Microfilament finished product rewinding parameter optimization method and device, electronic equipment and medium

Abstract

The invention provides a micro-filament finished product rewinding parameter optimization method and device, electronic equipment and a medium, and the method comprises the steps: collecting a plurality of sets of technological parameters of micro-filament finished product rewinding and the actual stop point number of times when a preset pay-off length corresponds to each set of technological parameters, and constructing a training data set; based on a plurality of training data subsets of the training data set, respectively training pre-constructed parameter optimization models to obtain a plurality of trained parameter optimization models; constructing a parameter search space; the parameter search space comprises a plurality of groups of constructed process parameters; based on each trained parameter optimization model, respectively processing each group of process parameters in the parameter search space, determining predicted stop point number distribution corresponding to each group of process parameters, determining target EI values corresponding to each group of process parameters in the parameter search space, screening out a target process parameter group corresponding to the target EI values meeting a preset selection condition, and determining the predicted stop point number distribution corresponding to each group of process parameters in the parameter search space; and determining a parameter optimization result of finished product rewinding of the fine filaments, so that a globally optimal parameter combination can be automatically searched.

Description

Technical Field

[0001] The present application relates to the technical field of micro-filament manufacturing, and in particular to a method, device, electronic equipment and medium for optimizing the rewinding parameters of a finished micro-filament product. Background Art

[0002] The rewinding process of microfilaments is crucial to the quality and pay-off performance of the finished wire. During the rewinding process, parameters such as tension, speed, wire spacing, ball reduction, wire length and wire diameter have a significant impact on the pay-off performance after rewinding. How to reduce the number of stops during the pay-off process and improve the pay-off performance by optimizing these process parameters is a major challenge in the current field. Traditional optimization methods often rely on experience and experiments, and have the problem of low efficiency. Summary of the invention

[0003] In view of this, the purpose of the present application is to provide a method, device, electronic device and medium for optimizing the rewinding parameters of finished micro-filament products, which can automatically search for the globally optimal parameter combination, improve the optimization efficiency, significantly reduce the number of wire pay-off stop points, and improve production efficiency.

[0004] The present application provides a method for optimizing the parameters of a finished microfilament rewinding, the method comprising:

[0005] Collect multiple sets of process parameters for rewinding the finished micro-filaments and the actual number of stop points when paying out the preset length corresponding to each set of process parameters to construct a training data set;

[0006] Based on the multiple training data subsets of the training data set, the pre-built parameter optimization models are trained respectively to obtain multiple trained parameter optimization models;

[0007] Constructing a parameter search space; the parameter search space includes multiple groups of constructed process parameters;

[0008] Based on each trained parameter optimization model, each group of process parameters in the parameter search space is processed respectively to determine the distribution of predicted stop times corresponding to each group of process parameters in the parameter search space;

[0009] Determine the target EI value corresponding to each set of process parameters in the parameter search space based on the predicted stop point number distribution corresponding to each set of process parameters in the parameter search space;

[0010] The target process parameter group corresponding to the target EI value that meets the preset selection conditions is screened out from the process parameters in the parameter search space to determine the parameter optimization result of the rewinding of the micro-filament finished product.

[0011] In some embodiments, in the method for optimizing the parameters of the finished microfilament rewinding, the following steps are repeated until a preset iteration stop condition is reached:

[0012] After selecting the target process parameter group corresponding to the target EI value that meets the preset selection conditions, a wire laying test is performed based on the target process parameter group to obtain the actual number of stop points corresponding to the target process parameter group;

[0013] Update the training data set based on the target process parameter group and the actual number of stop points corresponding to the target process parameter group;

[0014] The parameter optimization model is trained based on the updated training data set, and the parameter optimization results of the rewinding of the micro-filament finished product are updated.

[0015] In some embodiments, in the method for optimizing the parameters of the finished microfilament rewinding, the method collects multiple sets of process parameters for the finished microfilament rewinding and the actual number of stop points when the preset length of the wire is released corresponding to each set of process parameters to construct a training data set; including:

[0016] Collect multiple sets of process parameters for rewinding the finished micro-filaments and the actual number of stops when paying out the preset length corresponding to each set of process parameters;

[0017] constructing fusion parameters based on at least some of the process parameters in each set of process parameters;

[0018] Each set of process parameters, the fusion parameters corresponding to each set of process parameters, and the actual number of stop points are taken as a sample to construct a training data set.

[0019] In some embodiments, in the method for optimizing the parameters of the finished microfilament rewinding, the fusion parameters are constructed based on at least part of the process parameters in each set of process parameters; including:

[0020] Calculate a first ratio and a sum of the wire diameter and the wire spacing;

[0021] Calculate a second ratio of the ball row reduction amount and the wire diameter to the sum;

[0022] The first ratio and the second ratio are used as fusion parameters.

[0023] In some embodiments, in the method for optimizing the parameters of the finished microfilament rewinding, the pre-built parameter optimization models are trained based on the multiple training data subsets of the training data set to obtain multiple trained parameter optimization models, including:

[0024] Sampling the training data set, and repeating the sampling multiple times to obtain multiple training data subsets of the same size but different samples;

[0025] The pre-built parameter optimization models are trained respectively based on the multiple training data subsets to obtain multiple trained parameter optimization models.

[0026] In some embodiments, in the method for optimizing the parameters of the finished microfilament rewinding, the loss function of the parameter optimization model includes an error evaluation term, a tension penalty term, and a speed penalty term;

[0027] The tension penalty term is determined based on a preset minimum safety value of the tension;

[0028] The speed penalty term is determined based on a preset minimum safe speed.

[0029] In some embodiments, in the method for optimizing the parameters of the finished microfilament rewinding, the loss function of the parameter optimization model is calculated by the following method:

[0030] Calculate the first product of the tension penalty item weight and the tension penalty item, and the second product of the speed penalty item weight and the speed penalty item;

[0031] Calculate the sum of the error evaluation term, the first product, and the second product.

[0032] In some embodiments, a device for optimizing the rewinding parameters of a finished microfilament product is also provided, the device comprising:

[0033] A collection module is used to collect multiple sets of process parameters for rewinding the finished micro-filaments and the actual number of stops when paying out the preset length corresponding to each set of process parameters, and to construct a training data set;

[0034] A training module, used for respectively training the pre-built parameter optimization models based on multiple training data subsets of the training data set to obtain multiple trained parameter optimization models;

[0035] A construction module, used to construct a parameter search space; the parameter search space includes multiple groups of constructed process parameters;

[0036] A processing module, configured to process each set of process parameters in the parameter search space based on each trained parameter optimization model, and determine a distribution of predicted stop times corresponding to each set of process parameters in the parameter search space;

[0037] A first determination module is used to determine a target EI value corresponding to each set of process parameters in the parameter search space based on a distribution of predicted stop times corresponding to each set of process parameters in the parameter search space;

[0038] The second determination module is used to select a target process parameter group corresponding to a target EI value that meets a preset selection condition from the process parameters in the parameter search space, and determine a parameter optimization result for rewinding the micro-filament finished product.

[0039] In some embodiments, an electronic device is also provided, comprising: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate through the bus, and when the machine-readable instructions are executed by the processor, the steps of the method for optimizing the rewinding parameters of the finished micro-filament product are performed.

[0040] In some embodiments, there is also provided a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the method for optimizing the rewinding parameters of the finished micro-filament product are executed.

[0041] The embodiment of the present application provides a method, device, electronic device and medium for optimizing the parameters of the rewinding of a finished micro-filament product, which collects multiple groups of process parameters of the rewinding of the finished micro-filament product and the actual number of stop points when the preset length of the wire is released corresponding to each group of process parameters, and constructs a training data set; trains pre-constructed parameter optimization models based on multiple training data subsets of the training data set, and obtains multiple trained parameter optimization models; constructs a parameter search space; the parameter search space includes multiple groups of constructed process parameters; processes each group of process parameters in the parameter search space based on each trained parameter optimization model, and determines the distribution of the predicted number of stop points corresponding to each group of process parameters in the parameter search space; based on the distribution of the predicted number of stop points corresponding to each group of process parameters in the parameter search space, Determine the target EI value corresponding to each group of process parameters in the parameter search space; select the target process parameter group corresponding to the target EI value that meets the preset selection conditions from the process parameters in the parameter search space, and determine the parameter optimization result of the rewinding of the micro-filament finished product, so as to automatically search for the global optimal parameter combination, and accurately model the nonlinear relationship between each process parameter and the number of pay-off stop points through the model, so as to effectively optimize the process parameter combination, significantly reduce the number of pay-off stop points, and improve production efficiency; combine the efficiency function to iteratively search for the global optimal parameters, reduce the number of manual repeated experiments and adjustments, and greatly improve the optimization efficiency; improve the parameter search efficiency through adaptive adjustment of the efficiency function, especially under complex process conditions, and quickly find the best process combination. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0043] Figure 1 A flow chart showing the method for optimizing the parameters of the finished microfilament rewinding according to an embodiment of the present application is shown;

[0044] Figure 2 A flow chart of a method for constructing a training data set according to an embodiment of the present application is shown;

[0045] Figure 3 A flow chart of a method for respectively training pre-built parameter optimization models based on multiple training data subsets of the training data set according to an embodiment of the present application is shown;

[0046] Figure 4 A flow chart showing another method for optimizing parameters of rewinding of finished microfilament products according to an embodiment of the present application is shown;

[0047] Figure 5 A schematic diagram of the structure of the device for optimizing the rewinding parameters of the finished microfilament according to an embodiment of the present application is shown;

[0048] Figure 6 A schematic structural diagram of an electronic device described in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0049] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It should be understood that the drawings in the present application only serve the purpose of explanation and description and are not used to limit the scope of protection of the present application. In addition, it should be understood that the schematic drawings are not drawn in real proportion. The flowchart used in this application shows the operations implemented according to some embodiments of the present application. It should be understood that the operations of the flowchart can be implemented out of sequence, and the steps without logical context can be reversed in order or implemented simultaneously. In addition, those skilled in the art can add one or more other operations to the flowchart under the guidance of the content of the present application, or remove one or more operations from the flowchart.

[0050] In addition, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

[0051] It should be noted that the term "comprising" will be used in the embodiments of the present application to indicate the existence of the features declared thereafter, but does not exclude the addition of other features.

[0052] The rewinding process of microfilaments is crucial to the quality and pay-off performance of the finished wire. During the rewinding process, parameters such as tension, speed, wire spacing, ball reduction, wire length and wire diameter have a significant impact on the pay-off performance after rewinding. How to reduce the number of stops during the pay-off process and improve the pay-off performance by optimizing these process parameters is a major challenge in the current field. Traditional optimization methods often rely on experience and experiments, and have the problem of low efficiency.

[0053] Moreover, due to the complexity of the process and the significant mutual influence between parameters, empirical adjustments often cannot guarantee the best process results, especially in terms of improving the stability of the pay-off process and reducing the number of stops. Too many stops not only affect the production efficiency during use, but may also lead to a decline in the quality of the finished product.

[0054] Based on this, the embodiments of the present application provide a method, device, electronic device and medium for optimizing the rewinding parameters of finished micro-filament products, which collect multiple groups of process parameters for rewinding finished micro-filament products and the actual number of stop points when releasing a preset length corresponding to each group of process parameters, and construct a training data set; based on multiple training data subsets of the training data set, respectively train pre-constructed parameter optimization models to obtain multiple trained parameter optimization models; construct a parameter search space; the parameter search space includes multiple groups of constructed process parameters; based on each trained parameter optimization model, respectively process each group of process parameters in the parameter search space, and determine the distribution of the predicted number of stop points corresponding to each group of process parameters in the parameter search space; based on the predicted number of stop points corresponding to each group of process parameters in the parameter search space, the predicted number of stop points corresponding to each group of process parameters in the parameter search space is determined. The target EI value corresponding to each set of process parameters in the parameter search space is determined; the target process parameter group corresponding to the target EI value that meets the preset selection conditions is screened out from the process parameters in the parameter search space, and the parameter optimization result of the rewinding of the micro-filament finished product is determined, so that the global optimal parameter combination can be automatically searched. The nonlinear relationship between each process parameter and the number of pay-off stop points is accurately modeled by the model, and the process parameter combination can be effectively optimized, the number of pay-off stop points is significantly reduced, and the production efficiency is improved; the iterative search for the global optimal parameters is carried out in combination with the efficiency function, which reduces the number of manual repeated experiments and adjustments and greatly improves the optimization efficiency; the parameter search efficiency is improved through the adaptive adjustment of the efficiency function, especially under complex process conditions, the optimal process combination can be quickly found.

[0055] Please refer to Figure 1 , Figure 1 The flowchart of the method for optimizing the rewinding parameters of the microfilament finished product according to the embodiment of the present application is shown as follows: Figure 1 As shown, the method for optimizing the rewinding parameters of the finished microfilament product includes the following steps S101-S105:

[0056] S101, collecting multiple sets of process parameters for rewinding the finished micro-filaments and the actual number of stop points when paying out a preset length corresponding to each set of process parameters, and constructing a training data set;

[0057] S102, training pre-built parameter optimization models based on multiple training data subsets of the training data set respectively to obtain multiple trained parameter optimization models;

[0058] S103, constructing a parameter search space; the parameter search space includes multiple groups of constructed process parameters;

[0059] S104, processing each group of process parameters in the parameter search space respectively based on each trained parameter optimization model, and determining the distribution of predicted stopping times corresponding to each group of process parameters in the parameter search space;

[0060] S105, determining a target EI value corresponding to each set of process parameters in the parameter search space based on a distribution of predicted stop times corresponding to each set of process parameters in the parameter search space;

[0061] S106 , selecting a target process parameter group corresponding to a target EI value that meets a preset selection condition from the process parameters in the parameter search space, and determining a parameter optimization result for rewinding the microfilament finished product.

[0062] In step S101, multiple sets of process parameters for rewinding the finished microfilament and the actual number of stop points when paying out the preset length corresponding to each set of process parameters are collected to construct a training data set.

[0063] Finished microfilament products refer to microfilament materials that have undergone certain processing or production. These materials may be used in various applications such as electronics, textiles, medical, etc.

[0064] Rewinding generally refers to the process of taking a material (in this case, microfilament) off one reel or shaft and rewinding it onto another. This process may involve adjusting parameters such as tension and speed to ensure that the material is not damaged during the rewinding process and maintains specific quality and properties.

[0065] Process parameters refer to various settings or conditions that affect the production process. Different process parameters affect the quality and performance of microfilaments during the rewinding process.

[0066] In the process of reaching the preset pay-off length, the machine or process may stop due to various reasons (such as material breakage, abnormal tension, etc.). The "actual number of stops" here refers to the actual number of stops that occur in the process required to reach the preset pay-off length under each set of specific process parameters; the fewer the number of stops, the more stable the rewinding process is, and the smaller the impact on the quality and performance of the microfilament.

[0067] Each group of process parameters includes the following process parameters: rewinding tension, speed, wire spacing, wire pitch, ball reduction, winding length, and wire diameter.

[0068] As an example only, the preset length of the line release is 1000m.

[0069] The training dataset is a data set used to train a model or algorithm, which usually includes input features and output labels.

[0070] In the embodiment of the present application, the input feature is a process parameter, and the output label is the actual number of stop points.

[0071] Collect multiple sets of process parameters for rewinding of micro-filament finished products and the actual number of stops when releasing the preset length corresponding to each set of process parameters. Specifically, through historical data or a small amount of initial experiments, collect rewinding tension, speed, wire spacing, wire pitch, ball reduction, winding length, wire diameter, and the corresponding actual number of stops when releasing 1000 meters of wire.

[0072] Please refer to Figure 2 The method comprises collecting multiple sets of process parameters for rewinding the finished micro-filaments and the actual number of stop points when releasing the preset length corresponding to each set of process parameters to construct a training data set; comprising the following steps S201-S203:

[0073] S201, collecting multiple sets of process parameters for rewinding the finished microfilament product and the actual number of stop points when releasing the preset length corresponding to each set of process parameters;

[0074] S202, constructing fusion parameters based on at least part of the process parameters in each group of process parameters;

[0075] S203, taking each set of process parameters, the fusion parameters corresponding to each set of process parameters and the actual number of stop points as a sample to construct a training data set.

[0076] Specifically, constructing fusion parameters based on at least part of the process parameters in each set of process parameters includes:

[0077] Calculate a first ratio and a sum of the wire diameter and the wire spacing;

[0078] Calculate a second ratio of the ball row reduction amount and the wire diameter to the sum;

[0079] The first ratio and the second ratio are used as fusion parameters.

[0080] The fusion parameter may also be called a derived parameter or a derived feature.

[0081] Specifically, based on domain knowledge, fusion features are constructed: Where φ is the wire diameter, mm; R is the ball row reduction, mm; g is the row wire spacing, mm; r1 represents the first fusion feature; r2 represents the second fusion feature.

[0082] The fusion parameters can reflect the comprehensive impact of key process parameters such as wire diameter, wire spacing, ball row reduction, etc. on the rewinding process performance, thereby guiding parameter optimization in the production process and improving the overall performance of the product.

[0083] Based on this, in the training data set, the rewinding tension (T), speed (V), wire spacing (g), wire pitch (S), ball reduction (R), winding length (L), wire diameter (φ) and fusion features r1 and r2 are used as input parameters of the model, and the actual number of stops for 1000 meters of wire pay-off is used as the output parameter, which is the optimization target parameter.

[0084] When constructing the training data set, the input parameters need to be preprocessed.

[0085] The input parameters are subjected to data preprocessing, specifically as follows: 1) the missing data are filled with the mean value; 2) the input data are normalized to convert the value of each input parameter into the range of [0, 1].

[0086] Specifically, the normalization formula is: In the formula, Q represents the normalized data, x represents the original input data, and x max Characterizes the maximum value of the input data, x min Characterizes the minimum value of the input data.

[0087] After the training data set is constructed, the process parameters of the rewinding process can be optimized.

[0088] In the step S102, the pre-built parameter optimization models are trained respectively based on the multiple training data subsets of the training data set to obtain multiple trained parameter optimization models.

[0089] The parameter optimization model may adopt a neural network model, a random forest model or the like.

[0090] When the pre-constructed parameter optimization models are trained respectively based on multiple training data subsets of the training data set to obtain multiple trained parameter optimization models, specifically, the input parameters of the training data subsets are input into the pre-constructed parameter optimization model, the parameter optimization model processes the input parameters, determines the corresponding output results, and optimizes the parameter optimization model based on the output results and the loss function until the preset training stop condition is met to obtain multiple trained parameter optimization models.

[0091] The preset training stop condition may be reaching a preset number of training times, etc.

[0092] In some optional embodiments, the loss function of the parameter optimization model includes an error evaluation term, a tension penalty term, and a speed penalty term;

[0093] The tension penalty term is determined based on a preset minimum safety value of the tension;

[0094] The speed penalty term is determined based on a preset minimum safe speed.

[0095] Specifically, the loss function of the parameter optimization model is calculated by the following method:

[0096] Calculate the first product of the tension penalty item weight and the tension penalty item, and the second product of the speed penalty item weight and the speed penalty item;

[0097] Calculate the sum of the error evaluation term, the first product, and the second product.

[0098] The tension penalty term and the speed penalty term may be the same or different.

[0099] When the tension penalty term and the speed penalty term are the same, they can be represented by a symbol λ.

[0100] Specifically, the error evaluation item adopts mean square error, and the mean square error

[0101] Where n is the number of samples in the training data subset; y i represents the actual number of stops in the test, Characterizes the number of prediction stop points.

[0102] The loss function L = MSE + λ·(L tension +L speed ), L tension Represents the tension penalty term: L speed Represents the speed penalty term: Among them, T i represents the rewinding tension in the i-th sample, g; T min V represents the preset minimum safety value of tension, for example, 0.5g can be selected; too low tension may increase the risk of wire collapse, so the tension penalty term is increased; V i represents the winding speed in the i-th sample, m / min; V min It represents the preset minimum safety speed, for example, 60m / min can be selected; if the rewinding speed is too low, the production efficiency will be reduced, so the speed penalty item is increased; λ represents the penalty item weight.

[0103] In some optional embodiments, the value of λ is 0.01-0.2, and the specific value is determined by combining cross-validation and grid search to determine the optimal value.

[0104] Please refer to Figure 3 The method of training the pre-built parameter optimization models based on the multiple training data subsets of the training data set to obtain multiple trained parameter optimization models includes the following steps S301-S302:

[0105] S301, performing sampling processing on the training data set, repeating the sampling multiple times to obtain multiple training data subsets of the same size but different samples;

[0106] S302: training pre-built parameter optimization models based on the multiple training data subsets respectively to obtain multiple trained parameter optimization models.

[0107] Exemplarily, the training data set is sampled using the bootstrap method, the number of samples is consistent with the size of the initial data set, and the sampling is repeated 1000 times to obtain 1000 data subsets of the same size but different samples, and the data subsets are used to train to obtain 1000 different parameter optimization models.

[0108] In the step S103, a parameter search space is constructed; the parameter search space includes multiple groups of constructed process parameters.

[0109] The process parameter search space is as follows: rewinding tension [0.1-3g] step size 0.1g, speed [20-200m / min] step size 10m / min, wire arrangement interval (g) [0.01-1mm] step size 0.005mm, wire arrangement pitch [1-10mm] step size 0.1mm, ball row reduction [0.005-0.1mm] step size 0.002, winding length [100-2000m] step size 100, wire diameter [0.01-0.1mm] step size 0.001mm.

[0110] The fused features are obtained by calculation.

[0111] Based on the parameter search space, a large number of parameter combinations can be globally searched, so as to subsequently determine the globally optimal parameter combination.

[0112] In the step S104, each group of process parameters in the parameter search space is processed separately based on each trained parameter optimization model, and the predicted number of stop points corresponding to each group of process parameters in the parameter search space is determined; in the step S105, based on the predicted number of stop points corresponding to each group of process parameters in the parameter search space, the target EI value corresponding to each group of process parameters in the parameter search space is determined.

[0113] The stop point targets corresponding to all process parameter groups in the parameter search space are predicted, and 1000 models can obtain 1000 prediction results for each process parameter group.

[0114] The EI value corresponding to each process parameter group in the parameter search space is calculated based on the following formula:

[0115]

[0116] Among them, y represents the current objective function value to be evaluated, μ * represents the minimum number of actual stop points in the training data set, μ * -y represents the possible improvement of the target parameter; P(y / x′) represents the distribution of the predictions of each candidate target parameter; x′ represents the candidate data sample; represents the standard normal distribution of the predicted target parameter, φ(z) is the cumulative distribution function of the predicted target parameter; z = (μ * -μ) / σ, which represents the discreteness of the prediction of the target parameter; μ represents the mean of the number of stop points predicted by the machine learning model, and σ represents the variance of the number of stop points predicted by the machine learning model; Characterize uncertainty estimates of forecast results; Characterizes the expected improvement in the number of stops.

[0117] Here, the prediction target parameter is the predicted number of stop points.

[0118] In step S105, a target process parameter group corresponding to a target EI value that meets a preset selection condition is selected from the process parameters in the parameter search space to determine a parameter optimization result for rewinding the finished microfilament product.

[0119] The target EI value that meets the preset selection condition may be a target EI value that is greater than a preset optimization threshold, a preset number of target EI values ​​that are ranked first, and so on.

[0120] As an example only, the process parameters corresponding to the three largest EI values ​​are screened out as the target process parameters.

[0121] Please refer to Figure 4 In some embodiments, in the method for optimizing the parameters of the finished microfilament rewinding, the following steps S401-S403 are repeated until a preset iteration stop condition is reached:

[0122] S401, after screening out a target process parameter group corresponding to a target EI value that meets a preset selection condition, performing a wire laying test based on the target process parameter group to obtain an actual number of stop points corresponding to the target process parameter group;

[0123] S402, updating the training data set based on the target process parameter group and the actual number of stop points corresponding to the target process parameter group;

[0124] S403: training a parameter optimization model based on the updated training data set, and updating the parameter optimization result of the rewinding of the microfilament finished product.

[0125] That is, a target process parameter experiment is performed; the data obtained from the experiment is added to the updated data set, and the parameter optimization model is further optimized, and it is continuously iterated until the optimal process parameter combination is found.

[0126] The method for optimizing the parameters of the finished microfilament rewinding described in the embodiment of the present application was successfully applied to a microfilament rewinding production line. The number of stop points for 1000 meters of pay-off was reduced from 5 to 2 in just 4 iterations, which significantly improved the pay-off performance of the microfilament and significantly improved the quality of the finished product. The time for manual adjustment of process parameters was reduced, and the production cost was reduced.

[0127] Through the optimization method described in the embodiments of the present application, the optimization efficiency of the rewinding process is significantly improved, which is mainly manifested in: the nonlinear relationship between each process parameter and the number of stop points in the wire pay-off is accurately modeled through the random forest model, which can effectively optimize the process parameter combination, so that the number of stop points for 1000 meters of wire pay-off is significantly reduced, thereby improving production efficiency; the global optimal parameter is iteratively searched in combination with the efficiency function, reducing the number of manual repeated experiments and adjustments; the parameter search efficiency is improved through adaptive adjustment of the efficiency function, especially under complex process conditions, the best process combination can be quickly found; the optimal rewinding parameters can be found through 3 to 5 iterations; the system can automatically optimize the process parameters according to the initial experimental data, reduce manual intervention, realize intelligent process adjustment, and greatly reduce human errors and time costs in production; the fusion features constructed based on domain knowledge can better capture the complex interactions between process parameters, improve the prediction accuracy of the model, and thus further reduce the number of stop points.

[0128] Based on the same inventive concept, the embodiment of the present application also provides a device for optimizing the rewinding parameters of a finished micro-filament corresponding to the method for optimizing the rewinding parameters of a finished micro-filament. Since the principle of solving the problem by the device in the embodiment of the present application is similar to the method for optimizing the rewinding parameters of a finished micro-filament described above in the embodiment of the present application, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be repeated.

[0129] Please refer to Figure 5 , Figure 5 The schematic diagram of the structure of the device for optimizing the rewinding parameters of the finished microfilament according to the embodiment of the present application is shown, and the device comprises:

[0130] The acquisition module 501 is used to acquire multiple sets of process parameters for rewinding the finished micro-filaments and the actual number of stop points when releasing the preset length corresponding to each set of process parameters, and construct a training data set;

[0131] A training module 502 is used to train pre-built parameter optimization models based on multiple training data subsets of the training data set to obtain multiple trained parameter optimization models; each trained parameter optimization model corresponds to an uncertainty;

[0132] A construction module 503 is used to construct a parameter search space; the parameter search space includes multiple sets of constructed process parameters;

[0133] A processing module 504 is used to process each group of process parameters in the parameter search space based on each trained parameter optimization model, and determine the distribution of predicted stop times corresponding to each group of process parameters in the parameter search space;

[0134] A first determination module 505 is used to determine a target EI value corresponding to each set of process parameters in the parameter search space based on a distribution of predicted stop times corresponding to each set of process parameters in the parameter search space;

[0135] The second determination module 506 is used to select a target process parameter group corresponding to a target EI value that meets a preset selection condition from the process parameters in the parameter search space, and determine a parameter optimization result for rewinding the finished microfilament product.

[0136] In some embodiments, in the microfilament finished product rewinding parameter optimization device, the training module is further used to repeat the following steps until a preset iteration stop condition is reached:

[0137] After selecting the target process parameter group corresponding to the target EI value that meets the preset selection conditions, a wire laying test is performed based on the target process parameter group to obtain the actual number of stop points corresponding to the target process parameter group;

[0138] Update the training data set based on the target process parameter group and the actual number of stop points corresponding to the target process parameter group;

[0139] The parameter optimization model is trained based on the updated training data set, and the parameter optimization results of the rewinding of the micro-filament finished product are updated.

[0140] In some embodiments, in the device for optimizing the parameters of the finished microfilament rewinding, the acquisition module, when collecting multiple sets of process parameters of the finished microfilament rewinding and the actual number of stops when paying out the preset length of the wire corresponding to each set of process parameters, is used to construct a training data set:

[0141] Collect multiple sets of process parameters for rewinding the finished micro-filaments and the actual number of stops when paying out the preset length corresponding to each set of process parameters;

[0142] constructing fusion parameters based on at least some of the process parameters in each set of process parameters;

[0143] Each set of process parameters, the fusion parameters corresponding to each set of process parameters, and the actual number of stop points are taken as a sample to construct a training data set.

[0144] In some embodiments, in the microfilament finished product rewinding parameter optimization device, the acquisition module, when constructing the fusion parameters based on at least part of the process parameters in each group of process parameters, is specifically used to:

[0145] Calculate a first ratio and a sum of the wire diameter and the wire spacing;

[0146] Calculate a second ratio of the ball row reduction amount and the wire diameter to the sum;

[0147] The first ratio and the second ratio are used as fusion parameters.

[0148] In some embodiments, in the microfilament finished product rewinding parameter optimization device, the training module, when training pre-built parameter optimization models based on multiple training data subsets of the training data set to obtain multiple trained parameter optimization models, is specifically used to:

[0149] Sampling the training data set, and repeating the sampling multiple times to obtain multiple training data subsets of the same size but different samples;

[0150] The pre-built parameter optimization models are trained respectively based on the multiple training data subsets to obtain multiple trained parameter optimization models.

[0151] In some embodiments, in the microfilament finished product rewinding parameter optimization device, the loss function of the parameter optimization model includes an error evaluation term, a tension penalty term, and a speed penalty term;

[0152] The tension penalty term is determined based on a preset minimum safety value of the tension;

[0153] The speed penalty term is determined based on a preset minimum safe speed.

[0154] In some embodiments, in the microfilament finished product rewinding parameter optimization device, the loss function of the parameter optimization model is calculated by the following method:

[0155] Calculate the first product of the tension penalty item weight and the tension penalty item, and the second product of the speed penalty item weight and the speed penalty item;

[0156] Calculate the sum of the error evaluation term, the first product, and the second product.

[0157] By means of the optimization device described in the embodiment of the present application, the optimization efficiency of the rewinding process is significantly improved, which is mainly manifested in the following aspects: the nonlinear relationship between each process parameter and the number of stop points in the wire pay-off is accurately modeled by the random forest model, and the process parameter combination can be effectively optimized, so that the number of stop points for 1000 meters of wire pay-off is significantly reduced, thereby improving production efficiency; the global optimal parameter is iteratively searched in combination with the efficiency function, thereby reducing the number of manual repeated experiments and adjustments; the parameter search efficiency is improved through adaptive adjustment of the efficiency function, and the optimal process combination can be quickly found, especially under complex process conditions; the optimal rewinding parameters can be found through 3 to 5 iterations; the system can automatically optimize the process parameters according to the initial experimental data, reducing manual intervention, realizing intelligent process adjustment, and greatly reducing human errors and time costs in production; the fusion features constructed based on domain knowledge can better capture the complex interactions between process parameters, improve the prediction accuracy of the model, and thus further reduce the number of stop points.

[0158] Based on the same inventive concept, an electronic device corresponding to the method for optimizing the rewinding parameters of the finished micro-filament product is also provided in the embodiment of the present application. Since the principle of solving the problem by the electronic device in the embodiment of the present application is similar to the method for optimizing the rewinding parameters of the finished micro-filament product described above in the embodiment of the present application, the implementation of the electronic device can refer to the implementation of the method, and the repeated parts will not be repeated.

[0159] Please refer to Figure 6 , Figure 6 A schematic diagram of the structure of the electronic device described in an embodiment of the present application is shown; the electronic device 600 includes: a processor 602, a memory 601 and a bus, the memory 601 stores machine-readable instructions executable by the processor 602, and when the electronic device 600 is running, the processor 602 communicates with the memory 601 through the bus, and when the machine-readable instructions are executed by the processor 602, the steps of the micro-filament finished product rewinding parameter optimization method are executed.

[0160] Based on the same inventive concept, a computer-readable storage medium corresponding to the method for optimizing the rewinding parameters of the finished micro-filament product is also provided in the embodiment of the present application. Since the principle of solving the problem by the computer-readable storage medium in the embodiment of the present application is similar to the method for optimizing the rewinding parameters of the finished micro-filament product described above in the embodiment of the present application, the implementation of the computer-readable storage medium can refer to the implementation of the method, and the repeated parts will not be repeated.

[0161] A computer-readable storage medium stores a computer program, which, when executed by a processor, executes the steps of the method for optimizing the rewinding parameters of a finished microfilament product.

[0162] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, the specific working process of the system and device described above can refer to the corresponding process in the method embodiment, and will not be repeated in this application. In the several embodiments provided in this application, it should be understood that the disclosed system, device and method can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0163] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0164] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0165] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium that is executable by a processor. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a platform server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard drives, ROM, RAM, magnetic disks, or optical disks.

[0166] The above are only specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for optimizing the rewinding parameters of a finished microfilament product, characterized in that: The method comprises: Collect multiple sets of process parameters for rewinding the finished micro-filaments and the actual number of stop points when paying out the preset length corresponding to each set of process parameters to construct a training data set; Based on the multiple training data subsets of the training data set, the pre-built parameter optimization models are trained respectively to obtain multiple trained parameter optimization models; Constructing a parameter search space; the parameter search space includes multiple groups of constructed process parameters; Based on each trained parameter optimization model, each group of process parameters in the parameter search space is processed respectively to determine the distribution of predicted stop times corresponding to each group of process parameters in the parameter search space; Determine the target EI value corresponding to each set of process parameters in the parameter search space based on the predicted stop point number distribution corresponding to each set of process parameters in the parameter search space; The target process parameter group corresponding to the target EI value that meets the preset selection conditions is screened out from the process parameters in the parameter search space to determine the parameter optimization result of the rewinding of the micro-filament finished product.

2. The method for optimizing the rewinding parameters of the finished microfilament product according to claim 1, characterized in that: Repeat the following steps until the preset iteration stop condition is met: After selecting the target process parameter group corresponding to the target EI value that meets the preset selection conditions, a wire laying test is performed based on the target process parameter group to obtain the actual number of stop points corresponding to the target process parameter group; Update the training data set based on the target process parameter group and the actual number of stop points corresponding to the target process parameter group; The parameter optimization model is trained based on the updated training data set, and the parameter optimization results of the rewinding of the micro-filament finished product are updated.

3. The method for optimizing the rewinding parameters of the finished microfilament product according to claim 1, characterized in that: The method collects multiple sets of process parameters for rewinding the finished micro-filaments and the actual number of stop points when releasing the preset length corresponding to each set of process parameters to construct a training data set; including: Collect multiple sets of process parameters for rewinding the finished micro-filaments and the actual number of stops when paying out the preset length corresponding to each set of process parameters; constructing fusion parameters based on at least some of the process parameters in each set of process parameters; Each set of process parameters, the fusion parameters corresponding to each set of process parameters, and the actual number of stop points are taken as a sample to construct a training data set.

4. The method for optimizing the rewinding parameters of the finished microfilament product according to claim 3, characterized in that: Constructing fusion parameters based on at least part of the process parameters in each set of process parameters; comprising: Calculate a first ratio and a sum of the wire diameter and the wire spacing; Calculate a second ratio of the ball row reduction amount and the wire diameter to the sum; The first ratio and the second ratio are used as fusion parameters.

5. The method for optimizing the rewinding parameters of the finished microfilament product according to claim 1 or 3, characterized in that: The pre-built parameter optimization models are trained respectively based on the multiple training data subsets of the training data set to obtain multiple trained parameter optimization models, including: Sampling the training data set, and repeating the sampling multiple times to obtain multiple training data subsets of the same size but different samples; The pre-built parameter optimization models are trained respectively based on the multiple training data subsets to obtain multiple trained parameter optimization models.

6. The method for optimizing the rewinding parameters of the finished microfilament product according to claim 1, characterized in that: The loss function of the parameter optimization model includes an error evaluation term, a tension penalty term, and a speed penalty term; The tension penalty term is determined based on a preset minimum safety value of the tension; The speed penalty term is determined based on a preset minimum safe speed.

7. The method for optimizing the rewinding parameters of the finished microfilament product according to claim 6, characterized in that: The loss function of the parameter optimization model is calculated by the following method: Calculate the first product of the tension penalty item weight and the tension penalty item, and the second product of the speed penalty item weight and the speed penalty item; Calculate the sum of the error evaluation term, the first product, and the second product.

8. A device for optimizing the rewinding parameters of finished microfilament products, characterized in that: The device comprises: A collection module is used to collect multiple sets of process parameters for rewinding the finished micro-filaments and the actual number of stop points when paying out the preset length corresponding to each set of process parameters, and to construct a training data set; A training module, used for respectively training the pre-built parameter optimization models based on multiple training data subsets of the training data set to obtain multiple trained parameter optimization models; A construction module, used to construct a parameter search space; the parameter search space includes multiple groups of constructed process parameters; A processing module, configured to process each set of process parameters in the parameter search space based on each trained parameter optimization model, and determine a distribution of predicted stop times corresponding to each set of process parameters in the parameter search space; A first determination module is used to determine a target EI value corresponding to each set of process parameters in the parameter search space based on a distribution of predicted stop times corresponding to each set of process parameters in the parameter search space; The second determination module is used to select a target process parameter group corresponding to a target EI value that meets a preset selection condition from the process parameters in the parameter search space, and determine a parameter optimization result for rewinding the micro-filament finished product.

9. An electronic device, characterized in that include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate via the bus, and when the machine-readable instructions are executed by the processor, the steps of the method for optimizing the rewinding parameters of the finished micro-filament product as described in any one of claims 1 to 7 are performed.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which, when executed by a processor, executes the steps of the method for optimizing the rewinding parameters of a finished micro-filament product as claimed in any one of claims 1 to 7.

Images