Thin film transverse thickness uniformity control method based on multi-point temperature field dynamic decoupling
By establishing a multi-point temperature field coupling model and a dynamic decoupling control matrix, the problems of low film thickness adjustment efficiency and poor accuracy caused by thermal coupling of the mold head were solved, and high-precision, fast-response uniform control of the film's transverse thickness was achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SI CHUAN XIN YUAN YI SHI PIN KE JI YOU XIAN GONG SI
- Filing Date
- 2026-04-16
- Publication Date
- 2026-05-15
AI Technical Summary
In the prior art, the thermal coupling phenomenon in the heating area of the mold head leads to low efficiency in adjusting the transverse thickness of the film, poor control accuracy, and inability to adapt to changes in working conditions. Traditional PID control methods have slow response speeds and are difficult to achieve uniform control of the transverse thickness of the film.
By establishing a multi-point temperature field coupling model and calculating the dynamic decoupling control matrix, the independent adjustment of each heating zone is achieved by utilizing the synergistic effect of the PID controller and the decoupling matrix, eliminating the influence of thermal coupling, improving the adjustment response speed and control accuracy, and adapting to changes in operating conditions.
It achieves high-precision, fast-response uniform control of the film's lateral thickness, ensuring the stability and adaptability of the control system and avoiding repeated adjustments and system oscillations caused by thermal coupling.
Smart Images
Figure CN122043921A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of intelligent control technology, specifically to a method for controlling the lateral thickness uniformity of thin films based on dynamic decoupling of multi-point temperature fields. Background Technology
[0002] Plastic films play an important role in various industries such as packaging, electronics, and agriculture due to their excellent performance and wide range of applications. In the production process of plastic films, transverse thickness uniformity is one of the core indicators for measuring film quality, which directly affects the subsequent processing performance and use effect of the film.
[0003] Currently, plastic film production generally adopts extrusion molding process, in which molten plastic is evenly distributed on cooling rollers through the extrusion die of the extruder to form a film; and the die usually has multiple heating zones along the transverse direction, each zone is equipped with an independent heating element and temperature sensor, and the melt temperature is controlled by adjusting the heating power of each zone, which in turn affects the melt flow rate and the transverse thickness distribution of the final film.
[0004] However, due to the thermal conduction effect of the die head metal body, there is an inherent thermal coupling phenomenon between adjacent heating areas. Thus, when the power of a certain heating area is adjusted, the heat will not only change the temperature of this area, but also affect the temperature of adjacent areas through thermal conduction, resulting in changes in the thickness of multiple areas at the same time.
[0005] Traditionally, a single-point PID control method is used to solve this problem. However, this method has several drawbacks: adjusting the heating power of one area can cause reverse fluctuations in the thickness of adjacent areas, requiring repeated adjustments to reach a stable state, resulting in long adjustment times and slow response speeds; due to the inability to effectively compensate for coupling effects, thickness deviations are difficult to eliminate precisely, and lateral thickness uniformity is difficult to achieve an ideal level; when production speed, raw material batches, or ambient temperature change, the coupling characteristics change accordingly, and the controller with fixed parameters cannot adapt to these changes, further reducing the control effect.
[0006] Therefore, we propose a method that can adapt to changes in operating conditions and precisely control the lateral thickness uniformity of the thin film. Summary of the Invention
[0007] The purpose of this invention is to provide a method for controlling the transverse thickness uniformity of thin films based on dynamic decoupling of multi-point temperature fields, which solves the problems of low thickness adjustment efficiency, poor control accuracy, and inability to adapt to changes in working conditions caused by thermal coupling in the heating area of the die head in the prior art.
[0008] This invention is achieved through the following technical solution:
[0009] A method for controlling the transverse thickness uniformity of thin films based on dynamic decoupling of multi-point temperature fields, specifically including: Acquire real-time temperature data and corresponding film lateral thickness distribution data for multiple heating zones distributed laterally along the extrusion die. Based on real-time temperature data and thin film lateral thickness distribution data, a multi-point temperature field coupling model is established to describe the thermal coupling effect between multiple heating regions. Based on the multi-point temperature field coupling model, a dynamic decoupling control matrix is calculated to eliminate the thermal coupling effect between heating regions. Obtain the target thickness distribution curve and the real-time thickness distribution curve of the film in the lateral direction, compare the two distribution curves, and obtain the thickness deviation value of the film at each lateral position. Multiple thickness deviation values are input into the PID controller to calculate the initial temperature adjustment for each heating zone. The initial temperature regulation value is decoupled and transformed using a dynamic decoupling control matrix to obtain the final decoupled temperature regulation value. Based on the final temperature adjustment, the heating power of each heating zone is adjusted synchronously to achieve closed-loop uniformity control of the film's lateral thickness.
[0010] Furthermore, based on real-time temperature data and thin film lateral thickness distribution data, a multi-point temperature field coupling model is established to describe the thermal coupling effect between multiple heating regions. The specific steps are as follows: A preset excitation signal is applied to each heating zone, and temperature response data of each heating zone is collected. Based on temperature response data, the coupling transfer function between each heating region is calculated, and the transfer function matrix of the multi-input multi-output system is constructed. The expression of this matrix is as follows:
[0011] In the formula, The total number of heating areas, diagonal elements Indicates the first The dynamic characteristics of each heating zone itself, non-diagonal elements Description of the The heating zone for the first Thermal coupling strength of each heating zone.
[0012] Furthermore, the multiple input multiple output system is a controlled object that takes the heating power of multiple heating zones laterally distributed on the extrusion die as input and the measured value of the temperature sensor in the corresponding zone as output.
[0013] Furthermore, based on the multi-point temperature field coupling model, a dynamic decoupling control matrix is calculated to eliminate the thermal coupling effects between different heating regions. The specific steps are as follows: Based on the transfer function matrix, the dynamic decoupling control matrix is calculated using the feedforward compensation decoupling method. Furthermore, the principle for determining the dynamic decoupling control matrix is that the product matrix of the dynamic decoupling control matrix and the transfer function matrix is a diagonal matrix or a diagonally dominant matrix.
[0014] Furthermore, the feedforward compensation decoupling method can be a direct feedforward compensation method, a diagonal matrix decoupling method, or an identity matrix decoupling method.
[0015] Furthermore, the acquisition of the target distribution curve of the film's lateral thickness includes one or more combinations of a theoretical target curve calculated based on a theoretical model, a measured target curve based on actual measurements of a benchmark sample, a tolerance-based target curve based on a process window, a dynamic target curve based on adaptive learning, and a compensation target curve based on defect pattern recognition.
[0016] Furthermore, the target thickness distribution curve and real-time thickness distribution curve of the film's lateral thickness are obtained, and the two distribution curves are compared to obtain the thickness deviation value of the film at each lateral position. The calculation formula is as follows:
[0017] In the formula, For the first Target thickness value at each location For the first Real-time thickness value at each location.
[0018] Furthermore, the multiple thickness deviation values are input to the PID controller to calculate the initial temperature adjustment for each heating zone. The calculation formula is as follows:
[0019] In the formula, This is the proportionality coefficient. The integral time constant is... For integration variables, is the differential time constant.
[0020] Furthermore, the specific steps for using a dynamic decoupling control matrix to decouple the initial temperature regulation amount to obtain the decoupled final temperature regulation amount are as follows: The initial temperature adjustment of each heating zone is expressed in vector form, and the calculation formula is as follows:
[0021] In the formula, This represents the total number of heating zones. Using dynamic decoupling control matrix The initial adjustment vector is decoupled and transformed to obtain the final temperature adjustment vector. The calculation formula is:
[0022] In the formula, This is a dynamic decoupling control matrix.
[0023] In the time domain, the decoupling transformation is represented as a convolution to obtain the final temperature adjustment amount. The calculation formula is:
[0024] In the formula, For dynamic decoupling control matrix The Middle Line number The time-domain impulse response function corresponding to the column element, For the current moment, It is the integral variable.
[0025] Furthermore, the specific steps for synchronously adjusting the heating power of each heating zone based on the final temperature adjustment amount are as follows: The first Final temperature adjustment of each heating zone The formula for converting this to a heating power adjustment value is as follows:
[0026] in, For the first Power-temperature conversion coefficient for each heating zone; No. The actual output power of each heating zone is:
[0027] In the formula, For the first The reference power for each heating zone.
[0028] The technical solution of the present invention has at least the following advantages and beneficial effects: This invention discloses a method for controlling the transverse thickness uniformity of thin films based on dynamic decoupling of multi-point temperature fields. By establishing a multi-point temperature field coupling model and calculating a dynamic decoupling control matrix, the initial temperature adjustment output of the PID controller is decoupled and transformed, realizing independent adjustment and precise control of each heating region. Furthermore, it eliminates thermal coupling interference between adjacent regions, thereby improving the control accuracy and adjustment response speed of the transverse thickness uniformity of the thin film. At the same time, it enables the control system to adapt to changes in operating conditions, ensuring the long-term stability and reliability of thickness control.
[0029] In addition, by constructing the transfer function matrix of the multi-input multi-output system, a precise quantitative description of the coupling characteristics of the temperature field of the die head is achieved, providing an accurate mathematical model basis for subsequent decoupling control. The use of the feedforward compensation decoupling method to calculate the dynamic decoupling control matrix ensures the theoretical rigor and engineering feasibility of the control system. Furthermore, by using multiple methods to obtain target distribution curves, the control system can flexibly adapt to the thickness control requirements under different production scenarios. Furthermore, by utilizing the synergistic effect of the PID controller and the decoupling matrix, the simplicity and reliability of traditional PID control are maintained while overcoming its inherent weakness in handling coupled systems. Through vector-based decoupling transformation and time-domain convolution operations, synchronous and coordinated control of multi-region adjustment quantities is achieved, avoiding repeated adjustments and system oscillations caused by coupling. Attached Figure Description
[0030] Figure 1 This is a schematic diagram of a method for controlling the lateral thickness uniformity of thin films based on dynamic decoupling of multi-point temperature fields according to the present invention. Figure 2 This is a schematic diagram of a thin film lateral thickness uniformity control system based on dynamic decoupling of multi-point temperature fields according to the present invention. Figure 3 This is a schematic diagram of an electronic device structure according to the present invention. Detailed Implementation
[0031] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.
[0032] Example 1 like Figure 1 The method for controlling the transverse thickness uniformity of thin films based on dynamic decoupling of multi-point temperature fields, as shown, specifically includes: Acquire real-time temperature data and corresponding film lateral thickness distribution data for multiple heating zones distributed laterally along the extrusion die. At the process level, in the extrusion molding of plastic film, the temperature of the heating area of the extrusion die directly determines the viscosity and flow rate of the melt. Moreover, the temperature distribution in each lateral region of the extrusion die is strongly correlated with the lateral thickness distribution of the film. That is, the higher the temperature, the larger the melt flow and the thicker the film, while the lower the temperature, the smaller the melt flow and the thinner the film. Therefore, collecting real-time temperature data and lateral thickness distribution data is the prerequisite for establishing a "temperature-thickness" control relationship. In addition, the real-time temperature data of the transverse heating area of the extrusion die is obtained by arranging temperature sensing elements in the preset heating area of the die to achieve accurate and real-time temperature acquisition. Specifically: Temperature sensors (commonly thermocouples or platinum resistance thermometers) are embedded in each independent heating zone along the transverse direction of the extrusion die. The sensor probes are in direct contact with the die's metal body, ensuring accurate detection of the actual operating temperature of the heating zone. The sensors are positioned adjacent to and matched with the heating elements (such as heating rods or heating coils) to avoid blind spots. All temperature sensors are connected to a PLC control system, which converts the analog temperature signals from the sensors into digital signals via analog-to-digital conversion (A / D). Real-time continuous acquisition is achieved according to a preset acquisition frequency (e.g., 100ms / time, 500ms / time, adjustable according to production conditions). The temperature data of each heating zone is stored according to the zone number, forming a real-time dataset of the temperature distribution of the transverse heating zone of the die. During the acquisition process, the temperature data is calibrated in real time to eliminate errors caused by ambient temperature and sensor drift, ensuring the accuracy of the temperature data for each heating zone and providing a reliable basis for subsequent thermal coupling characteristic analysis.
[0033] The lateral thickness distribution data of the thin film is obtained by scanning the entire lateral width of the film using a thickness detection device. The detected lateral positions correspond one-to-one with the heating area of the die head, achieving thickness data matching from a specific heating area of the die head to a specific lateral position of the thin film. In industry, laser thickness gauges or... X-ray thickness gauges (all non-contact type, avoiding scratching the film and not affecting continuous production), specifically: A non-contact thickness measurement device is installed at a critical film forming position, behind the melt extrusion end of the extrusion die and in front of the cooling roller (at this point, the film has been initially formed, its thickness is stable, and it has not completely cooled, thus accurately reflecting the influence of the die temperature on the thickness). The scanning direction of the measurement device is consistent with the transverse (width) direction of the film, and the scanning range covers the entire width of the film. Simultaneously, the measurement device performs equidistant fixed-point scanning of the film transversely according to the division spacing of the transverse heating area of the die. The film is divided into horizontal heating zones, corresponding to the horizontal division of the thin film. Each detection point is linked to a specific location, with the heating area number of the die head corresponding to the transverse detection point number of the film (e.g., the first heating area of the die head corresponds to the first transverse detection point of the film, the second heating area corresponds to the second detection point, and so on). The thickness detection equipment and the die head temperature acquisition system are synchronized in time. At the same time as collecting temperature data from each heating area of the die head, the thickness data of the corresponding transverse points of the film is collected. The thickness data is also transmitted to the PLC via industrial communication and stored according to the detection point number, forming a real-time dataset of the transverse thickness distribution of the film. This dataset and the temperature dataset can be accurately matched using "area / point number + acquisition time". It is important to note that the collected thickness data is finally filtered and denoised to eliminate detection errors caused by film jitter and vibrations in the production environment, ensuring the authenticity of the thickness data.
[0034] Furthermore, since the collected lateral thickness distribution data of the film corresponds one-to-one with the lateral temperature distribution data of the die head, that is, the temperature data of each heating area corresponds to the thickness data of a certain position of the film in the lateral direction, this distributed mapping relationship can realize the accurate traceability of the film thickness deviation. Thus, when a thickness deviation occurs at a certain lateral position of the film, the corresponding die head heating area can be directly located, and it can be determined whether the deviation is caused by the abnormal temperature of the area itself or by thermal coupling interference of adjacent areas. Therefore, it provides a basis for deviation location for the subsequent precise adjustment of the PID controller and the targeted compensation of the decoupling matrix. Furthermore, in actual production, the operating conditions of film production (production speed, raw material batch, ambient temperature) will constantly change, and the changes in operating conditions will directly change the thermal coupling characteristics of the die head heating area and the temperature-thickness mapping relationship. This step collects multi-point temperature and thickness data in real time and continuously, which can capture the parameter changes caused by changes in operating conditions in real time. The subsequent coupling model and decoupling matrix can be dynamically adjusted based on the real-time updated data, allowing the control system to adapt to changes in operating conditions.
[0035] Based on real-time temperature data and thin film lateral thickness distribution data, a multi-point temperature field coupling model is established to describe the thermal coupling effect between multiple heating regions. The specific steps are as follows: A preset excitation signal is applied to each heating zone, and temperature response data of each heating zone is collected. Since the extrusion die is made of metal, heat conduction is an inherent physical characteristic. Therefore, the thermal coupling between adjacent heating areas is a dynamic and regular heat conduction process. For example, when the power of a certain area is increased, the temperature will first rise on its own and then gradually affect the adjacent areas through heat conduction. Moreover, the temperature transfer has the characteristics of delay and attenuation. If only temperature data under natural production conditions is collected, it is impossible to distinguish whether the temperature change is due to self-heating or the temperature change caused by coupling. Therefore, by applying an excitation signal to each area separately, a single variable can be artificially created, thereby clearly capturing the dynamic characteristics of the temperature change in that area, such as the heating rate, steady-state value, etc., as well as the thermal coupling transfer characteristics to all other areas. The preset excitation signal is a known heating power adjustment signal, such as a step signal, a sine signal, or a pulse signal. The collected temperature response data is the system's output signal. By using single-area excitation and full-area acquisition, the dynamic correlation between the input of each heating area and the output of all areas can be accurately captured. Therefore, this step allows for the quantitative calculation of: ① the dynamic characteristics of each heating region's own heating / cooling, such as the steady-state gain and time constant under step excitation; ② the thermal coupling strength of each heating region to any other arbitrary region, such as the excitation region. Afterwards, the region Temperature variation range and region ③ The ratio of the amplitude of its own temperature change; ④ The dynamic transmission law of thermal coupling, such as the delay time and rise rate of temperature change in adjacent regions after excitation.
[0036] The specific calculation steps are as follows: Assume the extrusion die has transverse direction Each heating zone has an independent heating area, numbered as follows: ; When applying excitation to a single region, the reference heating power remains constant for all other regions to eliminate additional interference. Furthermore, the preset excitation signal is assumed to be a step power excitation. When an excitation is applied to a heating region, the step change in its heating power is the input excitation of the input-output system, and the formula is:
[0037] In the formula, For the first The reference heating power for each heating zone For the first The step power increment of each heating zone can be specifically set according to the thermal characteristics of the mold head. The power level is set at 5%~20% to avoid sudden power fluctuations that could cause melt temperature runaway; and the excitation termination condition is: from Start data acquisition until the temperature of all heated areas reaches a steady state (temperature change rate < 0.1℃ / s), and record the steady-state time as _____. .
[0038] For the After applying excitation to each heating zone, the real-time temperature of all heating zones is simultaneously collected to form... Group 1 temperature response time series; 1st Each heating zone ( , and when The time is the incentive zone. (When the coupling region is in) Temperature response data at any time The formula is:
[0039] The acquired temperatures need to be preprocessed using moving average filtering and steady-state correction. After applying excitation sequentially to all heating regions, the temperature response data of all regions are integrated into a temperature response acquisition matrix. In the matrix, rows represent excitation regions, columns represent response regions, and elements are the temperature response time series of the corresponding regions. The formula is:
[0040] In the formula, For the first When an incentive is applied to the region, the first... The measured temperature of each region at steady state; For the first Before applying incentives to the first region, the first The steady-state reference temperature of each region; For the first When an incentive is applied to the region, the first... The temperature response increment of each region at steady state.
[0041] Based on temperature response data, the coupling transfer function between each heating region is calculated, and the transfer function matrix of the multi-input multi-output system is constructed. The expression of this matrix is as follows:
[0042] In the formula, The total number of heating areas, diagonal elements Indicates the first The dynamic characteristics of each heating zone itself, non-diagonal elements Description of the The heating zone for the first The thermal coupling strength of each heating zone; Among them, the th in the transfer function matrix Line number Column elements The specific calculation formula is as follows:
[0043] In the formula, For complex frequencies, For steady-state gain, reflecting the first The incentive region for the first The magnitude of temperature influence in each response region This is the inertial time, which is the time it takes for the temperature response increment to rise to 63.2% of the steady-state value. The delay time represents the lag time for heat to be conducted through the metal body of the mold head, that is, the moment when the temperature response increment begins to change significantly after the excitation is applied, which is greater than 2% of the temperature response increment.
[0044]
[0045]
[0046] in, and All solutions were obtained using linear interpolation.
[0047] The constructed transfer coefficient matrix can transform the inherent, dynamic thermal coupling effect of the extrusion die metal body into a set of verifiable complex frequency domain parameters based on industrial measurements. This allows the design of the subsequent dynamic decoupling control matrix to break free from general assumptions and specifically counteract cross-regional thermal coupling interference in actual production. Furthermore, its first-order inertial + pure delay classical model has a simple mathematical structure and can be directly implemented in a PLC system, balancing theoretical rigor and industrial applicability. It also accurately quantifies the dynamic mapping relationship between heating power and die temperature, which can be combined with previously collected temperature-thickness data to achieve precise mathematical conversion of "power adjustment → temperature change → thickness correction." This transforms the compensation for film lateral thickness deviation from fuzzy empirical adjustment to precise quantitative adjustment, effectively avoiding over-adjustment or under-adjustment and improving thickness control accuracy.
[0048] In addition, since changes in production conditions (production speed, raw material batch, ambient temperature) can alter the thermal coupling characteristics of the die head, the transfer function matrix can be dynamically updated based on new excitation-response measured data. The decoupling control matrix is then recalculated based on the updated matrix, ensuring that the entire thickness control system always conforms to the current operating conditions and guaranteeing the long-term stability and reliability of film thickness control.
[0049] In addition, the multiple input multiple output system is a controlled object that takes the heating power of multiple heating zones laterally distributed in the extrusion die as input and the actual measured value of the temperature sensor in the corresponding zone as output.
[0050] Based on the multi-point temperature field coupling model, a dynamic decoupling control matrix for eliminating the thermal coupling effects between different heating regions is calculated. The specific steps are as follows: Based on the transfer function matrix, the dynamic decoupling control matrix is calculated using the feedforward compensation decoupling method. Furthermore, the principle for determining the dynamic decoupling control matrix is that the product matrix of the dynamic decoupling control matrix and the transfer function matrix is a diagonal matrix or a diagonally dominant matrix. This anchors the core control requirement of eliminating cross-regional thermal coupling interference. Moreover, by using the output characteristics of the compensation link, the coupling characteristics of the system itself are offset in advance, so that the overall system after compensation eliminates the cross interference between input and output and there is no longer any cross-regional thermal coupling interference. Furthermore, if the product is a diagonal matrix, the input and output of the compensated system are strictly mapped one-to-one, meaning that adjusting the power of a certain heating zone will only change the temperature of that zone, completely eliminating cross-region thermal coupling interference. If the product is a diagonally dominant matrix, the coupling interference of the system is weakened to a negligible level, and the characteristics of the main control channel are much stronger than those of the cross-coupled channel, achieving approximately independent control.
[0051] Furthermore, the dynamic decoupling control matrix is calculated using the feedforward compensation decoupling method. After multiplying with the measured transfer coefficient matrix, a diagonal or diagonally dominant matrix is formed, which directly cancels the inherent thermal conduction coupling effect of the die head metal body. This transforms the multi-heating zone temperature control system of the extrusion die head from a multi-input multi-output system that originally interfered with each other into multiple approximately independent SISO (Single-Input Single-Output) systems. That is, when the heating power of a certain zone is adjusted, only the temperature of that zone will be changed, and it will hardly affect other zones, thus achieving interference-free independent and precise adjustment of each heating zone.
[0052] Among them, the feedforward compensation decoupling method can be the direct feedforward compensation method, the diagonal matrix decoupling method, or the identity matrix decoupling method. For the high-precision requirements of high-end thin film production, the identity matrix decoupling method can be selected to achieve strict one-to-one decoupling; for conventional thin film production, the direct feedforward compensation method can be selected to achieve a decoupling effect that meets the process requirements with a smaller amount of computation. The calculation formula for the feedforward compensation decoupling method is as follows:
[0053] The calculation formula for the diagonal matrix decoupling method is as follows:
[0054] In the formula, It is the inverse of the transfer function matrix. For the reason A diagonal matrix composed of diagonal elements; The calculation formula for the identity matrix decoupling method is as follows:
[0055] In the formula, It is an identity matrix.
[0056] The target thickness distribution curve and real-time thickness distribution curve of the film are obtained. By comparing the two distribution curves, the thickness deviation value of each lateral position of the film is obtained. Subsequently, by combining PID control, the closed-loop uniformity control of the film's lateral thickness can be achieved. The principle of closed-loop automatic control is to detect the actual value of the controlled object, compare it with the target value to obtain the deviation, and then output control commands based on the deviation, so as to eliminate the deviation and make the actual value match the target value. In this process, the lateral thickness of the thin film is the controlled object, the target thickness distribution curve is the control benchmark, the real-time distribution curve is the actual detected value, and the deviation between the two is the core control signal: the initial temperature adjustment of the subsequent PID controller, the decoupling transformation of the dynamic decoupling control matrix, and the synchronous adjustment of the power of each heating zone are all based solely on this deviation value. This step constructs a closed-loop sensing end for thickness control through the logic of "target-real-time-deviation," providing precise deviation guidance for all subsequent control actions, which is the fundamental premise for the realization of closed-loop control. Furthermore, this step abandons the method of single-point thickness comparison, using the lateral distribution curve as the carrier of the target value and the actual value, including the thickness of each position of the thin film in the lateral direction within the comparison range, and calculating the precise thickness deviation value of each lateral position. This allows the calculation of thickness deviation to accurately reflect the actual quality problem of the thin film, avoiding the process defect of single-point compliance but uneven overall distribution. It achieves full-dimensional and refined perception of the lateral thickness of the thin film, and can accurately capture the thickness deviation at any position in the lateral direction (such as thicker at the edge, thinner in the middle, local abrupt changes, etc.), transforming the perception of thickness deviation from single-point sampling to full-domain precise detection.
[0057] The target distribution curves include one or more combinations of theoretical target curves calculated based on theoretical models, measured target curves based on benchmark samples, tolerance target curves based on process windows, dynamic target curves based on adaptive learning, and compensation target curves based on defect pattern recognition. The design of these five target distribution curves is to meet the differentiated scenarios of different raw materials, product specifications, process requirements, and defect compensation needs in the film lamination industry, so that the target curve is no longer a single fixed value, but a dynamic control benchmark that can be flexibly adjusted according to actual production needs. Specifically, the theoretical target curve is calculated based on theoretical models of melt rheology and extrusion molding, adapting to the trial production stage of new formulations and products, and providing a theoretical benchmark for thickness control; the measured target curve is based on the actual thickness detection of benchmark samples, adapting to mass-produced and finalized products, ensuring that the product thickness is consistent with the benchmark product, and improving batch stability; the tolerance-based target curve is based on setting the allowable thickness deviation range based on the process window, adapting to conventional industrial-grade film production, and balancing control accuracy and production efficiency; the dynamic target curve is based on adaptive learning and real-time optimization, adapting to continuous production with constantly changing operating conditions, so that the target curve fits the actual production capacity; the compensation target curve is based on defect pattern recognition design, adapting to production scenarios with specific thickness defects, and specifically compensating for lateral local thickness deviations.
[0058] Therefore, when new products need to be trial-produced, a combination of theoretical target curves and tolerance-based target curves can be used, which relies on theoretical benchmarks and leaves room for process adjustments during trial production. When mass production of finalized products is required, a combination of measured target curves and dynamic target curves can be used, which ensures consistency with benchmark products and allows for adaptive learning to adapt to changes in operating conditions. When local thickness defects occur during production, compensation target curves can be superimposed to specifically compensate for thickness deviations at the defect locations.
[0059] Compared to the single, fixed thickness target value in traditional control, this design allows the thickness control benchmark to be flexibly adjusted according to production needs, adapting to the production of films with different raw materials, specifications, and process requirements, thus greatly improving the industrial versatility of this technical solution.
[0060] In addition, the formula for calculating the thickness deviation at each transverse position of the thin film is:
[0061] In the formula, For the first Target thickness value at each location For the first Real-time thickness value at each location.
[0062] Multiple thickness deviation values are input into the PID controller to calculate the initial temperature adjustment for each heating zone. The calculation formula is as follows:
[0063] In the formula, This is the proportionality coefficient. The integral time constant is... For integration variables, is the differential time constant.
[0064] The PID control converts the film thickness deviation into a temperature adjustment for the die head heating area: when the thickness deviation is positive (actual thickness < target thickness), the PID outputs a positive temperature adjustment to increase the temperature of the corresponding area and increase the melt flow rate to compensate for the thickness; when the thickness deviation is negative (actual thickness > target thickness), the PID outputs a negative temperature adjustment to decrease the temperature of the corresponding area and decrease the melt flow rate to compensate for the thickness. Through quantitative calculation, the thickness deviation value at each lateral position is... ,pass , and The synergistic effect of the three parameters is precisely converted into the initial temperature adjustment amount for the corresponding heating zone.
[0065] The initial temperature regulation value is decoupled and transformed using a dynamic decoupling control matrix to obtain the final decoupled temperature regulation value. The specific steps are as follows: First, the initial temperature adjustment values of each heating zone are represented as vectors. Using a standardized vector data structure, the independent adjustment values of multiple zones are transformed into a unified mathematical carrier capable of matrix operations. The calculation formula is as follows:
[0066] In the formula, This represents the total number of heating zones. Then, using the dynamic decoupling control matrix The initial regulation vector is decoupled to obtain the final temperature regulation vector. This step involves a reverse correction of the initial regulation—that is, for the initial regulation of each region, a correction term is added to offset the coupling interference from other regions. Ultimately, the corrected regulation, when applied to the mold, only produces a temperature change in the target region, eliminating cross-region interference. The calculation formula is as follows:
[0067] In the formula, It is a dynamic decoupling control matrix, and .
[0068] Finally, in the time domain, the decoupling transform is represented as a convolution to obtain the final temperature adjustment amount. The final temperature adjustment amount is the time-domain component in the final temperature adjustment amount vector. The purpose of the convolution operation is to dynamically superimpose historical influences, i.e., the current time-th... The final adjustment value of each region not only includes the immediate impact of its own initial adjustment value, but also incorporates the cumulative correction of the coupling effects of the initial adjustment values of all other regions at historical moments. This perfectly adapts to the dynamic and delayed characteristics of thermal coupling, ensuring the real-time performance and accuracy of the decoupling transformation. The calculation formula is as follows:
[0069] In the formula, For dynamic decoupling control matrix The Middle Line number The time-domain impulse response function corresponding to the column element, For the current moment, It is the integral variable.
[0070] This step relies on the matrix transformation principle of linear algebra, the decoupling theory of control engineering, and the time-domain to complex frequency domain conversion logic to correct the ideal uncoupled initial adjustment value output by the PID controller into a final adjustment value adapted to the actual thermal coupling conditions. This mathematically cancels cross-regional thermal interference, providing a precise basis for subsequent power synchronization adjustment. Furthermore, through decoupling transformation, components in the initial adjustment value that could cause coupling interference are precisely canceled, ensuring that the final temperature adjustment value, when applied to the mold head, can only change the temperature of the corresponding area, with almost no impact on other areas. For example, when adjusting the temperature of the third heating area to compensate for the thickness deviation of the corresponding position of the film, the decoupling transformation will superimpose a correction term to cancel the thermal conduction interference of this adjustment on the second and fourth areas, achieving the independent control effect of "adjusting A only changes A, without affecting B, C, and D".
[0071] Based on the final temperature adjustment, the heating power of each heating zone is adjusted synchronously to achieve closed-loop uniformity control of the film's lateral thickness. It is important to note that during synchronous adjustment, the power adjustments of all heating zones are executed simultaneously. This is because the decoupled final temperature adjustment already considers the synergistic relationship between the zones. Therefore, synchronous execution avoids new interference caused by sequential adjustments. For example, adjusting zone 1 first and then zone 2 might cause temporary coupling due to the time difference. This ensures that the temperature of each zone reaches the target value simultaneously, achieving independent and coordinated temperature control across multiple zones. The specific steps are as follows: The first Final temperature adjustment of each heating zone The formula for converting this to a heating power adjustment value is as follows:
[0072] in, For the first Power-temperature conversion coefficient for each heating zone; No. The actual output power of each heating zone is:
[0073] In the formula, For the first The reference power for each heating zone.
[0074] Furthermore, this method can generate adjustment commands when a thickness deviation in the film is detected, thus accelerating the response speed of the entire process of power adjustment, temperature change, and thickness compensation. This allows for timely suppression of deviation expansion (such as thickness fluctuations caused by sudden changes in raw material viscosity), preventing large-area defective films, improving the dynamic response capability of thickness control, and adapting to the real-time control requirements of continuous film production.
[0075] Example 2 In another embodiment, this method also includes an online adaptive update mechanism for the dynamic decoupling control matrix. By monitoring the dynamic characteristics of the system in real time, accurately determining changes in operating conditions, and rapidly iterating the decoupling matrix, the entire control system is upgraded from static decoupling to dynamic adaptive decoupling, further solidifying the long-term stability and accuracy of thin film lateral thickness uniformity control. First, real-time monitoring of the first Temperature change rate of each heating zone This is used to characterize the dynamic response speed of the temperature in a heating region. If a certain region... The sudden increase is essentially a direct manifestation of changes in the thermal conductivity of the die head (such as aging of the heating element or material accumulation in the die head) or a drift in the thermal coupling strength. No. Thickness change rate of each heating zone The real-time fluctuation rate of film thickness is used to characterize the film thickness. If the rate exceeds the limit, it means that the temperature-thickness mapping relationship has changed (such as raw material viscosity fluctuations or production speed adjustments), which indirectly reflects that the thermal coupling characteristics have deviated from the initial modeling state. And the spectral energy distribution of thickness deviation; in, The measured temperature of the heating area. The thickness of the film corresponding to the heating area; Calculate the energy percentage of the thickness deviation signal in the 0.01-0.1Hz frequency band. ,when When the preset threshold is exceeded, the system is determined to have low-frequency disturbances or changes in operating conditions. This design upgrades the operating condition determination from single-indicator monitoring to multi-dimensional feature recognition, avoiding false triggering of updates and is the core of adaptive update determination. The online update of the dynamic decoupling control matrix is triggered when any of the following conditions are met: There exists any Make ; There exists any Make ; ; in, , and These are corresponding preset thresholds, determined based on the statistical characteristics of historical operating condition changes. For high-precision thin film production, a lower threshold can be set to improve update sensitivity and ensure that even minor changes in operating conditions can be captured in a timely manner. For conventional industrial-grade thin film production, a higher threshold can be set to reduce update frequency, balancing control accuracy and production efficiency. The personalized threshold design allows the mechanism to adapt to thin film production scenarios with different raw materials, specifications, and process requirements, further enhancing the industrial versatility of the technical solution. After the update is triggered, normal production control is suspended, and a micro-amplitude pseudo-random binary sequence excitation signal is applied to the system under the current operating conditions. The amplitude of the excitation signal does not exceed 5% of the normal heating power, and the duration does not exceed 30 seconds. The micro-amplitude pseudo-random binary sequence specifically includes micro-amplitude characteristics: the excitation amplitude does not exceed 5% of the normal heating power and the duration is ≤30 seconds. The core is to avoid large fluctuations in melt temperature caused by excitation and ensure the continuity of film production (no large-area defective products are generated). Pseudo-random characteristics: Pseudo-random sequences can cover a wide frequency range and can quickly and comprehensively stimulate the dynamic response of the system. Compared with step excitation, it can obtain the complete coupling characteristics of the system in a shorter time and improve update efficiency. Reuse the modeling workflow: Re-execute the modeling steps of the multi-point temperature field coupling model (excitation-response acquisition-transfer function matrix construction) to ensure the updated model... Consistent with the initial modeling logic, this ensures theoretical coherence and data compatibility in decoupling matrix updates.
[0076] Re-execute the modeling steps for the multi-point temperature field coupling model to obtain the updated transfer function matrix. and based on Recalculate the dynamic decoupling control matrix .
[0077] Furthermore, this adaptive update mechanism enables the system to proactively resist interference: by monitoring and capturing characteristic changes caused by disturbances in real time, it quickly updates the decoupling matrix to offset the impact of disturbances and avoid the accumulation of thickness deviations. For example, if the ambient temperature continues to rise, causing the mold head to dissipate heat faster and the thermal coupling delay time to shorten, the system will trigger an update through low-frequency energy ratio monitoring, recalculate the decoupling matrix to adapt to the new delay characteristics, ensure long-term stability of thickness control accuracy, and significantly improve the robustness and reliability of the system.
[0078] Example 3 like Figure 2 The thin film lateral thickness uniformity control system shown specifically includes: The data acquisition module is used to acquire real-time temperature data and corresponding film lateral thickness distribution data of multiple heating zones distributed laterally along the extrusion die. The model building module establishes a multi-point temperature field coupling model to describe the thermal coupling effect between multiple heating regions, based on real-time temperature data and thin film lateral thickness distribution data. The model calculation module, based on a multi-point temperature field coupling model, calculates a dynamic decoupling control matrix to eliminate the thermal coupling effects between heating regions. The thickness deviation calculation module obtains the target thickness distribution curve and the real-time thickness distribution curve of the film in the lateral direction, compares the two distribution curves, and obtains the thickness deviation value of the film at each lateral position. The PID calculation module inputs multiple thickness deviation values into the PID controller to calculate the initial temperature adjustment of each heating zone. The decoupling transformation module uses a dynamic decoupling control matrix to decouple and transform the initial temperature regulation amount to obtain the final decoupled temperature regulation amount. The control module synchronously adjusts the heating power of each heating zone according to the final temperature adjustment amount, thereby achieving closed-loop uniformity control of the film's lateral thickness.
[0079] Example 4 As attached Figure 3 An electronic device shown includes: Processor, memory, communication interface; The memory is used to store the executable instructions of the processor; The processor is configured to execute the aforementioned thin film lateral thickness uniformity control method based on dynamic decoupling of multi-point temperature fields by executing the executable instructions.
[0080] A readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method for controlling the transverse thickness uniformity of thin films based on dynamic decoupling of multi-point temperature fields.
[0081] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
Claims
1. A method for controlling the transverse thickness uniformity of thin films based on dynamic decoupling of multi-point temperature fields, characterized in that, Specifically, it includes: Acquire real-time temperature data and corresponding film lateral thickness distribution data for multiple heating zones distributed laterally along the extrusion die. Based on real-time temperature data and thin film lateral thickness distribution data, a multi-point temperature field coupling model is established to describe the thermal coupling effect between multiple heating regions. Based on the multi-point temperature field coupling model, a dynamic decoupling control matrix is calculated to eliminate the thermal coupling effect between heating regions. Obtain the target thickness distribution curve and the real-time thickness distribution curve of the film in the lateral direction, compare the two distribution curves, and obtain the thickness deviation value of the film at each lateral position. Multiple thickness deviation values are input into the PID controller to calculate the initial temperature adjustment for each heating zone. The initial temperature regulation value is decoupled and transformed using a dynamic decoupling control matrix to obtain the final decoupled temperature regulation value. Based on the final temperature adjustment, the heating power of each heating zone is adjusted synchronously to achieve closed-loop uniformity control of the film's lateral thickness.
2. The method for controlling the transverse thickness uniformity of thin films based on dynamic decoupling of multi-point temperature fields according to claim 1, characterized in that: The multi-point temperature field coupling model, which describes the thermal coupling effect between multiple heating regions, is established based on real-time temperature data and thin film lateral thickness distribution data. The specific steps are as follows: A preset excitation signal is applied to each heating zone, and temperature response data of each heating zone is collected. Based on temperature response data, the coupling transfer function between each heating region is calculated, and the transfer function matrix of the multi-input multi-output system is constructed. The expression of this matrix is as follows: In the formula, The total number of heating areas, diagonal elements Indicates the first The dynamic characteristics of each heating zone itself, non-diagonal elements Description of the The heating zone for the first The thermal coupling strength of each heating zone It is a complex frequency.
3. The method for controlling the transverse thickness uniformity of thin films based on dynamic decoupling of multi-point temperature fields according to claim 2, characterized in that: The multiple input multiple output system is a controlled object that takes the heating power of multiple heating zones laterally distributed on the extrusion die as input and the measured value of the temperature sensor of the corresponding zone as output.
4. The method for controlling the transverse thickness uniformity of thin films based on dynamic decoupling of multi-point temperature fields according to claim 2, characterized in that: The dynamic decoupling control matrix for eliminating the thermal coupling effects between heating regions is calculated based on the multi-point temperature field coupling model. The specific steps are as follows: Based on the transfer function matrix, the dynamic decoupling control matrix is calculated using the feedforward compensation decoupling method. Furthermore, the principle for determining the dynamic decoupling control matrix is that the product matrix of the dynamic decoupling control matrix and the transfer function matrix is a diagonal matrix or a diagonally dominant matrix.
5. The method for controlling the transverse thickness uniformity of thin films based on dynamic decoupling of multi-point temperature fields according to claim 4, characterized in that: The feedforward compensation decoupling method is either the direct feedforward compensation method, the diagonal matrix decoupling method, or the identity matrix decoupling method.
6. The method for controlling the transverse thickness uniformity of thin films based on dynamic decoupling of multi-point temperature fields according to claim 1, characterized in that: The target distribution curve for obtaining the transverse thickness of the thin film includes one or more combinations of a theoretical target curve calculated based on a theoretical model, a measured target curve based on actual measurements of a benchmark sample, a tolerance-based target curve based on a process window, a dynamic target curve based on adaptive learning, and a compensation target curve based on defect pattern recognition.
7. The method for controlling the transverse thickness uniformity of thin films based on dynamic decoupling of multi-point temperature fields according to claim 1, characterized in that: The target thickness distribution curve and real-time thickness distribution curve of the film's lateral thickness are obtained, and the two distribution curves are compared to obtain the thickness deviation value of the film at each lateral position. The calculation formula is as follows: In the formula, For the first Target thickness value at each location For the first Real-time thickness values at each location. For the first extrusion die The spatial coordinates of each detection point.
8. The method for controlling the transverse thickness uniformity of thin films based on dynamic decoupling of multi-point temperature fields according to claim 7, characterized in that: The process involves inputting multiple thickness deviation values into a PID controller to calculate the initial temperature adjustment for each heating zone. The calculation formula is as follows: In the formula, This is the proportionality coefficient. The integral time constant is... For integration variables, The differential time constant is This refers to the current moment.
9. The method for controlling the transverse thickness uniformity of thin films based on dynamic decoupling of multi-point temperature fields according to claim 8, characterized in that: The specific steps for decoupling the initial temperature regulation amount using a dynamic decoupling control matrix to obtain the final decoupled temperature regulation amount are as follows: The initial temperature adjustment of each heating zone is expressed in vector form, and the calculation formula is as follows: In the formula, This represents the total number of heating zones. The initial temperature adjustment is in vector form; Using dynamic decoupling control matrix The initial adjustment vector is decoupled and transformed to obtain the final temperature adjustment vector. The calculation formula is: In the formula, For dynamic decoupling control matrix; In the time domain, the decoupling transformation is represented as a convolution to obtain the final temperature adjustment amount. The calculation formula is: In the formula, For dynamic decoupling control matrix The Middle Line number The time-domain impulse response function corresponding to the column element, In response to the heating zone, To stimulate the heating zone, For the current moment, It is the integral variable.
10. The method for controlling the transverse thickness uniformity of thin films based on dynamic decoupling of multi-point temperature fields according to claim 9, characterized in that: The specific steps for synchronously adjusting the heating power of each heating zone based on the final temperature adjustment amount are as follows: The first Final temperature adjustment of each heating zone The formula for converting this to a heating power adjustment value is as follows: in, For the first Power-temperature conversion coefficient for each heating zone; No. The actual output power of each heating zone is: In the formula, For the first The reference power for each heating zone.