Method for intelligent control of film thickness in preparation of electromagnetic shielding film

Abstract

The application discloses a kind of film thickness intelligent control methods in electromagnetic shielding film preparation, it is related to shielding film preparation technical field, comprising the following steps: setting multiple film thickness sensors;Electrostatic neutralization is applied in coating area;Microclimate control device is set in the periphery of coating platform;Non-contact deformation detection system is set, and the deformation time series data of each coating head unit is collected;Real-time value of film thickness and deformation is predicted;Compensation value is introduced to compensate film thickness deviation and deformation deviation, and control module executes displacement or pressure control according to compensation value to each coating head, and the input information of prediction model is updated synchronously;Cloud platform uses distributed algorithm to coordinate each control node, and is fused by asynchronous scheduling and prediction model;The application effectively inhibits the edge thickening / thinning phenomenon caused by coating head deformation, tension gradient.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of shielding film preparation, and particularly to a film thickness intelligent control method in electromagnetic shielding film preparation. BACKGROUND

[0002] In the precise preparation process of high-performance electromagnetic shielding film, film thickness uniformity directly affects the electromagnetic shielding effectiveness, mechanical flexibility and environmental stability of the material. The current mainstream coating process (such as slot coating and gravure coating) relies on a classic PID model based on feedback control to achieve film thickness regulation by adjusting coating gap, line speed and other parameters.

[0003] However, in the multi-layer film process, the film thickness regulation system faces challenges from edge effects and environmental disturbances (such as humidity and static electricity). In particular, under the influence of factors such as coating head deformation and tension gradient, the film thickness changes significantly at the edge of the coating head. In addition, the current control model usually lacks special compensation for edge effects and ignores the influence of environmental variables (such as humidity changes and charge accumulation) on film thickness, resulting in insufficient film thickness control accuracy. Since there is a close relationship between edge effects and environmental disturbances, these factors not only affect the middle of the film, but also cause greater fluctuations in the edge area of the film, further increasing the difficulty of film thickness control. SUMMARY

[0004] To solve the defects in the prior art, the present application provides a film thickness intelligent control method in electromagnetic shielding film preparation.

[0005] To solve the above technical problems, the present application provides the following technical solutions:

[0006] The present application provides a film thickness intelligent control method in electromagnetic shielding film preparation, comprising the following steps:

[0007] A plurality of film thickness sensors are arranged at the center area and the edge area of the coating platform to collect real-time two-dimensional coordinate film thickness time series data of the center area and the edge area;

[0008] An electrostatic neutralization is applied to the coating area, and the electrostatic neutralization process is controlled by a charge disturbance matrix obtained by an electrostatic inductor to control surface charge accumulation;

[0009] A microclimate control device is arranged around the coating platform to maintain the set temperature and relative humidity by a PID control algorithm and shield external air flow disturbances;

[0010] A non-contact deformation detection system is arranged to collect deformation time series data of each coating head unit;

[0011] The film thickness time sequence data and the deformation time sequence data are uploaded to a cloud platform based on an edge computing node, a mapping relationship between film thickness deviation and deformation deviation is established through a multi-model prediction module, and real-time values of the film thickness and the deformation are predicted;

[0012] A compensation value is introduced to compensate the film thickness deviation and the deformation deviation, a control module performs displacement or pressure control on each coating head according to the compensation value, and input information of a prediction model is updated synchronously;

[0013] The cloud platform adopts a distributed algorithm to coordinate each control node, and fuses through asynchronous scheduling and a prediction model.

[0014] The present application has the following advantages:

[0015] 1. The present application realizes directional correction of the edge film thickness deviation by arranging high-precision non-contact sensors in the center and the edge area of the coating platform, and combining an adaptive weight compensation mechanism of the edge feedback unit.

[0016] 2. In the present application, an electrostatic neutralization device and a microclimate control device are integrated, so that the system can actively offset the influence of humidity fluctuation and electrostatic accumulation on the film thickness.

[0017] 3. In the present application, the cloud platform based on edge computing realizes multi-physical field coupling analysis of film thickness-deformation-environmental parameters through a multi-model prediction module and a distributed control architecture. BRIEF DESCRIPTION OF DRAWINGS

[0018] The accompanying drawings are included to provide a further understanding of the present application, and constitute a part of the specification, illustrate the present application, and are used to explain the present application, and do not constitute a limitation on the present application. In the drawings:

[0019] Figure 1 It is a flowchart of the film thickness intelligent control method of the present application. DETAILED DESCRIPTION

[0020] The preferred embodiments of the present application are described below with reference to the accompanying drawings, and it should be understood that the preferred embodiments described herein are intended for illustrative purposes only and are not intended to limit the present application.

[0021] Embodiment one

[0022] As shown in the figure, a film thickness intelligent control method in electromagnetic shielding film preparation includes the following steps: Figure 1

[0023] A plurality of film thickness sensors are arranged at the center area and the edge area of the upper end of the coating platform, and the sensors are non-contact optical interference sensors (based on interference fringe shift, suitable for transparent or semi-transparent film materials) or laser triangular reflection sensors (which can be used for complex surfaces or non-transparent film materials) to collect real-time two-dimensional coordinate film thickness time sequence data at the center area and the edge area.

[0024] Preferably, the number of distribution points is not less than 9, 3-5 points are arranged in the center area, and 4-6 points are arranged in the edge area according to equal arc length distribution, which ensures that the axial and radial film thickness gradients can be monitored in real time.

[0025] An electrostatic neutralization is applied to the coating area, and the electrostatic neutralization process is controlled by the charge disturbance matrix obtained by the electrostatic inductor to control the accumulation of surface charge.

[0026] The charge disturbance matrix E(x, y, t) is a three-dimensional time sequence field describing the change of surface charge density at two-dimensional coordinates (x, y) per unit time, and the film thickness fluctuation matrix H(x, y, t) is obtained according to the above sensor, which represents the real-time change of film thickness at time t and coordinates (x, y). Since electrostatic disturbance can induce film shift / adsorption, which in turn affects local thickness, a causal mapping model can be established between the two, and used for film thickness prediction and compensation control, such as:

[0027]

[0028] Where ∇E is the charge distribution gradient, is the charge change rate, indicating the disturbance intensity, usually ≤ 200 nC / (cm²·s), and ε is the system modeling error, usually ≤ 0.05 μm.

[0029] The following key response thresholds are set:

[0030] Charge disturbance rate threshold (severe disturbance determination):

[0031] If , immediately increase the output power of the ion wind rod or increase the discharge frequency (by 10-30%), preferably δ1=150 nC / (cm²·s);

[0032] ​​Spatial gradient hotspot judgment (charge accumulation):

[0033] ▽E>δ2, then dynamically adjust the plasma electrode output frequency (±10%) or pulse voltage amplitude (±200V), preferably, δ2=500~V / m 2 .

[0034] Set up microclimate control devices around the coating platform to maintain the set temperature and relative humidity through PID control algorithm and shield external air flow disturbance;

[0035] Set up non-contact deformation detection system, adopt laser interferometer or structured light three-dimensional reconstruction equipment to collect deformation time series data of each coating head unit, optionally, the present application sets a set of laser interferometer in front, side and back of the coating head to collect deformation of key structure surface in real time; set up structured light projection sensor in special structure area (such as warping prone area) to carry out 3D profile reconstruction, measurement frequency ≥100Hz, collection accuracy ≤0.2μm, all data is uploaded to edge node for real-time fitting and difference compensation;

[0036] Upload film thickness time series data and deformation time series data to cloud platform based on edge computing node, establish mapping relationship between film thickness deviation and deformation deviation through multi-model prediction module, and predict real-time values of film thickness and deformation;

[0037] Introduce compensation value to compensate film thickness deviation and deformation deviation, and control module executes displacement or pressure control on each coating head according to the compensation value, and synchronously updates input information of prediction model;

[0038] Cloud platform adopts distributed algorithm to coordinate each control node, realizes film thickness distribution space consistency and numerical stability through asynchronous scheduling and prediction model fusion.

[0039] Further, the traditional control system is mainly designed for center film thickness uniformity, and lacks a special strategy for film thickness fluctuation in edge area, therefore, the present application proposes a scheme for film thickness deviation in edge area, introduces edge area weight coefficient and executes closed loop compensation control, which is crucial for improving film thickness consistency.

[0040] The film thickness data collected by the film thickness sensor is transmitted to the corresponding feedback unit for processing, and each feedback unit executes error closed loop compensation control according to the film thickness deviation in the region;

[0041] The feedback unit located in the edge area introduces an edge area weight coefficient for compensating the edge film thickness deviation, and the weight coefficient is adaptively calculated based on film speed and edge charge gradient.

[0042] The film thickness deviation is represented as follows:

[0043] ;

[0044] Then, the edge region weight coefficient w is introduced. e The deviation compensation formula is revised to:

[0045] ;

[0046] Among them, C i (t) represents the control compensation amount used to drive local adjustments of the coating head, by introducing an edge region weighting coefficient w. e It can "amplify" edge deviations, making the system pay more attention to edge anomalies and promptly suppress excessive film thickness deviations.

[0047] Charge tends to accumulate or dissipate in the edge region, generating an electric field gradient, which in turn changes the flow of the coating liquid and the film thickness. The charge gradient is defined as:

[0048] ;

[0049] Among them, Q edge (t) represents the surface charge of the edge region, Q center (t) represents the surface charge of the central region, and d represents the distance from the center to the edge.

[0050] Combining film velocity and edge charge gradient, w e It can be designed as:

[0051] ;

[0052] Where we0 is the basic weight, typically taken as ~1.0 to ~2.0, and k q denoted as the charge gradient gain coefficient, typically 0.5 to 1.5, k. v The coating speed gain coefficient is typically 0.05 to 0.2, |G e (t)| is the absolute value of the charge gradient, V coat (t) represents the coating speed.

[0053] If the edge charge gradient is large (e.g., exceeding 10), -6 C / m) means that a stronger electrostatic field is generated at the edge, requiring greater weight compensation. When the coating speed is fast, the liquid kinetic energy increases, and small disturbances can easily cause film thickness fluctuations, requiring an increase in w. e .

[0054] Furthermore, the microclimate control device includes a temperature and humidity sensor group, a temperature adjustment module, a humidification module, and a flexible sealing cover, and integrates a PID control module to control the temperature and humidity parameter settings as follows:

[0055] ;

[0056] Among them, Tset (t) is the target temperature setting value (unit: ℃) at the current time, T ref is the reference temperature, which can generally be set according to the solvent characteristics of the film material and the requirements of the workshop. For example, for water-based coating solvents, the appropriate temperature is generally T ref ~ 30℃, and for organic solvent type coating, the appropriate temperature can be set at ~ 30℃ ~ 35℃ to promote solvent evaporation and inhibit bubble generation, e(t) is the deviation value of the detected temperature from the reference temperature, k p is the proportional coefficient, which is used to quickly correct the temperature difference, and generally has a value range of ~ 0.5 ~ 3.0, k i is the integral coefficient, which is used to eliminate steady-state deviation, and generally has a value range of 0.01 ~ 0.5, k d is the differential coefficient, which is used to predict the deviation trend, and generally has a value of 0.001 ~ 0.1.

[0057] The formula calculates the output power required by the temperature adjustment module by collecting the deviation e(t) of the temperature sensor data from the reference value in real time, so as to realize rapid pulling back of the temperature to the set value. At the same time, humidity adjustment is also applicable, and the humidity control target is relative humidity RHset(t). The formula can be analogously expressed in actual production, and k p , k i , k d can be adjusted online according to the film thickness sensor feedback data to further match the requirements of different materials or coating rates.

[0058] Further, the electrostatic neutralization uses an ion wind rod or a plasma electrode array, and a control unit of the electrostatic neutralization periodically generates positive and negative ion pairs to form an electrostatic field control area.

[0059] Optionally, ion wind rods are arranged on both sides of the coating head, and bipolar ions are released in a high-voltage alternating current mode. A plasma electrode array is arranged at four corners of the coating platform to cover the electrostatic hot spots in the edge area. All devices have adjustable voltage / frequency characteristics and support dynamic control strategies.

[0060] The ion wind rod or the plasma electrode array can periodically release positive and negative ion pairs without contacting the film material to neutralize the static charges accumulated in the film material and its surrounding environment. An electrostatic inductor is arranged below or in the edge area of the coating platform to monitor the surface potential change in real time, output a charge disturbance matrix, and finally input the electrostatic disturbance data into a prediction model system to establish a joint mapping relationship with dependent variables such as film thickness and deformation.

[0061] The system obtains a group of disturbance charge time series data through the electrostatic inductor to form the following charge disturbance matrix:

[0062] ;

[0063] wherein e ij (t) represents the charge potential value of the i-th electrostatic sensor in the j-th direction at time t, m is the number of sensors, and n is the sensing dimension (generally 3: X-axis, Y-axis, and Z-axis).

[0064] The disturbance matrix is input into the electrostatic predictor model to obtain a predicted film thickness value:

[0065] ;

[0066] wherein, represents the predicted film thickness value of the i-th region, F(t) is the deformation time series data, and θ is a set of model weights, including weight matrices, bias terms, and attention mechanism parameters in the neural network.

[0067] Similarly, deformation data can also be predicted:

[0068] ;

[0069] wherein S(t) is the film thickness time series data, φ is a set of model weights, and f1 and f2 are film thickness prediction network functions and deformation prediction network functions, respectively.

[0070] Further, in order to improve the response speed and adjustment accuracy of the film thickness control system from the source, the deformation of the coating head needs to be introduced into the film thickness control formula for joint modeling compensation, and the influence on the film thickness needs to be corrected in real time, so as to realize double feedback regulation of edge disturbance and structural deformation.

[0071] The film thickness deviation is represented as follows:

[0072] ;

[0073] wherein S i (t) represents the actual film thickness value of the i-th region, represents the predicted film thickness value of the i-th region;

[0074] The deformation deviation is represented as follows:

[0075] ;

[0076] wherein F j (t) represents the actual deformation data of the j-th coating head, represents the predicted deformation data of the j-th coating head;

[0077] The film thickness deviation and the deformation deviation are jointly included in the calculation of the compensation control value, and the control formula is as follows:

[0078] ;

[0079] Wherein, β is the deformation influence weight factor, which plays a role in quantifying the sensitivity adjustment of deformation to the total control, and its value needs to balance the following two aspects:

[0080] Deformation dominance: if deformation plays a major role in film thickness fluctuation, β can be increased;

[0081] System robustness: too large β may cause excessive correction, causing system oscillation;

[0082] Preferably, β ∈ [0.1, 0.4], which is determined by the proportion of film thickness disturbed by deformation in actual process (empirical test or modeling), if the film head deformation is the main control factor, β can be set to > 0.3, otherwise a smaller value can be taken, C(t) is the total control compensation value.

[0083] In actual process, the small deformation of the coating head will cause the instantaneous change of the film thickness, especially in the edge area, therefore, the change trend of the deformation data Fj(t) has a certain time sequence correlation with the film thickness data Si(t), this formula can effectively predict and inhibit the local film thickness abnormality caused by mechanical structure change through the linear superposition compensation of deformation deviation ΔF j (t) and film thickness deviation ΔS i (t).

[0084] This control strategy models the film thickness control as a weighted deviation superposition compensation model, wherein:

[0085] characterizes the main driving term of local film thickness error;

[0086] is the deformation disturbance compensation term.

[0087] Further, a dynamic weight factor is introduced in the control formula, which is as follows:

[0088] ;

[0089] Wherein, α i is the dynamic film thickness weight factor of the i-th area, γ j is the dynamic deformation weight factor of the j-th coating head.

[0090] Wherein, by introducing dynamic weight factors α i and γ j , a differentiated response adjustment mechanism for film thickness error and deformation error is realized to enhance the system adaptability and edge compensation ability.

[0091] Due to different regional film thickness fluctuation intensity (such as frequent fluctuation in edge area), different deformation rate of different coating head units (affected by heat, mechanical vibration, etc.), therefore, introduce α i and γ j Make the control system can dynamically adjust the response strength according to the real-time data characteristics, ensure that the control strength of each area matches the disturbance characteristics, improve the control accuracy and energy efficiency ratio.

[0092] Further, the dynamic film thickness weight factor α i The formula is as follows:

[0093] ;

[0094] Where, δ i (t) is the time series standard deviation of the film thickness of the i-th region, the standard deviation of the time series data of the region reflects the fluctuation amplitude of the region, if the standard deviation is large, it means that the fluctuation of the region is more intense, and the direct compensation strength of the instantaneous deviation of the region needs to be reduced to avoid introducing high-frequency noise interference, for example, the time series standard deviation of the film thickness is ~0.01-0.1~μm, and α i The range is about ~0.9-0.99;

[0095] The dynamic deformation weight factor γ j The formula is as follows:

[0096] ;

[0097] Where, θ j (t) is the change rate of the deformation of the j-th coating head, which represents the instantaneous deformation rate of the mechanical mechanism, when the deformation changes too fast, the direct compensation may cause system response overshoot or oscillation, for example, the change rate of the deformation is ~0.001-0.02~mm / s, so γ j The range is about ~0.98-0.999.

[0098] Using the inverse form ~1 / (1+x)~can ensure that the value of α i and γ j is in the interval ~ (0, 1], when the standard deviation or change rate is close to zero, the weight tends to ~1, which means that the compensation is completely released; when the standard deviation or change rate increases, the weight automatically decreases, realizing adaptive adjustment.

[0099] In combination with the multi-layer film process scene, the typical application steps of dynamic weight calculation can be divided into the following links:

[0100] Real-time acquisition of multi-point regional film thickness data by optical interference type film thickness sensor, sampling frequency can reach ~1~kHz~above; deformation sensor synchronously collects coating head deformation time series data;

[0101] In each feedback cycle (e.g. 1 second), based on the last n sampling values (n≈1000), calculate the standard deviation of the film thickness δ i (t) of each area, and the deformation rate θ j (t) of each coating head unit, which can be obtained by difference between adjacent time points;

[0102] According to δ i (t) and θ j (t), update α i and γ j in real time, which is completed locally in the edge computing node, with a delay of less than 10 ms;

[0103] Calculate the total compensation value C(t) and output it to the coating head execution unit to adjust the displacement or apply a small amount of pressure to correct the local film thickness and ensure smooth transition of the film thickness in and out of the area;

[0104] Upload the latest film thickness and deformation data to the cloud prediction model to optimize the next cycle and to form an adaptive closed loop.

[0105] Further, Kalman filtering is used to process the film thickness data:

[0106] ;

[0107] Wherein, is the film thickness state estimation value at time k in Kalman filtering, is the film thickness estimation value after filtering at the last time, z k is the actual observed film thickness data at time k, K k is the Kalman filtering gain, which is determined by the system and observation covariance, and H is the observation matrix.

[0108] In the present application, the center and edge film thickness data collected by the film thickness sensor (optical interference type or laser reflection type) will still be affected by environmental disturbances, electronic noise, equipment jitter, etc. under high-precision detection conditions, resulting in high-frequency errors, nonlinear disturbances and low-frequency drift. Direct use of such data will interfere with compensation control.

[0109] Therefore, Kalman filtering is used to fuse the film thickness measurement value z k and the predicted value of each sensing point (or sub-area), and output the filtering value closer to the "true film thickness state", thereby improving the control accuracy and stability of the compensation algorithm.

[0110] Further, in the manufacturing process of high-precision multi-layer film materials such as electromagnetic shielding films, the consistency of film thickness control directly determines the electromagnetic shielding effectiveness, mechanical properties and adhesion stability of the final product. However, the consistency of film thickness distribution faces two significant challenges:

[0111] Regional differences: there is often a systematic deviation in film thickness between the center and the edge region;

[0112] Coupling disturbance: coating head deformation, environmental disturbance (such as charge accumulation, humidity change) often affects multiple regions at the same time, leading to regional linkage diffusion of film thickness error.

[0113] Existing control methods are mostly based on local feedback control, lacking global collaborative optimization mechanism for inter-regional influence.

[0114] Therefore, a regional adjustment optimization objective function is introduced, and a mathematical relationship between film thickness control and compensation cost is established to form a distributed global optimal control strategy. The cloud platform uses a distributed algorithm to execute regional adjustment optimization according to the following objective function:

[0115] ;

[0116] Where C ij (t) is the control compensation amount of the i-th region to the j-th region, with a value range of [-0.5~μm,+0.5~μm], corresponding to sub-micron level compensation ability, α ij is the control cost coefficient, determined according to the energy consumption model and response time of the execution element, the greater the cost, the slower the response, usually with a value range of [0.01,1], β ij is the coupling cost coefficient, related to the sensitivity of the film material and the response speed of the structure, the larger the value, the more severe the adjustment is suppressed, usually with a value range of [0.1,5], Penalty(C ij (t)) is the penalty function of the control amount, used to avoid system oscillation caused by excessive control input.

[0117] represents the cost of i→j control action, α ij is larger, representing higher control cost, which reflects the goal of minimizing energy consumption or execution cost;

[0118] represents the influence of the current control input on the stability of the system, where Penalty(·) is a designed penalty function (such as a quadratic function or an exponential function, which can be selected as (C ij ) 2 ), used to suppress system instability caused by excessive control action.

[0119] Further, based on the film thickness data acquisition and control mechanism, the system control process is constructed as a closed-loop feedback architecture, through real-time data updating and iterative operation of the prediction model, realizing the dynamic consistency of multi-layer film thickness and deformation control;

[0120] The control process runs in a cyclic feedback mode, achieving multiple error corrections, so that the multi-layer structure film thickness remains consistent and accurate in the entire preparation process;

[0121] The control process is divided into three layers:

[0122] Local feedback layer: for each sensing area, the film thickness deviation and deformation deviation closed-loop control is executed, which specifically includes calculating the film thickness deviation ΔS i of each sub-area separately, combining the deformation data ΔF j in the area, adjusting the pressure or displacement of the corresponding coating head through the control module, and compensating the local film thickness in real time;

[0123] Regional coordination layer: fusion of adjacent area deviation and prediction trend, implementation of boundary transition control, measures include: fusion of boundary transition zone film thickness and deformation trend, use of dynamic weight factor α i , γ j balance control strength, realize smooth transition compensation of edge effect, eliminate the "film thickness step" phenomenon;

[0124] Global coordination layer: the cloud platform schedules control tasks according to the following global error function:

[0125] ;

[0126] Where w i is the regional priority weight, w i ∈[0.1,1.0], the priority control key functional area, N is the total number of regions, determined according to the film width and the number of sensors, typical value N=10~100, E global is used to measure the overall film thickness deviation, used for global scheduling.

[0127] Although the film thickness of each region is locally compensated, it needs to be globally optimized and coordinated to prevent the overall deviation from being amplified due to mutual influence between regions, w i The parameter reflects the strengthened control of key functional areas (such as conductive layer, shielding layer) to ensure the film function performance.

[0128] Specific working steps

[0129] Real-time data acquisition: film thickness sensor, deformation detection equipment synchronously collect S i (t), F j (t data;

[0130] Prediction model calculation: predict film thickness , predict deformation , error calculation ΔS i , ΔF j ;

[0131] Multi-level feedback control:

[0132] Local layer: Perform sub-area film thickness and deformation compensation;

[0133] Collaborative layer: Adjacent area trend smoothing adjustment;

[0134] Global layer: Calculate the overall deviation according to E global , cloud platform issued coordination instructions;

[0135] Dynamic weight adaptive update: w i According to the real-time deviation level adjustment, high deviation area increases the weight, strengthens the control strength, low deviation area reduces the intervention, and improves the system stability.

[0136] Loop iteration: The whole process continues to circulate, forming a dynamic error correction closed loop, ensuring the thickness uniformity and stability in the process of multi-layer film superposition.

[0137] Finally, it should be noted that: the above only for the preferred embodiments of the present application, and not for the purpose of limiting the present application, although the foregoing embodiments of the present application are described in detail, for those skilled in the art, it still can be modified, or part of the technical features of the equivalent replacement. Any modification, equivalent replacement, improvement, etc. within the spirit and principles of the present application, shall be included within the scope of protection of the present application.

Claims

1. A method for intelligent control of film thickness in the preparation of electromagnetic shielding films, characterized in that, Includes the following steps: Multiple film thickness sensors are set up and arranged in the central and edge areas at the top of the coating platform to collect real-time two-dimensional coordinate film thickness time series data of the central and edge areas. Electrostatic neutralization is applied to the coating area, and the electrostatic neutralization process is controlled by a charge perturbation matrix obtained by an electrostatic sensor to control the accumulation of surface charge. A microclimate control device is installed around the coating platform to maintain the set temperature and relative humidity through a PID control algorithm, while shielding it from external airflow disturbances. A non-contact deformation detection system is set up to collect deformation time sequence data for each coating head unit; The film thickness time series data and deformation time series data are uploaded to a cloud platform based on edge computing nodes. The mapping relationship between film thickness deviation and deformation deviation is established through a multi-model prediction module to predict the real-time values ​​of film thickness and deformation. Compensation values ​​are introduced to compensate for film thickness deviation and deformation deviation. The control module performs displacement or pressure control on each coating head based on the compensation values ​​and updates the input information of the prediction model simultaneously. The cloud platform uses a distributed algorithm to coordinate the various control nodes and integrates asynchronous scheduling with the prediction model.

2. The method for intelligent thickness control in the preparation of an electromagnetic shielding film according to claim 1, characterized in that, The film thickness data collected by the film thickness sensor is transmitted to the corresponding feedback unit for processing. Each feedback unit performs error closed-loop compensation control based on the film thickness deviation in its region. The feedback unit located in the edge region introduces an edge region weight coefficient to compensate for edge film thickness deviation.

3. The intelligent thickness control method for electromagnetic shielding film preparation according to claim 1, characterized in that, The electrostatic neutralization uses an ion wind bar or a plasma electrode array, whose control unit periodically generates positive and negative ion pairs to form an electrostatic field control zone.

4. The intelligent thickness control method for electromagnetic shielding film preparation according to claim 1, characterized in that, The microclimate control device includes a temperature and humidity sensor group, a temperature adjustment module, a humidification module, and a flexible sealing cover, and integrates a PID control module to control the temperature and humidity parameter settings as follows: ； Among them, T set (t) represents the target temperature setpoint at the current moment, T ref The reference temperature is denoted as k, e(t) is the deviation between the detected temperature and the reference temperature, and k is the reference temperature. p k is the proportionality coefficient. i k is the integral coefficient. d is the differential coefficient.

5. The intelligent thickness control method for electromagnetic shielding film preparation according to claim 1, characterized in that, The film thickness deviation is expressed as follows: ； Among them, S i (t) represents the actual film thickness value of the i-th region. This is represented as the predicted film thickness value for the i-th region; The deformation deviation is expressed as follows: ； Among them, F j (t) represents the actual deformation data of the j-th coating head. This is represented as the predicted deformation data for the j-th coating head; The film thickness deviation and deformation deviation are both incorporated into the calculation of the compensation control value. The control formula is as follows: ； Where β is the deformation influence weighting factor, and C(t) is the overall control compensation value.

6. The intelligent thickness control method for electromagnetic shielding film preparation according to claim 5, characterized in that, A dynamic weighting factor is introduced into the control formula, as shown in the following formula: ； Where, α i γ is the dynamic film thickness weighting factor for the i-th region. j is the dynamic deformation weighting factor for the j-th coating head.

7. The method for intelligent thickness control in the preparation of an electromagnetic shielding film according to claim 6, characterized in that, The dynamic film thickness weighting factor α i It is expressed by the following formula: ； Where, δ i (t) represents the time series standard deviation of the film thickness in the i-th region; The dynamic deformation weighting factor γ j It is expressed by the following formula: ； Where, θ j (t) represents the rate of change of deformation of the j-th coating head.

8. The intelligent thickness control method for electromagnetic shielding film preparation according to claim 1, characterized in that, Kalman filtering was applied to the film thickness data. ； in, z is the membrane thickness state estimate at time k in the Kalman filter. k Let K be the actual observed film thickness data at time k. k The Kalman filter gain is determined by both the system and observation covariance, and H is the observation matrix.

9. The intelligent thickness control method for electromagnetic shielding film preparation according to claim 1, characterized in that, The cloud platform employs a distributed algorithm to perform regional adjustment optimization according to the following objective function: ； Among them, C ij (t) represents the control compensation amount of region i on region j, α ij To control the cost coefficient, β ij Penalty(C) is the coupling cost coefficient. ij (t) is the penalty function for the control quantity.

10. A method for intelligent control of film thickness in the preparation of an electromagnetic shielding film according to any one of claims 1-9, characterized in that, Based on the film thickness data acquisition and control mechanism, the system control process is constructed as a closed-loop feedback architecture. Through real-time data updates and iterative operation of the prediction model, dynamic consistency between multilayer film thickness and deformation control is achieved.

Images