from the pretreated spectral data set to construct a spectral feature vector set; based on the spectral feature vector set, extracting the time-space features of the glass cold end stress by using a time-space correlation analysis algorithm to obtain a stress time-space feature matrix; and collecting spectral data through the multispectral stress sensing array, which avoids the influence of traditional contact detection on the glass product and the production line, and improves the detection accuracy and real-time performance.

Description

Technical Field

[0001] This invention proposes a cold-end stress detection system and method for glass production lines, belonging to the field of glass production testing technology. Background Technology

[0002] In the glass manufacturing process, the cold end stage is crucial to glass quality. Stress is generated during the cooling process; uneven or excessive stress distribution can lead to cracks and breakage, severely impacting product quality and production efficiency. Current methods for detecting cold end stress in glass suffer from low accuracy, inability to comprehensively reflect stress distribution in real time, and difficulty in accurately locating areas of stress anomalies, failing to meet the demands of modern glass production for high-quality and efficient testing. Therefore, developing an efficient and accurate method for detecting cold end stress in glass production lines is of significant practical importance. Summary of the Invention

[0003] This invention provides a cold-end stress detection system and method for glass production lines to solve the problems mentioned in the background section above:

[0004] The present invention proposes a method for detecting cold-end stress in a glass production line, characterized in that the method comprises:

[0005] S1: Stress-related spectral data of the cold end area of ​​the glass production line is collected by a multispectral stress sensing array, and the collected spectral data is preprocessed to obtain a preprocessed spectral dataset.

[0006] S2: Extract spectral features from the preprocessed spectral dataset and construct a spectral feature vector set; based on the spectral feature vector set, use a spatiotemporal correlation analysis algorithm to extract the spatiotemporal features of the glass cold end stress and obtain the stress spatiotemporal feature matrix;

[0007] S3: Collect historical data on anomaly stress at the cold end of glass, perform feature clustering analysis on the spatiotemporal feature matrix of stress to obtain multiple stress feature clusters; combine the historical data on anomaly stress at the cold end of glass, perform anomaly feature matching on each stress feature cluster, screen out feature clusters with potential stress anomalies, and extract the stress anomaly feature vectors of the feature clusters.

[0008] S4: Based on historical glass cold end stress anomaly case data, construct a stress anomaly pattern feature library. Based on the stress anomaly feature vector and the stress anomaly pattern feature library, use a pattern matching algorithm to identify stress anomaly patterns and obtain a candidate stress anomaly pattern list. Calculate the similarity of each pattern in the candidate stress anomaly pattern list to generate a stress anomaly similarity vector.

[0009] S5: Collect data on stress-related factors at the cold end of the glass production line and construct a stress-related factor dataset; based on historical data of abnormal stress at the cold end of the glass and the stress-related factor dataset, use machine learning algorithms to establish a stress anomaly probability prediction model and obtain an initial stress anomaly probability vector; dynamically correct the initial stress anomaly probability vector based on the stress-related factor dataset to obtain stress anomaly probability distribution data.

[0010] S6: Based on the probability distribution data of stress anomalies, the candidate stress anomaly pattern list is probabilistically sorted to obtain a stress anomaly pattern ranking list; equipment layout data and glass transport path data of the glass production line are obtained; based on the stress anomaly pattern ranking list, equipment layout data, and glass transport path data, stress anomaly area location analysis is performed to obtain stress anomaly area location result data; based on the stress anomaly area location result data and the stress anomaly pattern feature library, stress anomaly cause analysis is performed to generate a cold end stress detection report for the glass production line.

[0011] The present invention proposes a cold-end stress detection system for a glass production line, comprising a memory, a processor, and a computer program stored in the memory and capable of running on the memory. The processor executes the program to implement a cold-end stress detection method for a glass production line as described above.

[0012] The beneficial effects of this invention are: by acquiring spectral data through a multispectral stress sensing array, the impact of traditional contact testing on glass products and production lines is avoided, and the testing accuracy and real-time performance are improved.

[0013] By combining spectral feature extraction, spatiotemporal correlation analysis, and feature clustering algorithms, abnormal patterns in the stress distribution at the cold end of glass can be automatically identified, improving the intelligent judgment capability of the detection system.

[0014] By analyzing and modeling historical anomaly cases, an anomaly pattern library is formed, which includes spectral features, stress distribution features, and defect information. This provides a reliable basis for subsequent pattern matching and enhances the system's learning and adaptability.

[0015] By using pattern matching algorithms and similarity calculations, it is possible to effectively identify whether the current stress state matches a known anomaly pattern and quantify its similarity, thereby improving the accuracy and robustness of anomaly identification.

[0016] By introducing influencing factors such as process parameters and environmental parameters, and combining them with machine learning algorithms, a stress anomaly probability prediction model is established, which can be dynamically adjusted based on real-time data to improve the scientific validity and practicality of the prediction.

[0017] By combining equipment layout and glass transport path data, the specific area where stress anomalies occur can be accurately located, and cause analysis can be performed based on the feature library, which helps with rapid response and fault diagnosis.

[0018] The final stress test report not only includes anomaly information, but also provides cause analysis and regional location, providing strong data support for production process adjustment, equipment maintenance and quality control. Attached Figure Description

[0019] Figure 1 This is a diagram illustrating the steps of the method described in this invention;

[0020] Figure 2 For the present invention Figure 1 S2 step diagram. Detailed Implementation

[0021] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0022] One embodiment of the present invention, such as Figure 1 As shown, a method for detecting cold-end stress in a glass production line includes:

[0023] S1: Stress-related spectral data of the cold end area of ​​the glass production line is collected by a multispectral stress sensing array, and the collected spectral data is preprocessed to obtain a preprocessed spectral dataset.

[0024] S2: Extract spectral features from the preprocessed spectral dataset and construct a spectral feature vector set; based on the spectral feature vector set, use a spatiotemporal correlation analysis algorithm to extract the spatiotemporal features of the glass cold end stress and obtain the stress spatiotemporal feature matrix;

[0025] S3: Collect historical data on anomaly stress at the cold end of glass, perform feature clustering analysis on the spatiotemporal feature matrix of stress to obtain multiple stress feature clusters; combine the historical data on anomaly stress at the cold end of glass, perform anomaly feature matching on each stress feature cluster, screen out feature clusters with potential stress anomalies, and extract the stress anomaly feature vectors of the feature clusters.

[0026] S4: Based on historical glass cold end stress anomaly case data, construct a stress anomaly pattern feature library. Based on the stress anomaly feature vector and the stress anomaly pattern feature library, use a pattern matching algorithm to identify stress anomaly patterns and obtain a candidate stress anomaly pattern list. Calculate the similarity of each pattern in the candidate stress anomaly pattern list to generate a stress anomaly similarity vector.

[0027] S5: Collect data on stress-related factors at the cold end of the glass production line and construct a stress-related factor dataset; based on historical data of abnormal stress at the cold end of the glass and the stress-related factor dataset, use machine learning algorithms to establish a stress anomaly probability prediction model and obtain an initial stress anomaly probability vector; dynamically correct the initial stress anomaly probability vector based on the stress-related factor dataset to obtain stress anomaly probability distribution data.

[0028] S6: Based on the probability distribution data of stress anomalies, the candidate stress anomaly pattern list is probabilistically sorted to obtain a stress anomaly pattern ranking list; equipment layout data and glass transport path data of the glass production line are obtained; based on the stress anomaly pattern ranking list, equipment layout data, and glass transport path data, stress anomaly area location analysis is performed to obtain stress anomaly area location result data; based on the stress anomaly area location result data and the stress anomaly pattern feature library, stress anomaly cause analysis is performed to generate a cold end stress detection report for the glass production line.

[0029] The working principle of the above technical solution is as follows: a multispectral stress sensing array is used to collect spectral data of the cold end area of ​​the glass production line. These data are closely related to the stress state of the glass. The collected raw spectral data is preprocessed to eliminate noise, enhance the signal, etc., to obtain a higher quality preprocessed spectral dataset, which provides a reliable data foundation for subsequent analysis.

[0030] Spectral features are extracted from the preprocessed spectral dataset to construct a set of spectral feature vectors. These feature vectors reflect different aspects of the cold-end stress of the glass. Using a spatiotemporal correlation analysis algorithm, the spatiotemporal features of the cold-end stress of the glass are extracted from the set of spectral feature vectors to form a stress spatiotemporal feature matrix. This step helps to understand the distribution and variation patterns of stress in time and space.

[0031] Historical data on anomaly stress at the cold end of glass was collected, including spectral characteristics, stress distribution, and corresponding glass defect information during the anomaly, providing a reference for anomaly feature matching. Feature clustering analysis was performed on the spatiotemporal feature matrix of stress to obtain multiple stress feature clusters. Combining the historical case data, anomaly feature matching was performed on each cluster to screen out feature clusters with potential stress anomalies, and stress anomaly feature vectors of these clusters were extracted.

[0032] Based on historical case data, a stress anomaly pattern feature library is constructed, which contains spectral features, stress distribution features, and glass defect features of different stress anomaly patterns. Using a pattern matching algorithm, a list of candidate stress anomaly patterns is identified based on the stress anomaly feature vector and the stress anomaly pattern feature library. The similarity of each pattern in the candidate pattern list is calculated to generate a stress anomaly similarity vector, which is used to evaluate the degree of matching between each pattern and the actual stress anomaly.

[0033] Data on stress-related factors, such as process parameters and environmental parameters, are collected at the cold end of the glass production line to construct a stress-related factor dataset. Using machine learning algorithms, a stress anomaly probability prediction model is established based on historical case data and the stress-related factor dataset to obtain an initial stress anomaly probability vector. The initial probability vector is then dynamically corrected based on the stress-related factor dataset to obtain more accurate stress anomaly probability distribution data.

[0034] Based on the probability distribution data of stress anomalies, the candidate stress anomaly pattern list is probabilistically sorted to obtain a stress anomaly pattern ranking list, which facilitates prioritizing the handling of high-probability anomaly patterns. Equipment layout data and glass transport path data of the glass production line are acquired, and combined with the stress anomaly pattern ranking list, stress anomaly area location analysis is performed to determine the specific location of stress anomalies. Based on the stress anomaly area location results and the stress anomaly pattern feature library, stress anomaly cause analysis is conducted, generating a cold-end stress detection report for the glass production line, providing decision support for production line maintenance and optimization.

[0035] The effect of the above technical solution is that, through multispectral stress sensing array and spatiotemporal correlation analysis algorithm, stress data in the cold end area of ​​glass production line can be captured and analyzed more accurately, reducing the omissions or misjudgments caused by the shortcomings of traditional detection methods.

[0036] By using machine learning and pattern matching algorithms, stress anomaly patterns can be automatically identified, significantly reducing reliance on manual inspection and judgment and improving detection efficiency.

[0037] By combining historical data and stress-related factor data, a stress anomaly probability prediction model can be established to predict potential stress anomalies at the cold end of the glass in advance, allowing for preventative measures to be taken and thus reducing glass defects and downtime risks during production.

[0038] By accurately locating areas of abnormal stress and analyzing the causes of these abnormalities, we can improve production processes and optimize equipment layout, thereby enhancing the stability of glass production lines and ensuring more reliable glass quality.

[0039] By promptly identifying and locating stress anomalies, measures can be taken early to prevent the production of substandard glass, thereby reducing defect rates and production losses.

[0040] This method comprehensively considers influencing factors such as environmental parameter data, and can dynamically adjust the stress anomaly probability prediction model to better adapt to production changes under different environmental conditions.

[0041] By performing cluster analysis and pattern recognition on stress characteristics, data support can be provided for subsequent process optimization, and more scientific production adjustment strategies can be formulated based on historical cases and actual production conditions.

[0042] In one embodiment of the present invention, S1 includes:

[0043] S11. In the critical area of ​​the cold end of the glass production line, a multispectral stress sensor array is uniformly arranged; and the arranged multispectral stress sensor array is calibrated; its spectral response is measured under different stress conditions, and a calibration curve between spectral response and stress is established.

[0044] S12. During the glass operation, the multispectral stress sensing array collects stress-related spectral data of the glass surface in real time; the collected spectral data is transmitted to the data processing system in real time for preliminary storage and management.

[0045] S13. Denoise the collected raw spectral data; normalize the denoised raw spectral data; then smooth the normalized spectral data to obtain the preprocessed spectral dataset.

[0046] The working principle of the above technical solution is as follows: Multispectral stress sensor arrays are uniformly arranged in key areas of the cold end of the glass production line, such as cooling rollers and conveyor belts. These areas are where glass stress changes are significant, and arranging the sensor array ensures accurate capture of stress changes on the glass surface. The arranged multispectral stress sensor array is calibrated to ensure the accuracy of its measurement results. The calibration process includes measuring the spectral response of the sensor array under different stress conditions and establishing a calibration curve between the spectral response and stress. This curve will serve as the basis for subsequent conversion between spectral data and stress.

[0047] The glass production line is started, allowing the glass to pass through the cold end area at normal production speed. During the glass's operation, a multispectral stress sensing array collects stress-related spectral data of the glass surface in real time. The collected spectral data is transmitted to the data processing system in real time for initial storage and management. This step ensures the timeliness and integrity of the data, providing a foundation for subsequent data processing and analysis.

[0048] The acquired raw spectral data undergoes denoising to eliminate noise and interference. Denoising improves the signal-to-noise ratio, making subsequent analysis more accurate. The denoised raw spectral data is then normalized. Normalization unifies data from different magnitudes to the same scale, facilitating subsequent comparison and analysis.

[0049] The normalized spectral data is smoothed to further eliminate minor fluctuations and noise. Smoothing makes the spectral data smoother, improving its stability and reliability. After these processing steps, a preprocessed spectral dataset is obtained. This dataset contains denoised, normalized, and smoothed spectral data, providing a high-quality data foundation for subsequent feature extraction and stress analysis.

[0050] The effect of the above technical solution is that by uniformly arranging a multispectral stress sensing array in the critical area of ​​the cold end and calibrating it, the spectral response under different stress conditions is ensured to be accurate and stable, thereby providing high-quality stress data.

[0051] By performing denoising, normalization, and smoothing of spectral data, the interference from environmental noise, equipment errors, and other factors is effectively reduced, making the final preprocessed spectral dataset more accurate and facilitating subsequent analysis.

[0052] During the glass production process, spectral data is collected and transmitted to the data processing system in real time, ensuring continuous monitoring of the stress condition in the cold end area of ​​the glass and timely detection of possible anomalies.

[0053] This method monitors the stress state of the cold end of the glass in real time and ensures that data acquisition does not interfere with the production process, which helps to make accurate detection without affecting production efficiency and optimizes the production process.

[0054] The automated acquisition and data processing of multispectral stress sensing arrays reduces the complexity of manual operation, improves the automation level of the production line, and reduces the risks associated with human operation.

[0055] The collected spectral data is transmitted and stored in the data processing system in real time, establishing a systematic data management and storage method, which facilitates subsequent data analysis and traceability, and provides reliable historical data support.

[0056] By following standardized data preprocessing steps, the reliability of the data is ensured in subsequent spectral feature extraction and spatiotemporal feature analysis, which helps to improve the accuracy of the final stress detection results.

[0057] One embodiment of the present invention, such as Figure 2 As shown, S2 includes:

[0058] S21. Extract stress-related spectral features from the preprocessed spectral dataset; perform dimensionality reduction on the extracted spectral features using a feature selection algorithm; construct a spectral feature vector set;

[0059] S22. Using a spatiotemporal correlation analysis algorithm, the spectral feature vector set is processed to extract the spatiotemporal characteristics of the glass cold end stress.

[0060] S23. Organize and store the extracted spatiotemporal features according to a preset format to construct a stress spatiotemporal feature matrix, where the rows of the matrix represent different time points or spatial locations, the columns represent different spatiotemporal features, and the elements in the matrix represent the spatiotemporal feature values ​​corresponding to the time and spatial location.

[0061] The working principle of the above technical solution is as follows: Stress-related spectral features are extracted from the preprocessed spectral dataset. These features include, but are not limited to, spectral peaks, peak and valley positions, spectral bandwidth, and spectral integral intensity. These features reflect different aspects of stress on the glass surface and are crucial for subsequent stress analysis. A feature selection algorithm is used to reduce the dimensionality of the extracted spectral features. Since the original spectral features may contain a large amount of redundant information, dimensionality reduction removes these redundant features, retaining only the most critical features for stress analysis, thereby improving the efficiency and accuracy of subsequent analysis. The dimensionality-reduced spectral features are organized according to a specific format to construct a spectral feature vector set. This vector set will serve as the foundational data for subsequent spatiotemporal correlation analysis.

[0062] Spatiotemporal correlation analysis algorithms, such as spatiotemporal clustering algorithms and spatiotemporal Markov models, are used to process the spectral feature vector set. These algorithms can uncover the temporal and spatial correlations of spectral features, thereby extracting the spatiotemporal characteristics of glass cold-end stress. The extracted spatiotemporal characteristics reflect the distribution patterns and trends of stress in time and space. These characteristics are of great significance for understanding the dynamic behavior of glass cold-end stress, predicting stress anomalies, and optimizing production processes.

[0063] The extracted spatiotemporal features are organized and stored according to a preset format. This step ensures the orderliness and accessibility of the spatiotemporal features, facilitating subsequent analysis and processing. Based on the organized and stored spatiotemporal features, a stress spatiotemporal feature matrix is ​​constructed. In this matrix, rows represent different time points or spatial locations, columns represent different spatiotemporal features, and the elements in the matrix represent the spatiotemporal feature values ​​for the corresponding time and spatial location. This matrix provides important data support for subsequent stress anomaly pattern recognition, probability prediction, and regional localization.

[0064] The effect of the above technical solution is that by extracting stress-related spectral features, stress information in the cold end region of glass can be captured more accurately, providing more representative data for subsequent analysis.

[0065] The extracted spectral features are reduced by using a feature selection algorithm, which reduces redundant data and makes data processing more efficient. At the same time, it avoids the computational burden and overfitting problems caused by high-dimensional data.

[0066] By using spatiotemporal correlation analysis algorithms, the spatiotemporal characteristics of glass cold-end stress can be extracted more effectively, and the distribution patterns and trends of stress in time and space can be comprehensively analyzed.

[0067] By extracting and analyzing spatiotemporal features, we can gain a deeper understanding of the dynamic changes and distribution characteristics of stress at the cold end of glass, helping to identify potential production problems and anomalies.

[0068] By employing automated spatiotemporal correlation analysis algorithms, the reliance on manual intervention is reduced, making the entire data processing workflow more intelligent and efficient, and minimizing the impact of human factors on the results.

[0069] The extracted spatiotemporal features were organized and stored according to a preset format, and a stress spatiotemporal feature matrix was constructed, which facilitates subsequent data query, analysis and decision-making, and improves the systematicness and operability of data management.

[0070] By constructing a spatiotemporal feature matrix, we can provide accurate data support for subsequent stress anomaly prediction and process optimization, and help to formulate more scientific and targeted production adjustment strategies.

[0071] In one embodiment of the present invention, step S23 includes:

[0072] Based on the physical meaning and analytical requirements represented by the spatiotemporal characteristics, they are divided into time-dimensional characteristics and spatial-dimensional characteristics, and an identifier is defined for each type of characteristic.

[0073] For spatiotemporal feature values ​​with different dimensions and ranges, a unified standardization method is used to process them, mapping the feature values ​​to specific numerical intervals.

[0074] Construct a time-space indexing system; in the time dimension, assign a unique index number to each time point in chronological order; in the spatial dimension, assign a corresponding index identifier to each spatial location based on the actual spatial layout of the glass cold end;

[0075] The dimensions of the stress spatiotemporal characteristic matrix in dynamic programming are determined as follows: in the time dimension, the number of rows of the matrix is ​​determined based on the time span of data acquisition and the time resolution requirements of analysis; in the spatial dimension, the number of columns of the matrix is ​​determined based on the degree of spatial division of the glass cold end.

[0076] Based on the time-space indexing system, the standardized spatiotemporal feature values ​​are filled into the corresponding positions of the matrix one by one; for missing data, interpolation methods are used for estimation and filling; after filling all feature values, a complete stress spatiotemporal feature matrix is ​​formed.

[0077] The quality of the constructed stress spatiotemporal feature matrix is ​​verified, and corrections and optimizations are made based on the verification results.

[0078] The stress spatiotemporal feature matrix, after quality verification and optimization, is stored in a specified database or file system, while key information of the matrix is ​​recorded to form a complete metadata record.

[0079] The working principle of the above technical solution is as follows: based on the physical meaning represented by the spatiotemporal characteristics (such as the rate of change of stress over time reflecting the dynamic characteristics of stress, and the spatial distribution gradient reflecting the non-uniformity of stress in space) and the analysis requirements (such as whether to focus on periodic fluctuations or local accumulation of stress), the spatiotemporal characteristics are divided into time-dimensional characteristics and spatial-dimensional characteristics. An identifier is defined for each type of characteristic to facilitate accurate identification and reference of different characteristics in subsequent matrix construction, data processing, and analysis.

[0080] Different spatiotemporal features may have different dimensions and value ranges. Directly using the original eigenvalues ​​for matrix construction and analysis can lead to inaccurate results or difficulty in comparison. By using a unified standardization method (such as max-min standardization to map eigenvalues ​​to the [0,1] interval) to process the eigenvalues, the influence of differences in dimensions and value ranges can be eliminated, making different features comparable in the matrix.

[0081] In the time dimension, a unique index number is assigned to each time point in chronological order, ensuring that each time point has a unique and definite position in the matrix, facilitating the analysis and tracking of stress characteristics over time. Based on the actual spatial layout of the glass cold end, grid partitioning and region coding are used to assign corresponding index identifiers to each spatial location, enabling accurate positioning of the spatial location within the matrix and facilitating the analysis of stress distribution and changes in different spatial regions. This indexing system serves as the positioning basis for matrix rows and columns, ensuring that spatiotemporal feature values ​​are accurately placed in their corresponding positions within the matrix, guaranteeing the correspondence between the matrix structure and the actual spatiotemporal feature data.

[0082] The number of rows in the matrix is ​​determined based on the time span of data acquisition and the required time resolution for analysis. The time span determines the time range covered by the matrix, while the required time resolution affects the accuracy of the representation of each time point in the matrix. The number of columns in the matrix is ​​determined based on the fineness of the spatial division of the glass cold end. The finer the spatial division, the more columns the matrix has, allowing for a more detailed reflection of the spatial distribution of stress. Sufficient expansion space is reserved to accommodate potential increases in data volume or new spatiotemporal feature analysis needs in the future, improving the matrix's adaptability and flexibility. Based on the time-space indexing system, standardized spatiotemporal feature values ​​are filled into the corresponding positions in the matrix, ensuring that the matrix fully represents the characteristics of stress at different times and spatial locations. For potential data gaps at certain time points or spatial locations, appropriate interpolation methods (such as linear interpolation, spline interpolation, etc.) are used for estimation and filling to ensure the integrity and continuity of the matrix data and avoid the impact of missing data on subsequent analysis results.

[0083] The constructed stress-spatiotemporal feature matrix undergoes quality verification to check for outliers (such as values ​​outside the reasonable range). Outliers are removed or replaced to ensure the accuracy and reliability of the matrix data. Inconsistencies are checked (such as contradictory feature values ​​at the same time point or spatial location), and any inconsistencies are traced and adjusted to ensure data logic and consistency. Efficient storage formats (such as sparse matrix storage, suitable when the matrix contains a large number of zero elements) and compression algorithms are employed to reduce storage space usage, improve data read / write efficiency, and facilitate subsequent storage, management, and analysis of the matrix data.

[0084] The stress spatiotemporal characteristic matrix, after quality verification and optimization, is stored in a designated database or file system to ensure data security and accessibility. Key information such as the matrix's construction time, data source, and processing parameters are recorded to form a complete metadata record. This metadata record facilitates subsequent understanding, traceability, and reuse of the matrix data, and allows users to understand the matrix's construction process and related background information.

[0085] The effect of the above technical solution is that by dividing spatiotemporal features into time-dimensional features and spatial-dimensional features, and defining an identifier for each type of feature, it is possible to more clearly analyze and understand the performance of stress changes in different dimensions, thereby enhancing the interpretability of the data.

[0086] By employing a unified standardized method to process spatiotemporal feature values ​​with different dimensions and ranges, a uniform scale for the feature values ​​is ensured, effectively avoiding deviations caused by differences in dimensions between different features, thereby improving the consistency of data analysis.

[0087] By constructing a complete time-space indexing system, the accurate positioning of spatiotemporal feature values ​​is ensured, the misalignment or chaos of data in the matrix is ​​avoided, and the systematicness and accuracy of data management are improved.

[0088] When dynamically planning the spatiotemporal feature matrix dimensions, the time span and spatial partitioning granularity are considered, and expansion space is reserved to ensure that the system can cope with future increases in data volume or new analysis needs, thereby improving the system's flexibility and future adaptability.

[0089] For time points or spatial locations where data is missing, reasonable interpolation methods are used to estimate and fill the missing data, which effectively reduces the impact of missing data on the quality of the spatiotemporal feature matrix and ensures the integrity and continuity of the data.

[0090] By performing quality verification on the constructed stress spatiotemporal characteristic matrix, outliers or data inconsistencies can be detected and corrected in a timely manner, thereby improving the accuracy and reliability of the data and ensuring the effectiveness of the analysis results.

[0091] By optimizing the matrix storage structure and adopting a sparse matrix storage format and compression algorithm, the storage space occupied is reduced, the data read and write efficiency is improved, and the performance and response speed of large-scale data processing are further enhanced.

[0092] The optimized stress spatiotemporal feature matrix is ​​stored in a specified database or file system, and key information such as matrix construction time, data source and processing parameters are recorded to form a complete metadata record, ensuring that the data is traceable and manageable.

[0093] In one embodiment of the present invention, S3 includes:

[0094] S31. Collect data on stress anomalies at the cold end of the glass production line over a period of time, establish a historical database of stress anomalies at the cold end of the glass; and organize and classify the collected historical case data.

[0095] S32. Use clustering algorithms to perform feature clustering analysis on the stress spatiotemporal feature matrix;

[0096] S33. Based on historical data of abnormal stress at the cold end of glass, perform abnormal feature matching for each stress feature cluster; compare the features of the cluster with the abnormal features in historical cases, and calculate the similarity; the similarity is obtained using the following formula:

[0097]

[0098] Wherein, the cluster feature vector of A is [ , , ..., ]; The historical anomaly feature vector of B is [ , , ..., ]; n represents the feature dimension (such as time, location, spectral features, etc.);

[0099] S34. Set a similarity threshold to filter out feature clusters with potential stress anomalies; for clusters with similarity exceeding the threshold, determine whether there are stress anomalies.

[0100] S35. Extract stress anomaly feature vectors from the selected feature clusters; select the most representative features from the clusters to construct stress anomaly feature vectors.

[0101] The working principle of the above technical solution is as follows: Data on stress anomalies occurring at the cold end of the glass production line over a past period is collected. This data forms a crucial foundation for subsequent analysis. The case data includes basic information such as the time, location, glass specifications, and batch number of the stress anomaly, as well as key data such as the spectral characteristics, stress distribution, and corresponding glass defect information at the time of the anomaly, comprehensively reflecting the various characteristics and impacts of stress anomalies. A historical database of glass cold end stress anomaly cases is established, centrally storing and managing the collected case data for convenient subsequent querying, retrieval, and analysis. The collected historical case data is organized and classified according to the type and severity of the stress anomaly. This classification method helps to quickly locate stress anomaly cases of specific types or severity during subsequent analysis, improving analytical efficiency and focus.

[0102] Clustering algorithms are employed to perform feature clustering analysis on the spatiotemporal feature matrix of stress. Clustering algorithms can group stress spatiotemporal data with similar characteristics together to form different clusters, thereby revealing the inherent patterns and structure of stress spatiotemporal features. Based on the characteristics and distribution patterns of stress spatiotemporal features, the number of clusters and the similarity measurement method are determined. The selection of the number of clusters needs to comprehensively consider the complexity of the data and the analytical requirements, while the similarity measurement method determines how to measure the degree of similarity between different data points. Clustering effectiveness indices, such as the silhouette coefficient and the Davies-Bouldin index, are used to evaluate the quality of clustering. These indices can reflect the quality of clustering results from different perspectives, helping to select the optimal clustering scheme. The characteristics and distribution of each cluster are analyzed to understand the characteristics of different stress spatiotemporal feature patterns. Through in-depth analysis of the clusters, potential patterns and anomalies in stress spatiotemporal features can be discovered, providing a basis for subsequent anomaly feature matching.

[0103] Based on historical data on anomaly stress at the cold end of glass, anomaly feature matching is performed on each stress feature cluster. The features of each cluster are compared with the anomaly features in historical cases, and similarity is calculated. This step aims to determine whether the features of the current cluster are similar to the anomaly features in historical cases, thereby identifying potential stress anomalies. Similarity calculation is a crucial step in anomaly feature matching. A reasonable similarity measurement method can accurately measure the degree of similarity between cluster features and historical anomaly features, providing a basis for subsequent screening and judgment. A similarity threshold is set to filter out feature clusters with potential stress anomalies. The similarity threshold needs to be adjusted according to the actual situation and analysis requirements, ensuring that truly potential anomaly clusters are selected while avoiding false positives. For clusters with similarity exceeding the threshold, it is determined whether stress anomalies exist. This step, based on the similarity matching results, performs a preliminary anomaly judgment on the clusters, providing direction for subsequent anomaly feature vector extraction and analysis.

[0104] Stress anomaly feature vectors are extracted from the selected feature clusters, and the most representative features are chosen from each cluster. These representative features summarize the main characteristics and anomaly patterns of the clusters, providing crucial information for subsequent stress anomaly pattern recognition and analysis. The selected representative features are then used to construct a stress anomaly feature vector, which serves as an important input for subsequent stress anomaly pattern recognition and analysis. By constructing feature vectors, complex spatiotemporal stress characteristics can be simplified into a form that is easier to process and analyze, improving analytical efficiency and accuracy.

[0105] The effect of the above technical solution is that by collecting and organizing historical data on abnormal stress at the cold end of glass, a comprehensive database has been established, providing high-quality reference data for subsequent anomaly analysis and enabling more accurate identification of the patterns of stress anomalies.

[0106] By organizing and classifying historical case data, and dividing them according to the type and severity of stress anomalies, the data processing flow was simplified, making subsequent analysis work more systematic and efficient.

[0107] Clustering algorithms are used to perform clustering analysis on the spatiotemporal feature matrix of stress, which can effectively identify the distribution patterns of different stress features. By using clustering validity indicators, the reliability and accuracy of the clustering results are guaranteed, and the depth and precision of the analysis are improved.

[0108] By performing anomaly feature matching on each stress feature cluster and comparing it with historical cases, potential stress anomaly patterns can be quickly discovered, improving the sensitivity and reliability of anomaly detection.

[0109] By setting a similarity threshold, feature clusters of potential stress anomalies can be filtered out. With reasonable threshold settings, the risk of false positives and false negatives can be effectively reduced, and the accuracy and practicality of anomaly detection can be improved.

[0110] By extracting stress anomaly feature vectors from the selected feature clusters, the most representative features can be selected to accurately extract abnormal features and provide precise basis for subsequent anomaly diagnosis and early warning.

[0111] This solution can handle new types of anomalies or new analytical needs that may arise in the future. With the increase in data volume and the optimization of algorithms, the system has strong scalability and flexibility, and can adapt to production environments of different sizes and complexities.

[0112] By automating the clustering and anomaly feature matching process, the need for manual intervention is reduced, work efficiency is improved, and the probability of human error is lowered, thereby ensuring the stability and production efficiency of the production line.

[0113] By effectively identifying potential stress anomalies and predicting their characteristics, preventative measures can be taken in advance to avoid major failures, thereby improving the overall reliability and production quality of the production line.

[0114] In one embodiment of the present invention, step S4 includes:

[0115] S41. Based on historical glass cold-end stress anomaly case data, construct a stress anomaly pattern feature library; standardize the data in the feature library to unify the dimensions and ranges of different features;

[0116] S42. Select a pattern matching algorithm and optimize the algorithm parameters based on the characteristics and distribution patterns of stress anomaly features; train the algorithm and adjust its parameters using stress anomaly feature vectors and corresponding stress anomaly patterns from historical case data.

[0117] S43. Based on the stress anomaly feature vector and the stress anomaly pattern feature library, use the selected pattern matching algorithm to perform stress anomaly pattern recognition; match the stress anomaly feature vector with each pattern in the feature library and calculate the similarity.

[0118] S44. Based on the similarity calculation results, obtain a list of candidate stress anomaly patterns; perform similarity calculation on each pattern in the list of candidate stress anomaly patterns to generate a stress anomaly similarity vector.

[0119] The working principle of the above technical solution is as follows: A stress anomaly pattern feature library is constructed based on historical glass cold-end stress anomaly case data. This feature library comprehensively covers the spectral features, stress distribution features, and glass defect features corresponding to different stress anomaly patterns. These features describe the manifestations of stress anomalies from different perspectives, providing a rich information foundation for subsequent pattern recognition. Since different features in the feature library may have different dimensions and ranges, directly using the raw data for matching and analysis will lead to inaccurate results. Therefore, the data in the feature library is standardized to unify the dimensions and ranges of different features, making different features comparable in subsequent pattern recognition and improving the accuracy and reliability of the recognition.

[0120] Based on the requirements of stress anomaly pattern recognition, appropriate pattern matching algorithms are selected, such as cosine similarity matching and Euclidean distance matching. These algorithms can measure the similarity between two vectors from different perspectives and are suitable for matching stress anomaly feature vectors. The parameters of the selected algorithm are optimized according to the characteristics and distribution patterns of stress anomaly features. Different parameter settings affect the algorithm's performance and recognition effect; optimizing the parameters allows the algorithm to better adapt to the characteristics of stress anomaly features, improving recognition accuracy and efficiency. The algorithm is trained using stress anomaly feature vectors and corresponding stress anomaly patterns from historical case data. During training, the algorithm's parameters are continuously adjusted to enable it to more accurately identify stress anomaly patterns in historical cases, improving its generalization ability and recognition accuracy.

[0121] Based on stress anomaly feature vectors and a stress anomaly pattern feature library, a selected pattern matching algorithm is used for stress anomaly pattern recognition. The stress anomaly feature vectors are matched with various patterns in the feature library, and similarity is calculated to measure the closeness between the feature vectors and each pattern. Similarity calculation is the core step in pattern recognition; a reasonable similarity measurement method can accurately reflect the similarity between the feature vectors and patterns, providing a basis for subsequent pattern selection and decision-making.

[0122] Based on the similarity calculation results, a list of candidate stress anomaly patterns is obtained. The patterns in the list are sorted from highest to lowest similarity; patterns with higher similarity are more likely to correspond to the current stress state as stress anomalies. This sorting method helps to quickly locate the most probable stress anomaly pattern, improving the efficiency of fault diagnosis. A similarity similarity vector is generated for each pattern in the candidate stress anomaly pattern list. The elements in the vector represent the similarity value between the current stress state and each candidate stress anomaly pattern. These similarity values ​​intuitively reflect the degree of proximity between the current stress state and each candidate pattern, providing important quantitative basis for subsequent fault diagnosis and decision-making.

[0123] The effect of the above technical solution is that by constructing a stress anomaly pattern feature library, the system can more comprehensively store and organize the spectral features, stress distribution features and glass defect features of different stress anomaly patterns, providing detailed and standardized data support for pattern recognition, thereby greatly improving the accuracy of pattern recognition.

[0124] By standardizing the feature library data, the dimensions and ranges of different features are unified, effectively reducing interference between different features, making pattern matching simpler and more efficient, and avoiding processing difficulties caused by data inconsistency.

[0125] By selecting appropriate pattern matching algorithms (such as cosine similarity and Euclidean distance) and optimizing them according to the characteristics and distribution patterns of stress anomalies, it is possible to better adapt to different types of stress anomaly patterns, thereby improving the flexibility and robustness of pattern recognition.

[0126] By optimizing algorithm parameters and training with historical case data, the system can complete stress anomaly pattern recognition in a shorter time, significantly improving processing efficiency and reducing response latency to real-time data.

[0127] By generating a list of candidate stress anomaly patterns based on similarity calculation results and sorting them by similarity, the most similar patterns can be matched first, thereby reducing the risk of misidentification and ensuring that the identified anomaly patterns are more consistent with the current stress state.

[0128] By generating stress anomaly similarity vectors, accurate basis is provided for subsequent fault diagnosis and decision-making, ensuring accurate judgment and timely response to abnormal patterns during the decision-making process, and improving the reliability and effectiveness of the decision-making process.

[0129] By accumulating rich historical data in the feature library and continuously optimizing the pattern matching algorithm, the system can continuously evolve to adapt to new production scenarios and complex changes, thereby improving the long-term maintainability and adaptability of the system.

[0130] The automated stress anomaly pattern recognition process reduces reliance on manual judgment, alleviates operator workload, lowers the rate of human error, and further improves production line stability.

[0131] Through effective pattern matching and similarity calculation, the system can predict and warn of potential stress anomalies in advance, thereby helping to take timely preventive measures, reduce production line downtime, and improve production line operating efficiency and product quality.

[0132] In one embodiment of the present invention, S43 includes:

[0133] The stress anomaly feature vector is pre-aligned with the feature vectors of each mode in the feature library; the current stress anomaly feature vector is rearranged according to the definition and order of the features in the feature library to ensure that the feature order of the two is consistent; for features that exist in the feature library but are missing from the current feature vector, a specific filling strategy is used to supplement them.

[0134] The pre-aligned stress anomaly feature vector and the pattern feature vector in the feature library are split into multiple dimensions according to feature type; for features of different dimensions, corresponding weights are assigned according to their importance in stress anomaly pattern recognition.

[0135] The selected pattern matching algorithm is used to calculate the similarity of the feature vectors in each dimension after splitting. The similarity value under each dimension is obtained by calculating the similarity of each dimension.

[0136] The similarity values ​​calculated by each dimension are fused according to the pre-assigned weights to obtain the comprehensive similarity value between the stress anomaly feature vector and each pattern in the feature library; the fused similarity value is comprehensively evaluated to initially select several candidate patterns with the top 20% similarity.

[0137] Based on the similarity distribution of different stress anomaly patterns in historical data and the characteristics of the current stress state, the similarity threshold is dynamically adjusted. The initially selected candidate patterns are compared with the dynamic threshold, and candidate patterns with similarity higher than the threshold are further selected as the final set of candidate stress anomaly patterns.

[0138] The working principle of the above technical solution is as follows: The stress anomaly feature vector is pre-aligned with the feature vectors of each mode in the feature library. Based on the definition and order of features in the feature library, the current stress anomaly feature vector is rearranged to ensure consistent feature order. This operation ensures that elements at the same position in different feature vectors represent the same feature meaning in subsequent similarity calculations, avoiding calculation errors caused by inconsistent feature order. For features present in the feature library but missing from the current feature vector, a specific imputation strategy (such as imputing the average or median of the feature in historical data) is used. The presence of missing values ​​affects the accuracy of similarity calculations; through a reasonable imputation strategy, the integrity of the feature vector can be restored as much as possible, reducing the impact of missing values ​​on the results.

[0139] The pre-aligned stress anomaly feature vectors and pattern feature vectors from the feature library are split into multiple dimensions according to feature type (spectral features, stress distribution features, and glass defect features). Different feature types have different roles and meanings in stress anomaly pattern recognition. By splitting, features of different dimensions can be analyzed and processed separately, improving the targeting and accuracy of recognition. For features of different dimensions, corresponding weights are assigned according to their importance in stress anomaly pattern recognition. Weight allocation can be determined based on expert experience, historical data analysis, or machine learning algorithms (such as entropy weighting). For example, spectral features may be more critical for the recognition of certain stress anomaly patterns, so they are given higher weights; while glass defect features may have a smaller impact in certain situations, so they are given relatively lower weights. Reasonable weight allocation can highlight the role of important features in similarity calculation and improve the reliability of recognition results. Selected pattern matching algorithms (such as cosine similarity matching, Euclidean distance matching, etc.) are used to calculate the similarity of the split feature vectors of each dimension. For the spectral feature dimension, due to the continuity and complexity of spectral data, further data processing (such as Fourier transform to extract frequency domain features) can be performed on the spectral features first, and then the similarity can be calculated to better capture the key information in the spectral features.

[0140] For the stress distribution feature dimension, considering the spatial distribution characteristics of stress, a spatial distribution-based similarity measurement method can be used, such as calculating the spatial correlation coefficient of stress distribution, to accurately measure the similarity between stress distribution features. For the glass defect feature dimension, appropriate similarity calculation methods are adopted according to the type and degree of the defect, such as encoding the defect type and then calculating the similarity of the encoding, or setting different similarity thresholds according to the degree of the defect. Through dimensional similarity calculation, similarity values ​​for each dimension are obtained, providing a basis for subsequent comprehensive evaluation.

[0141] The similarity values ​​calculated dimensionally are fused according to pre-assigned weights to obtain a comprehensive similarity value between the stress anomaly feature vector and each pattern in the feature library. During the fusion process, different fusion strategies, such as weighted average and weighted geometric average, can be used; the most suitable method is selected based on the actual situation. Simultaneously, considering the correlation between similarities across dimensions, methods such as principal component analysis (PCA) can be used to reduce the dimensionality of the similarity vector, extracting key information before fusion to reduce the impact of redundant information and improve the accuracy and stability of the fusion results. The fused similarity values ​​are comprehensively evaluated, and several candidate patterns in the top 20% of similarity are initially selected. The purpose of this initial selection is to narrow down the candidate range and improve the efficiency of subsequent analysis.

[0142] Based on the similarity distribution of different stress anomaly patterns in historical data and the characteristics of the current stress state (such as stress magnitude and trend), the similarity threshold is dynamically adjusted. Dynamically adjusting the threshold makes the screening process more flexible and accurate, adapting to the stress anomaly pattern recognition needs under different conditions. The initially selected candidate patterns are compared with the dynamic threshold, and candidate patterns with similarity higher than the threshold are further selected as the final set of candidate stress anomaly patterns. The final set of candidate stress anomaly patterns can more accurately reflect the stress anomaly patterns that may correspond to the current stress state, providing strong support for subsequent fault diagnosis and decision-making.

[0143] The effect of the above technical solution is that by performing a pre-alignment operation on the feature vectors, the consistency of the current stress anomaly feature vector with the pattern feature vector in the feature library in terms of feature order is ensured, eliminating the matching error caused by the inconsistency in order, thereby greatly improving the accuracy of pattern matching.

[0144] By employing specific imputation strategies (such as using the mean or median from historical data to imput missing features), pattern recognition failures or misidentifications caused by missing features are avoided, ensuring the integrity of feature data and the stability of recognition.

[0145] By splitting the feature vector into multiple dimensions such as spectral features, stress distribution features, and glass defect features, and calculating the similarity of each dimension separately, the contribution of each feature dimension to the stress anomaly pattern can be captured more meticulously, thereby improving the overall recognition accuracy.

[0146] By assigning weights to features of different dimensions and dynamically adjusting the weights using expert experience or machine learning algorithms (such as the entropy weight method), the influence of features on pattern recognition can be adjusted more flexibly according to the different characteristics of stress anomaly patterns, thereby further improving recognition accuracy.

[0147] By calculating the similarity of features in different dimensions separately and then performing weighted fusion, the complexity of high-dimensional data processing is effectively reduced, interference from redundant information is minimized, and the efficiency of the pattern recognition process is ensured.

[0148] By employing fusion strategies such as weighted average or weighted geometric average, the similarity across different dimensions can be fused to comprehensively consider the influence of each dimension, resulting in a more comprehensive and reasonable similarity value and improving the reliability of pattern recognition.

[0149] By employing dimensionality reduction methods such as principal component analysis (PCA), the similarity vector is dimensionality reduced, which reduces the interference of redundant information, further enhances the system's adaptability to different stress states, and improves the flexibility of pattern recognition.

[0150] By comprehensively evaluating the similarity value after fusion, candidate patterns with high similarity can be accurately screened, reducing the probability of false screening and ensuring that the screened candidate patterns are more in line with the current stress state.

[0151] By dynamically adjusting the similarity threshold, the screening criteria can be flexibly adjusted according to the characteristics of the current stress state, making pattern screening more targeted and improving the accuracy and reliability of pattern recognition.

[0152] By initially screening out the top 20% of candidate patterns based on the similarity distribution, the range of candidate patterns can be quickly narrowed down, improving the efficiency of subsequent diagnosis and decision-making and reducing unnecessary calculations and analysis.

[0153] In one embodiment of the present invention, step S5 includes:

[0154] S51. Collect stress-related factor data at the cold end of the glass production line; and establish a stress-related factor data acquisition system to collect and store the factor data in real time;

[0155] S52. Based on historical data of abnormal stress at the cold end of glass and a dataset of stress-related factors, a stress anomaly probability prediction model is established using machine learning algorithms.

[0156] S53. Perform preprocessing and feature engineering on the data; integrate stress correlation factor data and stress anomaly case data, and extract features that have an important impact on stress anomaly prediction;

[0157] S54. Divide the processed data into a training set and a test set. Use the training set to train the machine learning model and adjust the model parameters. Use the test set to evaluate and verify the model's performance. Use the trained stress anomaly probability prediction model to obtain the initial stress anomaly probability vector based on the current stress correlation factor dataset.

[0158] S55. Based on the real-time changes in the stress correlation factor dataset, dynamically correct the initial stress anomaly probability vector; obtain the stress anomaly probability distribution data.

[0159] The working principle of the above technical solution is as follows: Collect stress-related factor data, such as process parameter data and environmental parameter data, from the cold end of the glass production line. Process parameters, such as cooling rate, temperature gradient, and glass thickness, directly affect the stress state of the glass; environmental parameters, such as temperature, humidity, and air pressure, also have a certain impact on the stress of the glass. Comprehensive collection of these factor data provides a rich information foundation for subsequent stress anomaly probability prediction; a stress-related factor data acquisition system is established to achieve real-time data acquisition and storage. Real-time acquisition allows for timely access to the latest data from the cold end of the glass production line, reflecting the current state of stress-related factors; stored data provides historical data support for subsequent data analysis and model training.

[0160] Based on historical data on cold-end stress anomalies in glass and a dataset of stress-related factors, a probability prediction model for stress anomalies was established using machine learning algorithms. The historical data includes the occurrence of stress anomalies and corresponding related factors; machine learning algorithms can uncover the potential relationship between these two factors, thereby constructing a model capable of predicting the probability of stress anomalies. Machine learning algorithms possess powerful data processing and pattern recognition capabilities, enabling them to learn complex patterns and regularities from large amounts of data. Different machine learning algorithms are suitable for different types of data and problems; selecting an appropriate algorithm can improve the predictive performance of the model.

[0161] Stress-related factor data and stress anomaly case data are integrated for unified analysis and processing. The integrated data includes information on whether stress anomalies occurred and the corresponding related factors, providing a complete dataset for subsequent feature extraction and model training. Data cleaning is performed to remove noise and outliers, ensuring data quality. Missing values ​​are handled using methods such as mean imputation, median imputation, or model-based prediction imputation to avoid their impact on model training. Feature encoding is performed, converting categorical variables into numerical variables for processing by machine learning algorithms. Features with significant impact on stress anomaly prediction are extracted, and the most representative features are selected using feature selection methods to reduce data dimensionality and redundant information, improving model training efficiency and predictive performance. The processed data is divided into training and test sets. The training set is used for model training, allowing the model to learn patterns and regularities in the data; the test set is used to evaluate and validate the model's performance, checking its performance on unseen data.

[0162] The machine learning model is trained using a training set, and its parameters are adjusted to better fit the training data. By continuously adjusting the parameters, the model's performance is optimized, improving its prediction accuracy. The model's performance is then evaluated and validated using a test set, and the best-performing model is selected as the stress anomaly probability prediction model. Evaluation metrics can include accuracy, recall, and F1 score; appropriate metrics are selected based on actual needs. Using the trained stress anomaly probability prediction model, an initial stress anomaly probability vector is obtained based on the current stress-related factor dataset. The elements in the vector represent the initial probability of different stress anomaly patterns occurring in different regions or under different stress-related factor conditions. These initial probabilities reflect the likelihood of different stress anomaly patterns occurring under the current stress-related factor conditions.

[0163] The initial stress anomaly probability vector is dynamically corrected based on real-time changes in the stress correlation factor dataset. The stress correlation factor data is constantly changing, and these changes can affect the probability of stress anomalies. By monitoring these changes in real time and adjusting the stress anomaly probability vector accordingly, the current stress anomaly probability distribution can be more accurately reflected. After dynamic correction, the stress anomaly probability distribution data is obtained. This data provides real-time stress anomaly early warning information for glass production line operators, helping them to promptly identify potential stress anomaly problems and take corresponding measures for adjustment and optimization, ensuring the quality and stability of glass production.

[0164] The above technical solution achieves the following results: by collecting process and environmental parameter data from the cold end of the glass production line and establishing a stress correlation factor data acquisition system, real-time data acquisition and storage are realized, ensuring the comprehensiveness and timeliness of the data, and providing more accurate input data for subsequent stress anomaly probability prediction.

[0165] By using historical glass cold-end stress anomaly case data and stress correlation factor dataset, combined with machine learning algorithms, a stress anomaly probability prediction model was established, which effectively reduced the errors that may be caused by traditional experience-based judgment methods and improved the reliability and accuracy of prediction.

[0166] By integrating stress-related factor data and historical anomaly case data through data preprocessing and feature engineering, and performing operations such as cleaning, missing value handling, and feature encoding, the model effectively extracted features that have a significant impact on stress anomaly prediction, enhanced its ability to process complex data, and improved the input quality of the model.

[0167] By dividing the processed data into training and testing sets, the model is trained using the training set and its performance is validated using the testing set. This allows for dynamic adjustment of model parameters and selection of the optimal-performing model, thereby improving the model's adaptability and ensuring its predictive accuracy under different stress states.

[0168] By using methods such as cross-validation to evaluate the performance of machine learning models, we can ensure that the models perform consistently on training and testing data. This effectively reduces the risk of overfitting and gives the models stronger generalization ability, enabling them to adapt to more unknown data scenarios.

[0169] By dynamically correcting the initial stress anomaly probability vector based on changes in the real-time stress correlation factor dataset, more real-time and accurate stress anomaly probability distribution data are obtained, improving the system's ability to quickly respond to changes and adjust prediction results in actual production processes.

[0170] By dynamically adjusting the stress anomaly probability vector and combining it with real-time data changes, timely adjustments can be made based on the actual situation of the production line, thereby improving the flexibility of the prediction model and ensuring that the early warning of stress anomaly patterns can accurately reflect the current production status.

[0171] By automatically generating stress anomaly probability prediction results through machine learning models, the frequency and reliance on human intervention are reduced, subjective errors caused by human judgment are minimized, and the system's automation level and work efficiency are improved.

[0172] By combining machine learning and real-time data analysis, more scientific and reliable data support can be provided for production decisions, enhancing enterprises' decision-making capabilities in predicting stress anomalies and adjusting production, and improving the stability and efficiency of the overall production line.

[0173] Thanks to real-time data acquisition and dynamic correction, the predictive model can respond promptly to stress anomalies, providing faster early warning and response times for the production line and reducing the impact of potential stress anomalies on glass quality and production efficiency during the production process.

[0174] In one embodiment of the present invention, step S6 includes:

[0175] S61. Sort the candidate stress anomaly mode list by probability according to the stress anomaly probability distribution data; sort the stress anomaly probabilities from high to low to obtain the stress anomaly mode sorting list.

[0176] S62. Obtain equipment layout data and glass transport path data of the glass production line; perform stress anomaly area location analysis based on the stress anomaly mode sorting list, equipment layout data, and glass transport path data.

[0177] S63. Based on the stress anomaly area location data and the stress anomaly pattern feature library, conduct stress anomaly cause analysis; generate a glass production line cold end stress detection report based on the analysis results.

[0178] The working principle of the above technical solution is as follows: The candidate stress anomaly mode list is ranked probabilistically based on stress anomaly probability distribution data. The stress anomaly probability distribution data reflects the likelihood of different stress anomaly modes occurring under current stress-related factors. Ranking the stress anomaly probabilities from high to low visually demonstrates which stress anomaly modes are more likely to occur. This results in a stress anomaly mode ranking list, which arranges the candidate modes according to their stress anomaly probability. This provides a priority order for subsequent stress anomaly area location and cause analysis, enabling operators to focus on stress anomaly modes with higher probabilities.

[0179] Acquire equipment layout data and glass transport path data for the glass production line. Equipment layout data includes the location and parameters of cooling equipment, conveying equipment, and testing equipment, reflecting the physical structure and equipment configuration of the production line. Glass transport path data includes the movement trajectory and speed of the glass on the production line, helping to understand the flow of glass during production. Based on the stress anomaly pattern ranking list and the equipment layout and glass transport path data, perform stress anomaly area localization analysis. By analyzing the relationship between stress anomaly patterns and the equipment layout and glass transport path, determine the specific areas where stress anomalies may occur. For example, some stress anomaly patterns may be related to the location of specific cooling or conveying equipment, or to specific stages in the glass transport process. Spatial analysis and path tracing methods can be used for localization analysis. Spatial analysis can visually display the spatial relationship between stress anomaly patterns and equipment locations, while path tracing can help identify areas where glass may be affected by stress during transport. Determining the specific areas where stress anomalies may occur provides clear target locations for subsequent stress anomaly cause analysis and treatment.

[0180] Based on the stress anomaly area location data and the stress anomaly pattern feature library, a stress anomaly cause analysis is conducted. Combining factors such as equipment operating status, process parameter settings, and environmental conditions, possible causes of stress anomalies are analyzed. For example, equipment malfunctions may lead to uneven cooling, resulting in stress anomalies; unreasonable process parameter settings, such as excessively fast or slow cooling rates, may also affect the stress state of the glass; changes in environmental conditions, such as fluctuations in temperature and humidity, may also affect the stress of the glass. By comprehensively analyzing these factors, the root cause of the stress anomaly can be identified. A stress detection report for the cold end of the glass production line is generated based on the analysis results. The report should include a stress anomaly pattern ranking list, stress anomaly area location results, stress anomaly cause analysis, and corresponding handling suggestions. The stress anomaly pattern ranking list allows operators to understand the probability of different stress anomaly patterns occurring; the stress anomaly area location results clarify the specific locations where stress anomalies may occur; the stress anomaly cause analysis reveals the root causes of stress anomalies; and the handling suggestions provide operators with specific solutions and measures. The test report should be clear, accurate, and comprehensive, providing a basis for decision-making in the maintenance and management of the glass production line, helping operators to take timely measures to eliminate abnormal stress, and ensuring the quality and stability of glass production.

[0181] The effect of the above technical solution is that by ranking the candidate stress anomaly modes according to the probability distribution data of stress anomalies, high-probability stress anomaly modes can be accurately identified, thereby effectively improving the accuracy and pertinence of stress anomaly prediction.

[0182] By sorting stress anomaly probabilities from high to low to form a stress anomaly pattern ranking list, the screening process for anomaly patterns is simplified, the complexity of manual judgment is reduced, and work efficiency is improved.

[0183] By combining equipment layout data, glass transmission path data, and a stress anomaly pattern sorting list, spatial analysis and path tracing methods were used to locate and analyze stress anomaly areas, ensuring accurate identification and location of anomaly areas and providing more precise data support for subsequent processing.

[0184] By analyzing the location of stress anomalies, potential stress anomalies can be quickly identified, enabling timely troubleshooting and production line adjustments. This significantly improves the production line's response speed and reduces equipment downtime.

[0185] By combining factors such as equipment operating status, process parameter settings, and environmental conditions for cause analysis, the possible causes of stress anomalies can be identified in a short time, avoiding the time wasted in traditional manual diagnosis and improving analysis efficiency.

[0186] By combining multiple factors (such as equipment status, environmental parameters, process settings, etc.) to conduct a comprehensive analysis of the causes of stress anomalies, the comprehensiveness and scientific nature of the analysis results are enhanced, and the one-sided analysis of a single factor is avoided.

[0187] By generating a cold-end stress detection report that includes stress anomaly pattern ranking, stress anomaly area location results, and stress anomaly cause analysis, a clear, accurate, and comprehensive decision-making basis is provided for production line maintenance and management, thereby improving the reliability of decision-making.

[0188] By generating detailed stress detection reports, potential stress anomalies can be identified in advance, providing a reference for developing preventive and remedial measures and reducing the risk of production line failures during operation.

[0189] By ranking stress anomaly patterns using machine learning algorithms, and combining this with automated regional location analysis and cause analysis, the automation level of the entire system is significantly enhanced, the reliance on manual operation is reduced, and the intelligence level of the production line is improved.

[0190] By comprehensively analyzing stress anomaly patterns, equipment layout, and transmission paths, the accuracy and responsiveness of the fault early warning system have been enhanced, ensuring that the production line can obtain early warning information in a timely manner before anomalies occur, thus avoiding potential production interruptions.

[0191] In one embodiment of the present invention, S62 includes:

[0192] The potential correlation between each stress anomaly mode and different equipment is analyzed; based on the analysis results, an equipment-stress anomaly mode correlation matrix is ​​constructed. The rows of the matrix represent different stress anomaly modes, the columns represent different equipment and their parameters, and the matrix elements represent the correlation strength between the equipment parameters and the stress anomaly mode.

[0193] Based on the glass transport path data, the movement trajectory of the glass on the production line is divided into multiple critical path segments; for each path segment, the probability of stress anomaly mode occurring in that path segment is evaluated; a path segment-stress anomaly probability assessment model is established, and the probability of various stress anomaly modes occurring in each path segment is calculated by combining the stress anomaly mode ranking list and the equipment-stress anomaly mode correlation matrix.

[0194] A spatial topology model of the glass production line is constructed using equipment layout data to clarify the spatial positional relationships, connection relationships, and glass flow direction between equipment; stress anomaly modes are projected in space to analyze their correlation with the spatial distribution of equipment.

[0195] A path tracing algorithm is used to simulate the movement of glass on the production line. Combined with a stress anomaly pattern feature library, the propagation of stress inside the glass is simulated. Through simulation, the specific distribution location of stress anomalies on the glass and the range of areas that may be affected are determined.

[0196] The results of the correlation analysis between equipment and stress anomaly mode, the glass transmission path segmentation and stress impact assessment, the spatial topology analysis, and the path tracing and stress propagation simulation are integrated; a weighted comprehensive evaluation method is adopted to assign corresponding weights to each analysis result according to the importance and reliability of different analysis results;

[0197] Through comprehensive calculations, the specific areas where stress anomalies may occur are identified, and these areas are prioritized.

[0198] The working principle of the above technical solution is as follows: For each stress anomaly mode, a thorough analysis is conducted on its potential correlation with different equipment. For example, for stress anomalies caused by uneven cooling, the focus is on key parameters of cooling equipment (such as cooling air knives, cooling water tanks, etc.), including position, angle, and air / water flow speed adjustment range. These parameters directly affect the cooling effect; improper settings can easily lead to uneven cooling, thus causing stress anomalies. For stress anomalies caused by uneven force during conveying, parameters such as roller spacing, roller material, and transmission speed stability of the conveying equipment are considered. Changes in these parameters affect the stress on the glass during conveying, thereby triggering stress anomalies. Based on the above analysis results, a device-stress anomaly mode correlation matrix is ​​constructed. The rows of the matrix represent different stress anomaly modes, the columns represent different equipment and their parameters, and the matrix elements represent the correlation strength between the equipment parameter and the stress anomaly mode. This matrix format can intuitively display the correlation between equipment parameters and stress anomaly modes, providing basic data support for subsequent analysis and localization.

[0199] Based on glass transport path data, the movement trajectory of the glass on the production line is divided into multiple critical path segments. For example, starting from the glass entering the cooling zone, the path is divided into cooling stage, transition stage, and inspection stage. Different path segments have different process characteristics and stress conditions, which have different impacts on the occurrence of stress anomalies. For each path segment, the probability of stress anomaly modes occurring in that segment is assessed. Factors such as the glass's residence time, speed variation, and stress conditions in that path segment are considered. For example, in the cooling stage, the glass moves slowly and is subjected to cooling; if the cooling is uneven, the probability of stress anomalies occurring in that path segment is high. In the inspection stage, if the inspection equipment applies improper force to the glass, it may also trigger stress anomalies. Combining the stress anomaly mode ranking list and the equipment-stress anomaly mode correlation matrix, a path segment-stress anomaly probability assessment model is established to calculate the probability of various stress anomaly modes occurring in each path segment, thereby determining the potential distribution of stress anomalies along the glass transport path.

[0200] A spatial topology model of the glass production line is constructed using equipment layout data to clarify the spatial relationships and connections between equipment, as well as the flow direction of glass between equipment. This model can intuitively display the spatial layout of the production line and the flow path of glass, providing a foundation for analyzing the relationship between stress anomalies and equipment spatial distribution. Stress anomaly patterns are projected spatially to analyze their correlation with equipment spatial distribution. For example, if a certain stress anomaly pattern shows a spatial distribution around a certain cooling device, it can be preliminarily determined that this device may have a problem causing stress anomalies. Spatial clustering algorithms (such as the DBSCAN algorithm) are used to cluster the spatial distribution of different stress anomaly patterns to identify high-incidence areas or equipment clusters of stress anomalies, further narrowing down the possible area where stress anomalies may occur.

[0201] A path tracing algorithm is employed to simulate the movement of glass on the production line, recording the equipment the glass passes through at different times and the forces and temperatures it experiences. By simulating the glass's movement, the stress conditions and temperature changes throughout the production process can be understood, providing detailed data for analyzing the generation of stress anomalies. Combined with a stress anomaly pattern feature library, the propagation process of stress within the glass is simulated. For example, for stress anomalies caused by temperature gradients, the propagation path and intensity changes of stress from high-temperature regions to low-temperature regions are simulated based on the glass's thermal conductivity and temperature variations. Through simulation, the specific distribution location of stress anomalies on the glass and the potential affected area are determined, providing more accurate information for locating stress anomaly regions.

[0202] The results of the correlation analysis between equipment and stress anomaly modes, the glass transport path segmentation and stress impact assessment, the spatial topology analysis, and the path tracing and stress propagation simulation are integrated. A weighted comprehensive evaluation method is used to assign corresponding weights to each analysis result based on its importance and reliability. For example, the correlation analysis results between equipment and stress anomaly modes may have high reliability and can be assigned a larger weight; while the path tracing and stress propagation simulation results have relatively low reliability when the data is incomplete and are assigned a smaller weight. Through comprehensive calculation, specific areas where stress anomalies may occur are identified, and these areas are prioritized. High-priority areas indicate a higher probability of stress anomalies and more severe impacts, requiring priority investigation and handling. This prioritization helps operators allocate resources rationally and improves the efficiency of troubleshooting and handling.

[0203] The effect of the above technical solution is that by analyzing the potential correlation between each stress anomaly mode and different equipment and its parameters, a device-stress anomaly mode correlation matrix is ​​constructed, which can more accurately identify the key equipment and parameters that cause stress anomalies and improve the accuracy of anomaly mode diagnosis.

[0204] Dividing the production line into path segments based on glass transport path data and assessing the probability of stress anomalies in each path segment can reduce prediction errors caused by incomplete path analysis and improve the reliability of anomaly prediction.

[0205] By constructing a spatial topology model of the glass production line and combining it with spatial clustering algorithms to perform cluster analysis on the spatial distribution of different stress anomaly patterns, it is possible to effectively identify high-incidence areas or equipment clusters of stress anomalies, thereby enhancing the comprehensiveness and accuracy of spatial analysis of the production line.

[0206] By simulating the movement of glass on the production line using a path tracing algorithm and combining it with a stress anomaly pattern feature library, the propagation of stress on the glass can be simulated more accurately, improving the accuracy of predicting the location and extent of stress anomalies.

[0207] By using a weighted comprehensive evaluation method to assign different weights to different analysis results based on their importance and reliability, the interference of data noise can be effectively reduced, thereby improving the effectiveness and reliability of the comprehensive analysis.

[0208] By comprehensively considering the correlation analysis results between equipment and stress anomaly modes, the probability assessment of stress anomalies in path segments, the spatial topology analysis results, and the path tracing simulation results, we can more accurately determine the areas where stress anomalies may occur, and prioritize them to ensure that high-risk areas are investigated and dealt with in a timely manner.

[0209] By prioritizing specific areas where stress anomalies may occur, production line maintenance personnel can address higher-risk areas first, reducing the impact of malfunctions on the production line and improving the accuracy and timeliness of production line maintenance.

[0210] By comprehensively analyzing multiple factors, a clear priority ranking of stress anomaly areas can be provided, which can reduce the decision-making cost for maintenance personnel when dealing with faults and improve the efficiency of fault diagnosis and handling.

[0211] By combining multiple analysis methods (correlation analysis, path segment evaluation, spatial analysis, simulation analysis, etc.), the intelligence and predictive ability of the entire system are enhanced, human intervention is reduced, and the automation level of the system is improved.

[0212] By conducting detailed analysis of areas with abnormal stress and prioritizing them, the production line can identify potential fault areas in advance and perform preventative maintenance before a failure occurs, thereby reducing the frequency of production line failures and improving management reliability.

[0213] In one embodiment of the present invention, a cold-end stress detection system for a glass production line includes a memory, a processor, and a computer program stored in and executable on the memory. The processor executes the program to implement a cold-end stress detection method for a glass production line as described above.

[0214] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for detecting cold-end stress in a glass production line, characterized in that, The method includes: S1: Stress-related spectral data of the cold end area of ​​the glass production line is collected by a multispectral stress sensing array, and the collected spectral data is preprocessed to obtain a preprocessed spectral dataset. S2: Extract spectral features from the preprocessed spectral dataset and construct a spectral feature vector set; based on the spectral feature vector set, use a spatiotemporal correlation analysis algorithm to extract the spatiotemporal features of the glass cold end stress and obtain the stress spatiotemporal feature matrix; S3: Collect historical data on anomaly stress at the cold end of glass, perform feature clustering analysis on the spatiotemporal feature matrix of stress to obtain multiple stress feature clusters; combine the historical data on anomaly stress at the cold end of glass, perform anomaly feature matching on each stress feature cluster, screen out feature clusters with potential stress anomalies, and extract the stress anomaly feature vectors of the feature clusters. S4: Based on historical glass cold end stress anomaly case data, construct a stress anomaly pattern feature library. Based on the stress anomaly feature vector and the stress anomaly pattern feature library, use a pattern matching algorithm to identify stress anomaly patterns and obtain a candidate stress anomaly pattern list. Calculate the similarity of each pattern in the candidate stress anomaly pattern list to generate a stress anomaly similarity vector. S5: Collect data on stress-related factors at the cold end of the glass production line and construct a stress-related factor dataset; based on historical data of abnormal stress at the cold end of the glass and the stress-related factor dataset, use machine learning algorithms to establish a stress anomaly probability prediction model and obtain an initial stress anomaly probability vector; dynamically correct the initial stress anomaly probability vector based on the stress-related factor dataset to obtain stress anomaly probability distribution data. S6: Based on the probability distribution data of stress anomalies, the candidate stress anomaly pattern list is probabilistically sorted to obtain a stress anomaly pattern ranking list; equipment layout data and glass transport path data of the glass production line are obtained; based on the stress anomaly pattern ranking list, equipment layout data, and glass transport path data, stress anomaly area location analysis is performed to obtain stress anomaly area location result data; based on the stress anomaly area location result data and the stress anomaly pattern feature library, stress anomaly cause analysis is performed to generate a cold end stress detection report for the glass production line.

2. The method for detecting cold-end stress in a glass production line according to claim 1, characterized in that, S1 includes: S11. In the critical area of ​​the cold end of the glass production line, a multispectral stress sensor array is uniformly arranged; and the arranged multispectral stress sensor array is calibrated; its spectral response is measured under different stress conditions, and a calibration curve between spectral response and stress is established. S12. During the glass operation, the multispectral stress sensing array collects stress-related spectral data of the glass surface in real time; the collected spectral data is transmitted to the data processing system in real time for preliminary storage and management. S13. Denoise the collected raw spectral data; normalize the denoised raw spectral data; then smooth the normalized spectral data to obtain the preprocessed spectral dataset.

3. The method for detecting cold-end stress in a glass production line according to claim 1, characterized in that, S2 includes: S21. Extract stress-related spectral features from the preprocessed spectral dataset; perform dimensionality reduction on the extracted spectral features using a feature selection algorithm; construct a spectral feature vector set; S22. Using a spatiotemporal correlation analysis algorithm, the spectral feature vector set is processed to extract the spatiotemporal characteristics of the glass cold end stress. S23. Organize and store the extracted spatiotemporal features according to a preset format to construct a stress spatiotemporal feature matrix, where the rows of the matrix represent different time points or spatial locations, the columns represent different spatiotemporal features, and the elements in the matrix represent the spatiotemporal feature values ​​corresponding to the time and spatial location.

4. The method for detecting cold-end stress in a glass production line according to claim 3, characterized in that, S23 includes: Based on the physical meaning and analytical requirements represented by the spatiotemporal characteristics, they are divided into time-dimensional characteristics and spatial-dimensional characteristics, and an identifier is defined for each type of characteristic. For spatiotemporal feature values ​​with different dimensions and ranges, a unified standardization method is used to process them, mapping the feature values ​​to specific numerical intervals. Construct a time-space indexing system; in the time dimension, assign a unique index number to each time point in chronological order; in the spatial dimension, assign a corresponding index identifier to each spatial location based on the actual spatial layout of the glass cold end; The dimensions of the stress spatiotemporal characteristic matrix in dynamic programming are determined as follows: in the time dimension, the number of rows of the matrix is ​​determined based on the time span of data acquisition and the time resolution requirements of analysis; in the spatial dimension, the number of columns of the matrix is ​​determined based on the degree of spatial division of the glass cold end. Based on the time-space indexing system, the standardized spatiotemporal feature values ​​are filled into the corresponding positions of the matrix one by one; for missing data, estimation and filling are performed; after filling all feature values, a complete stress spatiotemporal feature matrix is ​​formed. The quality of the constructed stress spatiotemporal feature matrix is ​​verified, and corrections and optimizations are made based on the verification results. The stress spatiotemporal feature matrix, after quality verification and optimization, is stored in a specified database or file system, while key information of the matrix is ​​recorded to form a complete metadata record.

5. The method for detecting cold-end stress in a glass production line according to claim 1, characterized in that, The S3 includes: S31. Collect data on stress anomalies at the cold end of the glass production line over a period of time, establish a historical database of stress anomalies at the cold end of the glass; and organize and classify the collected historical case data. S32. Use clustering algorithms to perform feature clustering analysis on the stress spatiotemporal feature matrix; S33. Based on historical data of abnormal stress at the cold end of glass, perform abnormal feature matching for each stress feature cluster; compare the features of the clusters with the abnormal features in historical cases and calculate the similarity. S34. Set a similarity threshold to filter out feature clusters with potential stress anomalies; for clusters with similarity exceeding the threshold, determine whether there are stress anomalies. S35. Extract stress anomaly feature vectors from the selected feature clusters; select the most representative features from the clusters to construct stress anomaly feature vectors.

6. The method for detecting cold-end stress in a glass production line according to claim 1, characterized in that, The S4 includes: S41. Based on historical glass cold-end stress anomaly case data, construct a stress anomaly pattern feature library; standardize the data in the feature library to unify the dimensions and ranges of different features; S42. Select a pattern matching algorithm and optimize the algorithm parameters based on the characteristics and distribution patterns of stress anomaly features; train the algorithm and adjust its parameters using stress anomaly feature vectors and corresponding stress anomaly patterns from historical case data. S43. Based on the stress anomaly feature vector and the stress anomaly pattern feature library, use the selected pattern matching algorithm to perform stress anomaly pattern recognition; match the stress anomaly feature vector with each pattern in the feature library and calculate the similarity. S44. Based on the similarity calculation results, obtain a list of candidate stress anomaly patterns; perform similarity calculation on each pattern in the list of candidate stress anomaly patterns to generate a stress anomaly similarity vector.

7. The method for detecting cold-end stress in a glass production line according to claim 1, characterized in that, The S5 includes: S51. Collect stress-related factor data at the cold end of the glass production line; and establish a stress-related factor data acquisition system to collect and store the factor data in real time; S52. Based on historical data of abnormal stress at the cold end of glass and a dataset of stress-related factors, a stress anomaly probability prediction model is established using machine learning algorithms. S53. Perform preprocessing and feature engineering on the data; integrate stress correlation factor data and stress anomaly case data, and extract features that have an important impact on stress anomaly prediction; S54. Divide the processed data into a training set and a test set. Use the training set to train the machine learning model and adjust the model parameters. Use the test set to evaluate and verify the model's performance. Use the trained stress anomaly probability prediction model to obtain the initial stress anomaly probability vector based on the current stress correlation factor dataset. S55. Based on the real-time changes in the stress correlation factor dataset, dynamically correct the initial stress anomaly probability vector; obtain the stress anomaly probability distribution data.

8. The method for detecting cold-end stress in a glass production line according to claim 1, characterized in that, The S6 includes: S61. Sort the candidate stress anomaly mode list by probability according to the stress anomaly probability distribution data; sort the stress anomaly probabilities from high to low to obtain the stress anomaly mode sorting list. S62. Obtain equipment layout data and glass transport path data of the glass production line; perform stress anomaly area location analysis based on the stress anomaly mode sorting list, equipment layout data, and glass transport path data. S63. Based on the stress anomaly area location data and the stress anomaly pattern feature library, conduct stress anomaly cause analysis; and generate a glass production line cold end stress detection report based on the analysis results.

9. The method for detecting cold-end stress in a glass production line according to claim 8, characterized in that, S62 includes: The potential correlation between each stress anomaly mode and different equipment is analyzed; based on the analysis results, an equipment-stress anomaly mode correlation matrix is ​​constructed. The rows of the matrix represent different stress anomaly modes, the columns represent different equipment and their parameters, and the matrix elements represent the correlation strength between the equipment parameters and the stress anomaly mode. Based on the glass transport path data, the movement trajectory of the glass on the production line is divided into multiple critical path segments; for each path segment, the probability of stress anomaly mode occurring in that path segment is evaluated; a path segment-stress anomaly probability assessment model is established, and the probability of various stress anomaly modes occurring in each path segment is calculated by combining the stress anomaly mode ranking list and the equipment-stress anomaly mode correlation matrix. A spatial topology model of the glass production line is constructed using equipment layout data to clarify the spatial positional relationships, connection relationships, and glass flow direction between equipment; stress anomaly modes are projected in space to analyze their correlation with the spatial distribution of equipment. A path tracing algorithm is used to simulate the movement of glass on the production line. Combined with a stress anomaly pattern feature library, the propagation of stress inside the glass is simulated. Through simulation, the specific distribution location of stress anomalies on the glass and the range of affected areas are determined. The results of the correlation analysis between equipment and stress anomaly mode, the glass transmission path segmentation and stress impact assessment, the spatial topology analysis, and the path tracing and stress propagation simulation are integrated; a weighted comprehensive evaluation method is adopted to assign corresponding weights to each analysis result according to the importance and reliability of different analysis results; Through comprehensive calculations, the specific areas where stress anomalies occur are determined, and these areas are prioritized.

10. A cold-end stress detection system for a glass production line, characterized in that, The device includes a memory, a processor, and a computer program stored in and executable on the memory, wherein the processor executes the program to implement a method for detecting cold-end stress in a glass production line as described in any one of claims 1-9.

Images