A glass cup production line temperature field real-time monitoring control method and system

Abstract

The application provides a glass cup production line temperature field real-time monitoring control method and system, relates to the technical field of data acquisition and control, and comprises the following steps: acquiring drop material, mold state and multi-zone temperature data; combining temperature and drop material data to establish a drop material temperature viscosity forming window model, identify viscosity deviation and weight fluctuation; generating mold thermal memory results and performing risk prediction, obtaining a thermal defect risk thermodynamic map; constructing a cup type matching annealing curve based on the risk thermodynamic map; combining the model and analysis results to reconstruct the curve, obtaining a temperature coordination control strategy, so as to solve the technical problems that the existing glass cup production line cannot comprehensively analyze the forming risk of drop material viscosity state, weight fluctuation and mold thermal memory effect, resulting in fixed annealing curve and inability to cope with temperature field drift, achieve multi-link cooperation reconstruction through drop material-mold-annealing, realize dynamic and accurate control of the temperature field, inhibit forming thermal defects, and improve the technical effects of glass cup quality and production line stability.

Description

Technical Field

[0001] This invention relates to the field of data acquisition and control technology, specifically to a method and system for real-time monitoring and control of the temperature field in a glass production line. Background Technology

[0002] The core processes of a glass production line encompass dispensing, molding, annealing, and cooling. The coordinated control of the temperature field in each process directly determines product yield and quality consistency. Currently, the industry primarily relies on manual experience to set the temperature for each zone, supplemented by single-point thermocouples or infrared temperature measurement for fixed threshold alarms. This presents the following technical problems: First, the control of dispensing status is limited. Existing technologies focus primarily on whether the dispensing temperature falls within a preset range, neglecting the dynamic changes in viscosity response during temperature decay and the weight fluctuations caused by these processes. When the dispensing viscosity deviates from the optimal molding window, the flowability and uniformity of the droplet distribution deteriorate, and traditional methods struggle to promptly identify this interconnected deviation between temperature, viscosity, and weight. Second, the mold's thermal state exhibits a significant thermal memory effect. During continuous production cycles, the mold undergoes alternating changes in heat storage and thermal drift. Its temperature distribution is influenced not only by the current heating / cooling conditions but also by the thermal history of previous cycles. However, existing monitoring methods lack analysis of the correlation between the mold's heat storage state and thermal drift state, failing to distinguish between the active thermal changes of the mold itself and the passive effects of temperature fluctuations upstream and downstream of the production line. This results in delayed and indiscriminate mold temperature control. More critically, annealing control curves often use fixed process curves, lacking dynamic correlation with the front-end dripping state and the mold's thermal state. When dripping viscosity deviations, weight fluctuations, or mold thermal memory shifts occur, the fixed annealing heating, holding, and cooling curves cannot make corresponding adjustments, making it difficult to suppress molding thermal defects such as cup mouth deformation, cup wall thickness deviation, and cup bottom heat retention online. Summary of the Invention

[0003] This application provides a method and system for real-time monitoring and control of the temperature field in a glass production line, aiming to solve the technical problem that existing glass production lines cannot comprehensively analyze molding risks by considering the viscosity state of the dripping material, weight fluctuations, and the thermal memory effect of the mold, resulting in fixed annealing curves and an inability to cope with temperature field drift.

[0004] The first aspect of this application discloses a method for real-time monitoring and control of the temperature field in a glass production line. The method includes: acquiring dripping state data, mold state data, and multi-region temperature monitoring data of the glass production line; establishing a dripping temperature-viscosity forming window model based on the multi-region temperature monitoring data and the dripping state data, and identifying dripping viscosity deviation and dripping weight fluctuation states based on the dripping temperature-viscosity forming window model; generating mold thermal memory analysis results based on the mold state data, and performing glass forming process risk prediction based on the dripping viscosity deviation, the dripping weight fluctuation state, and the mold thermal memory analysis results to obtain a forming thermal defect risk heatmap; constructing a cup shape matching annealing control curve based on the forming thermal defect risk heatmap; and performing drift collaborative reconstruction of the cup shape matching annealing control curve based on the dripping temperature-viscosity forming window model and the mold thermal memory analysis results to obtain a production line temperature collaborative control strategy.

[0005] The second aspect of this application discloses a real-time temperature field monitoring and control system for a glass production line. This system is used in the aforementioned real-time temperature field monitoring and control method for a glass production line. The system includes: a data acquisition module for acquiring dripping state data, mold state data, and multi-zone temperature monitoring data from the glass production line; and a dripping identification module for establishing a dripping temperature-viscosity forming window model based on the multi-zone temperature monitoring data and the dripping state data, and identifying the dripping viscosity based on the dripping temperature-viscosity forming window model. Deviation state and drip weight fluctuation state; Risk prediction module: Generates mold thermal memory analysis results based on the mold state data, and performs risk prediction of the glass cup forming process based on the drip viscosity deviation state, the drip weight fluctuation state, and the mold thermal memory analysis results to obtain a molding thermal defect risk heat map; Curve construction module: Constructs a cup shape matching annealing control curve based on the molding thermal defect risk heat map; Collaborative reconstruction module: Performs drift collaborative reconstruction of the cup shape matching annealing control curve based on the drip temperature viscosity forming window model and the mold thermal memory analysis results to obtain a production line temperature collaborative control strategy.

[0006] One or more technical solutions provided in this application have at least the following beneficial effects:

[0007] First, real-time data collection is conducted on the dripping state, mold operation status, and temperature changes in various areas of the glass production line, forming a temperature field data foundation covering the dripping, forming, annealing, and cooling processes. Then, combining temperature change characteristics with the dripping operation status, a forming correlation model between dripping temperature and glass viscosity is established to identify forming instability phenomena such as abnormal glass viscosity and dripping weight fluctuations. Next, the heat storage and thermal drift of the mold during continuous production are analyzed to form mold thermal memory results. Combined with abnormal dripping conditions, this data is used to predict thermal defect risks during the glass forming process, such as cup rim deformation, wall thickness deviation, and bottom heat stagnation, generating corresponding risk thermal distributions. Then, based on the degree of thermal defect risk in different areas, targeted adjustments are made to the annealing heating, holding, and cooling processes to construct an annealing control curve that matches the cup shape characteristics. Finally, combining the dripping forming state and the mold's thermal history, the annealing curve is dynamically and collaboratively corrected, thereby outputting a temperature collaborative control strategy suitable for the current production conditions, achieving real-time optimized control of the temperature field of the glass production line.

[0008] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0009] Figure 1 This is a schematic flowchart of a real-time temperature field monitoring and control method for a glass production line, provided as an embodiment of this application.

[0010] Figure 2 This is a schematic diagram of a real-time temperature field monitoring and control system for a glass production line, provided as an embodiment of this application.

[0011] Figure labeling: Data acquisition module 11, dripping material recognition module 12, risk prediction module 13, curve construction module 14, collaborative reconstruction module 15. Detailed Implementation

[0012] This application provides a method and system for real-time monitoring and control of the temperature field in a glass production line, which solves the technical problem that existing glass production lines cannot comprehensively analyze molding risks by considering the viscosity of the dripping material, weight fluctuations, and the thermal memory effect of the mold, resulting in fixed annealing curves and an inability to cope with temperature field drift.

[0013] After introducing the basic principles of this application, various non-limiting embodiments of this application will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit this application.

[0014] Example 1, as Figure 1 As shown in the figure, this application provides a method for real-time monitoring and control of the temperature field in a glass production line, the method comprising: Acquire data on the dripping status, mold status, and multi-zone temperature monitoring of the glass production line.

[0015] In one embodiment, data such as temperature, weight, frequency, interval, shape, and location of the molten glass falling from the dripping zone are collected to form dripping state data, reflecting the stability of the molten glass before entering the mold. Simultaneously, data on the mold's temperature, opening / closing status, cooling status, usage cycle, heat accumulation, and local temperature difference within the mold cavity are collected during continuous production to form mold state data, used to determine if there is heat accumulation or uneven cooling in the mold. Furthermore, temperature data from multiple areas, including the dripping zone, forming zone, mold zone, annealing zone, and cooling zone, are collected to form multi-area temperature monitoring data. This data collection provides the foundation for subsequent dripping and forming state analysis, mold thermal memory assessment, and coordinated temperature control.

[0016] Furthermore, the method includes: Temperature data corresponding to the dripping zone, forming zone, mold zone, annealing zone and cooling zone in the glass production line are collected to generate the multi-zone temperature monitoring data.

[0017] Preferably, based on the process flow of the glass production line, the temperature monitoring range is divided into a dripping zone, a forming zone, a mold zone, an annealing zone, and a cooling zone. Corresponding temperature acquisition points are set up in each zone. Specifically, temperature acquisition points in the dripping zone can be located near the feed channel outlet, the shearing nozzle, and the dripping path to collect the glass melt outlet temperature, the dripping shearing temperature, and the temperature before dripping into the mold. Temperature acquisition points in the forming zone can be located at the dripping mold entry point, the pressing or blowing point, the cup rim forming point, the cup wall forming point, and the cup bottom forming point to collect glass melt temperature data. The system collects local temperatures of the glass preform during the forming process. Temperature collection points in the mold area can be set near the mold opening, mold cavity sidewalls, mold bottom, mold outer wall, and cooling channels to collect heat storage and cooling temperatures at different parts of the mold. Temperature collection points in the annealing zone can be arranged in sections according to the annealing furnace's inlet, heating, holding, slow cooling, and outlet sections to collect segmented furnace temperatures and the annealing temperature of the glass surface. Temperature collection points in the cooling zone can be set in the natural cooling section, air cooling section, and discharge detection section to collect temperature changes after the glass has completed annealing. Each temperature collection point can use an infrared thermometer, thermocouple, high-temperature thermal imager, or embedded temperature sensor for real-time sampling and continuously output temperature values ​​according to a preset sampling frequency. After receiving the temperature values ​​uploaded from each collection point, the system adds identification information such as collection time, collection area, collection point number, production cycle, mold number, droplet number, or glass number to each temperature value. It then performs unit standardization, time synchronization, abnormal temperature rejection, and short-term missing value compensation for the temperature values. Finally, the processed temperature data is aggregated according to the dimensions of region, time, and production cycle to form multi-region temperature monitoring data, including temperature sequences of the dripping zone, forming zone, mold zone, annealing zone, and cooling zone. This data enables subsequent steps to analyze the temperature transfer, temperature decay, heat retention, and temperature drift between different process zones during the glass production process.

[0018] Furthermore, the method includes: Collect the dripping temperature, dripping weight, dripping interval, and dripping morphology parameters in the dripping zone to generate the dripping state data.

[0019] Preferably, a drip detection unit is installed at the outlet of the feeding channel, below the shearing mechanism, and at the drop path before the drip enters the mold in the glass production line. The drip detection unit may include an infrared thermometer, a high-temperature thermal imager, a weighing sensor, a photoelectric trigger sensor, an industrial camera, and a timing module. When the feeding channel outputs molten glass and the shearing mechanism forms a single drop, the infrared thermometer or high-temperature thermal imager collects the surface temperature of the drop, the estimated temperature of the central area, and the temperature decay value of the drop during the drop's fall, as the drop temperature; the weighing sensor set at the feeding mechanism or the receiving calibration position collects the weight of each drop, as the drop weight; the photoelectric trigger sensor records the time point when two adjacent drops pass through the same detection position and calculates the difference between adjacent time points to obtain the drop drop interval; the industrial camera collects the outer contour image of the drop, and performs edge extraction, contour fitting, and size conversion on the image to obtain drop morphology parameters such as drop length, maximum diameter, minimum diameter, aspect ratio, roundness, and tail length. During the data collection process, each droplet is assigned a unique droplet number, which is then linked to the collection time, production cycle time, feed channel number, number of shearing operations, and corresponding mold number. Subsequently, the system performs unit standardization, time synchronization, outlier removal, and short-term missing data compensation on the collected droplet temperature, weight, droplet drop interval, and droplet morphology parameters. For example, it removes temperature spikes or weight anomalies that significantly exceed the preset process range and uses interpolation of adjacent droplet data to fill in short-term missing data. Finally, the processed data is aggregated according to the droplet number and time sequence to form droplet state data, including a droplet temperature sequence, droplet weight sequence, droplet drop interval sequence, and droplet morphology parameter sequence. This data is used for subsequent analysis of droplet temperature decay, viscosity response, weight fluctuation, and mold forming stability.

[0020] A temperature-viscosity forming window model for dripping material is established based on the multi-region temperature monitoring data and the dripping material state data. The dripping material viscosity deviation state and dripping material weight fluctuation state are then identified based on the temperature-viscosity forming window model.

[0021] In one embodiment, firstly, based on collected multi-region temperature monitoring data, a dripping temperature feature sequence is constructed, including data such as the feed channel temperature, dripping outlet temperature, and molding inlet temperature. This sequence is then used for time-coupled temperature difference calculation to determine the dripping temperature decay characteristics. Subsequently, combined with dripping state data, the viscosity state of the dripping material under the current temperature conditions is calculated, determining the dripping viscosity response characteristics. Finally, a correspondence between the dripping temperature decay characteristics and the dripping viscosity response characteristics is established, forming a dripping temperature-viscosity molding window model. During real-time judgment, the viscosity response characteristics of the current dripping material are compared with the allowable viscosity range. When the viscosity response characteristics are higher than the upper limit, the dripping material is determined to be in a high viscosity deviation state; when the viscosity response characteristics are lower than the lower limit, the dripping material is determined to be in a low viscosity deviation state. The deviation value, the number of dripping materials continuously deviating, and the corresponding dripping material number are recorded to form the dripping material viscosity deviation state. Simultaneously, the current dripping material weight is compared with the allowable dripping material weight range, and the average and range of the weights of multiple consecutive dripping materials are calculated. For example, for 10 consecutive dripping materials, when the weight of a single dripping material exceeds the allowable range, or the weight range of multiple consecutive dripping materials is greater than a preset weight fluctuation threshold, a dripping material weight fluctuation state is determined to exist, and the weight deviation, weight deviation, periodic fluctuation amplitude, and abnormal dripping material number are output to form the dripping material weight fluctuation state. Through the above processing, it is possible to determine in advance whether the dripping material meets the stable molding requirements, providing an accurate data foundation for subsequent molding thermal defect risk prediction and temperature collaborative control.

[0022] Furthermore, a temperature-viscosity molding window model for dripping material is established based on the multi-region temperature monitoring data and the dripping state data, including: Based on the multi-region temperature monitoring data, a dripping temperature feature sequence is constructed; time-coupled temperature difference calculation is performed based on the dripping temperature feature sequence to obtain the dripping temperature decay feature; viscosity response matching is performed based on the dripping temperature decay feature and the dripping state data to determine the dripping viscosity response feature; molding allowable feature binding is performed based on the dripping temperature decay feature and the dripping viscosity response feature to establish the dripping temperature viscosity molding window model.

[0023] Preferably, each droplet is analyzed as a single droplet, and a unique droplet number is assigned to each droplet. Based on the droplet number, the temperature of the feeding channel, the temperature of the droplet outlet, the temperature of the droplet falling path, and the temperature of the molding inlet are read sequentially from the temperature monitoring data of multiple areas. At the same time, the acquisition time points corresponding to the above temperatures are read. Then, according to the actual flow sequence of the droplet from the feeding channel to the molding inlet, the temperature of the feeding channel, the temperature of the droplet outlet, the temperature of the droplet falling path, and the temperature of the molding inlet are arranged in sequence, and each temperature value is bound to the corresponding acquisition time to form the droplet temperature characteristic sequence of the droplet. Subsequently, time-coupled temperature difference calculation is performed based on the dripping temperature characteristic sequence. Specifically, the first-stage temperature difference between the feeding channel temperature and the dripping outlet temperature is calculated, the second-stage temperature difference between the dripping outlet temperature and the dripping path temperature is calculated, and the third-stage temperature difference between the dripping path temperature and the forming inlet temperature is calculated. The temperature decay rate per unit time for each stage is obtained by combining the time interval between adjacent temperature acquisition time points. The first-stage temperature difference, the second-stage temperature difference, the third-stage temperature difference, the total temperature decay, and the temperature decay rate per unit time for each stage are then combined to form the dripping temperature decay characteristic. Next, the weight, drop interval, length, diameter, and drop point offset of the same drop are read from the drop state data. The glass material temperature-viscosity correspondence table obtained in advance through production calibration is called to convert the molding inlet temperature into the drop viscosity. The total temperature decay and the temperature decay rate per unit time are converted into the viscosity increase and viscosity increase rate. Then, combined with the drop weight, drop interval, drop length, drop diameter, and drop point offset, the weight deviation coefficient, cycle deviation coefficient, shape deviation coefficient, and drop point deviation coefficient are calculated respectively. Each deviation coefficient is obtained by subtracting from the corresponding standard value and normalizing it. The flow stability coefficient is obtained by weighted summation of the above deviation coefficients. By summing the drop viscosity, viscosity increase, viscosity increase rate, and flow stability coefficient, the drop viscosity response characteristics are obtained. Then, the qualified process parameters corresponding to the current glass cup shape are read, and the allowable temperature range of the feeding channel, the temperature range of the drip outlet, the temperature range of the drip path, the temperature range of the molding inlet, the total temperature decay range, the temperature decay rate range per unit time, the viscosity range upon entering the mold, the viscosity growth rate range, the drip weight range, and the drop point offset range are taken as the molding allowable conditions. These molding allowable conditions are then bound one-to-one with the drip temperature decay characteristics and the drip viscosity response characteristics to generate a drip temperature-viscosity molding window model. Through this process, the temperature decay, viscosity response, and molding allowable range of the drip from the feeding channel to the molding inlet can be unified into a single judgment model, enabling subsequent steps to accurately identify whether the drip deviates from the suitable temperature-viscosity window for molding.

[0024] Based on the mold state data, a mold thermal memory analysis result is generated. Based on the drop viscosity deviation state, the drop weight fluctuation state, and the mold thermal memory analysis result, a risk prediction of the glass forming process is performed to obtain a heat map of forming thermal defect risks.

[0025] In one embodiment, information such as mold surface temperature, mold cavity temperature difference, cooling status, opening and closing frequency, and continuous working cycle is first extracted from mold state data. This information is then organized into heat storage state data reflecting the accumulation of heat in the mold, and thermal drift state data reflecting the temperature shift of the mold with the production cycle. Subsequently, by combining temperature monitoring data from multiple areas such as the dripping area, molding area, and mold area, it is determined which changes in the current thermal state of the mold are caused by its own continuous operation and which changes are caused by temperature fluctuations in the surrounding area. This allows for the correction of the mold's heat storage and thermal drift, resulting in a mold thermal memory analysis result that more closely approximates the actual production state. After obtaining the mold thermal memory analysis result, it is used in conjunction with the dripping viscosity deviation state and the dripping weight fluctuation state for molding risk assessment. For example, abnormal dripping flowability may affect the roundness of the cup rim and the molding position, unstable dripping weight may cause uneven material distribution on the cup wall, and localized heat storage or thermal drift in the mold may exacerbate heat stagnation at the bottom of the cup, deviation in cup wall thickness, or deformation of the cup rim. The above processing methods assess the defect risk levels of areas such as the rim, wall, and bottom of the glass, and map the risk results of each area according to the structural location of the glass to create a heat map of molding thermal defect risks. This process combines historical heat accumulation in the mold, real-time temperature fluctuations, and abnormal dripping conditions to accurately display the location and risk intensity of potential thermal defects in various parts of the glass, providing a basis for subsequent annealing curve adjustments.

[0026] Furthermore, generating mold thermal memory analysis results based on the mold state data includes: Based on the mold state data, a mold heat storage state matrix and a mold thermal drift state matrix are constructed; based on the multi-region temperature monitoring data, the mold heat storage state matrix is ​​subjected to correlation influence identification to obtain the passive influence identification result of heat storage; based on the multi-region temperature monitoring data, the mold thermal drift state matrix is ​​subjected to correlation influence identification to obtain the passive influence identification result of thermal drift; based on the passive influence identification results of heat storage and thermal drift, the mold heat storage state matrix and the mold thermal drift state matrix are respectively compensated and corrected to obtain the mold thermal memory analysis result.

[0027] Preferably, the mold number is used as the main index, and the continuous production cycle is used as the time index. The mold status data is retrieved from the mold status data, including mold orifice temperature, mold cavity sidewall temperature, mold bottom temperature, cooling airflow, cooling water flow, mold opening and closing cycle, mold closing dwell time, and number of consecutive operations within a preset statistical period. Then, the mold orifice, mold cavity sidewall, and mold bottom are treated as matrix rows, and multiple consecutive production cycles are treated as matrix columns. The cumulative temperature value of each part under each production cycle, the temperature rise of adjacent cycles, the residual temperature after cooling, and the weighted value of the number of consecutive operations are written into the corresponding matrix. A mold thermal storage state matrix is ​​constructed to represent the degree of heat accumulation in different parts of the mold during continuous production. At the same time, using the reference stable temperature of the same mold as a reference value, the temperature drift of the mold opening, the temperature drift of the mold cavity sidewall, and the temperature drift of the mold bottom are calculated respectively. Combined with the changes in cooling air volume, cooling water flow rate, and opening and closing cycle, the offset values ​​of each part relative to the reference stable temperature under continuous cycle are written into the corresponding matrix unit to construct the mold thermal drift state matrix, which is used to represent the continuous shift trend of the mold temperature state with the production process.

[0028] After completing the matrix construction, the temperatures of the dripping zone, molding zone, annealing zone inlet, and cooling zone corresponding to the same time period for the mold are read from multi-region temperature monitoring data. These temperature data are then aligned with the mold's heat storage state matrix according to the production cycle. The impact of increased dripping zone temperature on the mold opening heat storage value, the impact of increased molding zone ambient temperature on the mold cavity sidewall heat storage value, and the impact of cooling zone temperature changes on the residual temperature at the mold bottom are calculated. The passive heating caused by external temperature changes is distinguished from the mold's own heat storage, generating a passive heat storage impact identification result. Each impact is obtained by multiplying the corresponding temperature deviation, time coupling weight, and heat transfer influence coefficient. The time coupling weight is calculated based on the time interval of heat impact transmission and the heat impact attenuation coefficient. Specifically, the heat impact attenuation coefficient is multiplied by the time interval from which the temperature change in the corresponding region is transmitted to the mold part to obtain the time attenuation amount, which is then negatively calculated and exponentially obtained. The generated passive heat storage impact identification result includes the passive impact area, passive impact intensity, corresponding production cycle, and affected mold part. Subsequently, the temperature monitoring data from the same multi-region area is aligned with the mold thermal drift state matrix. The influence of material supply temperature fluctuation, molding zone ambient temperature fluctuation, annealing inlet heat reflow, and cooling zone heat dissipation changes on mold orifice temperature drift, mold cavity sidewall temperature drift, and mold bottom temperature drift is calculated. The passive drift portion caused by changes in the external temperature field of the production line is identified in the mold thermal drift, and the passive influence identification result of thermal drift is generated. This passive influence identification result of thermal drift includes the thermal drift source area, drift influence direction, drift influence amplitude, and corresponding mold part. These influence quantities are obtained by multiplying the corresponding external temperature fluctuation, the drift transmission coefficient of the corresponding part, and the time coupling weight. The time decay amount is obtained by multiplying the attenuation coefficient of the thermal influence in the material supply area to the mold orifice by the time interval between the temperature change in the material supply area and the temperature response in the mold orifice. The time coupling weight is obtained by taking the negative value of the time decay amount and then performing an exponential operation.

[0029] Then, based on the identification results of passive heat storage effects, the mold heat storage state matrix is ​​compensated and corrected. That is, in each matrix cell, the passive heat storage effect caused by temperature changes in the dripping zone, forming zone, and cooling zone is deducted, while the active heat storage caused by the mold's continuous contact with high-temperature glass material, insufficient cooling, and cycle time accumulation is retained, resulting in the corrected mold heat storage state matrix. Simultaneously, based on the identification results of passive thermal drift effects, the mold thermal drift state matrix is ​​compensated and corrected. That is, in each matrix cell, the passive temperature shift caused by external temperature field fluctuations is deducted, while the actual thermal drift caused by mold material heat storage, cooling channel attenuation, and continuous opening and closing conditions is retained, resulting in the corrected mold thermal drift state matrix. Next, the corrected mold heat storage state matrix and the corrected mold thermal drift state matrix are fused together, and the thermal memory values ​​of the mold opening, mold cavity sidewall, and mold bottom are calculated respectively. Specifically, taking the current production cycle as the endpoint, the data of the most recent N production cycles are extracted from the corrected mold heat storage state matrix and the corrected mold thermal drift state matrix, and calculations are performed for the mold opening, mold cavity sidewall and mold bottom respectively, where N is a positive integer. For the die opening, read the die opening correction heat storage value and die opening correction thermal drift value from the most recent N production cycles. Divide the die opening correction heat storage value by the upper limit of die opening heat storage to obtain the die opening heat storage normalized value. Divide the absolute value of the die opening correction thermal drift value by the upper limit of die opening thermal drift to obtain the die opening thermal drift normalized value. Then, according to the rule that the closer to the current production cycle, the greater the weight, the more the die opening heat storage normalized value is weighted and summed to obtain the die opening heat storage memory component. The die opening thermal drift normalized value is weighted and summed to obtain the die opening thermal drift memory component. At the same time, count the number of cycles in the most recent N production cycles where the die opening heat storage normalized value or the die opening thermal drift normalized value exceeds the corresponding threshold, and divide this number by N to obtain the die opening abnormality persistence component. The mold opening thermal memory value is obtained by weighting and summing the mold opening thermal storage memory component, mold opening thermal drift memory component, and mold opening abnormality persistence component according to preset weights. Then, the same calculation method is used to calculate the mold cavity sidewall and mold bottom separately, yielding the mold cavity sidewall thermal memory value and mold bottom thermal memory value. The mold opening thermal memory value characterizes the degree of continuous heat storage and temperature drift accumulation at the mold opening location; the mold cavity sidewall thermal memory value characterizes the degree of continuous heat storage and temperature drift accumulation at the mold cavity sidewall location; and the mold bottom thermal memory value characterizes the degree of continuous heat storage and temperature drift accumulation at the mold bottom location. Finally, the mold opening thermal memory value, mold cavity sidewall thermal memory value, and mold bottom thermal memory value are combined to generate the mold thermal memory analysis result. This process eliminates the interference of external temperature fluctuations on mold condition judgment, accurately obtaining the heat accumulation and thermal drift state formed by the continuous production of the mold itself, providing a reliable basis for subsequent prediction of thermal defect risks in glass cup molding.

[0030] Furthermore, based on the deviation of the droplet viscosity, the fluctuation of the droplet weight, and the results of the mold thermal memory analysis, a risk prediction for the glass forming process is performed to obtain a heat map of forming thermal defects, including: Based on the deviation of the dripping material viscosity, the fluctuation of the dripping material weight, and the results of the mold thermal memory analysis, a cup rim deformation risk prediction is performed to obtain a cup rim deformation risk vector; based on the deviation of the dripping material viscosity, the fluctuation of the dripping material weight, and the results of the mold thermal memory analysis, a cup wall thickness deviation risk prediction is performed to obtain a cup wall thickness deviation risk vector; based on the deviation of the dripping material viscosity, the fluctuation of the dripping material weight, and the results of the mold thermal memory analysis, a cup bottom heat stagnation risk prediction is performed to obtain a cup bottom heat stagnation risk vector; the cup rim deformation risk vector, the cup wall thickness deviation risk vector, and the cup bottom heat stagnation risk vector are visualized and mapped to generate a molding thermal defect risk heat map.

[0031] Preferably, the glass preform corresponding to the current production cycle is used as the analysis object. The analysis reads the drip viscosity deviation state, drip weight fluctuation state, and mold thermal memory analysis results for that cycle. By performing drip flow offset analysis and mold entry eccentricity identification on the drip viscosity deviation state, the cup rim forming offset state is obtained. By performing cup rim material distribution analysis and edge thickness extrapolation on the drip weight fluctuation state, the cup rim weight imbalance state is obtained. By performing local thermal hysteresis analysis and mold rim cooling imbalance identification on the mold thermal memory analysis results, the cup rim thermal deformation state is obtained. These three states are then mapped to obtain the cup rim deformation risk vector. Subsequently, using the same prediction method, the cup wall thickness deviation risk is predicted based on the drip viscosity deviation state, drip weight fluctuation state, and mold thermal memory analysis results. By mapping the predicted cup wall forming offset state, the cup wall weight imbalance state, and the cup wall thermal thickness state, the cup wall thickness deviation risk vector is obtained. Similarly, the risk of heat stagnation at the bottom of the cup is predicted based on the deviation of the dripping viscosity, the fluctuation of the dripping weight, and the results of the mold thermal memory analysis. By mapping the predicted bottom forming offset, weight imbalance, and heat retention states, a heat stagnation risk vector is obtained. After obtaining the cup rim deformation risk vector, cup wall thickness deviation risk vector, and bottom heat stagnation risk vector, the glass structure is divided into the cup rim region, cup wall region, and cup bottom region. A two-dimensional unfolded coordinate diagram of the cup body is established. The cup rim deformation risk vector is then mapped to the cup rim region, the cup wall thickness deviation risk vector to the cup wall region, and the bottom heat stagnation risk vector to the cup bottom region. For each region, the risk level is determined based on the maximum value or weighted composite value of each component in the corresponding risk vector. Different display labels are assigned according to low risk, medium risk, and high risk, generating a molding thermal defect risk heat map that characterizes the distribution of thermal defect risks at the cup rim, cup wall, and cup bottom. Through the above processing, the effects of abnormal drip viscosity, drip weight fluctuation, and mold thermal memory on different parts of the cup can be transformed into intuitive regional risk results, providing a basis for the subsequent construction of cup-shaped matching annealing control curves, improving the accuracy of thermal defect prediction and the pertinence of subsequent annealing control.

[0032] Furthermore, based on the deviation of the dripping viscosity, the fluctuation of the dripping weight, and the results of the mold thermal memory analysis, a cup rim deformation risk prediction is performed to obtain a cup rim deformation risk vector, including: Based on the deviation of the dripping viscosity, dripping flow offset analysis and mold entry eccentricity identification are performed to obtain the cup rim forming offset state; based on the dripping weight fluctuation state, cup rim material distribution analysis and edge thickness extrapolation are performed to obtain the cup rim weight imbalance state; based on the mold thermal memory analysis results, local thermal hysteresis analysis and mold rim cooling imbalance identification are performed to obtain the cup rim thermal deformation state; the cup rim forming offset state, the cup rim weight imbalance state, and the cup rim thermal deformation state are correlated with risk mapping to generate the cup rim deformation risk vector.

[0033] Optionally, the analysis object is a single glass preform corresponding to the current production cycle. The analysis results of the drop viscosity deviation, drop weight fluctuation, and mold thermal memory are read. First, a drop flow offset analysis is performed based on the drop viscosity deviation. When the drop viscosity is higher than the upper limit of the allowable viscosity, it is determined that the drop flow is insufficient. The difference between the current drop viscosity entering the mold and the upper limit of the allowable viscosity is calculated to obtain the amount of viscosity exceeding the upper limit. The ratio of the amount of viscosity exceeding the upper limit to the allowable viscosity span is multiplied by the flow hysteresis conversion coefficient to obtain the flow hysteresis value. When the drop viscosity is lower than the lower limit of the allowable viscosity, the difference between the lower limit of the allowable viscosity and the current drop viscosity entering the mold is calculated to obtain the amount of viscosity exceeding the lower limit. The ratio of the amount of viscosity exceeding the lower limit to the allowable viscosity span is multiplied by the flow diffusion conversion coefficient to obtain the flow diffusion value. Subsequently, combining the droplet landing point offset and the preset cup rim center reference position, the distance deviation between the droplet entry center and the mold center is calculated using Euclidean distance. This distance deviation is taken as the entry eccentricity. The entry eccentricity is then divided by the allowable upper limit of entry eccentricity to obtain the normalized entry eccentricity value. The flow hysteresis value, flow diffusion value, and the normalized entry eccentricity value are then weighted and summed according to preset forming offset weights to obtain the cup rim forming offset value. By comparing the cup rim forming offset value with a preset offset level threshold, a low offset state, a medium offset state, or a high offset state is determined. The cup rim forming offset value, offset level, droplet number, and production cycle are combined to form the cup rim forming offset state.

[0034] The system analyzes the material distribution at the cup opening based on the fluctuation of the dripping weight. The current dripping weight is subtracted from the standard dripping weight to obtain the current dripping weight deviation. When the current dripping weight deviation is greater than zero, the direction of weight deviation is determined as the heavier direction; when the current dripping weight deviation is less than zero, the direction of weight deviation is determined as the lighter direction; when the current dripping weight deviation is equal to zero, the direction of weight deviation is determined as the no-deviation direction. The system takes the absolute value of the current dripping weight deviation to obtain the weight deviation amplitude, and calculates the weight fluctuation amount by comparing the weight deviation amplitude with the upper limit of the allowable weight deviation. When the ratio is greater than 1, the weight fluctuation amount is limited to 1. Simultaneously, the system reads the dripping weight over N consecutive production cycles, determines the maximum and minimum dripping weights, subtracts the minimum dripping weight from the maximum dripping weight to obtain the continuous dripping weight range, and calculates the continuous weight fluctuation coefficient by comparing the continuous dripping weight range with a preset weight fluctuation threshold. When the continuous weight fluctuation coefficient is greater than 1, the continuous weight fluctuation coefficient is limited to 1. Next, the weight fluctuation is multiplied by the first thickness influence weight to obtain the cup rim thickness influence value corresponding to the single drop weight deviation. The continuous weight fluctuation coefficient is multiplied by the second thickness influence weight to obtain the cup rim thickness influence value corresponding to continuous weight fluctuation. Finally, the two cup rim thickness influence values ​​are added together to obtain the cup rim edge thickness estimation value. By combining the weight deviation direction, weight fluctuation, continuous drop weight range, continuous weight fluctuation coefficient, and cup rim edge thickness estimation value, a cup rim weight imbalance state is generated.

[0035] Based on the mold thermal memory analysis results, local thermal hysteresis analysis is performed. The mold opening thermal memory value is compared with the mold opening thermal memory threshold. When the mold opening thermal memory value is greater than the mold opening thermal memory threshold, the continuous heat storage judgment result is determined to be continuous heat storage; when the mold opening thermal memory value is less than or equal to the mold opening thermal memory threshold, the continuous heat storage judgment result is determined to be non-continuous heat storage. The system calculates the ratio of the mold opening thermal memory value to the upper limit of the mold opening thermal memory value to obtain the mold opening thermal hysteresis degree value. When the ratio is greater than 1, the mold opening thermal hysteresis degree value is limited to 1. At the same time, the real-time temperature of multiple detection points around the mold opening is read, and the temperature difference between adjacent detection points is calculated. Then, the maximum temperature difference is selected from the multiple temperature differences of adjacent detection points, and the ratio of the maximum temperature difference to the preset mold opening temperature difference threshold is calculated to obtain the mold opening cooling unevenness value. When the ratio is greater than 1, the mold opening cooling unevenness value is limited to 1. When the maximum temperature difference exceeds the preset mold opening temperature difference threshold, the mold opening cooling imbalance judgment result is determined to indicate the presence of cooling imbalance; when the maximum temperature difference is less than or equal to the preset mold opening temperature difference threshold, the mold opening cooling imbalance judgment result is determined to indicate the absence of cooling imbalance. Next, the mold opening thermal hysteresis value is multiplied by the first thermal deformation influence weight to obtain the cup opening thermal deformation influence value corresponding to the mold opening thermal hysteresis. The mold opening cooling unevenness value is multiplied by the second thermal deformation influence weight to obtain the cup opening thermal deformation influence value corresponding to the mold opening cooling imbalance. The two cup opening thermal deformation influence values ​​are then added together to obtain the cup opening thermal deformation value. By combining the mold opening thermal memory value, the mold opening continuous heat storage judgment result, the mold opening thermal hysteresis value, the mold opening cooling imbalance judgment result, the mold opening cooling unevenness value, and the cup opening thermal deformation value, the cup opening thermal deformation state is generated.

[0036] Then, a risk mapping is performed on the cup rim forming offset state, cup rim weight imbalance state, and cup rim thermal deformation state. Specifically, the cup rim forming offset value, the cup rim edge thickness projection value, and the cup rim thermal deformation value are normalized to obtain normalized values ​​for forming offset, edge thickness projection, and thermal deformation. These are then weighted according to preset weights to obtain the comprehensive cup rim deformation risk value. When the forming offset normalized value is high, the cup rim eccentricity risk component is increased; when the edge thickness projection normalized value is high, the cup rim edge thickness unevenness risk component is increased; and when the thermal deformation normalized value is high, the cup rim roundness deformation risk component is increased. The system arranges the cup rim eccentricity risk component, the cup rim edge thickness unevenness risk component, and the cup rim roundness deformation risk component in a fixed order to generate a cup rim deformation risk vector. Through the above processing, the effects of drip flow deviation, cup rim material imbalance, and mold rim thermal deformation can be uniformly transformed into calculable cup rim deformation risk results, providing a data basis for the subsequent generation of molding thermal defect risk heat maps and improving the accuracy of cup rim defect prediction.

[0037] Based on the aforementioned heat map of molding thermal defects, a cup-shaped matching annealing control curve is constructed.

[0038] In one embodiment, the molding thermal defect risk heat map is first used to identify high-risk areas in the cup rim, wall, and bottom. For example, if the rim is high-risk, more precise temperature compensation and cooling control are needed during annealing; if the wall is high-risk, the annealing process should promote temperature uniformity across the wall thickness; and if the bottom is high-risk, adjustments to the heat release and slow cooling process at the bottom are required. Subsequently, annealing adjustment methods are determined for different areas. For the high-risk rim area, annealing adjustment parameters can be set to limit the edge cooling rate and reduce uneven shrinkage at the rim through temperature curve correction. For the high-risk wall area, annealing adjustment parameters can be set to gradually reduce the temperature difference at different heights and circumferential positions of the wall. For the high-risk bottom area, annealing adjustment parameters can be set to control the cooling slope, thereby reducing residual stress and cracking risk at the bottom. Finally, the annealing adjustment parameters corresponding to the rim, wall, and bottom of the glass are matched with the structural characteristics of the target glass shape. The heating, holding, and cooling stages of the annealing process are reconstructed as segmented curves to form a glass shape-matched annealing control curve suitable for the current glass shape and risk distribution. Through this process, the annealing curve can be specifically adjusted according to the defect risk in different areas of the glass, improving the matching degree between annealing control and the glass shape structure and actual forming state.

[0039] Furthermore, based on the aforementioned molding thermal defect risk heat map, a cup-shaped matching annealing control curve is constructed, including: Based on the heat map of molding thermal defects, high-risk areas at the cup rim, cup wall, and cup bottom are identified. For the high-risk areas at the cup rim, annealing compensation, edge cooling limitation, and roundness correction are performed to obtain rim annealing adjustment parameters. For the high-risk areas at the cup wall, wall thickness equalization, slow heat diffusion release, and temperature difference reduction are performed to obtain cup wall annealing adjustment parameters. For the high-risk areas at the cup bottom, heat stagnation release, slow bottom cooling, and stress reduction are performed to obtain cup bottom annealing adjustment parameters. Based on the rim annealing adjustment parameters, the cup wall annealing adjustment parameters, and the cup bottom annealing adjustment parameters, segmented curve reconstruction is performed on the annealing heating section, annealing holding section, and annealing cooling section to generate the cup shape matching annealing control curve.

[0040] Preferably, when constructing the cup-shaped matching annealing control curve based on the molding thermal defect risk heat map, the structural partition template of the current cup shape is first called to divide the glass cup into a rim area, a wall area, and a bottom area. The risk levels of the rim area, wall area, and bottom area are then read from the molding thermal defect risk heat map according to this structural partition template. The risk level of the rim area is then compared with the high-risk threshold for the rim. If the risk level of the rim area is greater than the high-risk threshold, the rim area is marked as a high-risk area. Similarly, the risk level of the wall area is compared with the high-risk threshold for the wall. If the risk level of the wall area is greater than the high-risk threshold for the wall, the wall area is marked as a high-risk area. Finally, the risk level of the bottom area is compared with the high-risk threshold for the bottom. If the risk level of the bottom area is greater than the high-risk threshold for the bottom, the bottom area is marked as a high-risk area. After identifying the high-risk area at the rim of the cup, the risk level of the rim area is read, and the difference between the risk level and the high-risk threshold is taken as the excess risk amount. The excess risk amount is then calculated as a ratio to the high-risk threshold to obtain the rim risk enhancement coefficient. The system multiplies the rim risk enhancement coefficient by a preset rim annealing compensation benchmark value to obtain the rim annealing compensation amount, which is used to increase the target temperature at the end of the annealing heating section. It also multiplies the rim risk enhancement coefficient by a preset edge cooling limit benchmark value to obtain the edge cooling limit amount, which is used to reduce the cooling rate of the rim area before entering the holding section. Finally, it multiplies the rim risk enhancement coefficient by a preset roundness correction benchmark value to obtain the roundness correction amount, which is used to extend the temperature equalization time of the rim area in the initial stage of annealing. The rim annealing compensation amount, edge cooling limit amount, and roundness correction amount are then combined to generate the rim annealing adjustment parameters.

[0041] After identifying high-risk areas on the cup wall, the system reads the risk level of each area and uses the difference between that risk level and the high-risk threshold as the excess risk amount. The excess risk amount is then compared to the high-risk threshold to obtain the cup wall risk enhancement coefficient. The system multiplies this coefficient by a preset wall thickness balance benchmark value to obtain the wall thickness balance adjustment amount; it also multiplies it by a preset heat diffusion mitigation benchmark value to obtain the heat diffusion mitigation amount; and it multiplies it by a preset temperature difference reduction benchmark value to obtain the temperature difference reduction amount. These three values ​​are then combined to generate the cup wall annealing adjustment parameters. Similarly, after identifying high-risk areas on the cup bottom, the system calculates the excess risk amount and compares it to the high-risk threshold to obtain the cup bottom risk enhancement coefficient. The system multiplies the cup bottom risk enhancement coefficient with the preset heat release benchmark value to obtain the heat release amount; multiplies the cup bottom risk enhancement coefficient with the preset bottom slow cooling benchmark value to obtain the bottom slow cooling adjustment amount; multiplies the cup bottom risk enhancement coefficient with the preset stress reduction benchmark value to obtain the stress reduction amount; and then combines the heat release amount, the bottom slow cooling adjustment amount, and the stress reduction amount to generate the cup bottom annealing adjustment parameters.

[0042] After obtaining the annealing adjustment parameters for the cup rim, cup wall, and cup bottom, read the standard annealing control curve corresponding to the current cup type. This standard annealing control curve includes the annealing heating section, the annealing holding section, and the annealing cooling section. The system superimposes the annealing compensation at the mouth of the annealing heating section onto the target temperature at the end of the annealing heating section, uses the edge cooling limitation to reduce the rate of temperature change between the end of the annealing heating section and the inlet of the annealing holding section, and uses the roundness correction to extend the temperature equalization time at the end of the annealing heating section, thus completing the reconstruction of the annealing heating section. Similarly, it superimposes the wall thickness equalization adjustment onto the holding time of the annealing holding section, uses the heat diffusion slow release to reduce the temperature drop rate in the latter part of the annealing holding section, and uses the temperature difference reduction to correct the allowable temperature fluctuation range of the annealing holding section, thus completing the reconstruction of the annealing holding section. Finally, it superimposes the heat retention release to the duration of the latter part of the annealing holding section, uses the bottom slow cooling adjustment to reduce the cooling slope of the annealing cooling section, and superimposes the stress reduction to the slow cooling duration of the annealing cooling section, thus completing the reconstruction of the annealing cooling section. Finally, according to the time sequence of the annealing heating section, annealing holding section, and annealing cooling section, the reconstructed segmented curves are connected to form a continuous temperature-time curve, generating a cup-shaped matching annealing control curve. Through the above processing, the annealing heating, holding and cooling processes can be quantitatively modified according to the actual high risk level of the cup rim, cup wall and cup bottom, so that the annealing control curve matches the current cup shape and thermal defect risk distribution, thereby improving the stress release effect of the annealing process and the quality stability of the finished glass cup.

[0043] Based on the drip temperature-viscosity molding window model and the mold thermal memory analysis results, the cup-shaped matching annealing control curve is reconstructed by drifting to obtain the production line temperature collaborative control strategy.

[0044] In one embodiment, the current temperature, viscosity, and molding suitability of the dripping material are first determined based on the dripping temperature-viscosity molding window model, and their potential impact on subsequent annealing temperature, holding time, and cooling rate are analyzed to generate feedforward compensation information corresponding to the dripping state. Subsequently, the impact of mold heat storage, thermal drift, and cooling imbalance on the thermal state of the molded cup is determined based on mold thermal memory analysis results, generating feedforward compensation information corresponding to the mold thermal state. Then, the two types of feedforward compensation information are integrated and used to adjust the heating, holding, and cooling parameters in the cup-shaped matching annealing control curve. This allows the annealing curve to adjust collaboratively with changes in dripping temperature and viscosity and mold thermal drift, ultimately outputting a production line temperature collaborative control strategy. Through this process, the annealing control can adapt to front-end temperature drift and mold thermal state changes in advance, improving the consistency of temperature control between different processes on the production line.

[0045] Furthermore, based on the dripping temperature-viscosity molding window model and the mold thermal memory analysis results, the cup-shaped matching annealing control curve is reconstructed by drifting to obtain a production line temperature collaborative control strategy, including: Based on the drip temperature-viscosity molding window model, the cup-shaped matching annealing control curve is mapped with annealing feedforward influence to obtain a first feedforward compensation vector; based on the mold thermal memory analysis results, the cup-shaped matching annealing control curve is mapped with annealing feedforward influence to obtain a second feedforward compensation vector; based on the first and second feedforward compensation vectors, the cup-shaped matching annealing control curve is reconstructed with control parameter linkage to output the production line temperature collaborative control strategy.

[0046] Preferably, when mapping the annealing feedforward influence of the cup-shaped matching annealing control curve according to the drip temperature and viscosity molding window model, the current drip temperature deviation, viscosity deviation, temperature decay deviation, and weight fluctuation value are first read from the drip temperature and viscosity molding window model, and then the ratios are calculated with the allowable deviation thresholds corresponding to the current cup shape to obtain the drip temperature influence coefficient, viscosity influence coefficient, temperature decay influence coefficient, and weight fluctuation influence coefficient. Subsequently, the mold inlet temperature influence coefficient is multiplied by the temperature compensation benchmark value of the heating section to obtain the target temperature compensation amount of the heating section. The viscosity influence coefficient is multiplied by the time compensation benchmark value of the holding section to obtain the holding time compensation amount. The temperature decay influence coefficient is multiplied by the heating rate limit benchmark value to obtain the heating rate limit amount. The weight fluctuation influence coefficient is multiplied by the temperature equalization time compensation benchmark value to obtain the temperature equalization time compensation amount. Then, the target temperature compensation amount of the heating section, the heating rate limit amount, the holding time compensation amount, and the temperature equalization time compensation amount are arranged in a preset order to generate the first feedforward compensation vector, which is used to characterize the advance correction requirements of the annealing control curve due to the fluctuations in dripping temperature, viscosity, and weight. When mapping the annealing feedforward effect of the cup-shaped matching annealing control curve based on the mold thermal memory analysis results, the thermal memory values ​​of the mold opening, mold cavity sidewall, and mold bottom are read and compared with the thermal memory thresholds of the mold opening, mold cavity sidewall, and mold bottom, respectively. When the corresponding thermal memory value is greater than the corresponding thermal memory threshold, the difference between the thermal memory value and the thermal memory threshold is taken as the thermal memory excess. Then, the ratio of the thermal memory excess to the corresponding thermal memory threshold is calculated to obtain the thermal memory enhancement coefficient of the mold opening, the thermal memory enhancement coefficient of the mold cavity sidewall, and the thermal memory enhancement coefficient of the mold bottom. Next, the mold opening thermal memory enhancement coefficient is multiplied by the cup opening equalization time compensation benchmark value to obtain the cup opening equalization time compensation amount; the mold cavity sidewall thermal memory enhancement coefficient is multiplied by the cup wall temperature difference reduction benchmark value to obtain the cup wall temperature difference reduction amount; the mold cavity sidewall thermal memory enhancement coefficient is multiplied by the cup wall heat preservation extension benchmark value to obtain the cup wall heat preservation extension amount; the mold bottom thermal memory enhancement coefficient is multiplied by the cooling rate correction benchmark value to obtain the cooling rate correction amount; the mold bottom thermal memory enhancement coefficient is multiplied by the bottom slow cooling time compensation benchmark value to obtain the bottom slow cooling time compensation amount. Then, the cup opening equalization time compensation amount, cup wall temperature difference reduction amount, cup wall heat preservation extension amount, cooling rate correction amount, and bottom slow cooling time compensation amount are arranged in a preset order to generate a second feedforward compensation vector, which is used to characterize the advance correction requirements of continuous heat storage and thermal drift of the mold on the annealing control curve.

[0047] After obtaining the first and second feedforward compensation vectors, the control parameters of the cup-shaped matching annealing control curve are reconstructed in a coordinated manner. Specifically, the target temperature, heating rate, holding temperature, holding time, cooling rate, and slow cooling time of the cup-shaped matching annealing control curve are read first. Then, the target temperature compensation amount of the heating segment in the first feedforward compensation vector is superimposed on the target temperature of the heating segment, the heating rate limit is used to reduce the heating rate of the heating segment, and the holding time compensation amount and temperature equalization time compensation amount are superimposed on the holding time of the holding segment. At the same time, the cup mouth equalization time compensation amount in the second feedforward compensation vector is superimposed on the temperature equalization time at the end of the heating segment, the cup wall temperature difference reduction amount is used to reduce the temperature fluctuation range of the holding segment, the cup wall holding extension amount is superimposed on the holding time of the holding segment, the cooling rate correction amount is used to reduce the cooling rate of the cooling segment, and the bottom slow cooling time compensation amount is superimposed on the slow cooling time of the cooling segment. To avoid over-correction of a single parameter, upper and lower limits are imposed on each corrected control parameter during the linkage reconstruction. When the corrected target temperature of the heating segment exceeds the maximum allowable annealing temperature for the current cup shape, it is limited to the maximum allowable annealing temperature. When the corrected heating rate is lower than the minimum allowable heating rate for the current cup shape, it is limited to the minimum allowable heating rate. When the corrected holding time exceeds the maximum allowable holding time of the equipment, it is limited to the maximum allowable holding time of the equipment. When the corrected cooling rate is lower than the minimum allowable cooling rate of the process, it is limited to the minimum allowable cooling rate of the process. After completing the constraints, the connection points between the heating, holding, and cooling segments are smoothed to ensure continuous temperature changes between adjacent segments, resulting in the annealing control curve after drift-coordinated reconstruction.

[0048] Finally, the reconstructed annealing control curve is converted into a production line temperature collaborative control strategy. This strategy includes commands for the target temperature of the annealing heating stage, the annealing heating rate, the annealing holding temperature, the annealing holding time, the annealing cooling rate, the annealing slow cooling time, mold cooling compensation, and the temperature stabilization of the dripping zone. Through this process, the effects of the leading edge deviation of the dripping temperature-viscosity molding window and the continuous cumulative effects of mold thermal memory are combined and applied to the annealing control curve. This allows the annealing parameters to be collaboratively adjusted according to the drift of the dripping state and the drift of the mold thermal state, improving the consistency of temperature control and the stability of molding quality in the glass production line.

[0049] Example 2, based on the same inventive concept as the real-time temperature field monitoring and control method for a glass production line in the foregoing examples, such as... Figure 2 As shown in the figure, this application provides a real-time monitoring and control system for the temperature field of a glass production line. The real-time monitoring and control system for the temperature field of a glass production line includes: Data acquisition module 11: Acquires dripping status data, mold status data, and multi-area temperature monitoring data of the glass production line; Dripping identification module 12: Establishes a dripping temperature-viscosity forming window model based on the multi-area temperature monitoring data and the dripping status data, and identifies the dripping viscosity deviation state and dripping weight fluctuation state based on the dripping temperature-viscosity forming window model; Risk prediction module 13: Generates mold thermal memory analysis results based on the mold status data, and performs risk prediction of the glass forming process based on the dripping viscosity deviation state, the dripping weight fluctuation state, and the mold thermal memory analysis results, and obtains a forming thermal defect risk heat map; Curve construction module 14: Constructs a cup shape matching annealing control curve based on the forming thermal defect risk heat map; Collaborative reconstruction module 15: Performs drift collaborative reconstruction of the cup shape matching annealing control curve based on the dripping temperature-viscosity forming window model and the mold thermal memory analysis results, and obtains a production line temperature collaborative control strategy.

[0050] Furthermore, the data acquisition module 11 is used to perform the following operation steps: Temperature data corresponding to the dripping zone, forming zone, mold zone, annealing zone and cooling zone in the glass production line are collected to generate the multi-zone temperature monitoring data.

[0051] Furthermore, the data acquisition module 11 is used to perform the following operation steps: Collect the dripping temperature, dripping weight, dripping interval, and dripping morphology parameters in the dripping zone to generate the dripping state data.

[0052] Furthermore, the dripping identification module 12 is used to perform the following operation steps: Based on the multi-region temperature monitoring data, a dripping temperature feature sequence is constructed; time-coupled temperature difference calculation is performed based on the dripping temperature feature sequence to obtain the dripping temperature decay feature; viscosity response matching is performed based on the dripping temperature decay feature and the dripping state data to determine the dripping viscosity response feature; molding allowable feature binding is performed based on the dripping temperature decay feature and the dripping viscosity response feature to establish the dripping temperature viscosity molding window model.

[0053] Furthermore, the risk prediction module 13 is used to perform the following operation steps: Based on the mold state data, a mold heat storage state matrix and a mold thermal drift state matrix are constructed; based on the multi-region temperature monitoring data, the mold heat storage state matrix is ​​subjected to correlation influence identification to obtain the passive influence identification result of heat storage; based on the multi-region temperature monitoring data, the mold thermal drift state matrix is ​​subjected to correlation influence identification to obtain the passive influence identification result of thermal drift; based on the passive influence identification results of heat storage and thermal drift, the mold heat storage state matrix and the mold thermal drift state matrix are respectively compensated and corrected to obtain the mold thermal memory analysis result.

[0054] Furthermore, the risk prediction module 13 is used to perform the following operation steps: Based on the deviation of the dripping material viscosity, the fluctuation of the dripping material weight, and the results of the mold thermal memory analysis, a cup rim deformation risk prediction is performed to obtain a cup rim deformation risk vector; based on the deviation of the dripping material viscosity, the fluctuation of the dripping material weight, and the results of the mold thermal memory analysis, a cup wall thickness deviation risk prediction is performed to obtain a cup wall thickness deviation risk vector; based on the deviation of the dripping material viscosity, the fluctuation of the dripping material weight, and the results of the mold thermal memory analysis, a cup bottom heat stagnation risk prediction is performed to obtain a cup bottom heat stagnation risk vector; the cup rim deformation risk vector, the cup wall thickness deviation risk vector, and the cup bottom heat stagnation risk vector are visualized and mapped to generate a molding thermal defect risk heat map.

[0055] Furthermore, the risk prediction module 13 is used to perform the following operation steps: Based on the deviation of the dripping viscosity, dripping flow offset analysis and mold entry eccentricity identification are performed to obtain the cup rim forming offset state; based on the dripping weight fluctuation state, cup rim material distribution analysis and edge thickness extrapolation are performed to obtain the cup rim weight imbalance state; based on the mold thermal memory analysis results, local thermal hysteresis analysis and mold rim cooling imbalance identification are performed to obtain the cup rim thermal deformation state; the cup rim forming offset state, the cup rim weight imbalance state, and the cup rim thermal deformation state are correlated with risk mapping to generate the cup rim deformation risk vector.

[0056] Furthermore, the curve construction module 14 is used to perform the following operation steps: Based on the heat map of molding thermal defects, high-risk areas at the cup rim, cup wall, and cup bottom are identified. For the high-risk areas at the cup rim, annealing compensation, edge cooling limitation, and roundness correction are performed to obtain rim annealing adjustment parameters. For the high-risk areas at the cup wall, wall thickness equalization, slow heat diffusion release, and temperature difference reduction are performed to obtain cup wall annealing adjustment parameters. For the high-risk areas at the cup bottom, heat stagnation release, slow bottom cooling, and stress reduction are performed to obtain cup bottom annealing adjustment parameters. Based on the rim annealing adjustment parameters, the cup wall annealing adjustment parameters, and the cup bottom annealing adjustment parameters, segmented curve reconstruction is performed on the annealing heating section, annealing holding section, and annealing cooling section to generate the cup shape matching annealing control curve.

[0057] Furthermore, the collaborative reconfiguration module 15 is used to perform the following operation steps: Based on the drip temperature-viscosity molding window model, the cup-shaped matching annealing control curve is mapped with annealing feedforward influence to obtain a first feedforward compensation vector; based on the mold thermal memory analysis results, the cup-shaped matching annealing control curve is mapped with annealing feedforward influence to obtain a second feedforward compensation vector; based on the first and second feedforward compensation vectors, the cup-shaped matching annealing control curve is reconstructed with control parameter linkage to output the production line temperature collaborative control strategy.

[0058] Through the foregoing detailed description of a real-time temperature field monitoring and control method for a glass production line, those skilled in the art can clearly understand the real-time temperature field monitoring and control system for a glass production line in this embodiment. Since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and relevant parts can be referred to the method section.

[0059] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for real-time monitoring and control of the temperature field in a glass production line, characterized in that, The method includes: Acquire data on the dripping status, mold status, and multi-zone temperature monitoring of the glass production line; A temperature-viscosity forming window model for dripping material is established based on the multi-region temperature monitoring data and the dripping material state data. The dripping material viscosity deviation state and the dripping material weight fluctuation state are then identified based on the temperature-viscosity forming window model. Based on the mold state data, generate mold thermal memory analysis results, and perform glass cup forming process risk prediction based on the drop viscosity deviation state, the drop weight fluctuation state, and the mold thermal memory analysis results to obtain a forming thermal defect risk heat map; Based on the aforementioned heat map of molding thermal defect risks, a cup-shaped matching annealing control curve is constructed. Based on the drip temperature-viscosity molding window model and the mold thermal memory analysis results, the cup-shaped matching annealing control curve is reconstructed by drifting to obtain the production line temperature collaborative control strategy.

2. The real-time monitoring and control method for the temperature field of a glass production line as described in claim 1, characterized in that, A temperature-viscosity molding window model for dripping material is established based on the multi-region temperature monitoring data and the dripping state data, including: Based on the multi-region temperature monitoring data, a dripping temperature characteristic sequence is constructed; Based on the dripping temperature characteristic sequence, time-coupled temperature difference calculation is performed to obtain the dripping temperature decay characteristics; Based on the dripping temperature decay characteristics and the dripping state data, viscosity response matching is performed to determine the dripping viscosity response characteristics; Based on the drip temperature decay characteristics and the drip viscosity response characteristics, the molding allowable characteristics are bound together to establish the drip temperature viscosity molding window model.

3. The real-time monitoring and control method for the temperature field of a glass production line as described in claim 1, characterized in that, Based on the mold state data, mold thermal memory analysis results are generated, including: Based on the mold state data, construct the mold heat storage state matrix and the mold thermal drift state matrix; Based on the multi-region temperature monitoring data, the correlation influence identification of the mold heat storage state matrix is ​​performed to obtain the heat storage passive influence identification result; Based on the multi-region temperature monitoring data, the correlation influence identification of the mold thermal drift state matrix is ​​performed to obtain the thermal drift passive influence identification result; Based on the identification results of passive effects of heat storage and passive effects of thermal drift, the mold heat storage state matrix and the mold thermal drift state matrix are compensated and corrected respectively to obtain the mold thermal memory analysis results.

4. The method for real-time monitoring and control of the temperature field in a glass production line as described in claim 1, characterized in that, Based on the deviation of the droplet viscosity, the fluctuation of the droplet weight, and the results of the mold thermal memory analysis, a risk prediction for the glass forming process is performed, and a heat map of forming thermal defects is obtained, including: Based on the deviation of the dripping viscosity, the fluctuation of the dripping weight, and the results of the mold thermal memory analysis, a cup mouth deformation risk prediction is performed to obtain a cup mouth deformation risk vector. Based on the deviation of the droplet viscosity, the fluctuation of the droplet weight, and the results of the mold thermal memory analysis, a cup wall thickness deviation risk prediction is performed to obtain a cup wall thickness deviation risk vector. Based on the deviation of the droplet viscosity, the fluctuation of the droplet weight, and the results of the mold thermal memory analysis, perform a cup bottom heat stagnation risk prediction to obtain a cup bottom heat stagnation risk vector. The cup rim deformation risk vector, the cup wall thickness deviation risk vector, and the cup bottom heat stagnation risk vector are visualized and mapped to generate the molding thermal defect risk heat map.

5. The real-time monitoring and control method for the temperature field of a glass production line as described in claim 4, characterized in that, Based on the deviation of the dripping viscosity, the fluctuation of the dripping weight, and the results of the mold thermal memory analysis, a cup rim deformation risk prediction is performed to obtain a cup rim deformation risk vector, including: Based on the deviation of the droplet viscosity, droplet flow offset analysis and mold entry eccentricity identification are performed to obtain the cup mouth forming offset state. Based on the fluctuation of the dripping weight, the distribution of material at the cup rim and the deduction of the edge thickness are performed to obtain the weight imbalance state at the cup rim. Based on the results of the mold thermal memory analysis, local thermal hysteresis analysis and mold orifice cooling imbalance identification are performed to obtain the cup orifice thermal deformation state. A risk mapping is performed on the cup rim forming offset state, the cup rim weight imbalance state, and the cup rim thermal deformation state to generate the cup rim deformation risk vector.

6. The method for real-time monitoring and control of the temperature field in a glass production line as described in claim 1, characterized in that, Based on the aforementioned molding thermal defect risk heat map, a cup-shaped matching annealing control curve is constructed, including: The high-risk areas at the cup rim, cup wall, and cup bottom are identified based on the molding thermal defect risk heat map. Based on the high-risk areas at the cup rim, perform rim annealing compensation, edge cooling limitation, and roundness correction to obtain cup rim annealing adjustment parameters; Based on the high-risk areas of the cup wall, wall thickness equalization, heat diffusion slow release, and temperature difference reduction are performed to obtain cup wall annealing adjustment parameters; Based on the high-risk area at the bottom of the cup, heat release, slow cooling of the bottom, and stress reduction are performed to obtain the annealing adjustment parameters for the bottom of the cup. Based on the annealing adjustment parameters of the cup rim, the cup wall, and the cup bottom, the segmented curves of the annealing heating section, the annealing heat preservation section, and the annealing cooling section are reconstructed to generate the cup-shaped matching annealing control curve.

7. The method for real-time monitoring and control of the temperature field in a glass production line as described in claim 1, characterized in that, Based on the drip temperature-viscosity molding window model and the mold thermal memory analysis results, the cup-shaped matching annealing control curve is reconstructed by drifting to obtain a production line temperature collaborative control strategy, including: Based on the dripping temperature-viscosity forming window model, the cup-shaped matching annealing control curve is mapped to annealing feedforward influence to obtain the first feedforward compensation vector. Based on the mold thermal memory analysis results, the cup-shaped matching annealing control curve is mapped to annealing feedforward influence to obtain the second feedforward compensation vector. Based on the first feedforward compensation vector and the second feedforward compensation vector, the control parameters of the cup-shaped matching annealing control curve are reconstructed in a coordinated manner, and the production line temperature collaborative control strategy is output.

8. The method for real-time monitoring and control of the temperature field in a glass production line as described in claim 1, characterized in that, The method includes: Temperature data corresponding to the dripping zone, forming zone, mold zone, annealing zone and cooling zone in the glass production line are collected to generate the multi-zone temperature monitoring data.

9. The real-time monitoring and control method for the temperature field of a glass production line as described in claim 1, characterized in that, The method includes: Collect the dripping temperature, dripping weight, dripping interval, and dripping morphology parameters in the dripping zone to generate the dripping state data.

10. A real-time monitoring and control system for the temperature field of a glass production line, characterized in that, For implementing the real-time temperature field monitoring and control method for a glass production line according to any one of claims 1-9, the system comprises: Data acquisition module: Acquires dripping status data, mold status data, and multi-zone temperature monitoring data from the glass production line; Droplet identification module: Based on the multi-region temperature monitoring data and the droplet status data, a droplet temperature and viscosity forming window model is established, and the droplet viscosity deviation state and droplet weight fluctuation state are identified based on the droplet temperature and viscosity forming window model; Risk prediction module: Generates mold thermal memory analysis results based on the mold status data, and performs risk prediction for the glass forming process based on the drop viscosity deviation state, the drop weight fluctuation state, and the mold thermal memory analysis results, and obtains a heat map of forming thermal defect risks; Curve construction module: Based on the aforementioned molding thermal defect risk heat map, construct the cup-shaped matching annealing control curve; Collaborative Reconstruction Module: Based on the dripping temperature-viscosity molding window model and the mold thermal memory analysis results, the cup-shaped matching annealing control curve is reconstructed by drift to obtain the production line temperature collaborative control strategy.

Images