A collaborative control method for intensive gradient composite molding

Abstract

This invention discloses an intensive gradient composite molding collaborative control method, relating to the field of industrial production control technology. The method includes a composite production machine equipped with a material feeding station, a material feeding station 2, a correction station, and a die-cutting station. The collaborative control method further includes a data acquisition component, a data preprocessing component, and a composite analysis component. The data acquisition component acquires relevant data on the operation of the composite production machine, including material dynamic data, process coordination data, and compensation-related data. The acquired data is then input into the data preprocessing component for data cleaning. After cleaning, the data is input into the composite analysis component. Through the mutual cooperation between the above structures, this invention forms a collaborative control logic of material sensing, process optimization, and depth compensation, solving the problem of material and process parameter disconnect in the prior art and achieving overall collaborative stability of the molding process.

Description

Technical Field

[0001] This invention relates to the field of industrial production control technology, specifically to an intensive gradient composite molding collaborative control method. Background Technology

[0002] With the development of intelligent manufacturing technology, the gradient composite molding process that integrates feeding and die-cutting is increasingly widely used in fields such as electronic information, packaging and printing, and new energy. This type of process requires the simultaneous coordination of the dynamic characteristics of materials and the coupling relationship of process parameters to meet the molding processing with a certain precision. For example, in the die-cutting production of self-adhesive labels, slight fluctuations in the thickness of the roll material can lead to deviations in the die-cutting depth, which in turn affects the peeling performance of the label. Similarly, changes in the process speed can also change the die-cutting frequency, resulting in unstable pressure transmission and ultimately affecting the product qualification rate.

[0003] Currently, most gradient composite molding equipment adopts static parameter preset or single parameter feedback control modes. This mode is difficult to deal with the multi-parameter dynamic coupling problem under complex working conditions. On the one hand, dynamic changes in material properties, such as material softening due to temperature rise and slippage of the roll due to increased friction coefficient, are difficult to quantify in real time, and control commands are easily decoupled from actual working conditions. On the other hand, the coordinated optimization of process parameters, such as speed, vibration, and phase difference, are independent of each other, making it difficult to achieve overall process coordination and stability. When the characteristics of the roll and process parameters change, most of these simple control systems can only adjust the corresponding controller parameters, which do not adequately consider the indirect impact on product die-cutting, resulting in fluctuations in die-cutting quality. Therefore, the existing composite molding processing technology still needs to be improved. Summary of the Invention

[0004] The purpose of this invention is to provide a intensive gradient composite molding collaborative control method, which solves the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical method: an intensive gradient composite molding collaborative control method, comprising the following steps:

[0006] S1: Data Acquisition: Obtain real-time data on material properties and the process status during the operation of the composite production machine to provide basic input for subsequent formula calculations;

[0007] S2: Data Processing: Start the composite production machine for die-cutting. The data obtained during the process is cleaned by the data preprocessing component and then sent to the composite analysis component. The composite analysis component calculates the material dynamic factor, process synergy coefficient and dynamic depth compensation amount. S3: Data Correction: Iteratively correct the calculated dynamic depth compensation amount until the convergence condition is met, and output the final dynamic depth compensation amount to the execution component. The execution component drives the electromagnetic control component to operate and control the die-cutting accuracy of the die-cutting knife.

[0008] This invention provides the following technical solution: an intensive gradient composite molding collaborative control method, including a composite production machine, wherein the composite production machine is provided with a feeding station one, a feeding station two, a correction station and a die-cutting station, and the collaborative control method further includes a data acquisition component, a data preprocessing component and a composite analysis component;

[0009] The data acquisition component acquires relevant data on the operation of the composite production machine, including material dynamic data, process coordination data, and compensation-related data. The acquired data is then input into the data preprocessing component to clean the data. After cleaning, the data is input into the composite analysis component.

[0010] The composite analysis component:

[0011] Based on the material dynamic data, including the real-time average thickness of the roll, the standard thickness of the material, the real-time average tension of the roll, the standard tension of the material, the temperature influence coefficient, the difference between the real-time and standard temperatures, the friction coefficient influence coefficient, the change in the friction coefficient, the standard friction coefficient, the roll width, and the positive pressure of the guide roller on the roll, the material dynamic factor is output. The influence of the real-time physical properties of the roll on the die-cutting quality is quantified through the material dynamic factor.

[0012] Based on the real-time speed of the roll material, standard process speed, real-time average vibration acceleration of the die-cutting station, standard vibration acceleration, real-time phase angle between the die-cutting pattern and the roll material, standard phase angle, elastic strain change, standard elastic strain, material elastic modulus, die-cutting frequency change, standard die-cutting frequency, die-cutting pattern length, and influence coefficient in the process collaboration data, the process collaboration coefficient is output, and the process adjustment coefficient under the current working condition is quantified through the process collaboration coefficient.

[0013] Based on the depth compensation gain, real-time die-cutting pressure, standard die-cutting pressure, and weighting coefficients in the compensation-related data, a dynamic depth compensation amount is output, and the depth compensation value that the die-cutting blade needs to be adjusted is calculated using the dynamic depth compensation amount.

[0014] Optionally, the collaborative control method further includes a data optimization component and an execution component. The data optimization component outputs the depth compensation gain of the (i+1)th iteration based on the depth compensation gain of the i-th iteration, the die-cutting quality deviation of the i-th iteration, the learning rate, the minimum value of the compensation gain, and the maximum value of the compensation gain in the compensation-related data. The dynamic depth compensation amount is recalculated based on the depth compensation gain of the (i+1)th iteration until the iteration convergence condition is met, and then the dynamic depth compensation amount is output to the execution component.

[0015] Optionally, the composite analysis component includes a material quantification module, a process quantification module, and a compensation quantification module.

[0016] Optionally, the processing logic of the material quantification module is as follows:

[0017] The product of real-time thickness and tension is divided by the product of standard thickness and tension to form a thickness-tension coupling deviation term, which quantifies the fluctuation of the material's basic properties. The temperature deviation term and the relative deviation term of the friction coefficient are then superimposed on the reference value 1 to supplement the dynamic influence of temperature and friction on the material properties.

[0018] Optionally, the processing logic of the process quantization module is as follows:

[0019] The maximum relative deviation of the three parameters—velocity, vibration, and phase difference—is used as the denominator. Each sub-item is normalized, and then the relative deviation of each parameter is multiplied by the corresponding coefficient and divided by the normalized denominator to quantify the influence of velocity, vibration, and phase difference on the process state.

[0020] Optionally, the processing logic of the compensation quantization module is as follows:

[0021] The initial gain amplifies the weighted contributions of materials and processes, the compensation direction is distinguished by the sign function, and then multiplied with the absolute value of the relative deviation of the die-cutting pressure. The urgency of compensation is quantified by the pressure deviation, making the compensation more consistent with the actual die-cutting pressure state.

[0022] Optionally, the iterative convergence logic of the data optimization component is such that the iteration terminates when one of the following two conditions is met:

[0023] A1: When the absolute value of the die-cutting quality deviation in the i-th iteration is less than or equal to the preset threshold, the die-cutting accuracy is considered to be up to standard.

[0024] A2: When the number of iterations reaches the preset maximum, the iteration will be forcibly terminated.

[0025] Optionally, the die-cutting station includes a die-cutting blade, an electromagnetic control component, and a bottom shaft component. After the dynamic depth compensation is output to the execution component, the execution component drives the electromagnetic control component to operate, thereby controlling the die-cutting accuracy of the die-cutting blade.

[0026] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0027] I. This invention quantifies the fluctuation of basic material properties by measuring the relative deviations of thickness and tension. Combined with the coupled calculation of changes in temperature coefficient and friction coefficient, it transforms the synergistic effects of thickness, tension, temperature, and friction coefficient into a perceptible comprehensive factor. This achieves precise quantification of the dynamic properties of multiple material parameters, overcomes the limitations of perceiving a single material parameter, and provides accurate material state basis for subsequent process adjustments.

[0028] Second, this invention takes the relative deviation of speed and vibration and the effective change of phase difference as core parameters. It balances the magnitude difference of different parameters through normalization processing, and then combines the weighted coupling of speed coefficient, vibration coefficient and phase difference coefficient to transform the independent fluctuations of speed, vibration and phase difference into synergistic influence coefficients. As a whole, it realizes the quantification of the synergistic influence of multiple process parameters and solves the discreteness problem of independent adjustment of process parameters.

[0029] Third, by integrating parameters such as phase angle deviation, elastic strain coefficient, elastic strain change, and die-cutting frequency change, this invention quantifies the coupling effect of phase angle deviation with elastic strain and die-cutting frequency into an effective change, thereby achieving accurate quantification of the actual impact of phase difference. This solves the problem of existing technologies that only consider phase angle deviation and ignore its coupling effect with other process parameters, providing a reliable basis for accurate calculation of process coordination coefficients.

[0030] Fourth, this invention takes material dynamic factors and process synergy coefficients as core inputs, and combines the weighted allocation of material weights and process weights to transform the comprehensive influence of materials and processes into a precise depth compensation amount, thereby realizing dynamic and precise compensation of forming depth as a whole and solving the problem of inaccuracy of single parameter feedback compensation.

[0031] Fifth, this invention uses the depth deviation of visual detection as feedback, adjusts the iteration step size of the compensation gain through the learning rate, and combines the convergence control of the maximum number of iterations and the quality deviation threshold to transform the error of a single compensation into a closed-loop process of continuous optimization. Overall, it realizes the continuous optimization and accurate convergence of the compensation amount, and solves the limitation problem that static compensation cannot adapt to dynamic working conditions. Attached Figure Description

[0032] Figure 1 This is a flowchart of the present invention;

[0033] Figure 2 This is a schematic diagram of the composite analysis component of the present invention;

[0034] Figure 3 This is an isometric view of the present invention;

[0035] Figure 4 This is a schematic diagram of the die-cutting station of the present invention.

[0036] In the diagram: 1. Composite production machine; 2. Feeding station 1; 3. Feeding station 2; 4. Correction station; 5. Die-cutting station; 6. Die-cutting blade; 7. Electromagnetic control assembly; 8. Bottom shaft assembly. Detailed Implementation

[0037] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0038] Please see Figure 1 This invention provides a method for intensive gradient composite molding collaborative control, comprising the following steps:

[0039] S1: Real-time data on material properties and the process status of the composite production machine 1 during operation are obtained through the data acquisition component, providing basic input for subsequent formula calculations.

[0040] Specifically, please refer to Figures 1 to 4 The system includes a composite production machine 1, which is equipped with a feeding station 2, a feeding station 3, a correction station 4, and a die-cutting station 5. The die-cutting station 5 includes a die-cutting blade 6, an electromagnetic control component 7, and a bottom shaft component 8. The collaborative control method also includes a data acquisition component, a data preprocessing component, and a composite analysis component. The data acquisition component acquires relevant data on the operation of the composite production machine 1, including material dynamic data, process coordination data, and compensation-related data. The material dynamic data includes the real-time average thickness T of the roll material, which is obtained by directly collecting the thickness data at different positions of the roll material using a laser thickness gauge along the roll material conveying path and then calculating the average value. avg The material standard thickness T is pre-entered by process engineers according to material specifications and production requirements. ref The real-time average tension σ of the roll material is obtained by directly collecting data at the roll material tensioning mechanism using a tension sensor and calculating the average value of multiple sampling points. avg The material standard tension σ is pre-entered by process engineers based on material properties and production processes. ref The temperature influence coefficient k is determined through calibration experiments before the equipment leaves the factory; the real-time temperature difference ΔT between the real-time and standard temperatures is obtained by subtracting the standard temperature preset in the process database from the real-time temperature collected on the surface of the roll material by an infrared temperature sensor; and the friction coefficient influence coefficient λ is determined through calibration experiments before the equipment leaves the factory.

[0041] It also includes the standard friction coefficient μ pre-entered by process engineers. refThe roll width W, pre-entered by process engineers according to material specifications, is directly measured by a pressure sensor at the contact point between the guide roller and the roll, where the positive pressure F exerted by the guide roller on the roll is collected. normal And the change in friction coefficient Δμ, where the change in friction coefficient Δμ is first expressed by the formula μ act =σ avg ×W / F normal Calculate the real-time friction coefficient μ act Then subtract the preset standard friction coefficient μ ref get;

[0042] Process collaboration data includes the real-time speed V of the roll material, which is directly acquired at the roll material conveying mechanism via an encoder. act The standard process speed V, pre-entered by process engineers ref Standard vibration acceleration A ref Standard phase angle φ ref Standard elastic strain ε ref Material elastic modulus E, standard die-cutting frequency f ref The length L of the die-cut pattern is directly collected at the die-cutting station using a vibration sensor. The real-time average vibration acceleration A at the die-cutting station is obtained by calculating the average value of multiple sampling points. avg The real-time phase angle φ between the die-cutting pattern and the roll material is directly acquired at the die-cutting station through a vision system. act The elastic strain change Δε, the die-cutting frequency change Δf, and the influence coefficient;

[0043] The change in elastic strain Δε is first expressed by the formula ε act =σ avg / E Calculates real-time elastic strain ε act Subtract the preset standard elastic strain ε ref The change in die-cutting frequency Δf is obtained first through the formula f act =V act / L calculates the real-time die-cutting frequency f act Subtract the standard die-cutting frequency f preset in the process database. ref The influence coefficients, including velocity influence coefficient α, vibration influence coefficient β, phase difference influence coefficient γ, material elastic strain influence coefficient δ, and die-cutting frequency influence coefficient h, were determined through calibration experiments before the equipment left the factory.

[0044] The compensation-related data includes the depth compensation gain K determined by the calibration test before the equipment leaves the factory, and the real-time pressure F of the die-cutting blade directly collected by the pressure sensor at the die-cutting blade. act The standard die-cutting pressure F is pre-entered by process engineers according to production process requirements. ref and weighting coefficients, where the weighting coefficients include the material property weighting coefficient ω. M and process parameter weighting coefficient ωP All parameters were determined through calibration tests conducted before the equipment left the factory.

[0045] S2: Start the composite production machine 1 to perform die-cutting. The data obtained during the process is cleaned by the data preprocessing component and then sent to the composite analysis component. The composite analysis component calculates the material dynamic factor, process synergy coefficient and dynamic depth compensation amount.

[0046] Specifically, the collected data is input into the data preprocessing unit, where it is cleaned. After cleaning, the data is input into the composite analysis unit, which includes a material quantification module, a process quantification module, and a compensation quantification module.

[0047] The material quantification module is based on the real-time average thickness T of the cleaned roll material from the material dynamic data. avg Material standard thickness T ref Real-time average tension σ of roll material avg Material standard tension σ ref Temperature influence coefficient k, real-time temperature difference ΔT between real-time and standard temperatures, friction coefficient influence coefficient λ, friction coefficient change Δμ, standard friction coefficient μ ref The output material dynamic factor (MDF) quantifies the impact of the real-time physical properties of the roll material on the die-cutting quality. Specifically, the processing logic of the material quantification module is as follows:

[0048]

[0049] Furthermore, considering that thickness and tension are the core parameters affecting the transmission of die-cutting pressure, the relative deviation of their product can more comprehensively reflect the material's overall influence on the die-cutting depth. Therefore, the product of real-time thickness and tension is multiplied and then divided by the product of standard thickness and tension to form the thickness-tension coupling deviation term. This quantifies the fluctuations in the basic properties of the material. The temperature deviation term and the relative deviation term of the friction coefficient are then superimposed on the baseline value 1 to supplement the dynamic influence of temperature and friction on the material properties. This supplements the impact of temperature changes on material hardness and friction changes on the stability of the roll material conveying. The temperature influence coefficient k is set to 0.005, and the friction coefficient influence coefficient λ is set to 0.03. Since the influence of temperature on material properties is slow and linear, the temperature influence coefficient k is set to a small value. However, the sudden change in the friction coefficient will directly cause the roll material to slip, which has a more significant impact on the die-cutting stability. Therefore, the friction coefficient influence coefficient λ is set to a slightly larger value.

[0050] By quantifying the fluctuations in the basic properties of materials through the relative deviations of thickness and tension, and combining the coupled calculations of the changes in temperature coefficient and friction coefficient, the synergistic effects of thickness, tension, temperature, and friction coefficient are transformed into a perceptible comprehensive factor. The relative deviations of thickness and tension reflect the basic fluctuations in the physical properties of the material, while the changes in temperature and friction supplement the dynamic effects of the environment and contact characteristics. Overall, the precise quantification of the dynamic properties of multiple parameters of the material is achieved, which solves the limitations of perceiving a single material parameter and provides an accurate basis for the material state for subsequent process adjustments.

[0051] The composite analysis component also includes a process quantification module, which is based on the real-time speed V of the roll material in the process collaboration data. act Standard process speed V ref Real-time average vibration acceleration A at the die-cutting station avg Standard vibration acceleration A ref The process coordination coefficient (PSC) is output, and the process adjustment coefficient under the current operating condition is quantified through the PSC. Specifically, the processing logic of the process quantification module is as follows:

[0052]

[0053]

[0054] Furthermore, the maximum relative deviation of the three parameters—speed, vibration, and phase difference—is used as the denominator. Each component is normalized, and then the relative deviation of each parameter is multiplied by its corresponding coefficient and divided by the normalized denominator. This quantifies the impact of speed, vibration, and phase difference on the process state. Speed ​​affects the die-cutting frequency, vibration affects die-cutting stability, and phase difference affects pattern alignment accuracy. These three factors collectively determine the compensation requirements at the process level. α is the speed influence coefficient. Because the effect of speed on die-cutting depth is relatively indirect, and speed affects the die-cutting frequency, which in turn affects pressure transmission, its weight is relatively low. We take 0.1. Vibration directly affects the stability of the die-cutting blade, but considering that the vibration deviation is generally small and most existing equipment has a vibration reduction design, the weight is taken as 0.02, which is lower than the weight of the speed term. The phase difference mainly affects the pattern alignment rather than the depth, but if the alignment deviation is too large, it will indirectly lead to the failure of depth compensation. Therefore, γ is used as the phase difference influence coefficient, and the weight is taken as the minimum of 0.02. Since the three parameter terms have been normalized in the above formula, the role of α, β and γ in the formula is only to adjust the relative influence of the corresponding sub-terms. There is no need to normalize them again, and there is no need to satisfy the sum of 1.

[0055] Using the relative deviations of velocity and vibration, and the effective change in phase difference as core parameters, normalization is applied to balance the magnitude differences of different parameters. Then, by combining the weighted coupling of velocity coefficient, vibration coefficient, and phase difference coefficient, the independent fluctuations of velocity, vibration, and phase difference are transformed into synergistic influence coefficients. The relative deviations of velocity and vibration reflect the basic stability of process execution, and the effective change in phase difference Δφ... eff The addition of dynamic effects of motion synchronization, normalization, and weighting coefficients ensures a reasonable allocation of parameter influences, thereby quantifying the synergistic effects of multiple process parameters and solving the discreteness problem of independent adjustment of process parameters.

[0056] The effective change in phase difference Δφ eff The calculation logic is based on the real-time phase angle φ between the die-cutting pattern and the roll material in the process quantization module. act Standard phase angle φ ref Elastic strain change Δε, standard elastic strain ε ref Die-cutting frequency change Δf, standard die-cutting frequency f ref The influence coefficients δ (elastic strain of the material) and h (die-cutting frequency) are derived, specifically as follows:

[0057]

[0058] Furthermore, through sign(φ) act -φ ref Distinguish the direction of deviation to ensure the directional accuracy of the compensation command, through |φ act -φ ref | / φ ref The relative deviation is calculated, and then the effects of elastic strain and die-cutting frequency are superimposed to make the quantification of phase difference more consistent with actual working conditions. Here, sign(x) is the sign function, returning 1 when x > 0, -1 when x < 0, and 0 when x = 0. Furthermore, due to the effective change in phase difference Δφ... eff The computational logic essentially quantifies the influence of the phase angle through relative deviation rather than absolute deviation; therefore, the standard phase angle φ ref It is a non-zero reference value at the default setting.

[0059] By integrating parameters such as phase angle deviation, elastic strain coefficient δ, elastic strain change Δε, and die-cutting frequency change Δf, the coupling effect of phase angle deviation with elastic strain and die-cutting frequency is quantified into an effective change, realizing the accurate quantification of the actual impact of phase difference. This solves the problem of existing technologies that only consider phase angle deviation and ignore its coupling effect with other process parameters, providing a reliable phase difference basis for the accurate calculation of process coordination coefficient.

[0060] The composite analysis component also includes a compensation quantization module, which is based on the depth compensation gain K and the real-time pressure F of the die-cutting blade in the compensation-related data. act Standard die-cutting pressure F ref and weighting coefficients, where the weighting coefficients include the material property weighting coefficient ω. M and process parameter weighting coefficient ω P and weighting coefficient ω M ω P Output the dynamic depth compensation value (DDC). Calculate the depth compensation value that the die-cutting blade (6) needs to adjust using the dynamic depth compensation value (DDC). Specifically, the processing logic of the compensation quantization module is as follows:

[0061]

[0062] Furthermore, the weighted contributions of materials and processes are amplified using the initial gain, through the sign function sign(F) ref -F act Distinguish the compensation direction: increase the depth when the pressure is insufficient and decrease the depth when the pressure is excessive, thereby transforming the dynamic changes in materials and processes into a directional compensation trend. This ensures that the compensation direction is consistent with the actual pressure deviation, and then compares it with the absolute value of the relative deviation of the die-cutting pressure |(F ref -F act ) / F ref | By multiplying the products, the urgency of compensation is quantified using pressure deviation. The greater the pressure deviation, the greater the compensation required, making the compensation more closely match the actual die-cutting pressure state. Since the material dynamic factor (MDF) and the process synergy coefficient (PSC) have different influence mechanisms on the die-cutting depth, a weighting coefficient ω is used. M and ω P The contribution ratio is allocated to ensure that the compensation amount takes into account both the dynamic changes in materials and processes. Material properties are considered a core factor affecting die-cutting depth; therefore, the material property weight ω is determined. M The weight of parameters like speed and vibration is relatively high, so it is set to 0.7. These are secondary factors affecting die-cutting depth, and therefore have a lower weight of 0.3. The weighting coefficient ω... M With ω P The sum of these conditions is 1, and the depth compensation gain K, as the basic amplification factor for the initial compensation, determines the initial adjustment range of the compensation amount, which is set to 0.05.

[0063] Using the Material Dynamic Factor (MDF) and Process Synergy Coefficient (PSC) as core inputs, and combining the weighted allocation of material weights and process weights, the combined influence of materials and processes is transformed into a precise depth compensation amount. The MDF and PSC provide the state basis of materials and processes respectively, the material weights and process weights balance the influence ratio of the two, and the pressure deviation and initial gain correct the actual execution accuracy of the compensation amount. Overall, dynamic and precise compensation of forming depth is achieved, solving the problem of inaccuracy of single parameter feedback compensation.

[0064] S3: Iteratively correct the calculated dynamic depth compensation amount until the convergence condition is met, and output the final dynamic depth compensation amount to the execution component. The execution component drives the electromagnetic control component 7 to operate and control the die-cutting accuracy of the die-cutting blade 6.

[0065] Specifically, the compensation-related data also includes the die-cutting quality deviation e obtained by subtracting the target depth from the actual die-cutting depth detected by the vision system in the i-th iteration. i The learning rate η, determined through calibration experiments before the equipment leaves the factory, is the maximum number of iterations N set by process engineers according to production process requirements. max In addition to the quality deviation threshold Z, the collaborative control method also includes a data optimization component, which is based on the depth compensation gain K of the i-th iteration. i And the die-cutting quality deviation e in the i-th iteration of the compensation-related data i Learning rate η, minimum compensation gain K min and the maximum compensation gain K max Output the depth compensation gain K for the (i+1)th iteration. i＋1 :

[0066]

[0067] The die-cutting quality deviation is amplified by the learning rate η. i Then superimposed on the current gain K i Then, by limiting the gain range, the depth compensation gain K is iterated. If the die-cutting depth is less than e... i If the value is greater than 0, increase the gain K to increase the compensation amount. If the die-cutting depth is too deep, e i If the value is less than 0, the gain K is reduced to decrease the compensation amount. At the same time, the upper and lower limits are used to avoid over-adjustment due to excessive gain or under-adjustment due to insufficient gain.

[0068] Based on the depth compensation gain K of the (i+1)th iteration i＋1 The dynamic depth compensation (DDC) is recalculated until the iterative convergence condition is met:

[0069]

[0070] The iterative convergence logic of the data optimization component is as follows: the iteration will terminate when one of the following three conditions is met:

[0071] A1: When the absolute value of the die-cutting quality deviation in the i-th iteration is less than or equal to a preset threshold, the die-cutting accuracy is considered to be up to standard, i.e., |e i |≤Z;

[0072] A2: When the number of iterations reaches the preset maximum, the iteration is forcibly terminated, i.e., i≥N max To avoid system instability caused by excessive iteration.

[0073] Using depth bias detected visually as feedback, the iteration step size for compensating for the gain is adjusted by the learning rate η, combined with the maximum number of iterations N. max The convergence control with the quality deviation threshold Z transforms the error of a single compensation into a closed-loop process of continuous optimization. The depth deviation provides feedback on the compensation effect, while the learning rate η ensures the stability of the gain adjustment. The maximum number of iterations N... max The quality deviation threshold Z controls the convergence of the optimization process, achieving continuous optimization and precise convergence of the compensation amount, thus solving the limitation of static compensation in adapting to dynamic working conditions.

[0074] The collaborative control method also includes an execution component that, after stopping iteration, outputs a dynamic depth compensation (DDC) value. i Dynamic depth compensation DDC i The output is sent to the execution unit, which drives the electromagnetic control component 7 to operate, thereby controlling the die-cutting accuracy of the die-cutting blade 6. It is worth noting that adjusting the position of the die-cutting blade 6 through the electromagnetic control component 7 is a common technique in existing die-cutting devices. In this embodiment, the principle of this part will not be described in detail.

[0075] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for intensive gradient composite molding collaborative control, characterized in that: Includes the following steps: S1: Data Acquisition: Obtain real-time data on material properties and the process status of the composite production machine (1) during operation, providing basic input for subsequent formula calculations; S2: Data processing: Start the composite production machine (1) for die cutting. The data obtained during the processing is cleaned by the data preprocessing component and then sent to the composite analysis component. The composite analysis component calculates the material dynamic factor, process synergy coefficient and dynamic depth compensation amount. S3: Data correction: Iteratively correct the calculated dynamic depth compensation amount until the convergence condition is met, and output the final dynamic depth compensation amount to the execution component. The execution component drives the electromagnetic control component (7) to operate and control the die-cutting accuracy of the die-cutting knife (6).

2. The intensive gradient composite molding collaborative control method according to claim 1, characterized in that: The composite production machine (1) is provided with a feeding station 1 (2), a feeding station 2 (3), a correction station (4) and a die-cutting station (5). The collaborative control method also includes a data acquisition component, a data preprocessing component and a composite analysis component. The relevant data of the operation of the composite production machine (1) are obtained through the data acquisition component, including material dynamic data, process coordination data and compensation-related data. The collected data is then input into the data preprocessing component, and the data is cleaned by the data preprocessing component. After cleaning, the data is input into the composite analysis component. The composite analysis component: Based on the material dynamic data, including the real-time average thickness of the roll, the standard thickness of the material, the real-time average tension of the roll, the standard tension of the material, the temperature influence coefficient, the difference between the real-time and standard temperatures, the friction coefficient influence coefficient, the change in the friction coefficient, the standard friction coefficient, the roll width, and the positive pressure of the guide roller on the roll, the material dynamic factor is output. The influence of the real-time physical properties of the roll on the die-cutting quality is quantified through the material dynamic factor. Based on the real-time speed of the roll material, standard process speed, real-time average vibration acceleration of the die-cutting station, standard vibration acceleration, real-time phase angle between the die-cutting pattern and the roll material, standard phase angle, elastic strain change, standard elastic strain, material elastic modulus, die-cutting frequency change, standard die-cutting frequency, die-cutting pattern length, and influence coefficient in the process collaboration data, the process collaboration coefficient is output, and the process adjustment coefficient under the current working condition is quantified through the process collaboration coefficient. Based on the depth compensation gain, real-time pressure of the die cutter, standard die cutting pressure and weighting coefficient in the compensation-related data, the dynamic depth compensation amount is output, and the depth compensation value that the die cutter (6) needs to be adjusted is calculated through the dynamic depth compensation amount.

3. The intensive gradient composite molding collaborative control method according to claim 2, characterized in that: The collaborative control method further includes a data optimization component and an execution component. The data optimization component outputs the depth compensation gain of the (i+1)th iteration based on the depth compensation gain of the i-th iteration, the die-cutting quality deviation of the i-th iteration, the learning rate, the minimum value of the compensation gain, and the maximum value of the compensation gain in the compensation-related data. The dynamic depth compensation amount is recalculated based on the depth compensation gain of the (i+1)th iteration until the iteration convergence condition is met, and then the dynamic depth compensation amount is output to the execution component.

4. The intensive gradient composite molding collaborative control method according to claim 3, characterized in that: The composite analysis component includes a material quantification module, a process quantification module, and a compensation quantification module.

5. The intensive gradient composite molding collaborative control method according to claim 4, characterized in that: The processing logic of the material quantification module is as follows: The real-time thickness and tension are multiplied together and then divided by the product of the standard thickness and tension to form the thickness and tension coupling deviation term. This is used to quantify the fluctuation of the material's basic properties. The temperature deviation term and the relative deviation term of the friction coefficient are then superimposed on the reference value 1 to supplement the dynamic influence of temperature and friction on the material properties.

6. The intensive gradient composite molding collaborative control method according to claim 5, characterized in that: The processing logic of the process quantification module is as follows: The maximum relative deviation of the three parameters—velocity, vibration, and phase difference—is used as the denominator. Each sub-item is normalized, and then the relative deviation of each parameter is multiplied by the corresponding coefficient and divided by the normalized denominator to quantify the influence of velocity, vibration, and phase difference on the process state.

7. The intensive gradient composite molding collaborative control method according to claim 6, characterized in that: The processing logic of the compensation quantization module is as follows: The initial gain amplifies the weighted contributions of materials and processes, the compensation direction is distinguished by the sign function, and then multiplied with the absolute value of the relative deviation of the die-cutting pressure. The urgency of compensation is quantified by the pressure deviation, making the compensation more consistent with the actual die-cutting pressure state.

8. The intensive gradient composite molding collaborative control method according to claim 7, characterized in that: The iterative convergence logic of the data optimization component is such that the iteration terminates when one of the following two conditions is met: A1: When the absolute value of the die-cutting quality deviation in the i-th iteration is less than or equal to the preset threshold, the die-cutting accuracy is considered to be up to standard. A2: When the number of iterations reaches the preset maximum, the iteration will be forcibly terminated.

9. The intensive gradient composite molding collaborative control method according to claim 8, characterized in that: The die-cutting station (5) includes a die-cutting blade (6), an electromagnetic control component (7), and a bottom shaft component (8). After the dynamic depth compensation is output to the execution component, the execution component drives the electromagnetic control component (7) to operate, thereby controlling the die-cutting accuracy of the die-cutting blade (6).

Images