A material intelligent scheduling method and system for a film production line

Abstract

This invention relates to the field of intelligent manufacturing and production scheduling, specifically disclosing a method and system for intelligent material scheduling in a thin-film production line. The method includes: calculating the cumulative plastic strain energy density of the metal coating based on historical winding stress data, simultaneously constructing an unsteady disturbance field distribution map, identifying disturbance-electrostatic coupling sensitive states, and generating a candidate material queue; predicting the crack propagation rate and electrostatic discharge risk index based on the disturbance-crack propagation transfer function and the electrostatic accumulation-dissipation equilibrium model, and dynamically generating a composite scheduling tolerance window; screening schedulable materials based on the tension response characteristics predicted by the disturbance-tension transfer function, performing progressive unwinding, and triggering dynamic reconstruction through acoustic emission monitoring. This invention solves the problems of unwinding crack propagation and discharge damage in metal-coated thin films by tracing the winding stress history and sensing environmental disturbances and electrostatic coupling effects, achieving closed-loop intelligent scheduling with multi-physics risk perception and dynamic tolerance capabilities.

Description

Technical Field

[0001] This invention relates to the field of intelligent manufacturing and production scheduling technology, specifically to a material intelligent scheduling method and system for a thin film production line. Background Technology

[0002] In the mass production of high-end functional films such as automotive heat-insulating window films and paint protection films, the composite structure of metal coatings and polymer substrates is typically transported in rolls between automated warehouses and stations such as magnetron sputtering, coating, and slitting. Existing intelligent scheduling systems mainly rely on the first-in, first-out (FIFO) rules of ERP or MES systems, combined with current inventory locations and delivery deadlines, to perform physical handling via AGVs. This method only focuses on the static properties of the materials and the macroscopic environmental temperature and humidity, neglecting the unique winding stress-induced microcrack phenomenon in metal-coated films during the winding and storage process.

[0003] Specifically, the difference in elastic modulus between the metal layer and the polymer substrate leads to the accumulation of plastic strain within the metal layer due to winding tension, resulting in the depletion of the microcrack initiation barrier. Existing scheduling methods completely ignore the material's winding stress history and cannot distinguish the differences in mechanical vulnerability during unwinding between materials with the same current appearance but vastly different winding histories. More critically, there is a significant coupling effect between micro-vibrations in the production environment and the accumulated electrostatic charge on the film surface. When materials with historical stress defects are scheduled to workstations with high electrostatic potential and strong environmental disturbances, unwinding tension fluctuations are amplified by the electrostatic-vibration coupling field, triggering accelerated microcrack propagation and even interlayer discharge damage.

[0004] Existing scheduling systems employ fixed tension thresholds and electrostatic protection standards, failing to dynamically adjust the operating window based on the cumulative plastic strain energy density of the material. Furthermore, they cannot identify crack propagation during unwinding through acoustic emission monitoring and trigger dynamic reconfiguration of the production path, leading to batch-specific quality losses in high-value metal-coated films. Therefore, there is an urgent need for an intelligent scheduling method capable of tracing winding stress history, sensing environmental disturbances and electrostatic coupling fields, dynamically generating composite tolerance windows, and supporting real-time acoustic emission monitoring and closed-loop correction. Summary of the Invention

[0005] The purpose of this invention is to provide a material intelligent scheduling method and system for thin film production lines, in order to solve the technical problem that existing thin film production line scheduling methods ignore the stress history of metal coating winding and the effect of environmental disturbance-electrostatic coupling, leading to microcrack propagation and discharge damage during unwinding.

[0006] To solve the above-mentioned technical problems, the present invention specifically provides the following technical solution: A method for intelligent material scheduling in a thin film production line includes the following steps: S1. Obtain historical data of winding stress to calculate the cumulative plastic strain energy density of the metal coating, collect environmental micro-vibration and airflow fluctuation data to construct an unsteady disturbance field distribution map, identify disturbance-electrostatic coupling sensitive states and associate them with the cumulative plastic strain energy density, and generate a material candidate queue with multi-physics risk stratification labels. S2. Receive the material candidate queue, predict the crack propagation rate based on the multiphysics risk stratification label and the unsteady disturbance field distribution map, and dynamically generate a composite scheduling tolerance window; S3. Based on the composite scheduling tolerance window, select schedulable material rolls, generate and execute a progressive unwinding scheduling instruction set, and monitor actual crack propagation events through acoustic emission to trigger dynamic reconstruction.

[0007] As a preferred embodiment of the present invention, S1 specifically includes: S11. By deploying a stress sensor array at the winding station, the winding tension time spectrum of each material roll is collected in real time during the winding process, and the stress lock-in state caused by the inertia of the roll is recorded at the moment of winding stop; based on the elastoplastic strain energy accumulation model of the metal-substrate composite layer, the cumulative plastic strain energy density of the metal coating induced by the winding tension time spectrum and the stress lock-in state is calculated, and the cumulative plastic strain energy density is used to characterize the degree of consumption of the microcrack initiation barrier inside the metal layer. S12. Simultaneously, high-sensitivity accelerometers and piezoelectric airflow sensors deployed at multiple nodes of the production line are used to collect environmental micro-vibration spectrum data and airflow fluctuation data, and the real-time electrostatic potential value of each in-production film roll is obtained through a non-contact electrostatic voltmeter; the environmental micro-vibration spectrum data and airflow fluctuation data are spatially interpolated and superimposed to construct an unsteady disturbance field distribution map covering the key areas of the production line. S13. Based on the unsteady-state disturbance field distribution map and real-time electrostatic potential value, identify the target scroll sensitive to disturbance-electrostatic coupling, associate it with the cumulative plastic strain energy density of the metal coating, and generate a material candidate queue with multi-physics risk stratification labels.

[0008] As a preferred embodiment of the present invention, S13 specifically includes: S131. The environmental micro-vibration spectrum energy density and airflow fluctuation pressure amplitude of each spatial grid point in the unsteady disturbance field distribution map are weighted and fused to calculate the disturbance coupling strength index of each grid point; the real-time electrostatic potential value of each in-production film roll is multiplied with the disturbance coupling strength index of the grid point to obtain the disturbance-electrostatic coupling sensitivity coefficient of each in-production film roll. S132. Select at least one in-production thin film roll whose disturbance-electrostatic coupling sensitivity coefficient exceeds a preset sensitivity threshold, mark it as a target roll, and extract the cumulative plastic strain energy density of the metal coating corresponding to the target roll. S133. The cumulative plastic strain energy density of the metal coating and the disturbance-electrostatic coupling sensitivity coefficient are nonlinearly weighted and fused to generate a crack initiation probability level; based on the time gradient change rate of the disturbance-electrostatic coupling sensitivity coefficient, stress relaxation compensation requirements are generated; based on the ratio of the disturbance coupling strength index of the grid point where the target roll is located to the real-time electrostatic potential value, a disturbance-electrostatic coupling vulnerability level is generated. S134. Aggregate the crack initiation probability level, stress relaxation compensation requirement, and disturbance-electrostatic coupling vulnerability level into a multiphysics risk stratification label, and bind the multiphysics risk stratification label to the target scroll to output a material candidate queue with the multiphysics risk stratification label.

[0009] As a preferred embodiment of the present invention, S2 specifically includes: S21. Receive the material candidate queue, obtain the material properties and thickness parameters of the target film corresponding to the target roll, and determine the critical winding stress threshold and electrostatic dissipation time constant of the target film based on the material properties and thickness parameters. S22. Obtain the tension control capability parameters of each unwinding station on the production line and the tension sensitivity threshold of the downstream composite station; S23. For each material roll in the material candidate queue, based on the crack initiation probability level and cumulative plastic strain energy density, and combined with the unsteady disturbance field distribution map, predict the crack propagation rate and electrostatic discharge risk index of the material under the preset unwinding acceleration. S24. Calculate the safety margin between the crack propagation rate and the tension sensitivity threshold, match the electrostatic discharge risk index of Siping City with the electrostatic dissipation time constant through a time window, and dynamically generate a composite scheduling tolerance window.

[0010] As a preferred embodiment of the present invention, S23 specifically includes: S231. For each material roll in the material candidate queue, extract its corresponding crack initiation probability level and cumulative plastic strain energy density, and at the same time obtain the metal coating thickness and substrate elastic modulus parameters of the material roll. S232. Perform time-frequency domain decomposition on the environmental micro-vibration spectrum energy density time series and the airflow fluctuation pressure amplitude time series of the spatial grid points where the target scroll is located in the unsteady disturbance field distribution map, extract the disturbance main frequency component and the fluctuation amplitude envelope, and construct the disturbance time-varying excitation function; S233. Based on the cumulative plastic strain energy density and crack initiation probability level, combined with the metal coating thickness and substrate elastic modulus parameters, a disturbance-crack propagation transfer function is established. The disturbance time-varying excitation function is input into the disturbance-crack propagation transfer function to predict the crack propagation rate of the material under a preset unwinding acceleration. S234. Based on the fluctuation frequency and amplitude change rate of the airflow fluctuation pressure amplitude in the time-varying excitation function of the disturbance, combined with the electrostatic dissipation constant and surface resistivity of the material, an electrostatic accumulation-dissipation equilibrium model is constructed to predict the electrostatic discharge risk index of the material under a preset unwinding acceleration. The electrostatic discharge risk index includes the estimated peak electrostatic potential and the discharge probability level.

[0011] As a preferred embodiment of the present invention, S24 specifically includes: S241. Calculate the ratio of the crack propagation rate to the tension sensitivity threshold to obtain the safety margin coefficient; when the safety margin coefficient is lower than the preset safety threshold, activate the safety margin compensation mechanism, and determine the compression ratio of the tension fluctuation range based on the difference between the safety margin coefficient and the preset safety threshold. S242. Using the estimated peak electrostatic potential in the electrostatic discharge risk index as input, and the electrostatic dissipation time constant as the time window benchmark, calculate the dissipation time required for the estimated peak electrostatic potential to decay to the upper limit of allowable electrostatic accumulation, and match the dissipation time with the preset time window threshold to obtain the time window matching degree. S243. Based on the safety margin coefficient and the time window matching degree, construct a two-dimensional coupling constraint matrix, apply the compression ratio to the initial tension fluctuation range to generate an acceptable tension fluctuation range; and use the electrostatic accumulation margin corresponding to the time window matching degree as the upper limit of allowable electrostatic accumulation. S244. Based on the acceptable tension fluctuation range and the allowable electrostatic accumulation limit, and combined with the activation state of the safety margin compensation mechanism, a stress gradient release strategy is generated. The stress gradient release strategy includes the number of stress steps, the holding time of each stress level, and the stress reduction amplitude of adjacent levels. The acceptable tension fluctuation range, the allowable electrostatic accumulation limit, and the stress gradient release strategy are aggregated into a composite scheduling tolerance window.

[0012] As a preferred embodiment of the present invention, S3 specifically includes: S31. Obtain the real-time electrostatic residual value of the candidate material rolls waiting to be rolled up, and based on the unsteady disturbance field distribution map and the composite scheduling tolerance window, use the disturbance-tension transfer function to predict the tension response characteristics of each candidate material roll under the combined action of environmental micro-vibration spectrum data and airflow fluctuation data. S32. Using the real-time electrostatic residual value and tension response characteristics as constraints, select schedulable material rolls that simultaneously meet the acceptable tension fluctuation range and the allowable upper limit of electrostatic accumulation, and generate a progressive unwinding scheduling instruction set. S33. Execute the connection command between the schedulable material roll and the target roll, collect the interlayer peeling signal through the acoustic emission sensor array, and when the actual crack propagation event deviates from the matching threshold of the expected acoustic emission feature template, trigger the dynamic reconstruction of the scheduling path to guide the material roll to the backup low-stress path or temporary buffer station.

[0013] As a preferred embodiment of the present invention, S32 specifically includes: S321. Obtain the real-time electrostatic residual value of the candidate material rolls waiting to be rolled up, and based on the unsteady disturbance field distribution map and the composite scheduling tolerance window, use the disturbance-tension transfer function to predict the tension response characteristics of each candidate material roll under the combined action of environmental micro-vibration spectrum data and airflow fluctuation data. The tension response characteristics include the tension fluctuation amplitude prediction range and the fluctuation frequency principal component. S322. The real-time electrostatic residual value and the tension fluctuation amplitude prediction range in the tension response characteristics are used as two-dimensional constraints. They are compared with the upper limit of allowable electrostatic accumulation and the acceptable tension fluctuation range in the composite scheduling tolerance window, respectively, and at least one candidate material volume that meets both constraints is selected as a schedulable material volume. S323. For the selected schedulable material rolls, extract their corresponding stress relaxation compensation requirements and disturbance-electrostatic coupling vulnerability levels. Combine this with the stress gradient release strategy in the composite scheduling tolerance window to generate a progressive unwinding scheduling instruction set. The progressive unwinding scheduling instruction set includes: a target workstation sequence, a tension feedforward correction curve corresponding to the stress relaxation compensation requirements, an unwinding acceleration limit corresponding to the disturbance-electrostatic coupling vulnerability level, and a pre-constructed expected acoustic emission feature template based on the metal coating properties and cumulative plastic strain energy density of the schedulable material rolls. The progressive unwinding scheduling instruction set is then sent to the execution unit as a benchmark for subsequent unwinding processes and dynamic reconstruction.

[0014] As a preferred embodiment of the present invention, S33 specifically includes: S331. Execute the instruction to transport the schedulable material roll to the target roll for connection. During the actual unwinding process of the material roll, the acoustic emission signal of interlayer stripping is collected in real time by the acoustic emission sensor array deployed at the unwinding station, and the spectral characteristics and energy accumulation rate of the acoustic emission signal are extracted. S332. Perform pattern matching between the extracted spectral features and the expected acoustic emission feature template in the progressive unwinding scheduling instruction set, and calculate the spectral feature deviation rate and the energy accumulation rate deviation rate; when the spectral feature deviation rate exceeds the first preset threshold or the energy accumulation rate deviation rate exceeds the second preset threshold, it is determined that an actual crack propagation event deviation has occurred. S333. Obtain the cumulative plastic strain energy density and disturbance-electrostatic coupling vulnerability level of the current material roll, and calculate the dynamic reconstruction priority coefficient based on the severity of the deviation judgment; when the dynamic reconstruction priority coefficient exceeds the preset reconstruction threshold, trigger the dynamic reconstruction of the scheduling path, and calculate the backup low-stress path or temporary buffer station from the current position to the downstream composite station. S334. Switch the current material roll conveying path to the backup low-stress path or the guide temporary buffer station, and at the same time feed back the occurrence sequence, spectral characteristic deviation rate and energy accumulation rate deviation rate of the actual crack propagation event to the elastic-plastic strain energy accumulation model, update the calculation weight coefficient of the accumulated plastic strain energy density online, and output the updated model parameters.

[0015] A material intelligent scheduling system for a thin film production line, used to implement a material intelligent scheduling method for a thin film production line, comprising: The disturbance sensing module is used to collect environmental micro-vibration spectrum data and airflow fluctuation data of multiple key nodes in the production line, and simultaneously acquire the real-time electrostatic potential value of each film roll in production; based on the environmental micro-vibration spectrum data and airflow fluctuation data, an unsteady disturbance field distribution map is constructed, and based on the unsteady disturbance field distribution map and the real-time electrostatic potential value of each film roll in production, at least one target roll in a disturbance-electrostatic coupling sensitive state is identified; The tolerance matching module, connected to the disturbance sensing module, is used to obtain the material properties and thickness parameters of the target film corresponding to the target roll, and determine the critical winding stress threshold and electrostatic dissipation time constant of the target film based on the material properties and thickness parameters; and dynamically generate a composite scheduling tolerance window for the target roll that includes an acceptable tension fluctuation range and an allowable electrostatic accumulation upper limit by combining the critical winding stress threshold and the electrostatic dissipation time constant. The collaborative scheduling module, connected to the tolerance matching module, is used to obtain the real-time electrostatic residual value of the candidate material rolls waiting to be rolled up, and predict the tension response characteristics of each candidate material roll under the action of the unsteady disturbance field distribution map; using the real-time electrostatic residual value and tension response characteristics as constraints, at least one schedulable material roll falling within the composite scheduling tolerance window is selected, and an instruction to transport the schedulable material roll to the target roll for connection is generated and executed.

[0016] Compared with the prior art, the present invention has the following advantages: 1. By establishing a historical integral assessment of winding stress and constructing a distribution map of unsteady disturbance fields, the limitations of traditional static inventory management are overcome. This enables the quantitative characterization of the initiation barrier of microcracks in the metal layer and the accurate identification of disturbance-electrostatic coupling sensitive states, transforming scheduling decisions from time-priority to multi-field coupling risk-priority.

[0017] 2. By constructing a disturbance-crack propagation transfer function and a dynamic composite tolerance window, the limitations of fixed process parameters are overcome, and adaptive safety boundary adjustment based on cumulative plastic strain energy density and environmental disturbance intensity is realized. This enables high-defect-risk materials to automatically match low-stress stations and progressive unwinding strategies, significantly reducing the risk of microcrack propagation.

[0018] 3. By establishing a closed loop for acoustic emission mode matching and dynamic reconstruction, the lag defect of traditional open-loop scheduling is overcome, enabling real-time identification of crack propagation anomalies and single-roll-level path reconstruction. Combined with the online update mechanism of model parameters, the real-time quality control capability and production reliability of high-end metal coating films are significantly improved. Attached Figure Description

[0019] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating the method described in Embodiment 1 of the present invention.

[0021] Figure 2 This is a framework diagram of the system described in Embodiment 2 of the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] The concepts involved in this application will first be described with reference to the accompanying drawings. It should be noted that the following descriptions of various concepts are only for the purpose of making the content of this application easier to understand and do not constitute a limitation on the scope of protection of this application; furthermore, the embodiments and features in the embodiments of this application can be combined with each other unless otherwise specified. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0024] Example 1 like Figure 1 As shown, the present invention provides a method for intelligent material scheduling in a thin film production line, comprising the following steps: S1. Obtain historical data on winding stress to calculate the cumulative plastic strain energy density of the metal coating; collect environmental micro-vibration and airflow fluctuation data to construct an unsteady-state disturbance field distribution map; identify disturbance-electrostatic coupling sensitive states and correlate them with the cumulative plastic strain energy density; generate a candidate material queue with multi-physics risk stratification labels; specifically including: S11. Quantification of the microcrack initiation barrier in metal coatings based on the history integral of winding stress, specifically: S111. A strain gauge tension sensor array is installed on the surface of the roll support bearing seat at the winding station. The sensors are arranged at equal intervals along the axial direction of the roll to ensure the uniformity of tension distribution along the width of the roll. During the winding process, the output signal of the tension sensor array is continuously collected at fixed time intervals to record the sequence data of the winding tension changing over time, forming a winding tension time sequence spectrum.

[0025] At the moment of winding cessation, by monitoring the sudden change in tension data at the instant the reel speed returns to zero, the residual tension lock-in value caused by the rotational inertia of the reel is captured, and the stress lock-in state is recorded.

[0026] S112. Based on the composite layer structure of metal coating and polymer substrate, an elastoplastic strain energy accumulation model is established. The tension peak sequence in the winding tension time series spectrum and the residual tension locking value in the stress locking state are input into the elastoplastic strain energy accumulation model. The model includes the elastoplastic constitutive model of the metal layer, the viscoelastic stress transfer sub-model of the substrate, and the interlayer strain energy coupling accumulation sub-model. The elastoplastic constitutive model of the metal layer adopts the von Mises isotropic hardening criterion as its core theoretical framework, converting the instantaneous tension peak in the winding tension time series spectrum into the equivalent stress value of the metal coating. The initial yield strength and hardening modulus of the metal material are set as constitutive parameters. The numerical relationship between the equivalent stress and the current yield strength is compared to determine whether the metal layer has entered a plastic state. When the equivalent stress does not exceed the current yield strength, the elastic strain component is calculated based on Hooke's law, and the elastic strain energy is stored. When the equivalent stress exceeds the current yield strength, the yield surface radius is updated according to the isotropic hardening criterion, and the plastic strain increment generated by the stress exceeding the yield strength is calculated, accumulating the plastic strain component. By recording the accumulated plastic strain in the loading history, the current yield strength of the metal layer is updated in real time, reflecting the work hardening effect. Finally, the elastic strain tensor and plastic strain tensor of the metal layer at each moment during the winding process are output.

[0027] The viscoelastic stress transfer sub-model of the substrate employs a generalized Maxwell model composed of multiple Maxwell elements connected in parallel to describe the time-varying mechanical properties of the polymer substrate. Each Maxwell element consists of an elastic spring and a viscous damper connected in series, possessing independent elastic modulus and relaxation time constants. The winding tension time series spectrum is input into the model as a boundary condition. By solving the differential constitutive equations of each Maxwell element, the instantaneous stress response and long-term stress relaxation behavior of the substrate during the winding process are calculated. During the winding stage, the model calculates the initial stress generated by the winding deformation of the substrate and its stress relaxation curve that decays exponentially with time. In the stress-locked state after winding stops, the model re-establishes the equilibrium equations based on the residual deformation constraints caused by the inertia of the winding shaft, calculating the redistribution state of the stress inside the substrate and the continuous stress transfer rate to the metal layer at the interface. By superimposing the stress contributions of each Maxwell element, the time-varying stress transfer function of the substrate to the metal layer is obtained, reflecting the evolution of the efficiency of stress transfer from the substrate to the metal layer over time.

[0028] A coupling accumulator model of interlayer strain energy is used to establish the displacement compatibility equation at the interface between the metal coating and the polymer substrate, which forces the strain displacement at the interface between the lower surface of the metal layer and the upper surface of the substrate to remain continuous and consistent. The model receives the plastic strain component output from the elastoplastic constitutive model of the metal layer and the interfacial stress transfer rate output from the viscoelastic stress transfer model of the substrate, establishing a balance between the internal stress state of the metal layer and the constraint reaction force of the substrate. Based on the interfacial displacement compatibility condition, the equivalent stress state actually borne by the metal layer under the constraint of the substrate is calculated, distinguishing between the free deformation part and the constrained deformation part. The plastic strain energy density inside the metal layer is calculated by volume integral, and the plastic work done by the stress part enhanced by the constraint of the substrate is accumulated as the plastic strain energy density of the metal layer. Through time integration, the increments of plastic strain energy density generated at each moment during the winding process are accumulated to obtain the total cumulative plastic strain energy density of the metal coating throughout the entire winding history. This cumulative plastic strain energy density value characterizes the degree of consumption of the microcrack initiation barrier inside the metal layer; a higher density value indicates more severe accumulated plastic damage inside the metal layer and a lower critical stress threshold for microcrack initiation.

[0029] S12. Identification of sensitive regions based on multi-physics field coupling reconstructed from environmental micro-vibration and airflow disturbance fields, specifically: S121. High-sensitivity triaxial accelerometers are installed at key nodes such as the entrance of the magnetron sputtering zone, the middle section of the coating and drying zone, and the exit of the slitting and winding zone on the production line. The accelerometers are rigidly fixed to the surface of the equipment frame structure through magnetic bases to ensure rigid coupling with the vibration transmission path of the production line's mechanical structure. The accelerometers continuously collect three-dimensional vibration acceleration time-domain signals at a high sampling frequency. The time-domain signals are converted into frequency-domain representations through a fast Fourier transform algorithm to extract the vibration energy distribution characteristics of each node in different frequency bands, forming environmental micro-vibration spectrum data.

[0030] Piezoelectric airflow sensor arrays are arranged on the ceiling and side walls above each workstation on the production line. The sensor's windward surface is perpendicular to the main airflow direction. It monitors in real time the airflow velocity pulsation and instantaneous pressure fluctuation caused by the air conditioning system, the passage of logistics vehicles, or the exhaust of adjacent heavy equipment. It records the sequence of airflow pressure amplitude changes over time and generates airflow fluctuation data.

[0031] A non-contact electrostatic voltmeter is installed above the unwinding path of the film. Using the principle of rotating blade electric field induction or vibration capacitance detection, the surface charge distribution of the film is sensed without contact with the film surface, and the electrostatic potential value of each film roll in production is measured in real time.

[0032] S122. Establish a unified time synchronization benchmark, and align the environmental micro-vibration spectrum data, airflow fluctuation data, and electrostatic potential values ​​of each node in time and space to ensure that the data from different sensors have the same sampling time marker on the time axis; use the Kriging spatial interpolation algorithm to reconstruct the continuous field of the environmental micro-vibration spectrum data collected by discrete nodes, calculate the semi-variogram function based on the spatial distance and directional relationship between each node, and generate the spatial vibration spectrum estimate of the area where no sensors are deployed through the optimal linear unbiased estimation method; superimpose the pressure amplitude field in the airflow fluctuation data with the interpolated environmental micro-vibration spectrum energy density field, and fuse the vibration energy and airflow pressure contribution according to the preset weight coefficient to construct an unsteady disturbance field distribution map covering the key area of ​​the production line. This distribution map characterizes the disturbance intensity distribution at each spatial location and its evolution over time, providing basic data support for subsequent identification of disturbance-electrostatic coupling sensitive states.

[0033] S13. Target scroll multiphysics risk stratification based on perturbation-electrostatic coupling sensitive state identification, specifically: S131. Calculation of the disturbance-electrostatic coupling susceptibility coefficient, specifically: The distribution map of the unsteady disturbance field is divided into regular spatial grid cells according to the physical space of the production line, with each grid cell corresponding to a specific spatial coordinate position. For each grid cell, the energy density value of the environmental micro-vibration spectrum stored therein is extracted. This value is the weighted sum of the energy values ​​of the vibration spectrum in each characteristic frequency band. At the same time, the airflow fluctuation pressure amplitude of the grid cell is extracted. This value is the peak or effective value of the pressure fluctuation sequence. Weighting coefficients are set according to the relative importance of vibration and airflow to the membrane roll disturbance. The environmental micro-vibration spectrum energy density and the airflow fluctuation pressure amplitude are fused by a linear weighted summation formula to calculate the disturbance coupling strength index of each grid point.

[0034] The real-time spatial coordinates of each in-production film roll are obtained and mapped to the corresponding spatial grid cell. The disturbance coupling strength index of the grid point where the film roll is located is extracted. The real-time electrostatic potential value of each in-production film roll is read by a non-contact electrostatic voltmeter. The real-time electrostatic potential value of each in-production film roll is multiplied by the disturbance coupling strength index of its grid point. The larger the product result, the more the location is affected by the superposition of strong environmental disturbance and high electrostatic potential. The product value is the disturbance-electrostatic coupling sensitivity coefficient of each in-production film roll.

[0035] S132. Correlation between target roll identification and cumulative plastic strain energy density of metal coating, specifically: Set a disturbance-electrostatic coupling sensitivity threshold, and compare the disturbance-electrostatic coupling sensitivity coefficient of each in-production film roll with the sensitivity threshold; select at least one in-production film roll that exceeds the sensitivity threshold and mark it as the target roll; retrieve the unique identifier of the target roll from the production database, extract the cumulative plastic strain energy density value of the metal coating calculated in step S11, and establish a data association between the target roll and the cumulative plastic strain energy density.

[0036] S133. Generation of multiphysics risk stratification labels, specifically: Extract the cumulative plastic strain energy density value of the metal coating corresponding to each target roll from the database. This value characterizes the degree of consumption of the microcrack initiation barrier inside the metal layer. Obtain the disturbance-electrostatic coupling sensitivity coefficient corresponding to each target roll.

[0037] A nonlinear weighted fusion function is constructed, which adopts a power-law or exponential form. The cumulative plastic strain energy density of the metal coating is used as the basic risk factor, and the disturbance-electrostatic coupling sensitivity coefficient is used as the environmental amplification factor. The crack initiation probability level is generated through nonlinear calculation. The higher the value, the easier it is for microcracks to propagate under environmental disturbance. Calculate the time gradient rate of change of the disturbance-electrostatic coupling sensitivity coefficient of each target scroll over time, and calculate the sensitivity change rate per unit time using a difference algorithm or a sliding window averaging algorithm; classify the levels according to the magnitude of the time gradient rate of change. If the rate of change exceeds the positive threshold, it indicates that the environmental disturbance is rapidly increasing, generating a high-level stress relaxation compensation demand; if the change is stable, a low-level demand is generated. Calculate the ratio of the perturbation coupling strength index of the grid point where the target scroll is located to its real-time electrostatic potential value. Divide the perturbation-electrostatic coupling vulnerability level according to the numerical range of the ratio. Abnormally high or low ratios correspond to different vulnerability modes. The three indicators—crack initiation probability level, stress relaxation compensation requirement, and disturbance-electrostatic coupling vulnerability level—are encapsulated and combined to form a complete multiphysics risk stratification label.

[0038] S134. Multiphysics Risk Hierarchy Tag Aggregation and Material Candidate Queue Output, specifically: Establish a tag data structure and define storage fields for crack initiation probability level, stress relaxation compensation requirement, and disturbance-electrostatic coupling vulnerability level; write the three risk index values ​​generated in step S133 into the corresponding fields to complete the aggregation and encapsulation of multiphysics risk stratification tags; bind and associate the aggregated multiphysics risk stratification tags with the unique identifier of the target roll; organize all target rolls and their corresponding multiphysics risk stratification tags in sequence, and output a material candidate queue with multiphysics risk stratification tags.

[0039] S2. Receive the material candidate queue, predict crack propagation rate based on multi-physics risk stratification labels and unsteady-state disturbance field distribution maps, and dynamically generate a composite scheduling tolerance window; specifically including: S21. Critical parameter calibration based on material-thickness constitutive properties, specifically: S211. Receive the material candidate queue with multi-physics risk stratification tags output from step S13, parse the queue data structure to obtain the unique identification code of each target roll; through the production database query interface, retrieve the corresponding target film material attribute data based on the target roll identification code. The material attributes include metal coating type, metal coating thickness, polymer type of polymer substrate, substrate elastic modulus, substrate resistivity, and surface coating resistivity; simultaneously obtain the thickness parameters of the target film, including total thickness, substrate thickness, metal coating thickness distribution, and surface functional coating thickness.

[0040] S212. Based on the combination characteristics of metal coating type and polymer substrate type, a pre-established material-thickness-stress mapping database is invoked. This database calibrates the mechanical failure boundary of different material combinations under different thickness ratios through offline tensile tests and winding simulation experiments. According to the metal coating thickness and substrate elastic modulus values ​​of the target film, a linear interpolation algorithm or a nonlinear fitting algorithm based on machine learning is used to query and calculate the critical winding stress threshold corresponding to the specific material-thickness combination in the material-thickness-stress mapping database. This threshold characterizes the critical stress boundary at which the metal coating is about to undergo irreversible plastic deformation or interface delamination.

[0041] S213. Calculate the electrostatic dissipation time constant based on the electrostatic dissipation theory model. This model includes a surface charge accumulation sub-model, a bulk conductivity dissipation sub-model, and an environmental humidity correction sub-model. The surface charge accumulation sub-model describes the electrostatic charge injection rate caused by airflow friction and interlayer peeling during unwinding. The bulk conductivity dissipation sub-model describes the attenuation law of charge leakage to the ground terminal through the bulk resistance of the polymer substrate. The surface conductivity sub-model describes the dissipation path of charge diffusion from the film surface to the atmospheric environment. The environmental humidity correction sub-model adjusts the surface conductivity parameter in real time according to the current relative humidity level. Input the surface resistivity, substrate resistivity, and current environmental humidity data of the target film into the electrostatic dissipation theory model. By solving the charge conservation differential equation, calculate the time scale required for the electrostatic potential to decay from the initial value to the safe threshold level. This time scale is the electrostatic dissipation time constant. Use the calculated critical winding stress threshold and electrostatic dissipation time constant as inherent physical constraint parameters of the target film, and bind them with the target roll identification code and store them in the scheduling parameter cache area.

[0042] S22. Obtaining the tension characteristic parameters of the production line station, specifically: Through the production execution system interface, the system can query the equipment status data of each unwinding station on the production line in real time and obtain the tension control capability parameters of each unwinding station. The parameters include the maximum settable tension value, the minimum stable tension value, the tension control accuracy level, the tension response bandwidth, and the current actual tension setting value. At the same time, the system can obtain the process constraint data of the downstream laminating stations, which include the coating laminating station, the slitting and winding station, and the intermediate buffer station.

[0043] For each downstream lamination station, its tension sensitivity threshold is obtained. This threshold represents the maximum tension fluctuation range that the station can withstand from the upstream unwinding station. Exceeding this threshold will lead to quality defects such as interlayer bubbles, uneven coating thickness, or misalignment of the cutting end face.

[0044] Spatially correlate the tension control capability parameters of each unwinding station with the tension sensitivity threshold of the corresponding downstream composite station to establish a tension transmission constraint relationship diagram between station pairs, and clarify the effective tension control boundary of each unwinding station under different downstream station combinations; store these parameters and relationship diagrams in the scheduling decision cache area for use in step S23 when performing crack propagation prediction and safety margin calculation.

[0045] S23. Disturbance-crack coupling dynamics and electrostatic risk prediction, specifically: S231. Extraction of material crack risk parameters and material mechanical properties, specifically: The process iterates through each material roll entry in the material candidate queue, parses the entry data structure to obtain the unique identifier code for that material roll, accesses the multiphysics risk stratification tag storage area based on the identifier code, and extracts the crack initiation probability level value from the tag field. Using the data association index, it retrieves the cumulative plastic strain energy density value of the metal coating corresponding to the material roll from the cumulative plastic strain energy density database generated in step S11. It then calls the material master data interface of the production execution system, queries the material database based on the identifier code, and reads the metal coating thickness parameter and polymer substrate elastic modulus parameter of the material roll. Finally, it aligns and standardizes the extracted crack initiation probability level, cumulative plastic strain energy density, metal coating thickness, and substrate elastic modulus, encapsulating them into a crack prediction input parameter set for the material roll, providing basic data support for subsequent disturbance-crack propagation transfer function calculations.

[0046] S232. Construction of perturbation time-varying excitation function and extraction of time-frequency features of multiphysics field, specifically: The environmental micro-vibration spectrum energy density time series and airflow fluctuation pressure amplitude time series of the spatial grid points where the target scroll is located are extracted from the unsteady disturbance field distribution map. The two time series are decomposed in the time-frequency domain by applying short-time Fourier transform to convert the time-domain signal into a two-dimensional time-frequency representation to obtain the spectrum energy distribution at different times.

[0047] The peak detection algorithm identifies the frequency component with the highest energy concentration in the spectrum as the dominant perturbation frequency component, and extracts the phase information of this dominant frequency component as a function of time. Hilbert transform or moving average algorithm is applied to the time series of airflow fluctuation pressure amplitude to extract the amplitude envelope of the pressure fluctuation and characterize the time-varying law of the intensity of airflow pulsation.

[0048] A time-varying excitation function for disturbance is constructed, which includes a vibration excitation subfunction, an airflow excitation subfunction, a coupling modulation coefficient, and a time variable. The vibration excitation subfunction uses the frequency and amplitude of the dominant frequency component of the disturbance as parameters and outputs an equivalent excitation amplitude in the form of simple harmonic vibration. The airflow excitation subfunction uses the instantaneous value of the amplitude envelope as input and outputs a pulsed pressure excitation. The coupling modulation coefficient is dynamically adjusted according to the frequency proximity of the vibration and airflow, and the coupling strength is enhanced when the frequency difference between the two is less than a preset threshold. The outputs of the vibration excitation subfunction and the airflow excitation subfunction are weighted and superimposed through the coupling modulation coefficient to form a time-varying excitation function for disturbance with time as the independent variable and the equivalent disturbance amplitude as the dependent variable. This function fully describes the comprehensive excitation characteristics of the environmental disturbance at the target roll position as it evolves over time.

[0049] S233. Crack propagation dynamics prediction based on the coupling of cumulative damage and disturbance, specifically: A disturbance-crack propagation transfer function is established, which includes a stress intensity factor conversion sub-model, a Paris crack propagation rate sub-model, a disturbance amplification effect sub-model, and a strain energy release rate correction sub-model. The stress intensity factor conversion sub-model receives the cumulative plastic strain energy density and the crack initiation probability level, and converts the cumulative plastic strain energy density into the equivalent crack tip stress intensity factor reference value based on fracture mechanics theory. The initial equivalent crack length is determined by combining the crack initiation probability level. The Paris crack propagation rate sub-model adopts a modified Paris law, using the stress intensity factor amplitude as input. The variables, with the metal coating thickness and the elastic modulus of the substrate as material parameters, are used to calculate the crack propagation under a unit load cycle. The perturbation amplification effect sub-model receives the time-varying excitation function of the perturbation generated in step S232, analyzes the equivalent perturbation force amplitude time sequence, and calculates the dynamic tension fluctuation amplitude in combination with the preset unwinding acceleration to evaluate the dynamic amplification factor of the perturbation excitation on the stress field at the crack tip. The strain energy release rate correction sub-model integrates the cumulative plastic strain energy density and the current stress intensity factor level to calculate the energy release rate during crack propagation and correct the material constants in the Paris crack propagation rate sub-model.

[0050] The time-varying excitation function of the disturbance is input into the disturbance-crack propagation transfer function. The initial stress intensity factor level is determined by the stress intensity factor transformation sub-model. The dynamic stress intensity factor increment caused by the unwinding acceleration is calculated by the disturbance amplification effect sub-model. The total stress intensity factor time history is obtained by superimposing the results. This time history is input into the Paris crack propagation rate sub-model. Combined with the parameter correction of the strain energy release rate correction sub-model, the evolution curve of crack length over time is calculated by time integration. The slope of the curve is extracted as the predicted value of crack propagation rate.

[0051] S234. Dynamic prediction of electrostatic risk based on airflow triboelectric charging and multipath dissipation, specifically: An electrostatic accumulation-dissipation equilibrium model was established, comprising an airflow triboelectric electrostatic model, a surface conductivity dissipation sub-model, a bulk conductivity dissipation sub-model, and an environmental humidity-coupled correction sub-model. The time series of airflow fluctuation pressure amplitudes was extracted from the time-varying excitation function generated in step S232, and its fluctuation frequency and amplitude change rate were analyzed. The relative slip velocity between the thin film surface and the airflow was calculated using a preset unwinding acceleration. Based on the relative slip velocity and the airflow fluctuation pressure amplitude, the airflow triboelectric electrostatic model calculated the charge generation rate on the thin film surface, i.e., the electrostatic rate, using the triboelectric coefficient.

[0052] The surface conductivity dissipation sub-model calculates the dissipation rate of charge diffusion along the thin film surface to the grounding terminal or atmospheric environment based on the surface resistivity determined in step S21. The bulk conductivity dissipation sub-model calculates the dissipation rate of charge leakage perpendicularly through the polymer substrate to the back electrode or grounding layer based on the substrate bulk resistivity determined in step S21. The environmental humidity coupling correction sub-model receives real-time environmental humidity data and dynamically corrects the surface resistivity parameter in the surface conductivity dissipation sub-model according to the exponential relationship between humidity and surface conductivity. When humidity increases, the surface conductivity is increased to accelerate dissipation.

[0053] A charge conservation differential equation is established, using the charging rate output from the airflow triboelectric electron-generating model as the source term and the sum of the dissipation rates output from the modified surface conductivity dissipation sub-model and the bulk conductivity dissipation sub-model as the sink term. Solving this differential equation yields the cumulative charge curve over time. The charge is converted into an electrostatic potential value, and the maximum value in the cumulative curve is extracted as the predicted peak electrostatic potential. The predicted peak electrostatic potential is compared with a preset discharge critical threshold, and discharge probability levels are classified based on their similarity. Combining the predicted peak electrostatic potential with the discharge probability levels generates an electrostatic discharge risk index.

[0054] S24. Construction of dynamic tolerance window under two-dimensional coupling constraints, specifically: S241. Crack propagation safety margin assessment and dynamic compression within the tension range, specifically: Read the crack propagation rate value predicted in step S233, and simultaneously extract the tension sensitivity threshold of the corresponding downstream composite station from the station tension characteristic parameter cache stored in step S22; divide the crack propagation rate value by the tension sensitivity threshold, and define the ratio as the safety margin coefficient, which represents the adequacy of the current crack propagation level relative to the station's bearing capacity.

[0055] A preset safety threshold is set as the discrimination boundary, with a value ranging from 1.0 to 2.0, specifically based on material characteristics and historical defect rate statistics. The calculated safety margin coefficient is compared with the preset safety threshold. When the safety margin coefficient is lower than the preset safety threshold, it is determined that the current material has a risk of uncontrolled crack propagation under this workstation combination, and the safety margin compensation mechanism is immediately activated. The absolute value of the difference between the safety margin coefficient and the preset safety threshold is calculated and input into a preset mapping function or lookup table. The mapping function adopts a linear or non-linear relationship to convert the difference into a compression ratio of the tension fluctuation range. The larger the difference, the higher the compression ratio. The compression ratio represents the allowable fluctuation range that needs to be reduced based on the initial tension setting.

[0056] S242. Calculation of the matching degree of the electrostatic dissipation time window, specifically: Extract the peak electrostatic potential value from the electrostatic discharge risk index data structure generated in step S234; read the electrostatic dissipation time constant calculated and stored in step S21, and use this time constant as the time window reference value; set the upper limit of allowable electrostatic accumulation as the target safety level, which is determined based on the electrostatic breakdown resistance of the metal coating.

[0057] An exponential decay equation is established, using the estimated peak electrostatic potential as the initial value and the electrostatic dissipation time constant as the decay coefficient, to calculate the theoretical dissipation time required for the electrostatic potential to decay from the peak to the upper limit of allowable electrostatic accumulation. The calculated theoretical dissipation time is numerically compared with a preset time window threshold, which represents the maximum waiting time allowed by the production cycle. Through normalization calculation or piecewise function evaluation, the ratio or difference between the theoretical dissipation time and the preset time window threshold is converted into a time window matching degree index. The lower the matching degree value, the higher the electrostatic risk, and the more stringent the scheduling control is required.

[0058] S243. Construction of a two-dimensional coupling constraint matrix and generation of tolerance boundaries, specifically: A two-dimensional coordinate system is established, with the safety margin coefficient as one dimension and the time window matching degree as the other dimension, constructing a two-dimensional coupled constraint matrix. Each grid cell in the matrix corresponds to a specific combination of the safety margin coefficient range and the time window matching degree range. Each cell stores the scheduling strategy weight coefficient and tension control correction coefficient under this risk combination. Based on the currently calculated safety margin coefficient value and time window matching degree value, the corresponding grid cell in the matrix is ​​located, and the tension control correction coefficient stored in that cell is extracted as the compression ratio. The initial tension fluctuation range is obtained, which is the maximum tension range allowed by the equipment's physical capabilities or process specifications. The compression ratio is applied to the upper limit of the initial tension fluctuation range, and the upper limit is reduced through multiplication while keeping the lower limit unchanged to generate an acceptable tension fluctuation range. At the same time, the corresponding electrostatic accumulation margin coefficient is queried based on the time window matching degree value, and this margin coefficient is multiplied or added to the allowable electrostatic accumulation upper limit to calculate the actual allowable electrostatic accumulation upper limit value.

[0059] S244. Stress gradient release strategy generation and composite tolerance window aggregation, specifically: Read the upper and lower limits of the acceptable tension fluctuation range generated in step S243, and calculate the midpoint of the range as the benchmark tension setpoint; determine the activation status of the safety margin compensation mechanism in step S241. If the mechanism is activated, trigger the stress gradient release strategy generation process. Determine the number of stress steps based on the span of the acceptable tension fluctuation range and the crack initiation probability level. The larger the span or the higher the level, the more steps there are. Divide the acceptable tension fluctuation range into several levels and set the stress holding time for each level. This time is determined comprehensively based on the electrostatic dissipation time constant and production cycle constraints. Set the stress reduction amplitude between adjacent levels to ensure that the upper limit of the next level is equal to the lower limit of the previous level, forming a continuously decreasing stepped tension control curve; encapsulate the number of stress steps, the stress holding time for each level, and the stress reduction amplitude between adjacent levels into a stress gradient release strategy data structure.

[0060] The acceptable tension fluctuation range, the upper limit of allowed static electricity accumulation, and the stress gradient release strategy data structures are aggregated and encapsulated by structures or class objects to form a complete composite scheduling tolerance window. This window defines the multi-physics field safe operation boundary of the material in the current state.

[0061] S3. Based on a composite scheduling tolerance window, select schedulable material rolls, generate and execute a progressive unwinding scheduling instruction set, and monitor actual crack propagation events through acoustic emission to trigger dynamic reconfiguration; specifically including: S31. Real-time electrostatic state sensing and disturbance-tension dynamic response prediction, specifically: The real-time residual electrostatic value of candidate material rolls awaiting loading is acquired through an electrostatic monitoring device at the production line entrance. Simultaneously, the current environmental micro-vibration spectrum data and airflow fluctuation data corresponding to the spatial location of the target roll are extracted from the unsteady-state disturbance field distribution map data storage area. The stress gradient release strategy and acceptable tension fluctuation range parameters for the corresponding material roll are read from the composite scheduling tolerance window database.

[0062] A disturbance-tension transfer function is established, consisting of a disturbance input analytical sub-model, a roll dynamic response sub-model, a tension fluctuation synthesis sub-model, and an electrostatic-mechanical coupling correction module. The disturbance input analytical sub-model extracts the total vibration energy by performing frequency domain integration on the environmental micro-vibration spectrum data and superimposes the pressure amplitude time series from the airflow fluctuation data, generating an equivalent disturbance force time series through weighted fusion. The roll dynamic response sub-model establishes a torsional vibration differential equation with the roll rotational inertia, strip elastic modulus, and current roll diameter as system parameters, using the equivalent disturbance force time series as input excitation to solve for the roll angular acceleration dynamic response. The tension fluctuation synthesis sub-model, based on the kinematic coupling relationship between strip tension and roll angular acceleration, and considering the viscoelastic hysteresis characteristics of the polymer substrate, converts the angular acceleration response into a strip tension fluctuation time series, extracting the fluctuation amplitude range and dominant frequency components. The electrostatic-mechanical coupling correction module introduces the Coulomb force generated on the film surface by the real-time electrostatic residual value as an additional boundary condition, correcting the amplitude and phase shift of the tension fluctuation time series through mechanical equilibrium equations.

[0063] By solving the disturbance-tension transfer function, the tension response characteristics of each candidate material roll under the current disturbance environment and electrostatic state are predicted, including the predicted range of tension fluctuation amplitude and the principal component of fluctuation frequency.

[0064] S32. Dual-constraint filtering and progressive unwinding instruction set generation are as follows: S321. Real-time electrostatic sensing and disturbance-tension dynamic response prediction, specifically: Non-contact electrostatic voltmeters deployed in the material buffer area are used to collect the electrostatic potential values ​​on the surface of candidate material rolls awaiting loading in real time, serving as real-time electrostatic residual values. Simultaneously, unsteady-state disturbance field distribution maps are acquired from a distributed sensor network, and environmental micro-vibration spectrum data and airflow fluctuation data of the spatial grid points where the target roll is located are extracted.

[0065] A pre-built disturbance-tension transfer function is invoked, which includes four calculation modules: disturbance input analysis, roll dynamics response, tension fluctuation synthesis, and electrostatic-mechanical coupling correction. Environmental micro-vibration spectrum data and airflow fluctuation data are input into the disturbance input analysis module to generate an equivalent disturbance time series. The roll dynamics response module calculates the angular acceleration dynamic response, which is then converted into a material strip tension fluctuation time series by the tension fluctuation synthesis module. Finally, the electrostatic-mechanical coupling correction module introduces the Coulomb force effect of the real-time electrostatic residual value for boundary correction. Solving this function yields the tension response characteristics of each candidate material roll, specifically including the predicted range of tension fluctuation amplitude and the principal components of the fluctuation frequency.

[0066] S322. Two-dimensional constraint screening and schedulable material volume determination, specifically: A two-dimensional constraint screening mechanism is established, using real-time electrostatic residual value as the electrostatic constraint dimension and the tension fluctuation amplitude prediction range as the tension constraint dimension. The candidate material roll list is traversed, and the real-time electrostatic residual value of each candidate material roll is compared with the allowable electrostatic accumulation upper limit in the composite scheduling tolerance window. The upper and lower limits of the tension fluctuation amplitude prediction range are compared with the upper and lower limits of the acceptable tension fluctuation range to determine the inclusion relationship. Only candidate material rolls whose real-time electrostatic residual value does not exceed the allowable electrostatic accumulation upper limit and whose tension fluctuation amplitude prediction range falls entirely within the acceptable tension fluctuation range are retained and marked as schedulable material rolls. This screening process ensures that the selected material rolls are within the safe operating boundaries under both the current electrostatic state and the predicted tension response, satisfying the multiphysics safety constraints defined by the composite scheduling tolerance window.

[0067] S323. Generation and issuance of progressive unwinding scheduling instruction set, specifically: For the schedulable material rolls selected in step S322, their stress relaxation compensation requirements and disturbance-electrostatic coupling vulnerability levels are retrieved from the multiphysics risk stratification tagging data structure; corresponding stress gradient release strategy parameters are extracted from the composite scheduling tolerance window database. A progressive unwinding scheduling instruction set is constructed, which includes four core elements: the target workstation sequence determined by the downstream workstation capacity balance and order priority algorithm; the tension feedforward correction curve calculated by table lookup or interpolation based on the stress relaxation compensation requirement level; the unwinding acceleration limit value set according to the disturbance-electrostatic coupling vulnerability level; and the expected acoustic emission characteristic template obtained by matching the metal coating type and cumulative plastic strain energy density of the material roll from the historical acoustic emission database. This instruction set is distributed to the AGV scheduling controller and the unwinding workstation servo drive via industrial fieldbus or Ethernet communication protocol as benchmark reference data for subsequent unwinding process control and anomaly monitoring.

[0068] S33. Execution of online acoustic emission monitoring and dynamic path reconstruction, specifically: S331. Real-time acquisition and feature extraction of acoustic emission signals, specifically: Control commands are sent to the AGV scheduling controller and the unwinding station servo system to drive the automated guided vehicle to transport the scheduleable material roll from the buffer area to the target roll position, completing the mechanical connection and material threading operation. A broadband acoustic emission sensor is installed on the end surface of the roll support bearing seat at the unwinding station, and a contact-type acoustic guided wave sensor is installed on the surface of the guide roller, forming an acoustic emission sensor array. The broadband acoustic emission sensor detects the transient elastic wave signal generated by the peeling between the metal plating and the substrate layer using piezoelectric ceramic elements, while the contact-type acoustic guided wave sensor monitors the continuous acoustic emission activity of crack propagation through guided wave propagation characteristics. The sensor array is connected to a high-speed data acquisition card through a pre-charge amplifier and an anti-aliasing filter, with the sampling frequency set to more than ten times the dominant frequency of the acoustic emission signal to satisfy the Nyquist sampling theorem.

[0069] The unwinding procedure is initiated, and the unwinding acceleration limit control roller is started according to step S323. The unwinding tension is adjusted based on the tension feedforward correction curve. During the actual unwinding process of the material roll, the acoustic emission sensor array continuously collects the interlayer peeling acoustic emission signals. The time-domain signal is converted into a frequency-domain representation using a fast Fourier transform algorithm, and the spectral characteristics of specific frequency bands are extracted, including the main frequency peak, spectral centroid, and frequency band energy distribution. The acoustic emission energy accumulation rate per unit time is calculated using a short-time energy integration algorithm.

[0070] S332. Crack propagation mode matching and deviation determination, specifically: The real-time spectral features and energy accumulation rate extracted in step S331 are used for pattern matching with the expected acoustic emission feature template stored in the progressive unwinding scheduling instruction set generated in step S323. The expected acoustic emission feature template is constructed based on the metal coating properties and cumulative plastic strain energy density of the material roll, using statistical analysis of acoustic emission data from historical normal unwinding processes. It includes a standard spectral envelope and a normal energy accumulation rate curve. A cross-correlation algorithm is used to calculate the similarity deviation between the real-time spectral features and the standard spectral envelope, yielding the spectral feature deviation rate; the relative error between the real-time energy accumulation rate and the normal energy accumulation rate curve is calculated, yielding the energy accumulation rate deviation rate.

[0071] A first preset threshold is set to 20% to 25%, determined by statistical analysis of acoustic emission data from historical defect-free unwinding batches, using the mean of the spectral characteristic deviation rate plus three times the standard deviation. A second preset threshold is set to 40% to 45%, determined by the cumulative distribution function of the energy accumulation rate, using the upper limit of the 95% confidence interval. When the spectral characteristic deviation rate exceeds the first preset threshold of 20% to 25%, it indicates an abnormal change in the crack propagation mode; or when the energy accumulation rate deviation rate exceeds the second preset threshold of 40% to 45%, it indicates that the crack propagation rate exceeds the expected range. When either condition is met, it is determined that an actual crack propagation event has deviated from the expected acoustic emission characteristic template, triggering the abnormal response process.

[0072] S333. Dynamic refactoring priority assessment and path replanning, specifically: Obtain the cumulative plastic strain energy density and disturbance-electrostatic coupling vulnerability level of the material roll currently being unwound. Based on the spectral characteristic deviation rate and energy accumulation rate deviation rate calculated in step S332, calculate the deviation severity index using a weighted summation or weighted geometric average algorithm. Input the cumulative plastic strain energy density, disturbance-electrostatic coupling vulnerability level, and deviation severity index into a nonlinear weighted fusion function. This function uses power-law weighting or neural network mapping to calculate a dynamic reconstruction priority coefficient. This coefficient comprehensively and quantitatively represents the risk level of continuing to process the material at the current workstation.

[0073] The preset reconstruction threshold is set to a range of 0.65 to 0.85. This threshold is determined by statistically classifying and analyzing the distribution curves of dynamic reconstruction priority coefficients of defective batches scrapped due to microcrack propagation and normal unwinding batches in historical production data, and selecting the critical boundary value that optimizes the combined defect sample detection rate and normal sample misclassification rate. The calculated dynamic reconstruction priority coefficient is compared with the preset reconstruction threshold of 0.65 to 0.85. When the dynamic reconstruction priority coefficient exceeds the preset reconstruction threshold, the dynamic reconstruction process of the scheduling path is immediately triggered.

[0074] The system queries the production line station status database to obtain the current occupancy status, tension control capability parameters, and topological connection relationships between each downstream unwinding station. Using Dijkstra's shortest path algorithm or the A-Star heuristic search algorithm, it calculates the optimal safe path from the current position to the backup low-stress unwinding station or temporary buffer station. The path selection prioritizes low-stress stations with active tension attenuation function, electrostatic elimination device, or historical acoustic emission background noise level.

[0075] S334. Scheduling path switching execution and online model parameter update, specifically: A path switching command is sent to the AGV scheduling controller to control the automated guided vehicle to detach the current material roll from the current target roll and guide it to the new destination according to the backup low-stress path or temporary buffer station calculated in step S333, realizing dynamic switching of the conveying path. Simultaneously, the occurrence sequence of actual crack propagation events, the spectral characteristic deviation rate, and the energy accumulation rate deviation rate recorded in step S332 are written into the elastoplastic strain energy accumulation model established in step S11 through the model feedback interface. Using a recursive least squares algorithm or a Kalman filter algorithm, the residual sequence between the observed crack propagation deviation and the model prediction is used to update the calculation weight coefficients of the metal layer elastoplastic constitutive submodel and the interlayer strain energy coupling accumulation submodel in the elastoplastic strain energy accumulation model online, optimizing the model's prediction accuracy of the cumulative plastic strain energy density of subsequent material rolls. The updated calculation weight coefficients are stored in the model parameter library, and the updated model parameters are output for subsequent steps in S1, realizing online learning and adaptive evolution of the model.

[0076] Example 2 like Figure 2 As shown, a material intelligent scheduling system for a thin film production line is used to implement a material intelligent scheduling method for a thin film production line, including: A. Disturbance sensing module, specifically including: The winding stress history acquisition unit is used to acquire the winding stress history data of each material roll at the winding station. The winding stress history data includes the winding tension time sequence spectrum and the stress lock-in state at the winding stop time. The cumulative plastic strain energy calculation unit is used to calculate the cumulative plastic strain energy density of the metal coating induced by the historical data of the winding stress based on the elastoplastic strain energy accumulation model of the metal-substrate composite layer. The cumulative plastic strain energy density is used to characterize the degree of consumption of the microcrack initiation barrier inside the metal layer. The environmental disturbance acquisition unit is used to collect environmental micro-vibration spectrum data and airflow fluctuation data at multiple key nodes in the production line, and simultaneously acquire the real-time electrostatic potential value of each film roll in production. The disturbance field construction unit is used to construct an unsteady disturbance field distribution map based on environmental micro-vibration spectrum data and airflow fluctuation data; The multiphysics tag generation unit is used to identify at least one target roll in a disturbance-electrostatic coupling sensitive state based on the unsteady disturbance field distribution map and the real-time electrostatic potential value of each in-production film roll, and associate the cumulative plastic strain energy density of the metal coating with the target roll to generate a material candidate queue with multiphysics risk stratification tags. The multiphysics risk stratification tags include crack initiation probability level, stress relaxation compensation requirement, and disturbance-electrostatic coupling vulnerability level.

[0077] B. Tolerance matching module, connected to the disturbance sensing module, specifically includes: The attribute parameter acquisition unit is used to acquire the material properties and thickness parameters of the target film corresponding to the target roll; The threshold determination unit is used to determine the critical winding stress threshold and electrostatic dissipation time constant of the target thin film based on the material properties and thickness parameters. The station parameter acquisition unit is used to acquire the tension control capability parameters of each unwinding station on the production line and the tension sensitivity threshold of the downstream composite station. The risk prediction unit is used to predict the crack propagation rate and electrostatic discharge risk index of each material roll in the material candidate queue based on the crack initiation probability level and cumulative plastic strain energy density, combined with the unsteady disturbance field distribution map. The tolerance window generation unit is used to calculate the safety margin between the crack propagation rate and the tension sensitivity threshold, match the electrostatic discharge risk index with the electrostatic dissipation time constant, and dynamically generate a composite scheduling tolerance window that includes an acceptable tension fluctuation range, an upper limit for allowed electrostatic accumulation, and a stress gradient release strategy.

[0078] C. The collaborative scheduling module, connected to the tolerance matching module, specifically includes: The tension response prediction unit is used to obtain the real-time electrostatic residual value of the candidate material rolls waiting to be rolled up, and based on the unsteady disturbance field distribution map and the composite scheduling tolerance window, predict the tension response characteristics of each candidate material roll under the combined action of environmental micro-vibration spectrum data and airflow fluctuation data. A schedulable screening unit is used to screen out schedulable material rolls that simultaneously meet the acceptable tension fluctuation range and the allowable static electricity accumulation limit by using the real-time electrostatic residual value and the tension response characteristics as constraints. The instruction set generation unit is used to generate a progressive unwinding scheduling instruction set, which includes a target workstation sequence, a tension feedforward correction curve corresponding to the stress relaxation compensation requirement, an unwinding acceleration limit corresponding to the disturbance-electrostatic coupling vulnerability level, and an expected acoustic emission feature template. The instruction execution unit is used to execute instructions to transport the schedulable material roll to the target roll for connection; The dynamic reconfiguration unit is used to collect interlayer stripping acoustic emission signals through an acoustic emission sensor array deployed at the unwinding station during the actual unwinding process, extract spectral features and energy accumulation rate, and trigger dynamic reconfiguration of the scheduling path when the occurrence timing and energy accumulation rate of the detected actual crack propagation event deviate from the matching threshold of the expected acoustic emission feature template, so as to guide the current material roll to a backup low-stress path or a temporary buffer station.

[0079] As can be seen from the above description, the embodiments of the present invention achieve the following technical effects: The calculation of cumulative plastic strain energy density of metal coatings based on the historical integral of winding stress established by step S1 breaks through the limitation of traditional scheduling systems that only focus on the current storage state of materials, and realizes a quantitative assessment of the degree of consumption of the initiation barrier of microcracks inside the metal layer. The simultaneously constructed unsteady disturbance field distribution map and disturbance-electrostatic coupling sensitivity coefficient calculation integrate environmental micro-vibration, airflow fluctuation and electrostatic potential in spatiotemporal analysis, enabling the system to identify the mechanical vulnerability of specific rolls under specific disturbance-electrostatic coupling states. This realizes the leap from single material property management to multi-physics environment collaborative perception, and significantly improves the defect prevention capability of high-end metal coating films such as magnetron sputtered automotive window films.

[0080] The disturbance-crack propagation transfer function and electrostatic accumulation-dissipation equilibrium model established through step S2 breaks through the rigid limitations of traditional fixed tension thresholds and electrostatic protection standards. Based on a composite scheduling tolerance window dynamically generated by a two-dimensional coupled constraint matrix, which includes an acceptable tension fluctuation range, an allowable upper limit for electrostatic accumulation, and a stress gradient release strategy, the safety operating boundary can be adaptively adjusted according to the real-time changes in the internal cumulative plastic strain energy density of the material and the intensity of external environmental disturbances. This mechanism achieves deep integration of scheduling decisions and process control parameters, enabling high-defect-risk materials to automatically match low-stress stations and progressive unwinding strategies, significantly reducing batch scrap rates caused by microcrack propagation and improving the flexible manufacturing capability of the production line.

[0081] The closed-loop system for tension response prediction and real-time acoustic emission monitoring based on the perturbation-tension transfer function, established through step S3, realizes a complete control system of prediction-screening-verification-reconstruction. It performs pattern matching between real-time acoustic emission signals and expected acoustic emission feature templates, identifies crack propagation deviations using first and second preset thresholds, and triggers dynamic reconstruction of the production path based on dynamic reconstruction priority coefficients and preset reconstruction thresholds. This allows the system to guide materials to backup low-stress paths or temporary buffer stations when anomalies are detected. This closed-loop mechanism compresses quality control from traditional batch-level post-inspection to single-roll-level real-time intervention. Simultaneously, it achieves system self-learning by updating the elastoplastic strain energy accumulation model parameters online, significantly improving the production reliability and yield of high-end functional films.

[0082] The embodiments and / or implementation methods described above are merely preferred embodiments and / or implementation methods for implementing the technology of the present invention, and are not intended to limit the implementation methods of the technology of the present invention in any way. Any person skilled in the art may make some modifications or alterations to other equivalent embodiments without departing from the scope of the technical means disclosed in the present invention, but these should still be regarded as the technology or embodiments that are substantially the same as the present invention. This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. The above descriptions are only preferred embodiments of this application. It should be noted that due to the limitations of written expression, while there are objectively infinite specific structures, those skilled in the art can make several improvements, modifications, or changes without departing from the principles of this application, and can also combine the above technical features in an appropriate manner. These improvements, modifications, changes, or combinations, or the direct application of the inventive concept and technical solution to other situations without modification, should all be considered within the scope of protection of this application.

Claims

1. A method for intelligent material scheduling in a thin-film production line, characterized in that, include: Historical data on winding stress is obtained to calculate the cumulative plastic strain energy density of the metal coating. Environmental micro-vibration and airflow fluctuation data are collected to construct an unsteady disturbance field distribution map. Disturbance-electrostatic coupling sensitive states are identified and associated with the cumulative plastic strain energy density to generate a material candidate queue with multi-physics risk stratification labels. Receive the material candidate queue, predict the crack propagation rate based on the multiphysics risk stratification label and the unsteady disturbance field distribution map, and dynamically generate a composite scheduling tolerance window; Based on the composite scheduling tolerance window, scheduleable material rolls are selected, a progressive unwinding scheduling instruction set is generated and executed, and actual crack propagation events are monitored through acoustic emission to trigger dynamic reconstruction.

2. The intelligent material scheduling method for a thin film production line according to claim 1, characterized in that, The process of acquiring historical winding stress data to calculate the cumulative plastic strain energy density of the metal coating, collecting environmental micro-vibration and airflow fluctuation data to construct an unsteady disturbance field distribution map, identifying disturbance-electrostatic coupling sensitive states and associating them with the cumulative plastic strain energy density, and generating a material candidate queue with multi-physics risk stratification labels specifically includes: By deploying a stress sensor array at the winding station, the winding tension time spectrum of each material roll is collected in real time during the winding process, and the stress lock-in state is recorded at the moment of winding stop; based on the elastoplastic strain energy accumulation model of the metal-substrate composite layer, the cumulative plastic strain energy density of the metal coating induced by the winding tension time spectrum and the stress lock-in state is calculated. Simultaneously, high-sensitivity accelerometers and piezoelectric airflow sensors deployed at multiple nodes of the production line are used to collect environmental micro-vibration spectrum data and airflow fluctuation data, and real-time electrostatic potential values ​​of each in-production film roll are obtained through a non-contact electrostatic voltmeter; the environmental micro-vibration spectrum data and airflow fluctuation data are spatially interpolated and superimposed to construct an unsteady-state disturbance field distribution map; Based on the unsteady disturbance field distribution map and real-time electrostatic potential value, target scrolls sensitive to disturbance-electrostatic coupling are identified, and the cumulative plastic strain energy density of the metal coating is correlated to generate a material candidate queue with multi-physics risk stratification labels.

3. The intelligent material scheduling method for a thin film production line according to claim 2, characterized in that, Based on the unsteady-state disturbance field distribution map and real-time electrostatic potential values, target scrolls sensitive to disturbance-electrostatic coupling are identified, and the cumulative plastic strain energy density of the metal coating is correlated to generate a material candidate queue with multi-physics risk stratification tags, specifically including: The environmental micro-vibration spectrum energy density and airflow fluctuation pressure amplitude in the unsteady disturbance field distribution map are weighted and fused to calculate the disturbance coupling strength index; the real-time electrostatic potential value of each in-production film roll is multiplied with the disturbance coupling strength index to obtain the disturbance-electrostatic coupling sensitivity coefficient. At least one in-production thin film roll with a disturbance-electrostatic coupling sensitivity coefficient exceeding a preset sensitivity threshold is selected, marked as a target roll, and the cumulative plastic strain energy density of the metal coating corresponding to the target roll is extracted. The cumulative plastic strain energy density of the metal coating and the disturbance-electrostatic coupling sensitivity coefficient are nonlinearly weighted and fused to generate a crack initiation probability level; the stress relaxation compensation requirement is generated based on the time gradient change rate of the disturbance-electrostatic coupling sensitivity coefficient; and the disturbance-electrostatic coupling vulnerability level is generated according to the ratio of the disturbance coupling strength index of the grid point where the target scroll is located to the real-time electrostatic potential value. The crack initiation probability level, stress relaxation compensation requirement, and disturbance-electrostatic coupling vulnerability level are aggregated into a multiphysics risk stratification label, and the multiphysics risk stratification label is bound to the target scroll to output a material candidate queue with the multiphysics risk stratification label.

4. The intelligent material scheduling method for a thin film production line according to claim 3, characterized in that, Receive the material candidate queue, predict the crack propagation rate based on the multiphysics risk stratification label and the unsteady disturbance field distribution map, and dynamically generate a composite scheduling tolerance window, specifically including: Receive the material candidate queue, obtain the material properties and thickness parameters of the target film corresponding to the target roll, and determine the critical winding stress threshold and electrostatic dissipation time constant of the target film based on the material properties and thickness parameters; Obtain the tension control capability parameters of each unwinding station on the production line and the tension sensitivity threshold of the downstream compounding station. For each material roll in the material candidate queue, based on the crack initiation probability level and cumulative plastic strain energy density, combined with the unsteady disturbance field distribution map, the crack propagation rate and electrostatic discharge risk index of the material under the preset unwinding acceleration are predicted. The crack propagation rate and the tension sensitivity threshold are used to calculate the safety margin. The electrostatic discharge risk index of Siping City is matched with the electrostatic dissipation time constant to dynamically generate a composite scheduling tolerance window.

5. The intelligent material scheduling method for a thin film production line according to claim 4, characterized in that, For each material roll in the candidate material queue, based on the crack initiation probability level and cumulative plastic strain energy density, and combined with the unsteady disturbance field distribution map, the crack propagation rate and electrostatic discharge risk index of the material under a preset unwinding acceleration are predicted, specifically including: For each material roll in the material candidate queue, extract its corresponding crack initiation probability level and cumulative plastic strain energy density, and simultaneously obtain the metal coating thickness and substrate elastic modulus parameters of the material roll. The environmental micro-vibration spectrum energy density time series and the airflow fluctuation pressure amplitude time series of the spatial grid points where the target scroll is located in the unsteady disturbance field distribution map are decomposed in the time and frequency domain to extract the disturbance main frequency component and the fluctuation amplitude envelope, and construct the disturbance time-varying excitation function. Based on the cumulative plastic strain energy density and crack initiation probability level, combined with the metal coating thickness and substrate elastic modulus parameters, a disturbance-crack propagation transfer function is established. The disturbance time-varying excitation function is input into the disturbance-crack propagation transfer function to predict the crack propagation rate of the material under a preset unwinding acceleration. Based on the fluctuation frequency and amplitude change rate of the airflow fluctuation pressure amplitude in the time-varying excitation function of the disturbance, combined with the electrostatic dissipation constant and surface resistivity of the material, an electrostatic accumulation-dissipation equilibrium model is constructed to predict the electrostatic discharge risk index of the material under a preset unwinding acceleration.

6. The intelligent material scheduling method for a thin film production line according to claim 5, characterized in that, The crack propagation rate is compared with the tension sensitivity threshold for safety margin calculation. The electrostatic discharge risk index of Siping City is matched with the electrostatic dissipation time constant for time window matching to dynamically generate a composite scheduling tolerance window, specifically including: The ratio of the crack propagation rate to the tension sensitivity threshold is calculated to obtain the safety margin coefficient. When the safety margin coefficient is lower than the preset safety threshold, the safety margin compensation mechanism is activated, and the compression ratio of the tension fluctuation range is determined based on the difference between the safety margin coefficient and the preset safety threshold. Using the estimated peak electrostatic potential in the electrostatic discharge risk index as input, and the electrostatic dissipation time constant as the time window benchmark, the dissipation time required for the estimated peak electrostatic potential to decay to the upper limit of allowable electrostatic accumulation is calculated. The dissipation time is matched with the preset time window threshold to obtain the time window matching degree. Based on the safety margin coefficient and the time window matching degree, a two-dimensional coupling constraint matrix is ​​constructed, and the compression ratio is applied to the initial tension fluctuation range to generate an acceptable tension fluctuation range; the electrostatic accumulation margin corresponding to the time window matching degree is used as the upper limit of allowable electrostatic accumulation. Based on the acceptable tension fluctuation range and the allowable static electricity accumulation limit, and combined with the activation state of the safety margin compensation mechanism, a stress gradient release strategy is generated, and the acceptable tension fluctuation range, the allowable static electricity accumulation limit, and the stress gradient release strategy are aggregated into a composite scheduling tolerance window.

7. The intelligent material scheduling method for a thin film production line according to claim 6, characterized in that, Based on the composite scheduling tolerance window, schedulable material rolls are selected, a progressive unwinding scheduling instruction set is generated and executed, and actual crack propagation events are monitored via acoustic emission to trigger dynamic reconstruction. Specifically, this includes: The real-time electrostatic residual value of the candidate material rolls waiting to be rolled up is obtained, and based on the unsteady disturbance field distribution map and the composite scheduling tolerance window, the disturbance-tension transfer function is used to predict the tension response characteristics of each candidate material roll under the combined action of environmental micro-vibration spectrum data and airflow fluctuation data. Using the real-time electrostatic residual value and tension response characteristics as constraints, schedulable material rolls that simultaneously meet the acceptable tension fluctuation range and the allowable upper limit of electrostatic accumulation are selected, and a progressive unwinding scheduling instruction set is generated. The system executes the connection command between the schedulable material roll and the target roll, collects interlayer peeling signals through an acoustic emission sensor array, and triggers dynamic reconstruction of the scheduling path when the actual crack propagation event deviates from the matching threshold of the expected acoustic emission feature template, guiding the material roll to a backup low-stress path or a temporary buffer station.

8. The intelligent material scheduling method for a thin film production line according to claim 7, characterized in that, Based on the aforementioned real-time electrostatic residual value and tension response characteristics, schedulable material rolls that simultaneously meet the acceptable tension fluctuation range and the allowable upper limit of electrostatic accumulation are selected, and a progressive unwinding scheduling instruction set is generated, specifically including: The real-time electrostatic residual value of the candidate material rolls waiting to be rolled up is obtained, and based on the unsteady disturbance field distribution map and the composite scheduling tolerance window, the disturbance-tension transfer function is used to predict the tension response characteristics of each candidate material roll under the combined action of environmental micro-vibration spectrum data and airflow fluctuation data. The real-time electrostatic residual value and the tension fluctuation amplitude prediction range in the tension response characteristics are used as two-dimensional constraints. They are compared with the upper limit of allowable electrostatic accumulation and the acceptable tension fluctuation range in the composite scheduling tolerance window, respectively, and at least one candidate material volume that meets both constraints is selected as a schedulable material volume. For the selected schedulable material rolls, extract their corresponding stress relaxation compensation requirements and disturbance-electrostatic coupling vulnerability levels. Combined with the stress gradient release strategy in the composite scheduling tolerance window, generate a progressive unwinding scheduling instruction set. Send the progressive unwinding scheduling instruction set to the execution unit.

9. The intelligent material scheduling method for a thin film production line according to claim 8, characterized in that, The process involves executing the connection command between the schedulable material roll and the target roll, acquiring interlayer peeling signals through an acoustic emission sensor array, and triggering dynamic reconfiguration of the scheduling path when the actual crack propagation event deviates from the matching threshold of the expected acoustic emission feature template. This reconfiguration guides the material roll to a backup low-stress path or a temporary buffer station. Specifically, this includes: The instruction to transport the schedulable material roll to the target roll for connection is executed. During the actual unwinding process of the material roll, the acoustic emission signal of interlayer stripping is collected in real time by an acoustic emission sensor array deployed at the unwinding station, and the spectral characteristics and energy accumulation rate of the acoustic emission signal are extracted. The extracted spectral features are pattern matched with the expected acoustic emission feature templates in the progressive unwinding scheduling instruction set to calculate the spectral feature deviation rate and the energy accumulation rate deviation rate. When the spectral feature deviation rate exceeds the first preset threshold or the energy accumulation rate deviation rate exceeds the second preset threshold, it is determined that an actual crack propagation event deviation has occurred. Obtain the cumulative plastic strain energy density and disturbance-electrostatic coupling vulnerability level of the current material roll, and calculate the dynamic reconstruction priority coefficient based on the severity of the deviation judgment; when the dynamic reconstruction priority coefficient exceeds the preset reconstruction threshold, trigger the dynamic reconstruction of the scheduling path, and calculate the backup low-stress path or temporary buffer station from the current position to the downstream composite station. Switch the current material roll conveying path to a backup low-stress path or guide it to a temporary buffer station, and simultaneously update the calculation weight coefficient of the cumulative plastic strain energy density online.

10. A material intelligent scheduling system for a thin film production line, characterized in that, A material intelligent scheduling method for implementing a thin film production line according to any one of claims 1-9 includes: The disturbance sensing module is used to collect environmental micro-vibration spectrum data and airflow fluctuation data of multiple key nodes in the production line, and simultaneously acquire the real-time electrostatic potential value of each film roll in production; based on the environmental micro-vibration spectrum data and airflow fluctuation data, an unsteady disturbance field distribution map is constructed, and based on the unsteady disturbance field distribution map and the real-time electrostatic potential value of each film roll in production, at least one target roll in a disturbance-electrostatic coupling sensitive state is identified; The tolerance matching module, connected to the disturbance sensing module, is used to obtain the material properties and thickness parameters of the target film corresponding to the target roll, and determine the critical winding stress threshold and electrostatic dissipation time constant of the target film based on the material properties and thickness parameters; and dynamically generate a composite scheduling tolerance window for the target roll that includes an acceptable tension fluctuation range and an allowable electrostatic accumulation upper limit by combining the critical winding stress threshold and the electrostatic dissipation time constant. The collaborative scheduling module, connected to the tolerance matching module, is used to obtain the real-time electrostatic residual value of the candidate material rolls waiting to be rolled up, and predict the tension response characteristics of each candidate material roll under the action of the unsteady disturbance field distribution map; using the real-time electrostatic residual value and tension response characteristics as constraints, at least one schedulable material roll falling within the composite scheduling tolerance window is selected, and an instruction to transport the schedulable material roll to the target roll for connection is generated and executed.

Images