A closed-loop control method for high-temperature adhesive tape coating process

Abstract

This invention relates to a closed-loop control method for a high-temperature tape coating process in the field of next-generation information technology. The method includes: acquiring adhesive layer thickness distribution data during the coating process; comparing the adhesive layer thickness distribution data with a preset thickness standard range to generate parameter correction instructions for the coating equipment; determining an optimized configuration for the curing process based on a coating uniformity analysis report, wherein the optimized configuration is used to adjust the temperature distribution during the curing process; monitoring the temperature distribution state during the curing process based on the optimized configuration, and triggering a temperature adjustment mechanism based on dynamic data of temperature uniformity to acquire curing effect data; and analyzing the adjustability of process parameters based on the curing effect data and preset performance testing standards to determine directions for improving the continuity of the production process.

Description

Technical Field

[0001] This invention relates to the field of next-generation information technology, and in particular to a closed-loop control method for a high-temperature tape coating process. Background Technology

[0002] In modern industrial manufacturing, high-temperature adhesive tapes are a key material widely used in industries such as electronics, automotive, and aerospace. Their performance directly impacts product reliability and safety. In particular, adhesion and stability under high-temperature environments are crucial indicators of material quality. However, despite the undeniable importance of this field, existing production methods struggle to meet the demands for high efficiency and quality, necessitating technological innovation to drive industry progress.

[0003] Current high-temperature tape manufacturing methods generally suffer from fragmented production processes, with a lack of effective connections between different steps, resulting in long overall production cycles and significant resource waste. At a deeper level, this fragmentation is not only reflected in the dispersion of equipment and processes, but also in the difficulty of striking a balance between optimizing material performance and improving production efficiency. Bottlenecks in one stage often affect the entire process, thus limiting the product's market competitiveness.

[0004] Focusing on specific challenges, coating uniformity during the production process becomes the primary issue. Since the coating process directly determines the adhesion between the adhesive layer and the substrate, uneven coating thickness due to excessively fast coating speed or improper process control leads to decreased adhesion at high temperatures. This issue further affects the drying and curing process; if the coating quality in the preceding steps is unstable, it becomes difficult to achieve uniform curing of the adhesive layer during drying, ultimately resulting in the product's tackiness failing to meet ideal standards at extreme temperatures. These two factors are interconnected and together constitute the core technical obstacle in the manufacturing process.

[0005] Therefore, how to achieve continuous and efficient production processes while ensuring coating uniformity and curing effect has become a key issue in improving the performance and production efficiency of high-temperature tapes. Summary of the Invention

[0006] This invention provides a closed-loop control method for a high-temperature tape coating process, mainly comprising:

[0007] The process involves acquiring adhesive layer thickness distribution data during the coating process, comparing this data with a preset thickness standard range, and generating parameter correction instructions for the coating equipment. The operating parameters of the coating equipment are updated according to these instructions, and the corrected adhesive layer thickness distribution data is acquired. Uniformity testing is performed based on this data, generating a coating uniformity analysis report. Based on the report, an optimized configuration for the curing process is determined, which adjusts the temperature distribution during curing. The optimized configuration is then used to monitor the temperature distribution during curing and trigger adjustments based on dynamic temperature uniformity data. A temperature regulation mechanism is used to acquire curing effect data. Based on the curing effect data and preset performance test standards, the adjustability of process parameters is analyzed to determine the direction for improving the continuity of the production process. Based on the direction for improving the continuity of the production process, relevant data on the connection between equipment processes are integrated to obtain the collaborative operation status between each process. Optimization and adjustment are made for delays or interruptions in the connection links to generate a resource utilization efficiency improvement plan. Based on the resource utilization efficiency improvement plan, the optimized process data is transmitted to the production management system to obtain real-time monitoring results of the production cycle duration. Dynamic adjustments are made for cycle fluctuations to determine the final optimized configuration of material performance. Furthermore, the step of acquiring adhesive layer thickness distribution data during the coating process, and comparing the adhesive layer thickness distribution data with a preset thickness standard range to generate parameter correction instructions for the coating equipment, includes: acquiring a set of adhesive layer thickness control parameters associated with the target substrate type and coating speed from a pre-established coating parameter database, the parameter set including coating head pressure parameters and a preset adhesive layer thickness standard range; using a thickness sensor in a real-time monitoring system to collect a lateral thickness distribution data sequence of the adhesive layer during the coating process, the data sequence being indexed by a timestamp and position coordinates along the substrate width direction; comparing the thickness distribution data sequence with the preset thickness standard range in the parameter set, calculating the deviation of the thickness value at each position coordinate, and obtaining a thickness deviation distribution map composed of the deviations at each position; if the overall deviation of the thickness deviation distribution map exceeds a preset uniformity threshold, then based on the spatial distribution characteristics of the deviation in the thickness deviation distribution map, retrieving the corresponding corrected coating head pressure parameter from the coating parameter database, and updating the corrected coating head pressure parameter to the control instructions of the coating equipment.Furthermore, the step of updating the operating parameters of the coating equipment according to the parameter correction instructions and obtaining the corrected adhesive layer thickness distribution data, performing uniformity detection based on the corrected adhesive layer thickness distribution data, and generating a coating uniformity analysis report includes: issuing pressure adjustment instructions to the partition control unit of the coating head according to the parameter correction instruction sequence of the coating equipment; the instructions drive the actuator to change the pressure setpoint of each unit, completing the dynamic update of the operating parameters; simultaneously recording the timestamp and instruction version number of this parameter update, and generating a parameter change log; after the coating process continues to run and the parameters are stable, triggering a new round of thickness data acquisition, using the same scanning path and sampling frequency as the initial detection, obtaining the corrected thickness distribution matrix covering the entire substrate surface, with the row and column indices of the matrix aligned with the coordinates of the initial detection; calculating the arithmetic mean of all data points in the corrected thickness distribution matrix as the overall thickness mean, and calculating the standard deviation of these data points as the overall thickness standard deviation, while simultaneously... The corrected thickness distribution matrix is ​​compared with the initial thickness distribution matrix stored in the historical database point by point to obtain a thickness variation matrix. If the overall thickness standard deviation is lower than a preset convergence threshold, the coating uniformity is determined to have met the process requirements, and the iteration process terminates. If the overall thickness standard deviation is not lower than the convergence threshold, a preset abnormal region identification rule is applied to the thickness variation matrix. The rule is based on comparing the absolute value of the thickness variation with the preset variation threshold to identify the set of pixels whose variation exceeds the variation threshold. Connectivity analysis is then performed on the pixel set to obtain the set of location coordinates of the residual thickness abnormal regions and the average variation of each region. The parameter change log, the overall thickness mean, the overall thickness standard deviation, the convergence determination result, and the set of location coordinates and average variation of the residual thickness abnormal regions are integrated to generate an updated coating uniformity analysis report containing the number of iterations, the final process status, and a detailed data summary, according to a preset report template.Furthermore, determining the optimized configuration of the curing process based on the coating uniformity analysis report includes: obtaining the set of location coordinates and average variation of the residual thickness anomaly area from the updated coating uniformity analysis report; establishing a mapping table between the coating surface coordinates and the oven heating zone numbers based on the transfer path and relative position of the substrate between the coating station and the curing oven; using the mapping table, converting the coating surface coordinates of the residual thickness anomaly area into the corresponding target heating zone number set; for each zone in the target heating zone number set, querying a preset thickness-temperature compensation lookup table based on the average variation of the associated residual thickness anomaly area to obtain a basic temperature compensation value; obtaining the current set temperature of all heating zones in the oven to form a current temperature distribution vector; if the application of the basic temperature compensation value causes the temperature of any heating zone to exceed the process safety range, then limiting the compensation value for that zone. The process involves: generating a preliminary zone temperature adjustment instruction set based on the target heating zone number set and the temperature compensation value after amplitude limiting; using a heat conduction model to calculate the predicted temperature field distribution on the entire substrate curing surface after executing the preliminary zone temperature adjustment instruction set; the inputs of the heat conduction model include heating element layout parameters, substrate physical property parameters, and oven air velocity parameters; analyzing the predicted temperature field distribution, if there is a region where the temperature difference between adjacent zones exceeds a preset threshold, then that region is determined to be a local thermal stress risk region; if such a local thermal stress risk region exists, then a heat load balancing process is initiated. This heat load balancing process uses the preliminary zone temperature adjustment instruction set as input and minimizes the maximum adjacent zone temperature difference as the adjustment target, iteratively fine-tuning the temperature settings of non-target heating zones to generate an optimized zone temperature adjustment instruction set; the optimized zone temperature adjustment instruction set and the mapping relationship table together constitute the optimized configuration of the curing process.Furthermore, based on the optimized configuration of the curing process, the temperature distribution during the curing process is monitored, and a temperature adjustment mechanism is triggered based on the dynamic data of temperature uniformity to obtain curing effect data. This includes: acquiring real-time temperature readings from multiple temperature measurement points within the curing oven to form an original data sequence for temperature field monitoring; processing the original data sequence using a sliding window method to calculate the standard deviation and range of the temperature in each zone within each window period to complete uniformity quantification; comparing the result of the uniformity quantification with a preset uniformity threshold; if the comparison result shows that the temperature fluctuation in any region continuously exceeds the preset uniformity threshold, it is determined that the region has local overheating or underheating, and the determination includes the abnormal zone number and the direction of temperature deviation; matching the heating zone mapping relationship in the optimized configuration of the curing process based on the abnormal zone number to determine the target control zone and its required temperature adjustment direction; and using the heat conduction model in the optimized configuration of the curing process, with the current temperature distribution and the target control zone and its temperature adjustment direction as input, simulating different temperatures. The future temperature field evolution under the adjustment scheme is analyzed. The simulation output of the heat conduction model includes the predicted temperature field distribution and the temperature gradient changes between adjacent intervals. The scheme that makes the target zone temperature approach the set value and minimizes the temperature gradient change is selected, and a dynamic setting command containing the specific zone temperature set value is generated. The dynamic setting command is sent to the oven control system for execution. During execution, temperature field monitoring data is continuously acquired and compared with the predicted temperature field of the heat conduction model. If the deviation between the actual temperature and the predicted temperature exceeds the preset allowable deviation range, a risk warning is triggered. The risk warning triggers the correction of the dynamic setting command based on the deviation. After completing a full curing cycle, the curing degree detection data of the substrate is collected to form effect feedback data. A regression analysis method is used to establish a correlation model between the effect feedback data and the historical temperature data of the target control zone during the execution of the dynamic setting command. The output of the correlation model is used to pre-correct the basic temperature set value in the curing process optimization configuration before the start of the next curing process.Furthermore, the step of analyzing the adjustability of process parameters and determining the direction for improving the continuity of the production process based on the curing effect data and preset performance test standards includes: acquiring the degree of curing and high-temperature shear strength data of the material in a high-temperature, high-humidity, or corrosive medium simulation chamber, wherein the data constitutes a standardized extreme environment performance dataset; completing the adhesion performance test using the fixtures and loading rates specified in the high-temperature adhesion performance test standard; performing regression calculations using a multiple linear regression algorithm, with curing temperature, holding pressure, and curing time as independent variables, and the degree of curing and high-temperature shear strength from the standardized extreme environment performance dataset as dependent variables, to obtain a degree of curing regression equation and a high-temperature shear strength regression equation, which together constitute a correlation model between process parameters and key performance indicators; setting a degree of curing qualification threshold and a high-temperature shear strength qualification threshold based on the degree of curing regression equation and the high-temperature shear strength regression equation in the correlation model; and solving for the output value of both equations in the three-dimensional parameter space that simultaneously satisfies the condition that the output value is greater than or equal to... The set of all curing temperature, holding pressure, and curing time parameter combinations corresponding to the qualified threshold is defined as the adjustable safety boundary of the process parameter, and the projection interval of the set on the parameter coordinate axis is defined as the adjustable safety boundary of the process parameter. The actual recorded values ​​of the curing temperature, holding pressure, and curing time in continuous production batches are collected, and the difference between the maximum and minimum values ​​of the actual recorded values ​​of each parameter is calculated to obtain the historical fluctuation range. If the historical fluctuation range of a certain parameter completely exceeds its corresponding adjustable safety boundary, the parameter is determined to be a constraint on the continuity of the production process. For the process parameter determined to be a constraint, within the adjustable safety boundary, an orthogonal experimental design method is used to arrange multiple sets of process parameter combinations for small-batch trial production. The curing degree and high-temperature shear strength data of the trial production samples are obtained, and the average distance between the sample performance data and the median value of the corresponding qualified threshold under each set of parameter combinations is calculated. The scheme with the smallest average distance and whose parameter values ​​are farthest from the endpoint of the adjustable safety boundary is selected as the optimized setting value of the production process.Furthermore, based on the improvement direction of the production process continuity, relevant data on equipment process connectivity are integrated to obtain the collaborative operation status between each process, and optimization and adjustment are made for delays or interruptions in the connection links to generate a resource utilization efficiency improvement plan. This includes: obtaining equipment operation logs and material flow sequence from the production line control system, and obtaining process cycle time data and planned execution deviation from the manufacturing execution system; using a timestamp alignment method to synchronize the equipment operation logs, material flow sequence, process cycle time data, and planned execution deviation to generate a process collaboration time sequence dataset with a unified time base; calculating the difference between the material arrival time and the downstream equipment readiness time between adjacent processes based on the process collaboration time sequence dataset to obtain a first connection waiting time; if the first connection waiting time exceeds a preset threshold, it is determined that there is a connection delay; simultaneously, parsing the abnormal codes in the equipment operation logs and the shutdown records in the manufacturing execution system to identify fault alarm information that causes production flow interruption; obtaining adjacent process data from the warehousing system... The material inventory value of the inter-sequence buffer area; integrating the first connection waiting time, the fault alarm information, and the material inventory value, a resource occupancy map is constructed. The resource occupancy map is based on the process as nodes and the material flow as edges. The weight of each edge includes the first connection waiting time and the number of fault occurrences. By analyzing the weight distribution of the edges in the resource occupancy map, the bottleneck process identifier is determined. The theoretical output quantity per unit time of the upstream and downstream processes is obtained, and the ratio of the theoretical output quantity of the upstream process to the theoretical output quantity of the downstream process is calculated to obtain the capacity matching coefficient. Combining the bottleneck process identifier and the capacity matching coefficient, if the capacity matching coefficient of the upstream process of the bottleneck process is less than one and the material inventory value is lower than a preset inventory threshold, a dynamic scheduling instruction to adjust the process cycle data of the upstream process is generated. The dynamic scheduling instruction is sent to the production line control system. The instruction execution time is marked in the process collaboration time sequence dataset. Based on the marked dataset, the connection waiting time is recalculated and the resource occupancy map is updated to obtain a new process collaboration operation status and material inventory distribution.Furthermore, the step of transmitting optimized process data to the production management system according to the resource utilization efficiency improvement scheme, obtaining real-time monitoring results of production cycle duration, and dynamically adjusting for cycle fluctuations to determine the final optimized material performance configuration includes: obtaining optimized process data from the resource utilization efficiency improvement scheme, transmitting it to the production management system through data synchronization verification to obtain real-time monitored production cycle duration results; for the production cycle duration results, using cycle deviation assessment to extract the duration sequence from the results and calculate the standard deviation as the fluctuation value; if the fluctuation value exceeds a preset threshold, generating a dynamic adjustment instruction; according to the dynamic adjustment instruction, integrating equipment load balancing data to determine the simulated material properties parameters after cycle fluctuation adjustment through an average load distribution method; obtaining the simulated material properties parameters and performance threshold calibration data, constructing a configuration parameter iteration sequence, and judging the sequence stability through successive parameter updates; and using the configuration parameter iteration sequence, integrating optimized path planning to select the optimal path from the sequence to determine the final optimized material performance configuration. Furthermore, the step of comparing the adhesive layer thickness distribution data with a preset thickness standard range to generate parameter correction instructions for the coating equipment further includes: obtaining a thickness distribution matrix characterizing the coating uniformity, wherein the row and column indices of the thickness distribution matrix correspond to the two-dimensional coordinates of the substrate surface; mapping the values ​​of the thickness distribution matrix to a grayscale image, using an edge detection algorithm to identify significant changes in the thickness gradient in the grayscale image, and then using a region growing algorithm to aggregate adjacent pixels with similar thickness values ​​starting from the boundaries to obtain a set of thickness anomaly regions composed of multiple closed regions; for each closed region in the set of thickness anomaly regions, calculating the average value of all thickness data points within that region; comparing the average value with a preset local thickness allowable range, and if the average value exceeds the allowable range, then determining that the region... The domain is defined as an area of ​​abnormal thickness exceeding the limit, and the geometric center coordinates and thickness deviation value of this area are recorded. The thickness deviation value is the difference between the average value and the median value of the allowable range. Based on the geometric center coordinates of all areas of abnormal thickness exceeding the limit, the coordinates are mapped to the partition control unit numbers along the width direction of the coating head. For each mapped control unit number, the initial adjustment amount of the coating head pressure for that control unit is retrieved from a pre-established parameter correction relationship database, taking into account the sign and magnitude of its corresponding thickness deviation value. The initial adjustment amounts of the coating head pressure for all control unit numbers are integrated, and based on the characteristic that the actions of adjacent control units in the mechanical structure of the coating head influence each other, a linear interpolation method is used to smooth the transition of the initial adjustment amount, generating a coating equipment parameter correction instruction sequence containing the final target pressure values ​​of each control unit.

[0008] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

[0009] This invention discloses a method for optimizing the uniformity and curing effect of high-temperature adhesive tape coating. It matches coating parameters for different substrates and speeds using a pre-established database, and analyzes thickness distribution through real-time monitoring and image processing. When local thickness deviations exceed limits, the coating equipment parameters are dynamically adjusted to correct uniformity. Subsequently, the optimized coating data is linked with the curing module to adapt and adjust the drying temperature distribution. Real-time monitoring and adjustment ensure uniform temperature, thereby achieving a stable curing effect. Based on this, high-temperature adhesion performance test data is combined to analyze process adjustability and improve production continuity. Finally, the collaborative data of each process is integrated to improve resource utilization efficiency and dynamically adjust the production cycle, achieving optimized configuration of material properties. Detailed Implementation

[0010] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the embodiments of this specification. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of this specification.

[0011] This embodiment of a closed-loop control method for a high-temperature tape coating process may specifically include:

[0012] S101. By using a pre-established coating parameter database, relevant data on adhesive layer thickness control are obtained. Parameters are matched for different substrate types and coating speeds. A real-time monitoring system is used to collect data on the thickness distribution during the coating process to obtain a preliminary evaluation result of the coating uniformity.

[0013] Based on a pre-established coating parameter database, a set of adhesive layer thickness control parameters associated with the target substrate type and coating speed is obtained. This parameter set includes coating head pressure parameters and a preset standard range for adhesive layer thickness. Using a thickness sensor in a real-time monitoring system, a lateral thickness distribution data sequence of the adhesive layer is collected during the coating process. This data sequence is indexed by a timestamp and position coordinates along the substrate width direction. The thickness distribution data sequence is compared with the preset standard range for thickness in the parameter set, and the deviation of the thickness value at each position coordinate is calculated to obtain a thickness deviation distribution map composed of the deviations at each position. If the overall deviation of the thickness deviation distribution map exceeds a preset uniformity threshold, the corresponding corrected coating head pressure parameter is retrieved from the coating parameter database based on the spatial distribution characteristics of the deviation in the thickness deviation distribution map, and the corrected coating head pressure parameter is updated in the control commands of the coating equipment.

[0014] In one implementation, the construction of the coating parameter database begins with the systematic organization of historical production data.

[0015] Specifically, the database stores key parameters corresponding to past successful coating processes.

[0016] For example, these parameters include, but are not limited to, the die lip clearance setting of the coating head, the rotational speed or pressure of the adhesive supply pump, the substrate conveyor speed, the temperature distribution of each temperature zone of the oven, and the viscosity and solids content of the adhesive used.

[0017] It should be noted that each record in the database is associated with a specific substrate type and coating speed.

[0018] For example, the substrate type field will record details such as "polyester film," "electrolytic copper foil," or "ceramic substrate," along with key physical properties such as surface tension and roughness. The coating speed is recorded as a specific value, such as 10 meters, 20 meters, or 30 meters per minute. The database is organized using a relational structure, with the substrate type code and speed range serving as the primary index, allowing for quick retrieval of corresponding complete sets of coating parameters. Based on this database, the parameter matching process for different substrates and coating speeds is the core step in achieving precise control of the adhesive layer thickness.

[0019] Understandably, this matching is not a simple table lookup, but a dynamic, intelligent process that may involve compensation calculations.

[0020] In one possible implementation, the system first receives production order information, which specifies the type of substrate to be coated and the target coating speed. Using these two parameters as query criteria, the system searches for the closest historical record in the coating parameter database.

[0021] For example, when the substrate is "aluminum foil for lithium-ion batteries" and the target speed is 25 meters per minute, the system will search the database for all records of aluminum foil substrates and find data sets with speeds between 22 and 28 meters per minute. If a perfect match is found, that set of parameters will be directly used as the initial settings. However, more often than not, parameter interpolation or model calculations are required.

[0022] Specifically, the system incorporates an empirical model that describes the functional relationship between key coating parameters (such as die lip gap) and substrate properties and coating speed. This relationship is obtained through multiple regression analysis of a large amount of historical data.

[0023] For example, the model might reveal that for the same substrate, for every 5 meters per minute increase in coating speed, the die lip gap needs to increase linearly by a specific micrometer to maintain the same wet adhesive thickness. However, for different substrates, even at the same speed, the required adhesive supply pressure needs to be adjusted non-linearly due to differences in surface wettability. The matching algorithm uses these model relationships to fine-tune the retrieved approximate parameters, thereby generating a set of optimized coating parameter presets suitable for the current production conditions, which are then sent to the coating equipment's control system. To ensure the stability of the adhesive layer thickness during coating, a real-time monitoring system is used to continuously measure the thickness distribution.

[0024] In one embodiment, a non-contact online thickness gauge is installed at a suitable location behind the coating die head and before the oven inlet.

[0025] Preferably, the thickness gauge employs the laser triangulation principle or the spectral confocal principle. Its operation involves the transmitting unit projecting a laser beam onto the surface of the moving wet adhesive coating, while the receiving unit detects the position of the reflected light spot. Since the back of the substrate serves as a fixed reference surface, the absolute thickness of the coating can be calculated in real-time by determining the optical path difference or position difference between the laser reflected from the coating surface and the back of the substrate. This thickness gauge typically operates in a transverse scanning mode, with the measuring head reciprocating perpendicular to the substrate's operating direction (i.e., the width direction), thereby acquiring thickness distribution data along a transverse line. The scanning frequency is synchronized with the substrate's operating speed, ensuring the construction of a two-dimensional thickness distribution cloud map across the entire coating area. All thickness data is transmitted in real-time to the host computer monitoring software via industrial Ethernet. The software updates and stores this data several times per second, forming a thickness dataset correlated with time and spatial location. Based on the real-time acquired thickness data, the system automatically performs a preliminary assessment of the coating uniformity.

[0026] Specifically, the evaluation focuses on two dimensions: lateral uniformity and longitudinal stability. For lateral uniformity, the system takes a lateral thickness data curve obtained from a single scan and calculates its mean, maximum, minimum, and standard deviation.

[0027] For example, on a substrate with a width of 1000 mm, the thickness of 200 points is measured by scanning. The system first calculates the average thickness of these 200 points as the transverse average thickness at that scanning moment. Then, it calculates the deviation of each point's thickness value from the average value; the standard deviation quantitatively reflects the thickness fluctuation on that transverse strip. The smaller the standard deviation, the more uniform the transverse coating. For longitudinal stability, the system continuously analyzes the transverse average thickness obtained from all scans over a period of time (e.g., one minute), observing its trend over time and calculating its range and standard deviation. These calculations are all performed automatically, and the evaluation results are displayed on the monitoring interface in the form of numerical values ​​and real-time trend graphs. Operators can intuitively see the current coating uniformity status.

[0028] For example, whether the lateral thickness standard deviation exceeds the preset process upper limit, or whether the longitudinal average thickness shows a continuous drift trend. This initial assessment provides process engineers with immediate feedback to determine whether intervention to adjust coating parameters is necessary. In another embodiment, the parameter matching process can be further refined, especially for multilayer composite coatings or functional coatings.

[0029] For example, in the fabrication of optical films, it may be necessary to coat multiple layers with different refractive indices. In this case, the records in the coating parameter database are expanded to include not only substrate and velocity information, but also fields such as coating number and target optical thickness. During retrieval, the matching algorithm needs to simultaneously match the substrate, velocity, and coating number, and, based on the conversion relationship between the target optical thickness and physical thickness (involving the solid content and refractive index of the coating material), reverse-engineer the required wet adhesive thickness, and then further match the corresponding coating parameters. This process incorporates consideration of the material's optical properties, making parameter matching more accurate and specialized. The data acquisition of the real-time monitoring system can also employ more advanced configurations to improve reliability.

[0030] Preferably, multiple thickness measurement points can be deployed at key locations in the coating production line.

[0031] For example, a first thickness gauge is deployed at the oven inlet to monitor the wet film thickness, and a second thickness gauge is deployed at the oven outlet to monitor the dry film thickness. Data from both gauges is collected synchronously and analyzed in a correlated manner to calculate the coating's curing shrinkage rate, which is crucial for evaluating the stability of the oven process. Simultaneously, the dual-point measurements can cross-verify; if one gauge displays abnormal data, the system can use data from the other gauge and historical patterns to assist in the judgment, improving the robustness of the monitoring system. All thickness measurement data is timestamped and encoded, stored in a time-series database, providing a complete data foundation for subsequent in-depth process analysis and traceability. The algorithm for evaluating coating uniformity can also be customized and extended according to specific product requirements.

[0032] For example, for products particularly sensitive to edge effects, the evaluation algorithm additionally calculates the ratio of the average thickness within a specific width region (e.g., 50 mm on each side) of the substrate's two edges to the average thickness of the central region, and monitors this ratio as a key uniformity indicator. If this ratio deviates from the target value (e.g., 0.95 to 1.05), the system will issue a warning, indicating potential issues such as poor die lip adjustment or substrate belt misalignment. This targeted evaluation indicator makes the initial evaluation results more aligned with the quality control priorities of actual production. In an integrated implementation scenario, the above technical features work synergistically. When a new production task begins, the system automatically matches and issues coating parameters based on the substrate and speed. After coating starts, the real-time monitoring system immediately begins working, continuously collecting thickness data. The monitoring software interface simultaneously displays the preset thickness target curve, the real-time collected thickness distribution cloud map, and the calculated lateral uniformity and longitudinal stability indicators. Operators can easily grasp the overall coating uniformity without performing complex calculations. If the evaluation indicators remain consistently good, it indicates accurate parameter matching and stable process. If any indicator shows a deteriorating trend, the system will highlight it, reminding the operator to check the equipment status or consider initiating the parameter fine-tuning program. The entire implementation method constitutes a complete closed loop from preset, execution, monitoring to preliminary evaluation, providing systematic technical support for the precise control of adhesive layer thickness.

[0033] S102. Based on the evaluation results of coating uniformity, the collected thickness distribution data is analyzed using image processing algorithms to determine whether there is a local thickness deviation. If the deviation value exceeds the preset threshold, an adjustment command is generated to determine the parameter correction scheme of the coating equipment.

[0034] A thickness distribution matrix characterizing the coating uniformity is obtained, where the row and column indices of the thickness distribution matrix correspond to the two-dimensional coordinates of the substrate surface. The values ​​of the thickness distribution matrix are mapped to a grayscale image. An edge detection algorithm is used to identify significant changes in the thickness gradient boundaries in the grayscale image. Then, a region growing algorithm is used to aggregate adjacent pixels with similar thickness values ​​starting from the boundaries, resulting in a set of thickness anomaly regions composed of multiple closed regions. For each closed region in the set of thickness anomaly regions, the average value of all thickness data points within that region is calculated. The average value is compared with a preset local thickness allowable range. If the average value exceeds the allowable range, the region is determined to be an out-of-limit thickness anomaly region, and the geometric center coordinates and thickness deviation value of the region are recorded. The thickness deviation value is the difference between the average value and the value within the allowable range. Based on the geometric center coordinates of all out-of-limit thickness anomaly regions, the coordinates are mapped to the partition control unit number along the width direction of the coating head. For each mapped control unit number, combined with the sign and magnitude of its corresponding thickness deviation value, the initial adjustment amount of the coating head pressure for that control unit is retrieved from a pre-established parameter correction relationship database. The initial pressure adjustment of the coating head is integrated from all control unit numbers. Based on the characteristic that the actions of adjacent control units in the mechanical structure of the coating head affect each other, a linear interpolation method is used to smooth the initial adjustment, generating a coating equipment parameter correction instruction sequence containing the final target pressure values ​​of each control unit.

[0035] In one implementation, based on the coating thickness distribution evaluation results obtained from the real-time monitoring system, the system further calls the image processing algorithm module to perform depth spatial analysis on the original thickness distribution dataset.

[0036] It should be noted that the "image processing algorithm" here does not refer to processing optical images, but rather to simulating the spatial distribution of thickness data as a digital image for processing.

[0037] Specifically, the system organizes the continuously collected thickness measurements, indexed by time and spatial location, into a two-dimensional data matrix.

[0038] For example, with a thickness gauge width of 1 meter and a lateral scanning resolution of 1 millimeter, a single scan can obtain 1000 data points. These data points are arranged sequentially according to their actual positions along the width of the substrate, forming a row of a matrix. As time progresses and the substrate moves forward, the data rows from multiple scans are stacked vertically in chronological order, forming a thickness data matrix where rows represent time (or vertical position) and columns represent lateral positions. Each element of this matrix represents the coating thickness at a specific location. After normalizing this data matrix, it can be viewed as a grayscale image, where the grayscale value corresponds to the coating thickness. Based on the thickness data image constructed above, the core task of the image processing algorithm is to identify local anomalies that do not conform to the overall distribution pattern, i.e., local thickness deviations.

[0039] In one possible implementation, the algorithm first smooths and filters the thickness data image to eliminate random measurement noise. Then, edge detection or region growing algorithms are used to segment and identify potential deviation regions.

[0040] For example, the Sobel or Canny operator can be used to calculate the gradient of a data image. Regions with gradient values ​​significantly higher than the surrounding background may correspond to edges where thickness changes abruptly, such as the boundary of a horizontal thick or thin stripe. Another approach is to set a reference thickness value (such as a process target value or a real-time calculated moving average thickness), and then use thresholding to mark all connected pixels in the data image whose thickness is significantly higher or lower than this reference value, forming one or more "deviation region masks." After identifying the local deviation regions, the algorithm needs to quantify the degree of deviation.

[0041] Specifically, for each identified deviation area, the system calculates the difference between the average thickness within that area and the average thickness of the entire image (or a larger neighborhood surrounding that area), and uses the absolute value of this difference as the deviation amount for that local area. Simultaneously, the system can also calculate the proportion of the deviation area to the entire monitoring area to assess the scope of the deviation's impact. The system then compares the calculated local deviation amount with preset thresholds. These thresholds are typically preset based on product specifications and process requirements.

[0042] For example, in precision optical film coating, the absolute value threshold for local thickness deviation might be set to ±2% of the target thickness, while the area threshold for the deviation region might be set to 1% of the total area. When the deviation or area of ​​one or more identified local deviation regions exceeds their corresponding preset thresholds, the system determines that there is a process anomaly requiring intervention and automatically triggers the adjustment command generation process. The key to generating adjustment commands lies in determining the parameter correction scheme for the coating equipment. This requires establishing a mapping relationship from "deviation characteristics" to "equipment parameters."

[0043] In one embodiment, the system internally maintains a "deviation-parameter" association knowledge base. This knowledge base is not derived through complex physical models, but is built based on data mining and experience summaries from a large number of historical process adjustment cases.

[0044] For example, the knowledge base stores rules, such as: "If a transverse, continuous thickening stripe is detected in the central region of the coating die, prioritize checking and suggest fine-tuning the die lip gap setting in the central region of the coating die to reduce it by a specific number of micrometers"; "If a large area of ​​thinning is detected simultaneously in the edge regions on both sides of the substrate, suggest simultaneously fine-tuning the position of the baffles on both sides or the adhesive supply pressure." Each rule is associated with a typical deviation pattern feature (including the location, shape, size, and direction of the deviation area) and a set of suggested equipment parameter adjustments (including which parameter to adjust, the direction of adjustment, and the estimated adjustment amount). Therefore, when determining a correction scheme, the image processing algorithm module matches the currently identified deviation area features (such as location coordinates, shape descriptors, and deviation direction) with the rules in the knowledge base. The matching process may calculate the similarity between the current deviation feature and the features described by each rule in the knowledge base.

[0045] For example, by calculating the distance to the centroid of the deviation area and the difference in Hu moment of the area shape, the most matching rule entry is found. Subsequently, the system generates specific, executable adjustment instructions based on these rules. The instructions typically include the target adjustment device (e.g., "left zone die lip adjusting bolt"), the adjustment type (e.g., "gap reduction"), and a suggested adjustment amount (e.g., "rotate 5 degrees" or "increase pressure by 0.05 bar"). This suggested adjustment amount is an initial value. In practical applications, the system may adopt a gradual adjustment strategy, i.e., first issuing the instruction at half the suggested amount, observing the thickness feedback in the following monitoring cycles, and then deciding whether to proceed with further adjustments. In another, more refined implementation, for multi-layer co-extrusion coatings or products with complex lateral requirements, the image processing algorithm can further refine the analysis of local deviations.

[0046] For example, when coating a cover film for flexible circuit boards, a more uniform coating thickness is required in specific circuit areas. In this case, the system can import the product's CAD design drawing and overlay and register the thickness data image with the key area layers in the design drawing. The algorithm will specifically analyze the uniformity of the thickness distribution within these key areas and calculate its deviation from the design requirements. If the deviation exceeds the limit, the generated adjustment instructions may be more specific.

[0047] For example, fine-tuning the heater temperature within a specific lateral coordinate range alters the leveling characteristics of the adhesive in that area, thereby achieving precise local thickness correction. All generated adjustment commands are prominently displayed on the user interface and recorded in the process log. Operators can choose to manually confirm and execute the commands, or authorize the system to execute them automatically within specific safety boundaries. Through this closed-loop control method based on graphical spatial analysis and rule matching, the system can quickly pinpoint the root cause of coating uniformity issues and provide targeted parameter correction guidance, thus improving the efficiency and accuracy of process adjustments.

[0048] S103. The operating parameters of the coating equipment are dynamically updated by adjusting the command to obtain the corrected coating thickness data. The uniformity is tested again based on the correction results to obtain an updated coating uniformity analysis report.

[0049] According to the parameter correction instruction sequence of the coating equipment, a pressure adjustment instruction is issued to the partition control unit of the coating head. This instruction drives the actuator to change the pressure setpoint of each unit, completing the dynamic update of the operating parameters. Simultaneously, the timestamp and instruction version number of this parameter update are recorded, generating a parameter change log. After the coating process continues to run and the parameters stabilize, a new round of thickness data acquisition is triggered. Using the same scanning path and sampling frequency as the initial detection, a corrected thickness distribution matrix covering the entire substrate surface is obtained. The row and column indices of this matrix are aligned with the coordinates of the initial detection. The arithmetic mean of all data points in the corrected thickness distribution matrix is ​​calculated as the overall thickness mean, and the standard deviation of these data points is calculated as the overall thickness standard deviation. Simultaneously, the corrected thickness distribution matrix is ​​compared point-by-point with the initial thickness distribution matrix stored in the historical database to obtain a thickness variation matrix. If the overall thickness standard deviation is lower than a preset convergence threshold, the coating uniformity is determined to have met the process requirements, and the iteration process terminates. If the overall thickness standard deviation is not lower than the convergence threshold, a preset abnormal region identification rule is applied to the thickness variation matrix. This rule compares the absolute value of the thickness variation with a preset variation threshold to identify the set of pixels whose variation exceeds the threshold. Connectivity analysis is then performed on these pixel sets to obtain the set of location coordinates of the residual thickness abnormal regions and the average variation of each region. Integrating the parameter change log, the overall thickness mean, the overall thickness standard deviation, the convergence determination result, and the set of location coordinates and average variation of the residual thickness abnormal regions, an updated coating uniformity analysis report is generated according to a preset report template. This report includes the number of iterations, the final process status, and a detailed data summary.

[0050] In one implementation, the coating equipment parameter adjustment instructions generated by the system are sent to the programmable logic controller of the coating equipment via an industrial fieldbus or Ethernet protocol.

[0051] Specifically, the adjustment command is encapsulated as a specific control message, which includes the target actuator (such as the servo motor number of the die head adjusting bolt, the frequency controller address of the feed pump) and specific action parameters (such as rotation angle, pressure setpoint). After receiving the command, the equipment controller drives the corresponding actuator to complete the parameter update.

[0052] It should be noted that, to prevent equipment vibration, the system typically sets a stabilization waiting time after parameter adjustments, such as 30 seconds, until the coating process stabilizes again before triggering a new round of thickness data acquisition. Based on the parameter updates and confirmation of process stability, the system automatically instructs the thickness gauge to initiate a complete scanning measurement cycle. The data acquisition process for this scan is the same as the initial monitoring, acquiring a new coating thickness data matrix reflecting the corrected parameter state at a preset horizontal resolution and vertical travel speed. The system labels this newly acquired data matrix as the "corrected dataset" and associates it with the previously stored "uncorrected dataset" for comparative analysis. For the corrected thickness data, the system again invokes the uniformity detection process. The core of this process is consistent with the initial detection: first, necessary preprocessing (such as filtering and noise reduction) is performed on the data matrix; then, the same algorithm (such as threshold-based region segmentation or gradient analysis) is used to identify local thickness deviation areas.

[0053] Understandably, the deviation judgment threshold used in this test can be kept consistent with the initial threshold to assess whether the correction has brought the process back to the acceptable range; or it can be set to a more stringent threshold according to the process optimization goals to pursue a higher uniformity standard.

[0054] In one embodiment, after completing the secondary uniformity test, the system generates an updated coating uniformity analysis report. This report not only includes the corrected thickness distribution cloud map, the location and size of the identified deviation areas, and the calculated overall uniformity index (such as the thickness standard deviation), but more importantly, it compares the key data before and after the adjustment side by side.

[0055] For example, the report will display the changes in the thickness profile curve at the same horizontal position before and after the adjustment in chart form, visually showing the impact of parameter adjustment on the thickness distribution morphology. Simultaneously, the report will list the specific adjustment instructions issued, the execution time of the instructions, and compare the average thickness deviation in the same deviation area (if it still exists) before and after the adjustment, thereby quantifying the actual effect of this parameter correction. This report provides operators with clear feedback on the effect of process intervention and provides a basis for decision-making regarding whether further fine-tuning is needed.

[0056] Preferably, in complex scenarios such as multi-layer co-extrusion coating, secondary uniformity testing can be more targeted.

[0057] For example, if the initial adjustment instruction is for the heating temperature of a specific area of ​​the width, the secondary detection can prioritize analyzing the thickness changes of the temperature adjustment area and its adjacent areas, calculate the degree of improvement in thickness uniformity within the local area, and present the comparative analysis results of this key area separately in the report, making the evaluation more focused and efficient.

[0058] S104. Based on the updated coating uniformity analysis report, transmit the data to the curing process control module, obtain relevant parameters of drying temperature uniformity, adjust the temperature distribution to adapt to the fluctuation of coating quality, and determine the optimal configuration of the curing process.

[0059] The location coordinates and average variation of the residual thickness anomaly areas are obtained from the updated coating uniformity analysis report. A mapping table between the coating surface coordinates and the oven heating zone numbers is established based on the substrate's transfer path and relative position between the coating station and the curing oven. Using this mapping table, the coating surface coordinates of the residual thickness anomaly areas are converted into a corresponding target heating zone number set. For each zone in the target heating zone number set, a preset thickness-temperature compensation lookup table is consulted based on the average variation of the associated residual thickness anomaly areas to obtain a basic temperature compensation value. The current set temperature of all heating zones in the oven is obtained, forming a current temperature distribution vector. If the application of the basic temperature compensation value causes the temperature of any heating zone to exceed the process safety range, the compensation value for that zone is limited. A preliminary zone temperature adjustment instruction set is generated based on the target heating zone number set and the limited temperature compensation value. A heat conduction model is used to calculate the predicted temperature field distribution on the entire substrate curing surface after executing the preliminary zone temperature adjustment instruction set. The inputs to the heat conduction model include heating element layout parameters, substrate property parameters, and oven airflow parameters. Analyzing the predicted temperature field distribution, if there are areas where the temperature difference between adjacent zones exceeds a preset threshold, these areas are identified as localized thermal stress risk zones. If such localized thermal stress risk zones exist, a heat load balancing process is initiated. This heat load balancing process uses the initial zone temperature adjustment instruction set as input and minimizes the maximum adjacent zone temperature difference as the adjustment target. It iteratively fine-tunes the temperature settings of non-target heating zones to generate an optimized zone temperature adjustment instruction set. The optimized zone temperature adjustment instruction set, together with the mapping table, constitutes the optimized configuration for the curing process.

[0060] In one implementation, the system transmits the updated coating uniformity analysis report to the curing process control module downstream of the production line via a preset data interface protocol.

[0061] Specifically, the transmitted data packet not only contains overall uniformity indicators, but more importantly, it includes thickness distribution profile data of the coating in the width direction, as well as the location coordinates and thickness deviation of specific deviation areas identified by the system. After receiving the above data, the curing process control module first performs data parsing and feature extraction.

[0062] For example, the module will extract the areas in the coating thickness profile data that exceed the preset thickness tolerance zone and their corresponding lateral positions.

[0063] Understandably, during the drying and curing process, the solvent evaporation rate is closely related to the coating thickness. Uneven thickness leads to localized differences in curing rates, which in turn affects the consistency of the final film's performance. Therefore, the core task of the module is to generate a drying temperature field distribution scheme that matches the current coating thickness distribution.

[0064] Specifically, the module has a built-in temperature distribution adaptation and adjustment strategy. This strategy is based on a process knowledge base, and its core logic is as follows: for areas with thicker coatings, the heating temperature of the corresponding drying oven section is appropriately increased to accelerate the evaporation of solvents in the coating in that area and compensate for insufficient curing that may be caused by the increased thickness; conversely, for areas with thinner coatings, the heating temperature of the corresponding section is appropriately reduced to prevent overheating from causing coating deformation or damage to the substrate.

[0065] In one embodiment, the adaptation and adjustment process is not a simple linear correspondence. The module calculates the required temperature adjustment for each independent temperature control zone based on the severity of the thickness deviation (i.e., the amount of deviation), the width of the deviation area, and the thickness gradient of adjacent areas, using an empirical formula or lookup table.

[0066] For example, for a region located in the middle of the width, with a thickness that is 5% thicker and a width of 10 centimeters, the module may calculate that its corresponding three heating zones need to have their temperatures increased by 2-5 degrees Celsius respectively. The adjustment amount will decrease from the center of the region to the edge to achieve a smooth transition of the temperature field and avoid generating new uneven thermal stress.

[0067] Preferably, in multi-layer coating or functional coating scenarios, the temperature distribution adjustment also needs to take into account the curing characteristics of different coating materials.

[0068] For example, if the coating uniformity report shows fluctuations in the thickness of the barrier layer, and this layer is temperature-sensitive, the curing module will call upon a dedicated temperature-curing rate curve for this layer material when adjusting the temperature to perform more precise adjustment calculations. This ensures that while correcting thickness unevenness, the molecular structure of the functional layer is not damaged. After calculating the temperature setpoints for each zone based on the above adaptation adjustment strategy, the curing process control module generates a new drying temperature distribution configuration instruction. This instruction is sent to the temperature controllers of each zone in the curing oven via the industrial network. The temperature controllers adjust the power output of the heating elements according to the new instruction, thereby creating a temperature distribution inside the oven that matches the current coating quality fluctuations. The system typically monitors the stable state after temperature adjustment, and after the process stabilizes, an optional follow-up step is to collect performance data of the cured film layer (such as solvent residue and surface energy) to indirectly evaluate the improvement effect of this temperature adaptation adjustment on the final product quality.

[0069] S105. By optimizing the configuration of the curing process, the distribution of drying temperature is monitored in real time to obtain dynamic data on temperature uniformity. It is determined whether there is local overheating or underheating. If an abnormality is detected, the temperature adjustment mechanism is triggered to obtain stable curing effect data.

[0070] Real-time temperature readings from multiple temperature measurement points within the curing oven are acquired to form a raw data sequence for temperature field monitoring. The raw data sequence is processed using a sliding window method, calculating the standard deviation and range of the temperature in each zone within each window period to quantify uniformity. The result of the uniformity quantification is compared with a preset uniformity threshold. If the comparison result shows that the temperature fluctuation in any region consistently exceeds the preset uniformity threshold, it is determined that the region has local overheating or underheating. The determination includes the abnormal zone number and the direction of temperature deviation. Based on the abnormal zone number, the heating zone mapping relationship in the curing process optimization configuration is matched to determine the target control zone and its required temperature adjustment direction. Using the heat conduction model in the curing process optimization configuration, with the current temperature distribution and the target control zone and its temperature adjustment direction as input, the future temperature field evolution under different temperature adjustment schemes is simulated. The simulation output of the heat conduction model includes the predicted temperature field distribution and the temperature gradient changes between adjacent zones. The scheme that makes the target zone temperature approach the set value and minimizes the temperature gradient change is selected, generating a dynamic setting command containing the specific zone temperature set value. The dynamic setting command is sent to the oven control system for execution. During execution, temperature field monitoring data is continuously acquired and compared with the predicted temperature field of the heat conduction model. If the deviation between the actual temperature and the predicted temperature exceeds a preset allowable deviation range, a risk warning is triggered. The risk warning triggers a correction of the dynamic setting command based on the deviation. After completing a full curing cycle, the curing degree detection data of the substrate is collected to form effect feedback data. A regression analysis method is used to establish a correlation model between the effect feedback data and the historical temperature data of the target control zone during the execution of the dynamic setting command. The output of the correlation model is used to pre-correct the basic temperature setpoint in the curing process optimization configuration before the start of the next curing process.

[0071] In one implementation, to achieve real-time monitoring of the drying temperature distribution, a distributed temperature sensing network is deployed within the curing oven.

[0072] Specifically, the network consists of multiple non-contact infrared temperature probes arranged in a matrix along the width of the oven and the direction of the conveyor belt.

[0073] For example, a probe is installed at 20-centimeter intervals along the width direction and a group is installed at 1-meter intervals along the travel direction. Each probe independently measures the real-time temperature of the coating surface within its field of view and periodically transmits the temperature data to the central data processing unit via fieldbus. After receiving the temperature data streams from all probes, the central data processing unit performs dynamic data extraction and analysis of temperature uniformity.

[0074] Specifically, the unit first integrates temperature data from different spatial locations at the same time to generate a two-dimensional thermogram reflecting the current temperature field distribution inside the oven. Based on this thermogram, the unit calculates key uniformity indicators.

[0075] For example, the standard deviation of temperature at all measuring points along the width direction is calculated as a quantitative value of overall uniformity; at the same time, the continuous areas with the highest and lowest temperatures in the heat map are identified, and their location coordinates and temperature extremes are recorded.

[0076] It should be noted that the judgment of abnormal conditions such as local overheating or underheating relies on preset judgment rules.

[0077] In one embodiment, the system employs dual criteria. The first criterion is a static threshold criterion, which directly compares whether the real-time temperature at any measuring point exceeds the maximum allowable temperature or falls below the minimum allowable temperature. The second criterion is a dynamic trend and distribution criterion.

[0078] For example, the system continuously tracks changes in the aforementioned temperature standard deviation. If it is found to increase continuously over three consecutive sampling periods and exceed the historical statistical control upper limit, it is determined to be an abnormal trend of deteriorating uniformity. Alternatively, when the average temperature difference between the identified high-temperature and low-temperature areas continuously exceeds a set threshold (e.g., 15 degrees Celsius), it is determined to be a significant local heating / cooling imbalance. When the data processing unit confirms an anomaly based on any of the above criteria, it immediately triggers the temperature regulation mechanism. The core of this mechanism is a feedback controller.

[0079] For example, for a detected localized overheating area, the controller sends a command to the power regulator responsible for heating that area to reduce its output power by a certain percentage (e.g., 5% of the rated power each time). Simultaneously, to maintain overall thermal balance, the controller may instruct power regulators in adjacent areas to make compensatory fine-tuning adjustments, such as slightly increasing their power, to prevent insufficient temperature in adjacent areas due to localized cooling. The entire adjustment process is incremental and gradual; the controller continuously monitors the adjusted temperature field until the temperature at all measuring points returns to the process setting range and the uniformity index stabilizes.

[0080] Preferably, after the system completes a round of temperature adjustment and runs stably for a period of time, it will collect key performance parameters of the cured film as curing effect data.

[0081] For example, the amount of residual solvent in the film can be detected by sampling with an online mass spectrometer, or the surface energy of the film can be measured using a surface tension meter. These data are recorded and compared with historical data before adjustment to assess the contribution of this dynamic adjustment to the stability of the final product quality. However, this assessment is not part of the real-time adjustment cycle, but rather used for long-term process optimization and model correction.

[0082] S106. Based on the curing effect data and the test standards for high-temperature adhesion performance, obtain the performance data of the material under extreme environments, analyze the adjustability of process parameters based on the test results, and determine the direction for improving the continuity of the production process.

[0083] Acquire the degree of curing and high-temperature shear strength data of the material in a high-temperature, high-humidity, or corrosive medium simulation chamber. This data constitutes a standardized extreme environment performance dataset. Adhesion performance tests are performed using the fixtures and loading rates specified in the high-temperature adhesion performance test standard. A multiple linear regression algorithm is used, with curing temperature, holding pressure, and curing time as independent variables, and the degree of curing and high-temperature shear strength from the standardized extreme environment performance dataset as dependent variables. Regression calculations are performed to obtain a degree of curing regression equation and a high-temperature shear strength regression equation. These two equations together constitute a correlation model between process parameters and key performance indicators. Based on the degree of curing regression equation and the high-temperature shear strength regression equation in the correlation model, qualified thresholds for degree of curing and high-temperature shear strength are set. In a three-dimensional parameter space, solve for the set of all combinations of curing temperature, holding pressure, and curing time parameters that simultaneously satisfy the output values ​​of the two equations being greater than or equal to their corresponding qualified thresholds. The projection interval of this set on the parameter coordinate axes is defined as the adjustability safety boundary of the process parameters. The actual recorded values ​​of curing temperature, holding pressure, and curing time in continuous production batches are collected. The difference between the maximum and minimum values ​​of each parameter is calculated to obtain the historical fluctuation range. If the historical fluctuation range of a certain parameter completely exceeds its corresponding adjustable safety boundary, the parameter is determined to be a constraint on the continuity of the production process. For the process parameters determined to be constraints, within the adjustable safety boundary, an orthogonal experimental design method is used to arrange multiple sets of process parameter combinations for small-batch trial production. The curing degree and high-temperature shear strength data of the trial production samples are obtained, and the average distance between the sample performance data and the median of the corresponding qualified threshold under each parameter combination is calculated. The scheme with the smallest average distance and whose parameter values ​​are farthest from the endpoint of the adjustable safety boundary is selected as the optimized setting value for the production process.

[0084] In one embodiment, the curing effect data acquired by the system includes the film thickness uniformity, surface gloss, and crosslinking density obtained through offline laboratory testing.

[0085] It should be noted that the testing standard for high-temperature adhesion performance specifically refers to the industry-standard cross-cut test combined with high-temperature aging test.

[0086] For example, the cured coating sample was placed in a constant temperature oven set at 150 degrees Celsius for 500 hours to simulate a long-term high-temperature service environment. After treatment, the sample was removed and cooled to room temperature. Immediately afterward, a grid pattern with a spacing of one millimeter was drawn on the coating surface using a cross-cutting tool. Subsequently, adhesive tape of specified adhesion was tightly adhered to the grid area and quickly peeled off. The number of peeled squares was compared with a standard chart to quantify the adhesion level after high temperature, which was divided into five levels, from zero to five. Level zero represents no peeling, and level five represents a peeling area greater than 65%. Based on the above test results, the system performs an analysis of the adjustability of process parameters.

[0087] Specifically, this analysis aims to establish a correlation model between key process parameters and final high-temperature adhesion performance, and to assess the feasibility of compensating for performance fluctuations by adjusting these parameters.

[0088] In one embodiment, the analysis process first focuses on retrospectively analyzing batches that failed the high-temperature adhesion performance test. The system retrieves all historical data of process parameters recorded during the production of that batch, including but not limited to data on the uniformity of the set and actual temperature distribution in each temperature zone of the oven, conveyor belt speed, and coating coverage.

[0089] Furthermore, the system employs correlation analysis methods, such as calculating the Pearson correlation coefficient, to evaluate the statistical correlation strength between the historical fluctuations of each process parameter and the final high-temperature adhesion test level.

[0090] Understandably, for parameters with high correlation strength, their adjustability has a more significant impact on performance.

[0091] For example, analysis might reveal a strong negative correlation between the real-time average temperature at a specific location along the width (corresponding to previously detected areas of poor temperature uniformity) and the high-temperature adhesion grade of the sample taken at that location. This means that the greater the temperature fluctuation in that area, the worse the adhesion grade. In this case, the system further analyzes the historical operating range of this temperature parameter, compares it to the upper and lower limits allowed by the process specifications, and calculates the percentage of the current actual fluctuation range relative to the allowable range, thus quantifying the adjustable margin of the parameter. If the margin is sufficient, it is determined that the parameter has the adjustability to improve adhesion through fine-tuning; if the margin is close to its limit, it indicates that the parameter may be approaching the process window boundary, requiring consideration of other compensation measures. Based on the test results and the adjustability analysis conclusions, the system determines directions for improving the continuity of the production process.

[0092] Preferably, the improvement focuses on enhancing the robustness of process control and introducing preventative adjustment mechanisms.

[0093] In one embodiment, for parameters that are highly adjustable and critical to performance, such as power output in a specific temperature range, the system will add a feedforward compensation circuit to its control logic.

[0094] For example, when online monitoring detects a positive deviation in the local thickness of the substrate entering the temperature zone, the control unit can slightly increase the heating power in that area in advance based on a preset thickness-temperature compensation model. This counteracts the increased heat demand caused by the thicker coating, preventing insufficient curing and subsequent decrease in adhesion. On the other hand, for parameters with insufficient adjustable margins, improvements focus on optimizing upstream processes.

[0095] For example, if analysis reveals that the uniformity of coating amount is a bottleneck factor limiting the curing temperature adjustment effect and thus affecting adhesion, then the improvement direction will be to shorten the maintenance cycle of the coating head or introduce a higher precision coating amount closed-loop control system to reduce fluctuations from the source, thereby releasing a wider safe operating window for the parameter adjustment of the downstream curing process, and ensuring that the production process can still operate continuously and stably when dealing with minor changes in materials or environment.

[0096] S107. By improving the continuity of the production process, integrate relevant data on the connection between equipment and processes, obtain the collaborative operation status between each process, optimize and adjust the delays or interruptions in the connection links, and obtain a solution to improve the efficiency of resource utilization.

[0097] Equipment operation logs and material flow sequences are obtained from the production line control system, and process cycle time data and planned execution deviations are obtained from the manufacturing execution system. A timestamp alignment method is used to synchronize the equipment operation logs, material flow sequences, process cycle time data, and planned execution deviations, generating a process coordination time sequence dataset with a unified time base. Based on the process coordination time sequence dataset, the difference between the material arrival time and the downstream equipment readiness time between adjacent processes is calculated to obtain the first connection waiting time. If the first connection waiting time exceeds a preset threshold, a connection delay is determined. Simultaneously, abnormal codes in the equipment operation logs and stoppage records in the manufacturing execution system are parsed to identify fault alarm information causing production flow interruptions. Material inventory values ​​in the buffer area between adjacent processes are obtained from the warehousing system. The first connection waiting time, the fault alarm information, and the material inventory value are integrated to construct a resource occupancy graph. The resource occupancy graph uses processes as nodes and material flows as edges, with the weight of each edge including the first connection waiting time and the number of fault occurrences. By analyzing the weight distribution of the edges in the resource occupancy graph, bottleneck process identifiers are determined. The theoretical output per unit time of upstream and downstream processes is obtained, and the ratio of the theoretical output of the upstream process to that of the downstream process is calculated to obtain the capacity matching coefficient. Combining the bottleneck process identifier with the capacity matching coefficient, if the capacity matching coefficient of the upstream process of the bottleneck process is less than one and the material inventory value is lower than a preset inventory threshold, a dynamic scheduling instruction to adjust the cycle time data of the upstream process is generated. The dynamic scheduling instruction is sent to the production line control system. The instruction execution time is marked in the process coordination timing dataset. Based on the marked dataset, the connection waiting time is recalculated and the resource occupancy map is updated to obtain a new process coordination operation status and material inventory distribution.

[0098] In one implementation, the system integrates relevant data on equipment process connectivity through sensors and production execution systems deployed at key nodes of each process.

[0099] It should be noted that this data specifically includes the completion signal timestamps of upstream process equipment, the readiness status signals of downstream process equipment, and the operating data of the material conveying device connecting the two processes, as well as real-time inventory data of the work-in-process buffer. Based on the integrated data above, the system obtains the collaborative operating status between each process.

[0100] Specifically, the system achieves status assessment by calculating two key indicators: the matching degree of production cycle time between adjacent processes and the smoothness of material flow.

[0101] For example, production cycle time matching is obtained by comparing the actual output of upstream processes per unit time with the theoretical processing capacity of downstream processes. Material flow smoothness is quantified by monitoring the actual time required for work-in-process to move from the upstream equipment outlet to the downstream equipment inlet and comparing it with the theoretical shortest transmission time for that path.

[0102] Understandably, when the beat matching degree consistently falls below the set threshold or the smoothness of the flow deteriorates significantly, it indicates poor collaborative operation. The system then performs optimization adjustments to address delays or interruptions in the connection process.

[0103] In one embodiment, the optimization and adjustment process is based on a preset rule engine.

[0104] For example, when the system detects through real-time data streams that a mold change operation on an upstream injection molding machine has timed out, causing the material in the feed inlet buffer of the downstream painting process to run out, the rules engine is triggered. This engine has pre-set strategies to deal with this type of "upstream supply disruption risk".

[0105] Specifically, the strategy first automatically sends a command to the control system of the spraying process to slightly reduce its conveyor belt speed within a safe range, thereby extending the processing time for remaining materials and providing a buffer for upstream production to resume. Simultaneously, the system sends early warning information to the production scheduling terminal and upstream equipment operator terminals, indicating the estimated duration and scope of the supply delay.

[0106] Furthermore, in another implementation scenario, optimization adjustments may also involve the dynamic rearrangement of production orders.

[0107] For example, when the system diagnoses a material flow interruption on multiple assembly lines due to equipment failure in a certain inspection process, the optimization module will quickly simulate and calculate several order rescheduling schemes based on the priority of all current work-in-process orders, process dependencies, and the availability status of each assembly line. The scheme will assess the feasibility of temporarily rescheduling affected orders to other parallel or alternative process paths and recommend the scheme with the least impact on the overall delivery cycle and the lowest equipment changeover cost to the scheduler for confirmation and execution. Through the real-time diagnosis and automatic or assisted optimization of the connection links, the system can obtain solutions to improve resource utilization efficiency. The implementation of this solution reduces idle time caused by equipment waiting for materials, controls work-in-process inventory backlog, and thus improves the overall equipment utilization rate and production rhythm stability of the production line.

[0108] S108. Based on the resource utilization efficiency improvement plan, transmit the optimized process data to the production management system, obtain the real-time monitoring results of the production cycle duration, make dynamic adjustments for cycle fluctuations, and determine the final material performance optimization configuration.

[0109] Optimized process data is obtained from the resource utilization efficiency improvement scheme and transmitted to the production management system through data synchronization and verification to obtain real-time monitoring of production cycle duration results. For the production cycle duration results, a cycle deviation assessment is used to extract the duration sequence from the results and calculate the standard deviation as a fluctuation value. If the fluctuation value exceeds a preset threshold, a dynamic adjustment instruction is generated. Based on the dynamic adjustment instruction, equipment load balancing data is integrated to determine the simulated material property parameters after cycle fluctuation adjustment using an average load distribution method. The simulated material property parameters and performance threshold calibration data are obtained, and a configuration parameter iteration sequence is constructed. The stability of the sequence is judged by successively updating the parameters. Through the configuration parameter iteration sequence, optimized path planning is fused to select the optimal path from the sequence to determine the final optimized material performance configuration.

[0110] In one implementation, the system transmits optimized process data to the production management system based on a resource utilization efficiency improvement plan.

[0111] Specifically, these process data include equipment operating parameters, material flow rates, and process coordination indicators after prior optimization.

[0112] It should be noted that the transmission process is implemented through a dedicated interface, such as a data bus, to ensure that data is uploaded from the production site to the management system database in real time.

[0113] Understandably, this transmission is intended to provide a basis for subsequent monitoring and to avoid data delays affecting overall decision-making.

[0114] Furthermore, the system obtains real-time monitoring results of the production cycle duration.

[0115] For example, the production management system integrates the incoming process data and calculates the complete cycle time from raw material input to finished product output.

[0116] In one embodiment, monitoring results are obtained through timers and log recording modules deployed on the production line. These modules record the start and end timestamps of each process and summarize them into a periodic curve.

[0117] Specifically, real-time monitoring involves comparing the actual cycle with a standard benchmark. If the deviation exceeds a threshold, it is flagged as a fluctuation signal. In this way, the system can capture changes in production rhythm, providing data support for adjustments. For dynamic adjustments to cycle fluctuations, the system executes corresponding optimization strategies.

[0118] In one possible implementation, when monitoring results show an extended cycle, the system analyzes the cause of the fluctuation, such as equipment failure or material shortage.

[0119] Preferably, the adjustment process is based on a preset rule engine, for example, the engine defines a variety of response rules, including temporarily adding backup equipment or adjusting the sequence of procedures.

[0120] Specifically, the rules engine first assesses the scope of the fluctuation's impact, then simulates the feasibility of adjustment plans. For example, between the injection molding and assembly processes, if fluctuations in the injection molding cycle cause assembly delays, the system will automatically instruct the assembly line to switch to a backup material source and simultaneously notify the operator to check the injection molding equipment. Through these steps, adjustments ensure production continuity and reduce downtime caused by fluctuations. In terms of business operations, this dynamic adjustment maintains production line stability and avoids resource waste.

[0121] For example, in another implementation scenario, dynamic adjustments also involve optimizing personnel scheduling. When cyclical fluctuations stem from manpower shortages, the system extracts real-time manpower data from the management system and reassigns operators to critical processes based on the degree of fluctuation. This adjustment is achieved through priority queues, ensuring that high-priority orders are not affected.

[0122] Understandably, through the above dynamic adjustments, the system determines the final optimized configuration of material properties.

[0123] In one embodiment, the configuration process integrates the adjusted cycle data to evaluate the material's performance under optimized cycles.

[0124] Specifically, the system calculates material durability and strength indicators, such as adjusting the alloy composition ratio to improve material toughness based on the shortened cycle time.

[0125] It should be noted that this configuration is achieved through iterative simulation. First, the adjustment results are input, and then the configuration scheme is output, ensuring that the material properties match production requirements. Furthermore, in one implementation, the material performance optimization configuration considers the fusion of multiple factors.

[0126] For example, the system compares the results of periodic monitoring with historical material data and generates a configuration report. The report lists recommended material parameter adjustment values, such as increasing the proportion of a certain additive to adapt to the fast-cycle production environment.

[0127] For example, the final configuration involves a verification process. The system simulates a production scenario after configuration to check if it meets performance standards. If not, adjustments are made iteratively until optimization is achieved. Through this configuration, the production process achieves coordination between materials and timelines, improving overall efficiency.

[0128] The above are only some preferred embodiments of the present invention, but the present invention is not limited thereto, and many improvements and modifications can be made. Any improvements and modifications made based on the basic principles of the present invention should be considered to fall within the protection scope of the present invention.

Claims

1. A closed-loop control method for a high-temperature tape coating process, characterized in that, include: The process involves: acquiring adhesive layer thickness distribution data during the coating process; comparing this data with a preset thickness standard range to generate parameter correction instructions for the coating equipment; updating the operating parameters of the coating equipment based on these instructions and acquiring the corrected adhesive layer thickness distribution data; performing uniformity detection based on the corrected data to generate a coating uniformity analysis report; determining the optimized configuration for the curing process based on the coating uniformity analysis report, whereby the optimized configuration is used to adjust the temperature distribution during curing; monitoring the temperature distribution during curing based on the optimized configuration and triggering a temperature adjustment mechanism based on dynamic temperature uniformity data to acquire curing effect data. Based on the curing effect data and preset performance test standards, the adjustability of process parameters is analyzed to determine the direction for improving the continuity of the production process. Based on the direction for improving the continuity of the production process, relevant data on the connection between equipment processes are integrated to obtain the collaborative operation status between each process, and optimization adjustments are made for delays or interruptions in the connection links to generate a resource utilization efficiency improvement plan. Based on the resource utilization efficiency improvement plan, the optimized process data is transmitted to the production management system to obtain real-time monitoring results of the production cycle duration, and dynamic adjustments are made for cycle fluctuations to determine the final optimized configuration of material performance.

2. The closed-loop control method for a high-temperature tape coating process as described in claim 1, characterized in that, The process of acquiring adhesive layer thickness distribution data during the coating process, and comparing the adhesive layer thickness distribution data with a preset thickness standard range to generate parameter correction instructions for the coating equipment, includes: obtaining a set of adhesive layer thickness control parameters associated with the target substrate type and coating speed from a pre-established coating parameter database, the parameter set including coating head pressure parameters and a preset adhesive layer thickness standard range; using a thickness sensor in a real-time monitoring system to collect a lateral thickness distribution data sequence of the adhesive layer during the coating process, the data sequence being indexed by a timestamp and position coordinates along the substrate width direction; comparing the thickness distribution data sequence with the preset thickness standard range in the parameter set, calculating the deviation of the thickness value at each position coordinate, and obtaining a thickness deviation distribution map composed of the deviations at each position; if the overall deviation of the thickness deviation distribution map exceeds a preset uniformity threshold, then based on the spatial distribution characteristics of the deviation in the thickness deviation distribution map, retrieving the corresponding corrected coating head pressure parameter from the coating parameter database, and updating the corrected coating head pressure parameter to the control instructions of the coating equipment.

3. The closed-loop control method for a high-temperature tape coating process as described in claim 2, characterized in that, The process of updating the operating parameters of the coating equipment according to the parameter correction instructions, obtaining the corrected adhesive layer thickness distribution data, performing uniformity detection based on the corrected adhesive layer thickness distribution data, and generating a coating uniformity analysis report includes: issuing pressure adjustment instructions to the partition control unit of the coating head according to the parameter correction instruction sequence of the coating equipment; the instructions drive the actuator to change the pressure setpoint of each unit to complete the dynamic update of the operating parameters; simultaneously recording the timestamp and instruction version number of this parameter update and generating a parameter change log; after the coating process continues to run and the parameters are stable, triggering a new round of thickness data acquisition; using the same scanning path and sampling frequency as the initial detection, obtaining the corrected thickness distribution matrix covering the entire substrate surface; the row and column indices of the matrix are aligned with the coordinates of the initial detection; calculating the arithmetic mean of all data points in the corrected thickness distribution matrix as the overall thickness mean, and calculating the standard deviation of these data points as the overall thickness standard deviation; simultaneously performing point-by-point difference calculation between the corrected thickness distribution matrix and the initial thickness distribution matrix stored in the historical database to obtain a thickness change matrix; If the overall thickness standard deviation is lower than a preset convergence threshold, the coating uniformity is determined to have met the process requirements, and the iteration process terminates. If the overall thickness standard deviation is not lower than the convergence threshold, a preset abnormal region identification rule is applied to the thickness variation matrix. This rule compares the absolute value of the thickness variation with a preset variation threshold to identify the set of pixels whose variation exceeds the threshold. Connectivity analysis is then performed on the pixel set to obtain the set of location coordinates of the residual thickness abnormal regions and the average variation of each region. The parameter change log, the overall thickness mean, the overall thickness standard deviation, the convergence determination result, and the set of location coordinates and average variation of the residual thickness abnormal regions are integrated, and an updated coating uniformity analysis report containing the number of iterations, the final process status, and a detailed data summary is generated according to a preset report template.

4. A closed-loop control method for a high-temperature tape coating process as described in any one of claims 1-3, characterized in that, The step of determining the optimized configuration of the curing process based on the coating uniformity analysis report includes: obtaining the set of location coordinates and average change of the residual thickness anomaly area from the updated coating uniformity analysis report; establishing a mapping relationship table between the coating surface coordinates and the oven heating zone numbers based on the transfer path and relative position of the substrate between the coating station and the curing oven; using the mapping relationship table, converting the coating surface coordinates of the residual thickness anomaly area into the corresponding target heating zone number set; for each zone in the target heating zone number set, querying a preset thickness-temperature compensation lookup table based on the average change of the associated residual thickness anomaly area to obtain a basic temperature compensation value; obtaining the current set temperature of all heating zones in the oven to form a current temperature distribution vector; if the application of the basic temperature compensation value causes the temperature of any heating zone to exceed the process safety range, then the compensation value of that zone is limited; generating a preliminary zone temperature adjustment instruction set based on the target heating zone number set and the limited temperature compensation value. A heat conduction model is used to calculate the predicted temperature field distribution on the entire substrate curing surface after executing the preliminary zone temperature adjustment command set. The inputs of the heat conduction model include heating element layout parameters, substrate physical property parameters, and oven air velocity parameters. The predicted temperature field distribution is analyzed. If there is a region where the temperature difference between adjacent zones exceeds a preset threshold, this region is determined to be a local thermal stress risk region. If such a local thermal stress risk region exists, a heat load balancing process is initiated. This heat load balancing process uses the preliminary zone temperature adjustment command set as input and minimizes the maximum adjacent zone temperature difference as the adjustment target. Iterative fine-tuning is performed on the temperature settings of non-target heating zones to generate an optimized zone temperature adjustment command set. The optimized zone temperature adjustment command set and the mapping table together constitute the optimized configuration of the curing process.

5. A closed-loop control method for a high-temperature tape coating process as described in any one of claims 1-3, characterized in that, The process involves: monitoring the temperature distribution during the curing process according to the optimized configuration of the curing process; triggering a temperature adjustment mechanism based on dynamic data of temperature uniformity; and acquiring curing effect data. This includes: acquiring real-time temperature readings from multiple temperature measurement points within the curing oven to form a raw data sequence for temperature field monitoring; processing the raw data sequence using a sliding window method to calculate the standard deviation and range of temperatures in each zone within each window period to quantify uniformity; comparing the quantized uniformity result with a preset uniformity threshold; determining that the temperature fluctuation in any region consistently exceeds the preset uniformity threshold, indicating local overheating or underheating in that region, with the determination including the abnormal zone number and temperature deviation direction; matching the heating zone mapping relationship in the optimized configuration of the curing process based on the abnormal zone number to determine the target control zone and its required temperature adjustment direction; and simulating the future temperature field evolution under different temperature adjustment schemes using a heat conduction model in the optimized configuration of the curing process, with the current temperature distribution and the target control zone and its temperature adjustment direction as input. The simulation output of the heat conduction model includes the predicted temperature field distribution and temperature gradient changes between adjacent zones. The scheme that minimizes the temperature gradient change and approaches the target zone temperature is selected, and a dynamic setting instruction containing the specific zone temperature setting value is generated. The dynamic setting instruction is sent to the oven control system for execution. During execution, temperature field monitoring data is continuously acquired and compared with the predicted temperature field of the heat conduction model. If the deviation between the actual temperature and the predicted temperature exceeds a preset allowable deviation range, a risk warning is triggered, and the risk warning triggers a correction of the dynamic setting instruction based on the deviation. After completing a full curing cycle, the curing degree detection data of the substrate is collected to form effect feedback data. A regression analysis method is used to establish a correlation model between the effect feedback data and the historical temperature data of the target control zone during the execution of the dynamic setting instruction. The output of the correlation model is used to pre-correct the basic temperature setting value in the curing process optimization configuration before the start of the next curing process.

6. A closed-loop control method for a high-temperature tape coating process as described in any one of claims 1-3, characterized in that, The step of analyzing the adjustability of process parameters and determining the direction for improving the continuity of the production process based on the curing effect data and preset performance test standards includes: acquiring the degree of curing and high-temperature shear strength data of the material in a high-temperature, high-humidity, or corrosive medium simulation chamber, wherein the data constitutes a standardized extreme environment performance dataset; completing the adhesion performance test using the fixtures and loading rates specified in the high-temperature adhesion performance test standard; performing regression calculations using a multiple linear regression algorithm, with curing temperature, holding pressure, and curing time as independent variables, and the degree of curing and high-temperature shear strength from the standardized extreme environment performance dataset as dependent variables, to obtain a degree of curing regression equation and a high-temperature shear strength regression equation, which together constitute a correlation model between process parameters and key performance indicators; setting a degree of curing qualification threshold and a high-temperature shear strength qualification threshold based on the degree of curing regression equation and the high-temperature shear strength regression equation in the correlation model; and solving for the set of all combinations of curing temperature, holding pressure, and curing time parameters that simultaneously satisfy the output values ​​of the two equations being greater than or equal to their corresponding qualification thresholds in a three-dimensional parameter space, wherein the projection interval of the set on the parameter coordinate axis is defined as the adjustability safety boundary of the process parameters. The actual recorded values ​​of curing temperature, holding pressure, and curing time in continuous production batches are collected. The difference between the maximum and minimum values ​​of each parameter is calculated to obtain the historical fluctuation range. If the historical fluctuation range of a certain parameter completely exceeds its corresponding adjustable safety boundary, the parameter is determined to be a constraint on the continuity of the production process. For the process parameters determined to be constraints, within the adjustable safety boundary, an orthogonal experimental design method is used to arrange multiple sets of process parameter combinations for small-batch trial production. The curing degree and high-temperature shear strength data of the trial production samples are obtained, and the average distance between the sample performance data and the median of the corresponding qualified threshold under each parameter combination is calculated. The scheme with the smallest average distance and whose parameter values ​​are farthest from the endpoint of the adjustable safety boundary is selected as the optimized setting value of the production process.

7. A closed-loop control method for a high-temperature tape coating process as described in any one of claims 1-3, characterized in that, The process, based on the improvement direction of the production process continuity, integrates relevant data on equipment process connectivity, obtains the collaborative operation status between each process, and optimizes and adjusts for delays or interruptions in the connection links to generate a resource utilization efficiency improvement plan. This includes: obtaining equipment operation logs and material flow sequence from the production line control system, and obtaining process cycle time data and planned execution deviation from the manufacturing execution system; using a timestamp alignment method to synchronize the equipment operation logs, material flow sequence, process cycle time data, and planned execution deviation to generate a process collaboration sequence dataset with a unified time base; calculating the difference between the material arrival time and the downstream equipment readiness time between adjacent processes based on the process collaboration sequence dataset to obtain a first connection waiting time; if the first connection waiting time exceeds a preset threshold, it is determined that there is a connection delay; simultaneously, parsing the abnormal codes in the equipment operation logs and the shutdown records in the manufacturing execution system to identify fault alarm information causing production flow interruption; and obtaining the material inventory value of the buffer area between adjacent processes from the warehousing system. By integrating the first connection waiting time, the fault alarm information, and the material inventory value, a resource occupancy graph is constructed. This graph uses processes as nodes and material flows as edges, with each edge's weight including the first connection waiting time and the number of fault occurrences. The bottleneck process identifier is determined by analyzing the edge weight distribution in the resource occupancy graph. The theoretical output per unit time of upstream and downstream processes is obtained, and the ratio of the theoretical output of the upstream process to the theoretical output of the downstream process is calculated to obtain the capacity matching coefficient. Combining the bottleneck process identifier and the capacity matching coefficient, if the capacity matching coefficient of the upstream process of the bottleneck process is less than one and the material inventory value is lower than a preset inventory threshold, a dynamic scheduling instruction to adjust the process cycle time data of the upstream process is generated. The dynamic scheduling instruction is sent to the production line control system. The instruction execution time is marked in the process coordination time sequence dataset. Based on the marked dataset, the connection waiting time is recalculated, and the resource occupancy graph is updated to obtain a new process coordination operation status and material inventory distribution.

8. A closed-loop control method for a high-temperature tape coating process as described in any one of claims 1-3, characterized in that, The step of transmitting optimized process data to the production management system according to the resource utilization efficiency improvement scheme, obtaining real-time monitoring results of production cycle duration, and dynamically adjusting for cycle fluctuations to determine the final optimized material performance configuration includes: obtaining optimized process data from the resource utilization efficiency improvement scheme, transmitting it to the production management system through data synchronization verification to obtain real-time monitoring results of production cycle duration; for the production cycle duration results, using cycle deviation assessment to extract the duration sequence from the results and calculate the standard deviation as the fluctuation value; if the fluctuation value exceeds a preset threshold, generating a dynamic adjustment instruction; according to the dynamic adjustment instruction, integrating equipment load balancing data to determine the simulated material property parameters after cycle fluctuation adjustment through an average load distribution method; obtaining the simulated material property parameters and performance threshold calibration data, constructing a configuration parameter iteration sequence, and judging the sequence stability through successive parameter updates; and using the configuration parameter iteration sequence, integrating optimized path planning to select the optimal path from the sequence to determine the final optimized material performance configuration.