A digital-twin-based printing screen plate manufacturing parameter optimization system

Abstract

The present application relates to the technical field of printing screen plate manufacturing and quality control, in particular to a printing screen plate manufacturing parameter optimization system based on digital twinning, comprising: a data acquisition module for obtaining manufacturing process parameters, service environment data and maintenance strategy parameters; a multiscale mapping module for determining microscopic grain boundary orientation entropy index, mesoscopic residual stress tensor gradient index, macroscopic modal damping ratio and dynamic passivation layer thickness; a coupling failure risk assessment module for determining metastable state barrier index, chemical unlocking driving force index, acoustic vibration amplification coefficient and net failure driving force, and calculating a failure risk index; a collaborative optimization module for determining optimal manufacturing parameter combination and optimal maintenance strategy; and a closed-loop execution control module for calculating real failure risk; the present application improves the intrinsic reliability and service life of the printing screen plate by constructing a multiscale mapping module.

Description

Technical Field

[0001] This invention relates to the field of printing screen manufacturing and quality control technology, specifically to a printing screen manufacturing parameter optimization system based on digital twins. Background Technology

[0002] During the manufacturing and use of printing screens, their manufacturing parameters, service environment, and maintenance strategies jointly determine their final service life and printing quality. In the manufacturing stage, the screens are processed through processes such as laser precision cutting. In the use stage, the screens will experience various working conditions such as squeegee friction, chemical solvent cleaning, and acoustic vibration. In the maintenance stage, the screens need to be cleaned to ensure printing quality.

[0003] To ensure the reliability of stencils, it is usually necessary to optimize their manufacturing process parameters and formulate corresponding maintenance and cleaning strategies. Stencil failure is a complex process resulting from the combined effects of multiple factors in manufacturing, service, and maintenance. On the one hand, microscopic defects and residual stress generated during manufacturing can solidify as potential risks within the stencil, affecting its subsequent service performance. On the other hand, during service and maintenance, multi-physical field effects such as chemical passivation, acoustic vibration, and cleaning stress can interact with these potential risks, ultimately inducing catastrophic failure of the stencil. Existing technologies separate manufacturing and maintenance, making it difficult to accurately predict the health status of the stencil throughout its entire life cycle. In particular, they cannot explain or avoid the counterintuitive phenomenon that enhanced cleaning can actually accelerate stencil failure during certain critical periods, leading to unpredictable and unavoidable catastrophic hole blockage failures, which pose risks of yield loss and production line interruption. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a digital twin-based system for optimizing printing screen manufacturing parameters. Specifically, the technical solution of this invention includes:

[0005] The data acquisition module is used to acquire manufacturing process parameters, including laser power, scanning speed, and defocusing amount. The data acquisition module is also used to acquire service environment data, including the scraper triboelectric charge gradient, printing cycle count, acoustic field spectrum, and stencil intrinsic modes. The data acquisition module is also used to acquire maintenance strategy parameters, including normalized cleaning stress.

[0006] The multi-scale mapping module is used to determine the microscopic grain boundary orientation entropy index and the mesoscopic residual stress tensor gradient index based on manufacturing process parameters and using a preset multivariate nonlinear regression surrogate model. The multi-scale mapping module is also used to determine the macroscopic modal damping ratio based on the mesoscopic residual stress tensor gradient index. The multi-scale mapping module is also used to determine the dynamic passivation layer thickness based on the doctor blade triboelectric charge gradient and the number of printing cycles.

[0007] The coupled failure risk assessment module is used to determine the metastable barrier index based on the microscopic grain boundary orientation entropy index and the mesoscopic residual stress tensor gradient index; it is also used to determine the chemical unlocking driving force index based on the passivation layer thickness; it is also used to determine the acoustic vibration amplification factor based on the acoustic field spectrum and the damping ratio of the intrinsic and macroscopic modes of the mesh; it is also used to determine the net failure driving force, which is a logical combination of the chemical unlocking driving force index, the acoustic vibration amplification factor, the normalized cleaning stress, and the metastable barrier index; and it is also used to calculate the failure risk index by substituting the net failure driving force into the logistic function.

[0008] The collaborative optimization module is used to determine the optimal combination of manufacturing parameters and the optimal maintenance strategy under the constraint of target service life, with the failure risk index as the objective function.

[0009] The closed-loop execution control module is used to send the optimal combination of manufacturing parameters to the laser precision cutting equipment; the closed-loop execution control module is also used to collect service environment data and the current cleaning stress in real time, and calculate the actual failure risk; the closed-loop execution control module is also used to dynamically modify the maintenance strategy based on the actual failure risk and the preset risk threshold; the maintenance strategy includes executing the standard cleaning strategy; the maintenance strategy also includes executing the optimal maintenance strategy; the maintenance strategy also includes locking the equipment.

[0010] Preferably, the multi-scale mapping module determines the microscopic grain boundary orientation entropy index and the mesoscopic residual stress tensor gradient index, including:

[0011] The manufacturing process parameters are collected and normalized to obtain the normalized process parameters;

[0012] Using a pre-defined surrogate model, the orientation entropy index of micrograin boundaries is determined based on normalized process parameters;

[0013] Using a pre-defined surrogate model, the gradient exponent of the mesoscopic residual stress tensor is determined based on normalized process parameters.

[0014] Preferably, the multi-scale mapping module determines the macroscopic modal damping ratio, including:

[0015] Based on the gradient exponent of the mesoscopic residual stress tensor, the macroscopic modal damping ratio is determined by an exponential saturation model.

[0016] Preferably, the multi-scale mapping module determines the dynamic passivation layer thickness, including:

[0017] An initial model of the parabolic growth law based on the interface passivation reaction is established; the initial model characterizes that the passivation layer thickness is proportional to the square root of the number of printing cycles.

[0018] A correction term for the growth rate constant based on the scraper triboelectric charge gradient is introduced to correct the initial model;

[0019] The modified model was used to determine the dynamic passivation layer thickness.

[0020] Preferably, the coupling failure risk assessment module includes:

[0021] The metastable state barrier index is determined by weighted sum of the micrograin boundary orientation entropy index and the mesoscopic residual stress tensor gradient index.

[0022] The chemical unlocking driving force index is determined based on the normalized value of the passivation layer thickness relative to the reference critical thickness.

[0023] Based on the resonance amplification factor in the standard forced vibration theory, and combined with the sound field spectrum and the damping ratio of the intrinsic and macroscopic modes of the mesh plate, the acoustic vibration amplification coefficient is determined.

[0024] Preferably, the coupled failure risk assessment module calculates the failure risk index, including:

[0025] The chemical unlocking driving force index is defined as a positive factor;

[0026] The acoustic vibration amplification factor is defined as a positive factor;

[0027] The normalized cleaning stress is defined as a positive factor;

[0028] Define the metastability barrier index as a negative factor;

[0029] The net failure driving force is determined by combining positive and negative factors;

[0030] Substituting the net failure driving force into the logistic function, the failure risk index is calculated.

[0031] Preferably, the collaborative optimization module determines the optimal combination of manufacturing parameters and the optimal maintenance strategy, including:

[0032] The failure risk index is defined as the collaborative optimization objective;

[0033] Under the constraint of target service life, intelligent optimization algorithm is used to solve for the manufacturing process parameters that minimize the collaborative optimization objective, and the optimal combination of manufacturing parameters is obtained.

[0034] An intelligent optimization algorithm is used to solve for the normalized cleaning stress that minimizes the collaborative optimization objective, thus obtaining the optimal maintenance strategy.

[0035] Preferably, the closed-loop execution control module calculates the actual failure risk, including:

[0036] Real-time collection of service environment data;

[0037] Real-time acquisition of the current cleaning stress corresponding to the currently executing maintenance strategy;

[0038] The failure risk index model is invoked, and the actual failure risk is dynamically calculated by inputting service environment data and current cleaning stress.

[0039] Preferably, the closed-loop execution control module dynamically modifies the maintenance strategy, including:

[0040] When the actual risk of failure is less than the low-risk threshold, it is considered a safe period, and standard cleaning strategies are implemented.

[0041] When the actual failure risk is greater than or equal to the low-risk threshold and less than the high-risk threshold, it is determined to be a warning period, triggering counterintuitive maintenance to reduce the cleaning stress to the optimal value corresponding to the optimal maintenance strategy.

[0042] When the actual risk of failure is greater than or equal to the high-risk threshold, it is determined to be in the failure period, the equipment is locked, and the stencil is forcibly replaced.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] 1. This invention achieves precise mapping from easily obtainable macroscopic process parameters to difficult-to-measure internal physical states by constructing a multi-scale mapping module. For example, it maps laser parameters to microscopic grain boundary orientation entropy and mesoscopic residual stress. This approach replaces the traditional post-manufacturing inspection mode with pre-manufacturing prediction, providing key data inputs that conform to physical laws for risk assessment and forward-looking optimization throughout the entire life cycle, thereby improving the inherent reliability and service life of the printed circuit board.

[0045] 2. This invention innovatively establishes a coupled failure risk assessment module, which for the first time quantifies the complex coupling effect between manufacturing, service, and maintenance. This module calculates a risk index that can reproduce an S-shaped catastrophic failure curve by logically combining positive factors such as chemical unlocking driving force, acoustic vibration amplification coefficient, and normalized cleaning stress with a metastable state barrier index as a negative factor. This scientifically reveals and predicts the counterintuitive failure mechanism that enhanced cleaning actually accelerates pore blockage.

[0046] 3. This invention sets up a collaborative optimization module, which systematically couples the traditionally disconnected manufacturing and maintenance processes. This module takes minimizing the failure risk index throughout the entire life cycle as the unified optimization objective. Under the constraint of the target service life, it simultaneously solves the optimal combination of manufacturing parameters and the optimal maintenance strategy through intelligent optimization algorithms, breaking the limitations of local and passive management.

[0047] 4. This invention achieves real-time synchronization and dynamic control between the digital twin system and the physical entity through a closed-loop execution control module. The core advantage of this module lies in its counterintuitive maintenance logic: when the system judges that the failure risk has entered the warning period based on real-time data, it will proactively reduce the cleaning stress to the preset optimal value instead of blindly increasing it, thereby proactively avoiding catastrophic failures triggered by excessive cleaning stress at the critical point and solving an industry problem. Attached Figure Description

[0048] The present invention will be further explained below with reference to the accompanying drawings and embodiments:

[0049] Figure 1 This is a structural diagram of the system of the present invention. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments. Example 1:

[0051] Please see Figure 1 A digital twin-based system for optimizing printing screen manufacturing parameters includes:

[0052] The data acquisition module is used to acquire manufacturing process parameters, including laser power, scanning speed, and defocusing amount. The data acquisition module is also used to acquire service environment data, including the scraper triboelectric charge gradient, printing cycle count, acoustic field spectrum, and stencil intrinsic modes. The data acquisition module is also used to acquire maintenance strategy parameters, including normalized cleaning stress.

[0053] The multi-scale mapping module is used to determine the microscopic grain boundary orientation entropy index and the mesoscopic residual stress tensor gradient index based on manufacturing process parameters and using a preset multivariate nonlinear regression surrogate model. The multi-scale mapping module is also used to determine the macroscopic modal damping ratio based on the mesoscopic residual stress tensor gradient index. The multi-scale mapping module is also used to determine the dynamic passivation layer thickness based on the doctor blade triboelectric charge gradient and the number of printing cycles.

[0054] The coupled failure risk assessment module is used to determine the metastable barrier index based on the microscopic grain boundary orientation entropy index and the mesoscopic residual stress tensor gradient index; it is also used to determine the chemical unlocking driving force index based on the passivation layer thickness; it is also used to determine the acoustic vibration amplification factor based on the acoustic field spectrum and the damping ratio of the intrinsic and macroscopic modes of the mesh; it is also used to determine the net failure driving force, which is a logical combination of the chemical unlocking driving force index, the acoustic vibration amplification factor, the normalized cleaning stress, and the metastable barrier index; and it is also used to calculate the failure risk index by substituting the net failure driving force into the logistic function.

[0055] The collaborative optimization module is used to determine the optimal combination of manufacturing parameters and the optimal maintenance strategy under the constraint of target service life, with the failure risk index as the objective function.

[0056] The closed-loop execution control module is used to send the optimal combination of manufacturing parameters to the laser precision cutting equipment; the closed-loop execution control module is also used to collect service environment data and the current cleaning stress in real time, and calculate the actual failure risk; the closed-loop execution control module is also used to dynamically modify the maintenance strategy based on the actual failure risk and the preset risk threshold; the maintenance strategy includes executing the standard cleaning strategy; the maintenance strategy also includes executing the optimal maintenance strategy; the maintenance strategy also includes locking the equipment.

[0057] This invention provides a digital twin-based system for optimizing the manufacturing parameters of printed circuit boards. In this embodiment, the system is an integrated, closed-loop control system based on a digital twin model, including: a data acquisition module, a multi-scale mapping module, a coupling failure risk assessment module, a collaborative optimization module, and a closed-loop execution control module.

[0058] The data acquisition module aims to provide the full lifecycle data input required for the construction, mapping, evaluation, and control of the digital twin model. In this embodiment, the data acquisition module is used to acquire manufacturing process parameters, specifically including laser power, scanning speed, and defocusing amount. The data acquisition module is also used to acquire service environment data, specifically including the scraper triboelectric gradient, printing cycle count, acoustic field spectrum, and stencil intrinsic modes. The data acquisition module is also used to acquire maintenance strategy parameters, specifically including normalized cleaning stress.

[0059] The multi-scale mapping module aims to map easily obtainable process parameters and environmental data into key indices that characterize the internal physical state of the stencil, which are difficult to measure directly. In this embodiment, the multi-scale mapping module is used to determine the microscopic grain boundary orientation entropy index and the mesoscopic residual stress tensor gradient index based on manufacturing process parameters and using a preset multivariate nonlinear regression surrogate model. The multi-scale mapping module is also used to determine the macroscopic modal damping ratio based on the mesoscopic residual stress tensor gradient index. Furthermore, the multi-scale mapping module is used to determine the dynamic passivation layer thickness based on the doctor blade triboelectric charge gradient and the number of printing cycles.

[0060] The coupled failure risk assessment module aims to quantify the failure probability of the stencil under the coupled effects of multiple factors during manufacturing, service, and maintenance. In this embodiment, the coupled failure risk assessment module is used to determine the metastable barrier index based on the microscopic grain boundary orientation entropy index and the mesoscopic residual stress tensor gradient index. It is also used to determine the chemical unlocking driving force index based on the passivation layer thickness; to determine the acoustic amplification coefficient based on the acoustic field spectrum and the damping ratio of the stencil's intrinsic and macroscopic modes; and to determine the net failure driving force. The net failure driving force refers to a comprehensive index characterizing the net tendency of the system to deviate from the metastable state. In this embodiment, the net failure driving force is a logical combination of the chemical unlocking driving force index, the acoustic amplification coefficient, the normalized cleaning stress, and the metastable barrier index. Furthermore, the coupled failure risk assessment module is used to calculate the failure risk index by substituting the net failure driving force into the logistic function.

[0061] The purpose of the collaborative optimization module is to proactively seek optimization before and during the manufacturing of the stencil and to find the parameter combination that minimizes the risk of failure throughout its entire life cycle. In this embodiment, the collaborative optimization module is used to determine the optimal combination of manufacturing parameters and the optimal maintenance strategy under the constraint of the target service life, with the failure risk index as the objective function.

[0062] The closed-loop execution control module aims to apply the optimization results and risk warnings from the digital twin to the manufacturing equipment and maintenance processes in the physical world, forming a complete perception-analysis-decision-execution closed loop. In this embodiment, the closed-loop execution control module is used to send the optimal combination of manufacturing parameters to the laser precision cutting equipment. It is also used to collect service environment data and current cleaning stress in real time to calculate the actual failure risk. Furthermore, based on the actual failure risk and a preset risk threshold, the closed-loop execution control module dynamically adjusts the maintenance strategy. In this embodiment, the maintenance strategy includes executing a standard cleaning strategy, executing an optimal maintenance strategy, and locking the equipment.

[0063] By constructing a digital twin model that maps manufacturing parameters to internal states and service failure risks, and combining collaborative optimization and closed-loop control, a forward-looking and systematic management of the entire life cycle of printing screens is achieved. It not only pre-embeds highly reliable and optimal manufacturing parameters in the manufacturing stage, but also dynamically responds to the risks brought about by environmental changes and maintenance behaviors during the service stage, and dynamically adjusts maintenance strategies. This systematically solves the problem of the disconnect between traditional manufacturing and maintenance, and the difficulty in predicting and avoiding catastrophic failures, and significantly improves the service life of screens and printing quality. Example 2:

[0064] The multi-scale mapping module determines the microscopic grain boundary orientation entropy index and the mesoscopic residual stress tensor gradient index, including:

[0065] The manufacturing process parameters are collected and normalized to obtain the normalized process parameters;

[0066] Using a pre-defined surrogate model, the orientation entropy index of micrograin boundaries is determined based on normalized process parameters;

[0067] Using a pre-defined surrogate model, the gradient exponent of the mesoscopic residual stress tensor is determined based on normalized process parameters.

[0068] The specific process by which the multi-scale mapping module in this embodiment determines the microscopic grain boundary orientation entropy index and the mesoscopic residual stress tensor gradient index includes:

[0069] The step involves collecting and normalizing manufacturing process parameters to obtain normalized process parameters. The purpose of this step is to eliminate the differences in physical dimensions between different process parameters, enabling them to be processed and compared uniformly within the same model, thus ensuring the model's universality. In this embodiment, the manufacturing process parameters include laser power. Scan speed Defocus The normalization process is performed based on a preset process window reference value to obtain normalized process parameters.

[0070] The steps are to determine the micrograin boundary orientation entropy index based on normalized process parameters using a pre-defined surrogate model, and to determine the mesoscopic residual stress tensor gradient index based on normalized process parameters using a pre-defined surrogate model.

[0071] Here, the pre-defined surrogate model refers to a multivariate nonlinear regression model obtained by statistical fitting of a large amount of laser cutting experimental data. Its function is to replace high-cost physical detection with low-cost computation. In order to determine the microscopic grain boundary orientation entropy index, this embodiment uses the following surrogate model:

[0072]

[0073] in, The micrograin boundary orientation entropy index is a dimensionless parameter that characterizes the distribution of chemically active sites in the recast layer after laser cutting.

[0074] The normalized laser power, scanning speed, and defocusing amount are calculated from the previous normalization step;

[0075] The fitting coefficients are obtained through laser cutting experiments, for example, by collecting a set of samples containing manufacturing parameters. and its corresponding metallographically measured grain boundary orientation entropy index The calibration dataset, combined with sample data obtained from metallographic analysis, was fitted using least squares regression analysis; to determine the gradient exponent of the mesoscopic residual stress tensor, this embodiment uses the following surrogate model:

[0076]

[0077] in, The mesoscopic residual stress tensor gradient exponent is a dimensionless parameter that measures the risk of microcrack initiation sites in the pore wall.

[0078] The normalized process parameters are calculated from the preceding normalization steps;

[0079] The fitting coefficients are obtained through laser cutting experiments, for example, by collecting a set of samples containing manufacturing parameters. and its corresponding XRD measured stress gradient exponent The calibration dataset is determined by fitting sample data obtained from X-ray diffraction (XRD) stress tests;

[0080] By introducing normalization processing and a multivariate nonlinear regression surrogate model, a rapid and low-cost mapping from macroscopic manufacturing processes to microscopic and mesoscopic internal states is achieved. This enables the system to predict the initial chemical activity of the stencil under different process parameters before manufacturing. and microcrack risk The impact provides crucial initial state input for subsequent risk assessment and optimization, and has significant advantages in foresight and efficiency compared to traditional post-manufacturing inspection methods.

[0081] Example 3:

[0082] The multi-scale mapping module determines the macroscopic modal damping ratio, including:

[0083] Based on the gradient exponent of the mesoscopic residual stress tensor, the macroscopic modal damping ratio is determined by an exponential saturation model.

[0084] The specific process by which the multi-scale mapping module in this embodiment determines the macroscopic modal damping ratio includes: determining the macroscopic modal damping ratio based on the gradient exponent of the mesoscopic residual stress tensor through an exponential saturation model;

[0085] The purpose of this step is to establish the physical relationship between mesoscopic defects and macroscopic mechanical properties; the mesoscopic residual stress tensor gradient exponent. Increasing the value of introduces more microscopic defects, which enhance the structure's energy dissipation capacity during vibration, thereby increasing the macroscopic modal damping ratio. This embodiment uses an exponential saturation model to characterize this relationship because energy dissipation capacity does not grow indefinitely but tends to a saturation value. The exponential saturation model is expressed as follows:

[0086]

[0087] in, The macroscopic modal damping ratio is a dimensionless parameter that characterizes the structure's ability to dissipate vibration energy and is a key input for subsequent calculations of acoustic-vibration amplification coefficients.

[0088] Mesoscopic residual stress tensor gradient exponent;

[0089] Fit coefficients; This represents the upper limit of the damping ratio saturation. The rate at which saturation was reached was controlled; sample data obtained through mechanical impact modal testing, for example, by collecting a set of data containing known stress gradient exponents. and its corresponding mechanical hammer impact measured modal damping ratio The calibration dataset was fitted to determine the result;

[0090] It connects manufacturing parameters with macroscopic dynamic characteristics. The system can quantify the contribution of manufacturing defects to vibration suppression capability through the exponential saturation model, providing accurate and physically consistent parameter inputs for subsequent assessment of the resonance risk of the mesh in the acoustic field environment.

[0091] Example 4:

[0092] The multi-scale mapping module determines the dynamic passivation layer thickness, including:

[0093] An initial model of the parabolic growth law based on the interface passivation reaction is established; the initial model characterizes that the passivation layer thickness is proportional to the square root of the number of printing cycles.

[0094] A correction term for the growth rate constant based on the scraper triboelectric charge gradient is introduced to correct the initial model;

[0095] The modified model was used to determine the dynamic passivation layer thickness.

[0096] The specific process by which the multi-scale mapping module determines the dynamic passivation layer thickness includes:

[0097] An initial model based on the parabolic growth law of the interfacial passivation reaction is established. The purpose of this step is to establish a fundamental physical model for the evolution of the passivation layer thickness. This parabolic growth law is a classic model describing oxide layer growth under the control of interfacial reactions, and its initial model characterizes the passivation layer thickness. Number of printing cycles It is directly proportional to the square root, that is

[0098] To correct the initial model, a correction term for the growth rate constant based on the scraper triboelectric charge gradient is introduced. This step aims to incorporate electrochemical effects from the service environment into the model, addressing the problem of inaccurate predictions caused by the traditional parabolic law's failure to consider charge catalysis. This invention recognizes that the scraper triboelectric charge gradient... This will significantly accelerate the interface passivation reaction; therefore, this embodiment focuses on the basic growth rate constant. A correction term was introduced; the modified model was used to determine the dynamic passivation layer thickness. This model follows the parabolic growth law, and its modified formula is as follows:

[0099]

[0100] in, The dynamic passivation layer thickness is measured in meters (m).

[0101] The number of printing cycles is dimensionless.

[0102] Basic growth constant, in units of square meters per growth cycle Characterizes the base growth rate without charge catalysis;

[0103] The triboelectric charge gradient of the scraper was acquired by the data acquisition module, with dimensions in coulombs per square meter.

[0104] The charge catalytic coefficient, in units of square meters per coulomb, is Characterizes the catalytic strength of the charge gradient on the growth rate;

[0105] This is a correction term for the growth rate constant, and it is dimensionless;

[0106] It provides a dynamically evolving passivation layer thickness model that not only follows the basic parabolic growth law but also innovatively introduces the scraper triboelectric charge gradient, a key service environment factor; this enables thickness prediction... It more closely resembles real-world working conditions, providing dynamic and accurate input for subsequent precise calculations of chemical unlocking driving forces.

[0107] Example 5:

[0108] The coupling failure risk assessment module includes:

[0109] The metastable state barrier index is determined by weighted sum of the micrograin boundary orientation entropy index and the mesoscopic residual stress tensor gradient index.

[0110] The chemical unlocking driving force index is determined based on the normalized value of the passivation layer thickness relative to the reference critical thickness.

[0111] Based on the resonance amplification factor in the standard forced vibration theory, and combined with the sound field spectrum and the damping ratio of the intrinsic and macroscopic modes of the mesh plate, the acoustic vibration amplification coefficient is determined.

[0112] The specific calculation process of the coupling failure risk assessment module includes:

[0113] The metastable state barrier index is determined by weighted sum of the micrograin boundary orientation entropy index and the mesoscopic residual stress tensor gradient index.

[0114] The purpose of this step is to quantify the degree to which the microstructure of the mesh aperture wall is locked in a high-energy metastable state; the metastable barrier index is a custom dimensionless factor used to characterize the height of the barrier against failure of the system, also referred to as the lock-in barrier index in this embodiment, and the formula for calculating the weighted sum is as follows:

[0115]

[0116] in, The Asia-Pacific Stability Barrier Index;

[0117] The micrograin boundary orientation entropy index is calculated by the multi-scale mapping module;

[0118] The gradient exponent of the mesoscopic residual stress tensor is calculated by the multi-scale mapping module;

[0119] The weighting coefficients are determined through historical failure data analysis and expert experience. The determination logic lies in analyzing historical failure cases, such as a set of data on failed stencils, including their... The value is set by using statistical regression or expert system rules to determine the coefficient that best distinguishes between failure and failure.

[0120] The chemical unlocking driving force index is determined based on the normalized value of the passivation layer thickness relative to the reference critical thickness.

[0121] The purpose of this step is to quantify the ability of a slow chemical variable to trigger a fast mechanical variable to unlock; the chemical unlocking driving force index is a custom dimensionless factor that characterizes the chemical driving force that propels the system over the potential barrier, and the formula for calculating this normalized value is as follows:

[0122]

[0123] in, The chemical unlocking driving force index is related to the number of cycles. Dynamically changing;

[0124] The dynamic passivation layer thickness is calculated by the multi-scale mapping module;

[0125] Referring to the critical thickness, calibration is also performed using historical failure data and expert experience. The calibration logic lies in analyzing historical failure cases, such as the failure of the stencil at the time of failure. Value distribution, and setting a reasonable critical thickness value;

[0126] The weighting coefficients are also determined through calibration.

[0127] Based on the resonance amplification factor in the standard forced vibration theory, and combined with the sound field spectrum and the damping ratio of the intrinsic and macroscopic modes of the mesh plate, the acoustic vibration amplification coefficient is determined.

[0128] The purpose of this step is to quantify the amplification effect of the external acoustic field environment on the local dynamic stress of the hole wall; the acoustic amplification factor is a dimensionless factor that characterizes the external energy source that triggers stress unlocking.

[0129] This model originates from the resonance amplification factor in the standard forced vibration theory; when the external excitation frequency is [value missing], the sound field spectrum [value missing]. Approaching the system's natural frequency, the eigenmode of the stencil At this time, the amplitude will increase sharply, and the macroscopic modal damping will increase. This helps to suppress amplification, and the calculation formula is as follows:

[0130]

[0131] in, The acoustic vibration amplification factor is dimensionless.

[0132] The sound field spectrum, specifically the peak frequency, is acquired in real time by the data acquisition module.

[0133] The intrinsic modes of the PCB, specifically the first-order or higher-order modes of interest, are acquired or calibrated by the data acquisition module.

[0134] The macroscopic modal damping ratio is calculated by the multi-scale mapping module;

[0135] By customizing the metastable barrier index Chemical unlocking driving force index Harmony and acoustic vibration amplification factor Innovatively utilizing the lock-in effect derived from manufacturing and the unlocking effect derived from service Mathematical quantification was performed; this laid a solid and computable factor foundation for the subsequent construction of a risk model that can reflect the multi-physics coupling mechanism of stress locking-chemical unlocking and acoustic vibration triggering.

[0136] Example 6:

[0137] The coupled failure risk assessment module calculates the failure risk index, including:

[0138] The chemical unlocking driving force index is defined as a positive factor;

[0139] The acoustic vibration amplification factor is defined as a positive factor;

[0140] The normalized cleaning stress is defined as a positive factor;

[0141] Define the metastability barrier index as a negative factor;

[0142] The net failure driving force is determined by combining positive and negative factors;

[0143] Substituting the net failure driving force into the logistic function, the failure risk index is calculated.

[0144] The specific logic for calculating the failure risk index in the coupled failure risk assessment module includes:

[0145] The direction of action of each factor needs to be defined; the purpose of this step is to clarify the direction of contribution of each factor to system failure; a positive factor means that the larger its value, the more likely it is to drive the system to fail; a negative factor means that the larger its value, the more likely it is to maintain system stability and resist failure.

[0146] The chemical unlocking driving force index is defined as a positive factor;

[0147] The acoustic vibration amplification factor is defined as a positive factor;

[0148] The normalized cleaning stress is defined as a positive factor;

[0149] Define the metastability barrier index as a negative factor;

[0150] This approach identifies a key counterintuitive phenomenon: enhanced cleaning, at certain stages, can actually accelerate pore-clogging failure; therefore, normalized cleaning stress... in These are maintenance strategy parameters acquired by the data acquisition module. The reference stress is also considered a positive factor, and its contribution is expressed through the cleaning triggering coefficient. To scale;

[0151] By combining positive and negative factors, the net failure driver is determined. The purpose of this step is to integrate all positive and negative contributions to calculate a single comprehensive index characterizing the net trend of the system's deviation from metastability. In this embodiment, the net failure driver is defined as:

[0152]

[0153] in, The net failure driving force is dimensionless and varies with the number of cycles. and cleaning stress Dynamic changes;

[0154] These are the chemical unlocking driving force index, acoustic vibration amplification factor, and metastable state barrier index, respectively.

[0155] Normalized cleaning stress is obtained by the data acquisition module;

[0156] The cleaning trigger coefficient was calibrated based on a specific cleaning stress comparison experiment. Multiple control experiments were set up, applying different cleaning stresses. And observe its impact on failure risk The impact was determined through regression analysis. It should be noted that the linear combination model here is a first-order approximation of the multiphysics coupling effect. In a more precise model, cross terms between the factors can be introduced to characterize their nonlinear synergistic effect.

[0157] Substituting the net failure driving force into the logistic function, the failure risk index is calculated; the purpose of this step is to linearize the net failure driving force. It maps to a non-linear probability value that conforms to the characteristics of catastrophic failure; the logistic function is a... The type function can effectively reproduce the catastrophic characteristics of stencil failure, from slow accumulation to sudden outbreak. The calculation formula is as follows:

[0158]

[0159] in, The failure risk index is a dimensionless probability value between 0 and 1.

[0160] The net failure driving force is calculated from the previous step;

[0161] The curve steepness coefficient is a dimensionless coefficient, calibrated based on the S-shaped curve shape of historical failure data. The calibration logic involves collecting historical failure data points, such as those from different service cycles. corresponding failure percentage By using nonlinear fitting, the optimal method to reproduce the S-shaped curve is determined. value;

[0162] By constructing net failure drivers By substituting this into the logistic function, a failure risk index capable of reproducing the S-shaped catastrophic failure curve was successfully established. Model; this model not only couples manufacturing service The factors also innovatively incorporate counterintuitive cleaning stress. Incorporating it as a positive factor enables generalized, dynamic, and high-precision prediction of the entire lifecycle risk of the stencil. Furthermore, each mathematical model in this system is effective within the process window and experimental data range upon which its design is based. For extreme input parameters that exceed the range, such as scanning speed approaching zero or perfect matching of acoustic resonance frequencies, the system has built-in boundary protection logic, such as setting a reasonable physical upper limit on the calculation results, to prevent the model from outputting meaningless singular values, thus ensuring the robustness of the entire system.

[0163] Example 7:

[0164] The collaborative optimization module determines the optimal combination of manufacturing parameters and the optimal maintenance strategy, including:

[0165] The failure risk index is defined as the collaborative optimization objective;

[0166] Under the constraint of target service life, intelligent optimization algorithm is used to solve for the manufacturing process parameters that minimize the collaborative optimization objective, and the optimal combination of manufacturing parameters is obtained.

[0167] An intelligent optimization algorithm is used to solve for the normalized cleaning stress that minimizes the collaborative optimization objective, thus obtaining the optimal maintenance strategy.

[0168] The specific process by which the collaborative optimization module in this embodiment determines the optimal combination of manufacturing parameters and the optimal maintenance strategy includes:

[0169] The failure risk index is defined as the collaborative optimization objective; the purpose of this step is to clarify the direction of optimization; the optimization objective is to achieve the target service life. At that time, a generalized failure risk index was used, taking into account manufacturing parameters and maintenance strategies. Minimize the collaborative optimization objective function Defined as:

[0170]

[0171] in, Collaborative optimization objectives;

[0172] Generalized failure risk index;

[0173] Target service life, based on pre-set constraints according to production needs;

[0174] Normalized manufacturing process parameters are the variables to be optimized.

[0175] Normalized cleaning stress, which is the variable to be optimized;

[0176] Under the constraint of target service life, the manufacturing process parameters that minimize the collaborative optimization objective are solved by intelligent optimization algorithm to obtain the optimal combination of manufacturing parameters. Furthermore, the normalized cleaning stress that minimizes the collaborative optimization objective is solved by intelligent optimization algorithm to obtain the optimal maintenance strategy.

[0177] The purpose of this step is to use computational tools to search the multidimensional parameter space, while satisfying process constraints, to find a solution that satisfies the objective function. The smallest globally optimal solution;

[0178] Intelligent optimization algorithms refer to non-gradient optimization algorithms such as genetic algorithms and particle swarm optimization, which are good at handling... This is a complex optimization problem involving high dimensions and nonlinearity;

[0179] The optimization process simultaneously solves for manufacturing process parameters and normalized cleaning stress; the results are:

[0180] Optimal combination of manufacturing parameters:

[0181] Optimal maintenance strategy: After inverse normalization, we get ;

[0182] It will unify the traditionally separate processes of manufacturing and maintenance through a single, collaborative optimization goal. The system is coupled; through intelligent optimization algorithms, it can simultaneously find the inherently optimal solution. and the best after tomorrow The parameter combination works synergistically from both the root cause and process levels, breaking the limitations of traditional passive maintenance and achieving global minimization of risks throughout the entire life cycle of the network board.

[0183] Example 8:

[0184] The closed-loop execution control module calculates the actual failure risk, including:

[0185] Real-time collection of service environment data;

[0186] Real-time acquisition of the current cleaning stress corresponding to the currently executing maintenance strategy;

[0187] The failure risk index model is invoked, and the actual failure risk is dynamically calculated by inputting service environment data and current cleaning stress.

[0188] The specific process by which the closed-loop execution control module calculates the actual failure risk includes:

[0189] Real-time collection of service environment data;

[0190] Once the physical PCB is put into service, the data acquisition module continues to operate, acquiring real-time service environment data, such as the current sound field spectrum. scraper triboelectric charge gradient and the cumulative number of printing cycles ;

[0191] Real-time acquisition of the current cleaning stress corresponding to the currently executing maintenance strategy;

[0192] The purpose of this step is to capture the real-time impact of maintenance actions, a key intervention, on risk; the closed-loop execution control module monitors the cleaning strategy currently being executed by the printing press and obtains the current cleaning stress corresponding to that strategy. Subsequently, the value was normalized to ;

[0193] Call the failure risk index model, input service environment data and current cleaning stress, and dynamically calculate the actual failure risk;

[0194] The purpose of this step is to use a digital twin model to perform a real-time snapshot assessment of the current health status of the physical network panel; the system calls the established generalized failure risk index model. Real-time collection of service environment data and current cleaning stress will be used to monitor the environment. As input, the actual failure risk is dynamically calculated. :

[0195]

[0196] It achieves real-time synchronization between the digital twin model and the physical entity; by collecting service data and current cleaning stress in real time and substituting them into the generalized risk model, the system can dynamically calculate the actual failure risk; this makes risk assessment no longer a static prediction, but a dynamic, live data that is updated in real time with operating conditions and maintenance behavior, providing a basis for decision-making for subsequent dynamic early warning and closed-loop correction.

[0197] Example 9:

[0198] The closed-loop execution control module dynamically modifies and maintains the strategy, including:

[0199] When the actual risk of failure is less than the low-risk threshold, it is considered a safe period, and standard cleaning strategies are implemented.

[0200] When the actual failure risk is greater than or equal to the low-risk threshold and less than the high-risk threshold, it is determined to be a warning period, triggering counterintuitive maintenance to reduce the cleaning stress to the optimal value corresponding to the optimal maintenance strategy.

[0201] When the actual risk of failure is greater than or equal to the high-risk threshold, it is determined to be in the failure period, the equipment is locked, and the stencil is forcibly replaced.

[0202] The specific process of the closed-loop execution control module dynamically correcting and maintaining the strategy includes:

[0203] Based on actual failure risk Based on the assessment results, the system automatically performs tiered early warnings and predictive maintenance, forming a control closed loop and replacing passive maintenance based on fixed frequency; the system presets low-risk thresholds. and high risk threshold The method for determining the preset risk threshold is as follows: statistically analyze historical failure data. The risk curve defines the point where the slope of the S-shaped curve begins to increase significantly as the low-risk threshold. The steepest point where the slope of the S-curve is at its maximum or close to saturation is defined as the high-risk threshold. ;

[0204] When the actual risk of failure is less than the low-risk threshold, it is considered a safe period, and standard cleaning strategies are implemented.

[0205] Right now The system determines that the stencil is in a stable service state, and at this time, it executes the standard cleaning strategy. ;

[0206] When the actual failure risk is greater than or equal to the low-risk threshold and less than the high-risk threshold, it is determined to be a warning period, triggering counterintuitive maintenance to reduce the cleaning stress to the optimal value corresponding to the optimal maintenance strategy.

[0207] Right now The system determines that the stencil has entered a metastable critical point; at this point, the system identifies an intensified locking-unlocking conflict, affecting conventional cleaning stress. Instead, it becomes a positive factor triggering failure; therefore, the system immediately triggers counterintuitive maintenance: reducing cleaning stress and correcting it to the optimal value obtained through optimization. ,in Avoid providing the energy to trigger stress unlocking under cleaning stress.

[0208] When the actual failure risk is greater than or equal to the high-risk threshold, it is determined to be in the failure period, the equipment is locked and the stencil is forcibly replaced;

[0209] Right now The system determines that the stencil has experienced or is about to experience a catastrophic S-curve failure. At this time, the system executes the highest level of response: immediately lock the equipment, prohibit further production, and issue a command to forcibly replace the stencil to prevent catastrophic bridging defects from flowing into downstream processes.

[0210] By establishing a dynamic correction closed loop of three levels of risk and three maintenance strategies, true predictive maintenance is achieved. The counterintuitive maintenance strategy during the early warning period actively avoids catastrophic failures triggered by cleaning stress by reducing rather than increasing cleaning stress at the risk threshold. This systematically solves the industry problem that increasing cleaning actually accelerates clogging, ensuring the reliable operation of the stencil throughout its entire life cycle.

[0211] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A printing screen manufacturing parameter optimization system based on digital twin, characterized in that... ,include: The data acquisition module is used to acquire manufacturing process parameters, including laser power, scanning speed, and defocusing amount. The data acquisition module is also used to acquire service environment data, including the scraper triboelectric charge gradient, printing cycle count, acoustic field spectrum, and stencil intrinsic modes. The data acquisition module is also used to acquire maintenance strategy parameters, including normalized cleaning stress. The multi-scale mapping module is used to determine the microscopic grain boundary orientation entropy index and the mesoscopic residual stress tensor gradient index based on manufacturing process parameters and using a preset multivariate nonlinear regression surrogate model. The multi-scale mapping module is also used to determine the macroscopic modal damping ratio based on the mesoscopic residual stress tensor gradient index. The multi-scale mapping module is also used to determine the dynamic passivation layer thickness based on the doctor blade triboelectric charge gradient and the number of printing cycles. The coupled failure risk assessment module is used to determine the metastable barrier index based on the microscopic grain boundary orientation entropy index and the mesoscopic residual stress tensor gradient index; it is also used to determine the chemical unlocking driving force index based on the passivation layer thickness; it is also used to determine the acoustic vibration amplification factor based on the acoustic field spectrum and the damping ratio of the intrinsic and macroscopic modes of the mesh; it is also used to determine the net failure driving force, which is a logical combination of the chemical unlocking driving force index, the acoustic vibration amplification factor, the normalized cleaning stress, and the metastable barrier index; and it is also used to calculate the failure risk index by substituting the net failure driving force into the logistic function. The collaborative optimization module is used to determine the optimal combination of manufacturing parameters and the optimal maintenance strategy under the constraint of target service life, with the failure risk index as the objective function. The closed-loop execution control module is used to send the optimal combination of manufacturing parameters to the laser precision cutting equipment; the closed-loop execution control module is also used to collect service environment data and the current cleaning stress in real time, and calculate the actual failure risk; the closed-loop execution control module is also used to dynamically modify the maintenance strategy based on the actual failure risk and the preset risk threshold; the maintenance strategy includes executing the standard cleaning strategy; the maintenance strategy also includes executing the optimal maintenance strategy; the maintenance strategy also includes locking the equipment.

2. The printing screen manufacturing parameter optimization system based on digital twin as described in claim 1, characterized in that, The multi-scale mapping module determines the microscopic grain boundary orientation entropy index and the mesoscopic residual stress tensor gradient index, including: The manufacturing process parameters are collected and normalized to obtain the normalized process parameters; Using a pre-defined surrogate model, the orientation entropy index of micrograin boundaries is determined based on normalized process parameters; Using a pre-defined surrogate model, the gradient exponent of the mesoscopic residual stress tensor is determined based on normalized process parameters.

3. The printing screen manufacturing parameter optimization system based on digital twin according to claim 1, characterized in that, The multi-scale mapping module determines the macroscopic modal damping ratio, including: Based on the gradient exponent of the mesoscopic residual stress tensor, the macroscopic modal damping ratio is determined by an exponential saturation model.

4. The printing screen manufacturing parameter optimization system based on digital twin according to claim 1, characterized in that, The multi-scale mapping module determines the dynamic passivation layer thickness, including: An initial model of the parabolic growth law based on the interface passivation reaction is established; the initial model characterizes that the passivation layer thickness is proportional to the square root of the number of printing cycles. A correction term for the growth rate constant based on the scraper triboelectric charge gradient is introduced to correct the initial model; The modified model was used to determine the dynamic passivation layer thickness.

5. The printing screen manufacturing parameter optimization system based on digital twin according to claim 1, characterized in that, The coupling failure risk assessment module includes: The metastable state barrier index is determined by weighted sum of the micrograin boundary orientation entropy index and the mesoscopic residual stress tensor gradient index. The chemical unlocking driving force index is determined based on the normalized value of the passivation layer thickness relative to the reference critical thickness. Based on the resonance amplification factor in the standard forced vibration theory, and combined with the sound field spectrum and the damping ratio of the intrinsic and macroscopic modes of the mesh plate, the acoustic vibration amplification coefficient is determined.

6. A printing screen manufacturing parameter optimization system based on digital twin according to claim 1 or 5, characterized in that, The coupled failure risk assessment module calculates the failure risk index, including: The chemical unlocking driving force index is defined as a positive factor; The acoustic vibration amplification factor is defined as a positive factor; The normalized cleaning stress is defined as a positive factor; Define the metastability barrier index as a negative factor; The net failure driving force is determined by combining positive and negative factors; Substituting the net failure driving force into the logistic function, the failure risk index is calculated.

7. The printing screen manufacturing parameter optimization system based on digital twin according to claim 1, characterized in that, The collaborative optimization module determines the optimal combination of manufacturing parameters and the optimal maintenance strategy, including: The failure risk index is defined as the collaborative optimization objective; Under the constraint of target service life, intelligent optimization algorithm is used to solve for the manufacturing process parameters that minimize the collaborative optimization objective, and the optimal combination of manufacturing parameters is obtained. An intelligent optimization algorithm is used to solve for the normalized cleaning stress that minimizes the collaborative optimization objective, thus obtaining the optimal maintenance strategy.

8. The printing screen manufacturing parameter optimization system based on digital twin according to claim 1, characterized in that, The closed-loop execution control module calculates the actual failure risk, including: Real-time collection of service environment data; Real-time acquisition of the current cleaning stress corresponding to the currently executing maintenance strategy; The failure risk index model is invoked, and the actual failure risk is dynamically calculated by inputting service environment data and current cleaning stress.

9. The printing screen manufacturing parameter optimization system based on digital twin according to claim 8, characterized in that, The closed-loop execution control module dynamically modifies and maintains the strategy, including: When the actual risk of failure is less than the low-risk threshold, it is considered a safe period, and standard cleaning strategies are implemented. When the actual failure risk is greater than or equal to the low-risk threshold and less than the high-risk threshold, it is determined to be a warning period, triggering counterintuitive maintenance to reduce the cleaning stress to the optimal value corresponding to the optimal maintenance strategy. When the actual risk of failure is greater than or equal to the high-risk threshold, it is determined to be in the failure period, the equipment is locked, and the stencil is forcibly replaced.

Images