Parameter intelligent optimization method and system for stay wire film slitting

Abstract

The invention relates to the field of parameter optimization, in particular to an intelligent parameter optimization method and system for stay wire film slitting, and the method comprises the steps: obtaining a technological parameter group including winding tension, unwinding tension and slitting speed; a fitness function for evaluating the excellence degree of the technological parameter set is constructed, the fitness function comprises a transverse wrinkle index and a dynamic impact factor, and the transverse wrinkle index is in positive correlation with the difference value of the winding tension and the unwinding tension; the dynamic impact factor is in positive correlation with the slitting speed; based on the fitness function, using a particle swarm optimization algorithm to perform iterative optimization on the process parameter set to obtain a global optimal position; and taking the global optimal position as a final process parameter. According to the invention, the slitting effect of the stay wire film is improved.

Description

Technical Field

[0001] This invention relates to the field of parameter optimization, and in particular to a method and system for intelligent parameter optimization for wire drawing film slitting. Background Technology

[0002] Slitting of stretched film is a crucial step in its production and further processing. This process uses high-speed rotating cutters to precisely slit a wide master roll of film into multiple narrow sub-rolls. The slitting quality directly affects the yield and production efficiency of subsequent packaging, printing, and other processes. The process parameters for slitting, mainly including unwinding tension, winding tension, slitting speed, and cutter pressure, play a decisive role in the final slitting quality. Improper parameter settings can easily lead to quality defects such as wrinkles, uneven edges, and excessively tight or loose winding.

[0003] Currently, in actual production, when changing to new batches of film with different materials, thicknesses, or widths, the setting of process parameters mainly relies on manual settings by operators. Operators need to repeatedly test and fine-tune the machine according to the characteristics of the new material. The debugging process is time-consuming and labor-intensive, seriously affecting production efficiency.

[0004] Chinese patent document CN104062901B discloses a control system parameter optimization method based on orthogonal optimization particle swarm optimization algorithm. This control system parameter optimization method combines orthogonal optimization method with particle swarm optimization method. It determines the optimization range of each controller parameter, the sensitivity relationship between controller parameters and the weight relationship between various performance indicators through orthogonal experiments. On this basis, the particle swarm optimization algorithm is applied to perform optimization, and the optimization result is obtained through iterative iteration.

[0005] To improve slitting quality and automation, existing technologies typically employ machine learning models to predict slitting quality and combine them with intelligent optimization algorithms (such as particle swarm optimization, PSO) to find the optimal combination of process parameters. However, when applying PSO to parameter optimization in thin film slitting, it is prone to getting trapped in local optima, making it impossible to find a better parameter combination, thus affecting the optimization effect and resulting in poor slitting performance. Summary of the Invention

[0006] To address the problem that existing particle swarm optimization algorithms are prone to getting stuck in local optima, resulting in poor slitting performance, this invention provides an intelligent parameter optimization method and system for wire drawing film slitting.

[0007] In a first aspect, the present invention provides a method for intelligent parameter optimization for wire drawing film slitting, employing the following technical solution: Obtain a set of process parameters including winding tension, unwinding tension, and slitting speed; construct a fitness function to evaluate the quality of the process parameter set, the fitness function including a transverse wrinkle index and a dynamic impact factor, the transverse wrinkle index being positively correlated with the difference between winding tension and unwinding tension; the dynamic impact factor being positively correlated with slitting speed; based on the fitness function, use a particle swarm optimization algorithm to iteratively optimize the process parameter set to obtain the global optimal position; use the global optimal position as the final process parameters.

[0008] By constructing a fitness function containing a transverse wrinkling index and a dynamic impact factor, a mathematical model that can more comprehensively and accurately reflect the film slitting quality was established. This model not only considers the influence of the winding and unwinding tension difference on wrinkles but also quantifies the impact of slitting speed on edge quality. Based on this model, a particle swarm optimization algorithm is used for global optimization to find a combination of process parameters that achieves the optimal balance between production efficiency and product quality. This overcomes the shortcomings of manual setting, such as high subjectivity and low efficiency, as well as the insufficient optimization of conventional algorithms.

[0009] Preferably, the method for calculating the transverse wrinkle index is as follows: calculate the first product of the thickness, width and elastic modulus of the film material; calculate the difference between the winding tension and the unwinding tension, and the second product of the difference and the Poisson of the film material; and use the ratio of the second product to the first product as the transverse wrinkle index.

[0010] By combining the difference in unwinding and winding tension with the physical properties of the film itself, the quantitative assessment of wrinkle risk becomes more scientific and accurate, and it can be adapted to films of different specifications and materials.

[0011] Preferably, the dynamic impact factor is calculated as follows: the product of the density of the film material and the cutting speed is calculated; the difference between the preset reference elastic modulus and the elastic modulus of the film material is calculated; and the ratio of the product to the difference is used as the dynamic impact factor.

[0012] By combining slitting speed with the physical properties of the film material, the severity of the impact of the cutting tool on the film edge during high-speed slitting is scientifically quantified. This provides a clear quantitative indicator for evaluating edge quality, such as the risk of burrs and cracks, enabling the optimization process to more comprehensively weigh the efficiency gains and quality risks brought by speed, thereby finding a truly optimal speed.

[0013] Preferably, the expression for the fitness function is: ; ; In the formula, P represents the set of process parameters to be evaluated. Let V represent the fitness function with respect to the process parameter set P, V represent the slitting speed, I represent the transverse wrinkling index, and F represent the dynamic impact factor. Indicates winding tension. Indicates unwinding tension. The weighting coefficient represents the cutting speed. This represents the weighting coefficient of the horizontal wrinkle index. This represents the weighting coefficient of the dynamic impact factor.

[0014] By setting the cutting speed, representing production efficiency, as a reward, and the lateral wrinkling index and dynamic impact factor, representing the risk of quality defects, as penalty terms, and normalizing them using the tanh function, maximizing the fitness function value becomes equivalent to simultaneously pursuing high efficiency and high quality. Compared to simple linear weighting, the use of the tanh function helps balance the influence of different dimensional indicators, preventing any one indicator from excessively dominating the optimization direction, making the optimization process more stable and the results more balanced.

[0015] Preferably, the method for optimizing the process parameter set using the particle swarm optimization algorithm is as follows: the fitness function is used as the objective function of the particle swarm optimization algorithm, a particle swarm is randomly generated, and each particle represents a set of candidate process parameters; the position and moving speed of the particles are iteratively updated, and the fitness value of each particle is calculated in each iteration; each particle updates its speed and position based on its historical best position and the global best position of the entire swarm.

[0016] The preferred method for iteratively updating the particle's movement speed is as follows: ; ; in, For particles In the Inertia weights in the next iteration and Particles In the The second iteration and the first Movement speed in the next iteration and These are learning factors, namely, individual experience and social experience. and A random number between [0,1] For particles In the Position in the next iteration For the first In the next iteration, particles The optimal position of an individual For the first The global optimal position in the next iteration. This represents the preset maximum and minimum inertia weights, and K represents the preset total number of iterations.

[0017] The inertia weight decreases linearly from large to small, giving the algorithm a strong global exploration capability in the early stages of optimization, enabling it to extensively search the solution space to avoid missing optimal regions; while in the later stages of optimization, it has a strong local exploration capability, enabling it to perform fine-grained searches near the optimal solution to improve the accuracy of the solution. This effectively balances the algorithm's global and local search capabilities, improving convergence speed and the quality of the searched quantities.

[0018] Preferably, the method for optimizing the process parameter set using the particle swarm optimization algorithm further includes: setting a stagnation counter S and a stagnation threshold during the iterative optimization process. After each iteration, compare the current globally optimal fitness value with the fitness value of the previous iteration; if the fitness value has not increased, let S = S + 1; otherwise, reset S = 0; when S ≥ At that time, the globally optimal position of the current record. Apply a Gaussian perturbation to generate a new global optimum position. The new global optimal position Replace with Then proceed with the next iteration until the maximum number of iterations or the fitness value converges.

[0019] Preferably, the method for optimizing the process parameter set using the particle swarm optimization algorithm further includes: generating a global optimal position after each iteration, calculating the stability index of the global optimal position, using the difference between 1 and the stability index as the search radius decay factor, and adjusting the search radius using the radius decay factor to obtain the optimal search radius for the next iteration.

[0020] The search range for the next iteration is dynamically adjusted based on the stability index of the current optimal solution. If the current solution is stable (i.e., low risk of folding and shock), the search range is narrowed for more refined optimization; if the current solution is unstable (high risk), the search range is expanded to explore other safer regions. This intelligent adjustment mechanism enables the algorithm to achieve a dynamic balance between exploration and exploitation, making it more efficient and robust than fixed or unidirectional decreasing search strategies.

[0021] Preferably, the stability index is calculated as follows: the corresponding transverse wrinkle index and dynamic impact factor are calculated using the global optimal position, and the reciprocal of the sum of the transverse wrinkle index, the dynamic impact factor and 1 is used as the stability index.

[0022] By using the stability index, it is ensured that the adjustment of the search radius is based on a true assessment of the risk to the cutting quality, thus making the entire adaptive optimization process more scientific and effective.

[0023] Secondly, the present invention provides a parameter intelligent optimization system for wire drawing film slitting, which adopts the following technical solution: A parameter intelligent optimization system for wire drawing film slitting includes a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-described parameter intelligent optimization method for wire drawing film slitting is implemented.

[0024] The above-mentioned intelligent parameter optimization method for wire drawing film slitting is used to generate a computer program, which is stored in a memory so that it can be loaded and executed by a processor. Thus, a system can be made based on the memory and processor for convenient use.

[0025] The present invention has the following technical effects: This scheme constructs a fitness function that can simultaneously evaluate wrinkle risk and impact effect, and improves the traditional particle swarm optimization algorithm. First, it introduces a stagnation judgment and Gaussian perturbation mechanism to escape local optima. Second, it introduces an adaptive search radius adjustment strategy based on the stability index of the solution, so that the algorithm can efficiently and accurately find the globally optimal process parameters that take into account both production efficiency and slitting quality, avoid the particle swarm optimization algorithm from getting trapped in local optima, and improve the slitting effect of wire drawing films. Attached Figure Description

[0026] Figure 1 This is a flowchart of a method for intelligent parameter optimization for wire drawing film slitting according to the present invention. Detailed Implementation

[0027] 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, not all, of the embodiments of the present invention. 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.

[0028] This invention discloses an intelligent parameter optimization method for wire drawing film slitting, referring to... Figure 1 The process includes the following steps, as detailed below: S1: Obtain the process parameters of the slitting production line and the physical parameters of the packaging film.

[0029] Process parameters are collected in real time by sensors distributed on the production line. These parameters include winding tension, unwinding tension, and slitting speed. A parameter vector is constructed using multiple process parameters at the same time. The physical parameters of the film material are retrieved through the production management system. These physical parameters include film thickness, elastic modulus, Poisson's ratio, film width, and film density.

[0030] S2: Construct evaluation indicators for the quality of packaging film slitting, and construct a fitness function based on the evaluation indicators.

[0031] S21: Construct the horizontal wrinkle index.

[0032] During the slitting process, if the transverse compressive stress inside the film exceeds its critical value for resistance to instability, wrinkles are likely to occur. This compressive stress mainly originates from the Poisson effect, that is, when the film is subjected to longitudinal stretching, it will produce transverse shrinkage. When the transverse shrinkage is subject to external constraints such as roller friction, it will be transformed into internal compressive stress. The transverse wrinkle index is used to quantify the risk of wrinkles caused by tension.

[0033] The expression for the lateral wrinkle index is:

[0034] In the formula, I represents the transverse wrinkle index; Indicates winding tension; Indicates unwinding tension; Poisson's ratio represents the ratio of thin film materials; The thickness of the thin film material is expressed in meters (m); the width of the thin film material is expressed in meters (m); and the elastic modulus of the thin film material is expressed in Pa.

[0035] The greater the difference between the winding tension and the unwinding tension, the stronger the lateral shrinkage tendency caused by the Poisson effect, and the higher the risk of wrinkling. When the film material is thicker, wider, and has a larger elastic modulus, the film has stronger resistance to instability and is less prone to wrinkling under the same tension.

[0036] S22: Construct dynamic impact factors.

[0037] During high-speed slitting, the contact between the cutting tool and the thin film material is a dynamic impact process. High impact forces can cause burrs and cracks on the cut edges. This impact effect is closely related to the slitting speed and the toughness of the material itself. By constructing a dynamic impact factor, the dynamic impact intensity at the cutting edge during high-speed slitting can be quantified.

[0038]

[0039] In the formula, F represents the dynamic impact factor. Density of thin film materials, expressed in units of 1. V represents the cutting speed, in m / s; E This represents the elastic modulus of a thin film material, expressed in Pa. This represents the preset reference elastic modulus, in Pa. Its value is set manually according to the actual situation. For example, its value is 1 GPa, which is greater than the elastic modulus of the film material to ensure that the denominator is not 0. C represents the adjustment coefficient, with a value range of (0, 1), which is used to adjust the value range of the dynamic impact factor so that the value range of the dynamic impact factor is within the range of (0, 1).

[0040] According to the work-energy theorem, The larger the value, the greater the kinetic energy carried per unit mass of film, and the more intense the impact effect when colliding with the tool; it characterizes the toughness or energy absorption capacity of the material. The larger the elastic modulus, the harder and more brittle the material is, and the worse its ability to absorb impact energy through deformation, resulting in a larger value of dynamic impact factor; conversely, the smaller the elastic modulus, the softer the material is, and the stronger its ability to absorb impact energy through deformation, resulting in a smaller value of dynamic impact factor.

[0041] In summary, the dynamic impact factor represents the dynamic impact process between the cutting tool and the film material. The larger the value, the greater the impact force of the cutting tool on the film material, and the greater the possibility of poor slitting quality.

[0042] S23: Construct the fitness function.

[0043] The fitness function is expressed as follows:

[0044]

[0045] In the formula, P represents the set of process parameters to be evaluated. Let V represent the fitness function with respect to the process parameter set P, V represent the slitting speed, I represent the transverse wrinkling index, and F represent the dynamic impact factor. Indicates winding tension. Indicates unwinding tension. The weighting coefficient represents the cutting speed. This represents the weighting coefficient of the horizontal wrinkle index. This represents the weighting coefficient of the dynamic impact factor. The value of the weighting coefficient is set manually according to the actual situation, and the sum of the three weighting coefficients is 1.

[0046] This refers to the reward items in the film slitting process. During production, a higher cutting speed indicates a stronger production line capacity and higher work efficiency; while... The fitness function represents a penalty term in the film slitting process. Higher transverse wrinkle index and / or dynamic impact factor indicate poorer film slitting quality and lower yield. Therefore, the fitness function value is used to measure the excellence of a set of process parameters; a higher value indicates a better slitting effect produced by the corresponding set of process parameters.

[0047] S3: Use the particle swarm optimization algorithm for global optimization.

[0048] Using the fitness function as the objective function of the particle swarm optimization algorithm, the process parameters are initialized, and a particle swarm is randomly generated, with each particle representing a set of candidate process parameters. The position and velocity of the particles are iteratively updated, and the fitness value of each particle is calculated in each iteration. Each particle updates its velocity and position (i.e., updates the process parameters it represents) based on its individual historical best position and the global best position of the entire swarm.

[0049] In the In each iteration, the movement speed and position of all particles are updated. The update process is as follows: ; ; ; in, For particles In the Inertia weights in the next iteration and Particles In the The second iteration and the first Movement speed in the next iteration and These are learning factors, namely, individual experience and social experience. and A random number between [0,1] and Particles In the The second iteration and the first Position in the next iteration For the first In the next iteration, particles The optimal position of an individual For the first The global optimal position in the next iteration. This represents the preset maximum and minimum inertia weights, where the learning factor is the individual's experience. Learning factors of social experience The value is 2; random number and Maximum inertia weight is used to increase the search randomness of the particle swarm optimization algorithm. The value is 0.9, the minimum inertia weight. The value is 0.4, and K represents the preset total number of iterations, which is 100.

[0050] In the early stages of iterative optimization, a larger inertia weight is needed to enable particles to perform a wide-ranging search, exploring the entire parameter space. In the later stages of optimization, a smaller inertia weight is needed to allow particles to perform a fine-grained search near the optimal solution, accelerating convergence. Therefore, during iterative optimization, a stagnation counter S and a stagnation threshold are set. After each iteration, compare the current globally optimal fitness value with the fitness value from the previous iteration. If the fitness value has not increased, set S = S + 1; otherwise, reset S = 0. When S ≥ At that time, a perturbation mechanism is triggered, which involves adjusting the currently recorded globally optimal position. Apply a Gaussian perturbation to generate a new global optimum position. ,in: , It is a function with a mean of 0 and a standard deviation of . σ A Gaussian random number with a value of 0.1 is used to ensure that the perturbation is not too drastic, thus allowing for small-scale exploration around the current optimal solution. The new global optimum position is then determined. Replace with Then, the next iteration is performed until the maximum number of iterations or the fitness value converges, thus obtaining the final global optimum position.

[0051] A global optimal position is generated after each iteration, and the global optimal position is calculated. The stability index is expressed as:

[0052] In the formula, Indicates the first The stability index of the global optimal position in the next iteration. Indicates the first The transverse wrinkling index corresponding to the global optimal position in the next iteration. Indicates the first The dynamic impact factor corresponding to the global optimal position in the next iteration. A higher stability indicates better parameter stability in generating the global optimal position in the current iteration.

[0053] Then, based on stability, the search radius decay factor at iteration k+1 is calculated. Furthermore, the optimal search radius at iteration k+1 is obtained. :

[0054] in, This represents the optimal search radius at iteration k+1. These represent the preset maximum and minimum search radii, respectively. After obtaining the optimal search radius, during the k+1 iteration process, each particle is within the optimal search radius range. It should be noted that when the search range exceeds the limit range, the boundary of the limit range is forcibly taken.

[0055] If the overall process stability index corresponding to the currently found global optimum is high, it indicates a high-quality, low-risk process region. In this case, the search radius will automatically shrink, concentrating the particle swarm optimization's search range around this optimum. This allows the algorithm to perform a refined search within this favorable region, further optimizing performance, finding more precise optimal parameters, and achieving localized optimization.

[0056] If the current optimal solution may have a high fitness value (e.g., high speed) but a low stability index (meaning high folding or high impact risk), the search radius will automatically expand. This forces the particle swarm to jump out of the current dangerous local optimum and search for other safer and more stable feasible solutions in a wider range, achieving global exploration and effectively avoiding the algorithm from falling into a trap of seemingly optimal solutions that are actually unapplicable.

[0057] After the iteration is completed, the global optimal position (global optimal solution) is used as the final process parameter for cutting the packaging film, thereby realizing intelligent optimization of the parameters for cutting the string film.

[0058] This invention also discloses an intelligent parameter optimization system for wire drawing film slitting, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement an intelligent parameter optimization method for wire drawing film slitting according to the present invention.

[0059] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

[0060] The above are all preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Therefore, all equivalent changes made in accordance with the structure, shape and principle of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A method for intelligent parameter optimization in wire drawing film slitting, characterized in that, Including the following steps: Obtain a set of process parameters including winding tension, unwinding tension, and slitting speed; construct a fitness function to evaluate the quality of the process parameter set, the fitness function including a transverse wrinkle index and a dynamic impact factor, the transverse wrinkle index being positively correlated with the difference between winding tension and unwinding tension; the dynamic impact factor being positively correlated with slitting speed; based on the fitness function, use a particle swarm optimization algorithm to iteratively optimize the process parameter set to obtain the global optimal position; use the global optimal position as the final process parameters.

2. The intelligent parameter optimization method for wire drawing film slitting according to claim 1, characterized in that, The method for calculating the transverse wrinkle index is as follows: calculate the first product of the thickness, width and elastic modulus of the film material; calculate the difference between the winding tension and the unwinding tension, and the second product of the difference and the Poisson of the film material; and take the ratio of the second product to the first product as the transverse wrinkle index.

3. The intelligent parameter optimization method for wire drawing film slitting according to claim 1, characterized in that, The dynamic impact factor is calculated as follows: calculate the product of the density of the film material and the cutting speed; calculate the difference between the preset reference elastic modulus and the elastic modulus of the film material; and use the ratio of the product to the difference as the dynamic impact factor.

4. The intelligent parameter optimization method for wire drawing film slitting according to claim 3, characterized in that, The fitness function is expressed as follows: ； ； In the formula, P represents the set of process parameters to be evaluated. Let V represent the fitness function with respect to the process parameter set P, V represent the slitting speed, I represent the transverse wrinkling index, and F represent the dynamic impact factor. Indicates winding tension. Indicates unwinding tension. The weighting coefficient represents the cutting speed. This represents the weighting coefficient of the horizontal wrinkle index. This represents the weighting coefficient of the dynamic impact factor.

5. The intelligent parameter optimization method for wire drawing film slitting according to claim 1, characterized in that, The method for optimizing a set of process parameters using particle swarm optimization is as follows: the fitness function is used as the objective function of the particle swarm optimization algorithm, a particle swarm is randomly generated, and each particle represents a set of candidate process parameters; the position and movement speed of the particles are iteratively updated, and the fitness value of each particle is calculated in each iteration; each particle updates its speed and position based on its historical best position and the global best position of the entire swarm.

6. The intelligent parameter optimization method for wire drawing film slitting according to claim 5, characterized in that, The method for iteratively updating the particle's movement speed is as follows: ； ； in, For particles In the Inertia weights in the next iteration and Particles In the The second iteration and the first Movement speed in the next iteration and These are learning factors, namely, individual experience and social experience. and A random number between [0,1] For particles In the Position in the next iteration For the first In the next iteration, particles The optimal position of an individual For the first The global optimal position in the next iteration. This represents the preset maximum and minimum inertia weights, and K represents the preset total number of iterations.

7. The intelligent parameter optimization method for wire drawing film slitting according to claim 5, characterized in that, The method for optimizing the process parameter set using the particle swarm optimization algorithm further includes: setting a stagnation counter S and a stagnation threshold during the iterative optimization process. After each iteration, compare the current globally optimal fitness value with the fitness value of the previous iteration; if the fitness value has not increased, let S = S + 1; otherwise, reset S = 0; when S ≥ At that time, the globally optimal position of the current record. Apply a Gaussian perturbation to generate a new global optimum position. The new global optimal position Replace with Then proceed with the next iteration until the maximum number of iterations or the fitness value converges.

8. The intelligent parameter optimization method for wire drawing film slitting according to claim 1, characterized in that, The method of using particle swarm optimization algorithm to optimize process parameter sets also includes: generating a global optimal position after each iteration, calculating the stability index of the global optimal position, using the difference between 1 and the stability index as the search radius decay factor, and using the radius decay factor to adjust the search radius to obtain the optimal search radius for the next iteration.

9. The intelligent parameter optimization method for wire drawing film slitting according to claim 8, characterized in that, The stability index is calculated as follows: the corresponding transverse wrinkling index and dynamic impact factor are calculated using the global optimal position, and the reciprocal of the sum of the transverse wrinkling index, dynamic impact factor and 1 is used as the stability index.

10. A parameter intelligent optimization system for wire drawing film slitting, characterized in that, include: A processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement a method for intelligent parameter optimization for wire drawing film slitting according to any one of claims 1-9.

Images