Design method, device, equipment and storage medium for the core support body of stamping die

Abstract

This application relates to the field of automotive manufacturing process technology, and discloses a method, apparatus, equipment, and storage medium for designing the main body of a stamping die's support core. The method involves establishing a local coordinate system; obtaining a profile body by cutting through process supplementary surfaces; obtaining the working part through Boolean subtraction; constructing a conformal surface to generate a conformal body; and summing the working part and the conformal body through Boolean summation to obtain the initial support core structure. An LSSVM (Least Squares Support Vector Machine) nonlinear regression model is established using LSSVM, and the nonlinear regression model is used to predict the springback and surface roughness indices of the initial support core structure under the current process and structural parameters, obtaining the prediction results. Based on the prediction results, the process and structural parameters of the initial support core structure are updated using the NSGA-II algorithm to obtain the target support core structure. This method can automatically, efficiently, and accurately create the support core structure, significantly improving design efficiency.

Description

Technical Field

[0001] This application relates to the field of automotive manufacturing process design technology, and in particular to a design method, apparatus, equipment and storage medium for a stamping die core body. Background Technology

[0002] Springback is a key issue affecting the debugging of automotive stamping outer body panels. Due to stress release after flanging, the panels experience springback deformation of 5-10mm. Applying a clamping flanging structure can improve springback, controlling it to below 1mm. However, compared to conventional flanging structures, clamping flanging requires an additional support core structure. This structure is complex, requiring designers to repeatedly perform a series of intricate steps, including complex surface processing, curve processing, solid creation, Boolean operations, and synchronous modeling, to complete the support core. Existing technologies rely entirely on manual design, requiring designers with extensive experience. The existing technology is time-consuming and labor-intensive. In actual project design, the input process for mold design changes in real time according to customer requirements. When the process input changes, the support core structure requires extensive design modifications, or even redesign, wasting a significant amount of design work. Summary of the Invention

[0003] This application provides a design method for the material support core of a stamping die.

[0004] In a first aspect, embodiments of this application provide a method for designing a core support body for a stamping die, the method comprising the following steps:

[0005] Acquire process input data, calculate the global envelope size of the parting line, construct a reference plane in the vertical retraction direction, and establish a local coordinate system;

[0006] The orientation of the working part is determined based on the local coordinate system, the boundary line is extracted and the cutting edge width is calculated, the parting line is projected and fitted to form a closed area, the closed area is stretched to generate a contour body, the surface body is cut by process supplementary surface, and the working part is obtained by Boolean subtraction.

[0007] Extract the surface edge lines from the working part, construct the stretched conformal surface to generate the conformal body, and sum the working part and the conformal body through Boolean summation to obtain the initial material support core main structure;

[0008] An LSSVM nonlinear regression model was established using LSSVM least squares support vector machine. The regularization parameters and kernel function parameters of the LSSVM nonlinear regression model were optimized using the IWOA improved whale optimization algorithm to obtain the IWOA-LSSVM nonlinear regression model.

[0009] The IWOA-LSSVM nonlinear regression model is used to predict the springback and surface roughness of the initial support core structure under the current process parameters and structural parameters, and the prediction results are obtained.

[0010] Based on the prediction results, the process parameters and structural parameters are optimized in a coordinated manner using the NSGA-II algorithm. The process parameters and structural parameters of the initial material support core body structure are updated according to the optimization results to obtain the target material support core body structure.

[0011] In one possible implementation, the steps of acquiring process input data, calculating the global envelope size of the parting line, constructing a reference plane in the vertical retraction direction, and establishing a local coordinate system include:

[0012] Acquire process input data, including at least the original product sheet, process supplementary surfaces, parting lines, and retraction direction lines;

[0013] Identify the starting point of the parting line, calculate the maximum contour envelope size of the parting line in the global coordinate system, denoted as W for the width and L for the length, where L represents the working range of the flange. The envelope of the parting line in the global coordinate system is obtained as follows:

[0014]

[0015] -

[0016] -

[0017] in, , This represents the minimum / maximum coordinate values ​​of the parting line in the X direction, defining the spatial range of the parting line on the X-axis; , This represents the minimum / maximum coordinate values ​​of the parting line in the Y direction, defining the spatial range of the parting line on the Y-axis; , This represents the minimum / maximum coordinate value of the parting line in the Z direction, defining the spatial range of the parting line on the Z axis;

[0018] Construct a reference plane D at a certain distance perpendicular to the retraction direction line. Construct a local coordinate system on the reference plane, with the center of the envelope body set as the origin of the local coordinate system. The x-axis of the local coordinate system is parallel to the long side of the envelope body L, and the y-axis points to one side of the original product sheet body. According to the right-hand rule, the Z-axis is along the retraction direction.

[0019] In one possible implementation, the process of determining the orientation of the working part based on the local coordinate system, extracting the boundary line to calculate the cutting edge width, fitting the projection of the parting line to form a closed region, stretching the closed region to generate a contour body, cutting the profile body through process supplementary surfaces, and obtaining the working part through Boolean subtraction includes:

[0020] Based on the local coordinate system, the construction orientation of the working part is determined, the boundary line SL1 of the original product sheet is identified and extracted, the cutting edge width of the working part is determined, the parting line is discretized into several discrete points, and projected onto the reference plane D along the retraction direction. The parting line projection line SL1 is then fitted and constructed.

[0021] Search for the two points on the original product sheet boundary line that are closest to the starting point of the parting line. Cut off the boundary line to obtain a boundary cut-off line that is close to the length of the parting line. Discretize the boundary cut-off line into several discrete points. First, offset it by 20mm along the negative Y direction, and then project it onto the reference plane D along the back direction. Fit and construct the boundary line projection line SL3.

[0022] Identify the start and end points of the parting line projection line SL1 and the boundary line projection line SL3 to form a closed region abcd;

[0023] The closed area abcd is stretched to a certain height along the retraction direction to generate a contour body. The process supplement surface is cut and retained along the negative Z direction as the surface body. The parting line is offset along the local negative Z direction to obtain curve S1. Curve S1 is offset by the cutting edge height in the local negative Y direction to obtain curve S2.

[0024] Construct a mesh surface, and generate a back void body by stretching the mesh surface along the local negative Z direction. Perform Boolean subtraction between the void body and the surface body to obtain the working part.

[0025] In one possible implementation, the step of extracting the profile edge line from the working part, constructing an extruded conformal surface to generate a conformal body, and summing the working part and the conformal body using Boolean summation to obtain the initial material support core main structure includes:

[0026] Extract the surface edge line from the working part, discretize the boundary surface edge line into several discrete points, offset the drop height along the local negative Z direction, fit the offset discrete points to construct a smooth and uniform drop line, and construct a stretched uniform surface along the local negative Y direction.

[0027] Based on the conformal surface, a conformal body is constructed. Boolean operations are used to automatically sum the working part of the material support core body and the conformal body to obtain the initial material support core body structure.

[0028] In one possible implementation, the step of establishing an LSSVM nonlinear regression model using LSSVM (Less Squares Support Vector Machine) and optimizing the regularization parameters and kernel function parameters of the LSSVM nonlinear regression model using the IWOA (Improved Whale Optimization) algorithm to obtain the IWOA-LSSVM nonlinear regression model includes:

[0029] Retrieve the structural parameters, process parameters, and corresponding actual springback of the main body structure of the support core from the database. and surface roughness The structural parameters include at least the cutting edge width. Surface curvature Thickness of the body The process parameters include at least the stamping speed. and blank holder force ;

[0030] The input variables of the model are defined as the structural and process parameters of the support core, and the output variables are the springback and surface roughness. An LSSVM nonlinear regression model is built based on LSSVM (Less Squares Support Vector Machine). The RBF radial basis function is set as the kernel function of the model, and the calculation formula is as follows:

[0031]

[0032] in, The kernel function value is used to measure the current... With sample vector The higher the kernel function value, the greater the similarity between them. This represents a vector composed of structural parameters and process parameters. Indicates the first The input vector of each sample; This represents the kernel width parameter. Represents the input vector With sample vector The Euclidean distance is used to initially set the regularization parameters. The range is [0.01, 1000], kernel function The range is [0.01, 100];

[0033] The mathematical expression for the LSSVM nonlinear regression model is:

[0034]

[0035] in, Represents the weight vector in the high-dimensional feature space. Represents the weight vector norm A measure of model complexity. This represents the regularization parameter, used to balance model complexity and fitting error. Indicates the first The fitting error of each sample. This indicates the total number of samples. Indicates the first The actual output value of each sample This represents the bias term, and the core parameters of the model are solved through this optimization problem. Represents the weight vector transpose, This represents a high-dimensional mapping function used to transform a low-dimensional sample input vector. Mapped to a high-dimensional feature space.

[0036] In one possible implementation, the step of using the IWOA-LSSVM nonlinear regression model to predict the springback and surface roughness indices of the initial support core structure under the current process and structural parameters, and obtaining the prediction results, includes:

[0037] The improved IWOA whale optimization algorithm is obtained by introducing Logistic chaotic mapping to initialize the population on the traditional WOA whale optimization algorithm, converting chaotic variables into the parameter solution of the algorithm, and adding linearly decreasing inertia weights.

[0038] Using the root mean square error (RMSE) of the LSSVM nonlinear regression model as the objective function, the calculation formula is as follows:

[0039]

[0040] in, The root mean square error (RMSE) is an evaluation metric for the accuracy of model predictions; the smaller the value, the more accurate the model predictions. Indicates the number of samples in the test set. Indicates the sequence number of the test sample. This represents the model's predicted value. Representing the actual measured value, the regularization parameter... With kernel function parameters As optimization variables, the number of iterations is set to 200 and the population size is 30. The improved IWOA whale optimization algorithm is used to iteratively find the optimal combination of parameters until the root mean square error of RMSE is minimized.

[0041] Substituting the optimal parameter combination into the LSSVM nonlinear regression model, we obtain the IWOA-LSSVM nonlinear regression model.

[0042] Extract the process parameters and structural parameters of the initial material support core body structure. The process parameters include the cutting edge width. Surface curvature and body thickness Structural parameters include stamping speed and blank holder force The process parameters and structural parameters are organized into a feature vector that the model can recognize. For the feature vector Perform normalization processing;

[0043] The normalized feature vector Inputting the data into the IWOA-LSSVM nonlinear regression model, the actual rebound of the initial structure is predicted. and surface roughness The prediction results were obtained.

[0044] In one possible implementation, based on the prediction results, the process parameters and structural parameters are jointly optimized using the NSGA-II algorithm. The process parameters and structural parameters of the initial support core body structure are then updated according to the optimization results to obtain the target support core body structure, including:

[0045] Based on the prediction results, the NSGA-II algorithm is used to jointly optimize the process parameters and structural parameters, and a dual-objective optimization mathematical model is determined, expressed as follows:

[0046]

[0047] in, It is an abbreviation for constraint condition, which represents the range of restrictions that variables must satisfy during the optimization process, and the objective function is to minimize the rebound. , It represents the basic rebound amount, which is the rebound reference value of the support core under reference conditions; , , The coefficients of the IWOA-LSSVM nonlinear regression model represent the surface roughness minimization objective function. , , , The coefficients represent the fitting coefficients of the IWOA-LSSVM nonlinear regression model; the coordinated control of springback and surface quality is achieved through a dual objective function. Indicates the first The lower bound of each optimization variable. Indicates the first The upper limit of the number of optimization variables; Optimize the number of variables

[0048] Determine optimization variables The range of values ​​for each variable is set, and an optimization model for the NSGA-II algorithm is built. The prediction results of the IWOA-LSSVM model are used as the fitness function of the algorithm. The linear weighting method is used to select the optimal solution from the Pareto optimal solution set. The process parameters and structural parameters of the initial support core body structure are updated according to the optimal solution to obtain the target support core body structure.

[0049] Secondly, this application provides a design device for the core support body of a stamping die, the design device for the core support body of the stamping die comprising the following modules:

[0050] The local coordinate system establishment module is used to acquire process input data, calculate the global envelope size of the parting line, construct a reference plane in the vertical retraction direction, and establish a local coordinate system.

[0051] The working part generation module is used to determine the orientation of the working part based on the local coordinate system, extract the boundary line to calculate the cutting edge width, fit the parting line projection to form a closed area, stretch the closed area to generate a contour body, cut the surface body through the process supplement surface to obtain the working part after Boolean subtraction.

[0052] The material support core body generation module is used to extract the profile edge line from the working part, construct the stretched conformal surface to generate the conformal body, and sum the working part and the conformal body through Boolean summation to obtain the initial material support core body structure;

[0053] The regression model building module is used to build an LSSVM nonlinear regression model using LSSVM least squares support vector machine. The regularization parameters and kernel function parameters of the LSSVM nonlinear regression model are optimized using the IWOA improved whale optimization algorithm to obtain the IWOA-LSSVM nonlinear regression model.

[0054] The material support core performance prediction module is used to predict the springback amount and surface roughness index of the initial material support core body structure under the current process parameters and structural parameters using the IWOA-LSSVM nonlinear regression model, and obtain the prediction results;

[0055] The material support core body optimization module is used to perform collaborative optimization of process parameters and structural parameters using the NSGA-II algorithm based on the prediction results, and update the process parameters and structural parameters of the initial material support core body structure according to the optimization results to obtain the target material support core body structure.

[0056] Thirdly, this application provides a device for designing the core body of a stamping die, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor invokes the instructions in the memory to cause the device for designing the core body of a stamping die to perform the various steps of the above-described method for designing the core body of a stamping die.

[0057] Fourthly, this application provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the various steps of the above-described method for designing the core body of a stamping die.

[0058] Based on the method provided in this application, its beneficial effects are as follows: by using explicit parameter constraints and formula definitions to replace the traditional experience-based design mode, the fit error between the working part surface and the product is controlled within 0.03mm, and the contour deviation is ≤0.01mm, significantly reducing human design errors and improving the forming accuracy of the initial structure of the support core. By optimizing the LSSVM model parameters through the improved IWOA whale optimization algorithm, the relative prediction error of the constructed IWOA-LSSVM regression model is reduced, allowing for accurate prediction of the springback and surface roughness of the initial structure in advance. This avoids repeated rework during the trial molding stage in traditional design, reducing mold trial production time and material costs. The NSGA-II algorithm is used to achieve dual-objective optimization of springback and surface roughness. Combined with the linear weighted method to screen the optimal engineering solution, this ensures that the springback of the support core support area meets the acceptance threshold, while also considering the surface quality of the product contact area. This solves the shortcomings of traditional single-objective optimization, which often overlooks certain aspects, thus improving the pass rate of stamped products. It can automatically, efficiently, and accurately create the main structure of the support core, saving a significant amount of manual operation and significantly improving design efficiency. Attached Figure Description

[0059] Figure 1 This is a flowchart illustrating an embodiment of the first stamping die core body design method provided in this application.

[0060] Figure 2 This is a flowchart illustrating an embodiment of the design method for the material support core of the second stamping die provided in this application.

[0061] Figure 3 A schematic diagram of the structure of the first type of stamping die support core body design device provided in the embodiments of this application;

[0062] Figure 4 This is a structural schematic diagram of a material support core design device for a stamping die, provided as an embodiment of this application. Detailed Implementation

[0063] This application provides a design method for the material support core of a stamping die.

[0064] The technical solutions of the embodiments of this application will now be described with reference to the accompanying drawings. The terms "first," "second," "third," "fourth," etc., used in the specification, claims, and accompanying drawings of this application, are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "including" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to these processes, methods, products, or apparatuses.

[0065] It is understood that any part of this application concerning data acquisition or collection has been authorized by the user.

[0066] It is understood that the subject of this application may be a design device for the material support core of a stamping die, or a mobile terminal or server, and no specific limitation is made here.

[0067] The specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 , Figure 1 This is a flowchart illustrating an embodiment of a design method for the material support core of a stamping die provided in this application, including:

[0068] 101. Obtain process input data, calculate the global envelope size of the parting line, construct a reference plane in the vertical retraction direction, and establish a local coordinate system;

[0069] It is understandable that acquiring process input data includes at least the original product sheet, process supplementary surfaces, parting lines, and retraction direction lines;

[0070] Identify the starting point of the parting line, calculate the maximum contour envelope size of the parting line in the global coordinate system, denoted as W for the width and L for the length, where L represents the working range of the flange. The envelope of the parting line in the global coordinate system is obtained as follows:

[0071]

[0072] -

[0073] -

[0074] in, , This represents the minimum / maximum coordinate values ​​of the parting line in the X direction, defining the spatial range of the parting line on the X-axis; , This represents the minimum / maximum coordinate values ​​of the parting line in the Y direction, defining the spatial range of the parting line on the Y-axis; , This represents the minimum / maximum coordinate value of the parting line in the Z direction, defining the spatial range of the parting line on the Z axis;

[0075] Construct a reference plane D at a certain distance perpendicular to the retraction direction line. Construct a local coordinate system on the reference plane, with the center of the envelope body set as the origin of the local coordinate system. The x-axis of the local coordinate system is parallel to the long side of the envelope body L, and the y-axis points to one side of the original product sheet body. According to the right-hand rule, the Z-axis is along the retraction direction.

[0076] 102. Determine the orientation of the working part based on the local coordinate system, extract the boundary line to calculate the cutting edge width, fit the projection of the parting line to form a closed area, stretch the closed area to generate a contour body, cut the surface body through the process supplement surface, and obtain the working part through Boolean subtraction.

[0077] Understandably, based on the local coordinate system, the construction orientation of the working part is determined, the boundary line SL1 of the original product sheet is identified and extracted, the cutting edge width of the working part is determined, the parting line is discretized into several discrete points, projected onto the reference plane D along the retraction direction, and the parting line projection line SL1 is fitted and constructed.

[0078] Search for the two points on the original product sheet boundary line that are closest to the starting point of the parting line. Cut off the boundary line to obtain a boundary cut-off line that is close to the length of the parting line. Discretize the boundary cut-off line into several discrete points. First, offset it by 20mm along the negative Y direction, and then project it onto the reference plane D along the back direction. Fit and construct the boundary line projection line SL3.

[0079] Identify the start and end points of the parting line projection line SL1 and the boundary line projection line SL3 to form a closed region abcd;

[0080] The closed area abcd is stretched to a certain height along the retraction direction to generate a contour body. The process supplement surface is cut and retained along the negative Z direction as the surface body. The parting line is offset along the local negative Z direction to obtain curve S1. Curve S1 is offset by the cutting edge height in the local negative Y direction to obtain curve S2.

[0081] Construct a mesh surface, and generate a back void body by stretching the mesh surface along the local negative Z direction. Perform Boolean subtraction between the void body and the surface body to obtain the working part.

[0082] 103. Extract the profile edge lines from the working part, construct the stretched conformal surface to generate the conformal body, and sum the working part and the conformal body through Boolean summation to obtain the initial material support core main structure;

[0083] It is understandable that the boundary surface edge line is extracted from the working part, the boundary surface edge line is discretized into several discrete points, the drop height is offset along the local negative Z direction, the offset discrete points are fitted to construct a smooth and uniform drop line, and a stretched uniform surface is constructed along the local negative Y direction.

[0084] Based on the conformal surface, a conformal body is constructed. Boolean operations are used to automatically sum the working part of the material support core body and the conformal body to obtain the initial material support core body structure.

[0085] 104. An LSSVM nonlinear regression model was established using LSSVM least squares support vector machine. The regularization parameters and kernel function parameters of the LSSVM nonlinear regression model were optimized using the IWOA improved whale optimization algorithm to obtain the IWOA-LSSVM nonlinear regression model.

[0086] It is understandable that the structural parameters, process parameters, and corresponding actual rebound amount of the main body structure of the support core are obtained from the database. and surface roughness The structural parameters include at least the cutting edge width. Surface curvature Thickness of the body The process parameters include at least the stamping speed. and blank holder force ;

[0087] The input variables of the model are defined as the structural and process parameters of the support core, and the output variables are the springback and surface roughness. An LSSVM nonlinear regression model is built based on LSSVM (Less Squares Support Vector Machine). The RBF radial basis function is set as the kernel function of the model, and the calculation formula is as follows:

[0088]

[0089] in, The kernel function value is used to measure the current... With sample vector The higher the kernel function value, the greater the similarity between them. This represents a vector composed of structural parameters and process parameters. Indicates the first The input vector of each sample; This represents the kernel width parameter. Represents the input vector With sample vector The Euclidean distance is used to initially set the regularization parameters. The range is [0.01, 1000], kernel function The range is [0.01, 100];

[0090] The mathematical expression for the LSSVM nonlinear regression model is:

[0091]

[0092] in, Represents the weight vector in the high-dimensional feature space. Represents the weight vector norm A measure of model complexity. This represents the regularization parameter, used to balance model complexity and fitting error. Indicates the first The fitting error of each sample. This indicates the total number of samples. Indicates the first The actual output value of each sample This represents the bias term, and the core parameters of the model are solved through this optimization problem. Represents the weight vector transpose, This represents a high-dimensional mapping function used to transform a low-dimensional sample input vector. Mapped to a high-dimensional feature space.

[0093] Understandably, data on the process and structural parameters, production conditions, and corresponding performance of similar stamping die support cores from the past are collected as training samples for the model. The process and structural parameters cover structural parameters such as the surface curvature, cutting edge width, and conformal thickness of the working part of the support core, as well as process parameters such as stamping speed, die clearance, and sheet material. The performance data includes the actual springback and surface roughness data of the support core under the corresponding parameters. The sample data is then preprocessed to remove outliers, and parameters of different dimensions are standardized to ensure that the data are of the same order of magnitude, providing a reliable data foundation for model training.

[0094] Based on the fundamental principles of Least Squares Support Vector Machine (LSSVM), preprocessed process parameters and structural parameters are used as input variables, while springback and surface roughness are used as output variables to construct an initial LSSVM nonlinear regression model. This model achieves a nonlinear mapping between input parameters and output performance indicators by constructing an optimal classification hyperplane, thus establishing a preliminary correlation between process parameters, structural parameters, and the core performance of the support core.

[0095] The key parameters of the LSSVM model to be optimized are identified, namely the regularization parameter and the kernel function parameter. The regularization parameter mainly balances the model's fitting accuracy and generalization ability, while the kernel function parameter determines the model's fitting effect on nonlinear data. To address the problems of traditional whale optimization algorithms, such as being prone to getting trapped in local optima and slow convergence speed, improvements are made: First, a dynamic mutation mechanism is introduced, randomly mutating some individuals during iteration to increase population diversity; second, the position update formula is optimized, adjusting the search step size in conjunction with the iteration stages, expanding the global search range in the early stages and reducing the step size in the later stages to improve local search accuracy, forming an improved whale optimization algorithm (IWOA).

[0096] The prediction error of the LSSVM model is used as the fitness function of the IWOA algorithm. Basic parameters such as the number of iterations, population size, and search range are set, and the IWOA algorithm is launched to optimize the regularization and kernel function parameters. During the iteration process, the algorithm continuously updates the parameter combination by simulating three behaviors: whale encirclement, bubble net predation, and random search. After each iteration, the prediction error of the LSSVM model under the corresponding parameters is calculated until the preset number of iterations is reached or the error meets the accuracy requirement. The optimal parameter combination obtained at this point is the best parameter of the LSSVM model.

[0097] The optimal regularization parameters and kernel function parameters obtained by the IWOA algorithm are substituted into the initial LSSVM model to complete the parameter configuration of the model, and finally an IWOA-LSSVM nonlinear regression model with high-precision prediction capability is formed.

[0098] 105. The IWOA-LSSVM nonlinear regression model was used to predict the springback and surface roughness of the initial support core structure under the current process parameters and structural parameters, and the prediction results were obtained.

[0099] It is understandable that the improved IWOA whale optimization algorithm is obtained by introducing Logistic chaotic mapping to initialize the population on the traditional WOA whale optimization algorithm, converting chaotic variables into the parameter solution of the algorithm, and adding linearly decreasing inertia weights.

[0100] Using the root mean square error (RMSE) of the LSSVM nonlinear regression model as the objective function, the calculation formula is as follows:

[0101]

[0102] in, The root mean square error (RMSE) is an evaluation metric for the accuracy of model predictions; the smaller the value, the more accurate the model predictions. Indicates the number of samples in the test set. Indicates the sequence number of the test sample. This represents the model's predicted value. Representing the actual measured value, the regularization parameter... With kernel function parameters As optimization variables, the number of iterations is set to 200 and the population size is 30. The improved IWOA whale optimization algorithm is used to iteratively find the optimal combination of parameters until the root mean square error of RMSE is minimized.

[0103] Substituting the optimal parameter combination into the LSSVM nonlinear regression model, we obtain the IWOA-LSSVM nonlinear regression model.

[0104] Extract the process parameters and structural parameters of the initial material support core body structure. The process parameters include the cutting edge width. Surface curvature and body thickness Structural parameters include stamping speed and blank holder force The process parameters and structural parameters are organized into a feature vector that the model can recognize. For the feature vector Perform normalization processing;

[0105] The normalized feature vector Inputting the data into the IWOA-LSSVM nonlinear regression model, the actual rebound of the initial structure is predicted. and surface roughness The prediction results were obtained.

[0106] Understandably, by obtaining all the process and structural parameters of the initial material support core main structure, including the key dimensions of the working part surface, the extension range of the conformal body, and the specific width of the cutting edge, and clarifying the preset stamping process parameters corresponding to the initial structure, such as stamping pressure, sheet material feeding speed, and mold preheating temperature, these parameters are standardized according to the data preprocessing standards to form a parameter set that can be directly input into the model.

[0107] The processed current process parameters and structural parameters are input into the completed IWOA-LSSVM nonlinear regression model. The model will process the input parameters based on the trained nonlinear mapping relationship and output the springback value and surface roughness level that the initial support core structure may produce in actual application under the process parameters and structural parameters.

[0108] To ensure the reliability of the prediction results, a small number of known actual performance core process and structural parameters were selected and substituted into the model for verification. The deviation between the model's predicted values ​​and the actual values ​​was compared. If the deviation was within a preset reasonable range, the prediction results were confirmed as valid. If the deviation exceeded the range, additional sample data was needed to retrain the model until the prediction accuracy met the standard. Finally, the valid prediction results were compiled to identify the shortcomings of the current initial structure in terms of springback control and surface quality.

[0109] 106. Based on the prediction results, the process parameters and structural parameters are optimized in a coordinated manner using the NSGA-II algorithm. The process parameters and structural parameters of the initial material support core body structure are updated according to the optimization results to obtain the target material support core body structure.

[0110] Understandably, based on the prediction results, the NSGA-II algorithm is used to jointly optimize the process parameters and structural parameters, determining the dual-objective optimization mathematical model, the expression of which is:

[0111]

[0112] in, It is an abbreviation for constraint condition, which represents the range of restrictions that variables must satisfy during the optimization process, and the objective function is to minimize the rebound. , It represents the basic rebound amount, which is the rebound reference value of the support core under reference conditions; , , The coefficients of the IWOA-LSSVM nonlinear regression model represent the surface roughness minimization objective function. , , , The coefficients represent the fitting coefficients of the IWOA-LSSVM nonlinear regression model; the coordinated control of springback and surface quality is achieved through a dual objective function. Indicates the first The lower bound of each optimization variable. Indicates the first The upper limit of the number of optimization variables; Optimize the number of variables;

[0113] Determine optimization variables The range of values ​​for each variable was set, and an optimization model for the NSGA-II algorithm was built. The prediction results of the IWOA-LSSVM model were used as the fitness function of the algorithm. The linear weighting method was used to select the optimal solution from the Pareto optimal solution set. The process parameters and structural parameters of the initial support core body structure were updated according to the optimal solution to obtain the target support core body structure.

[0114] It is understandable that by setting a dual objective function for this collaborative optimization, namely minimizing the springback of the support core and optimizing the surface roughness of the workpiece, and by specifying the constraints, the process parameter constraints include the safe range of stamping speed and the process allowable range of die clearance, while the structural parameter constraints cover the strength and dimensional requirements of the working part of the support core and the assembly space limitations of the conformal body, etc., the optimized parameters are ensured to meet the actual production and assembly specifications.

[0115] First, the population is initialized by randomly generating multiple combinations of process and structural parameters that satisfy the constraints as the initial population. Then, the population is non-dominated and sorted. Based on the degree to which individuals satisfy the dual objective function, the population is divided into different non-dominated levels, with individuals at higher levels having better overall performance. Next, the crowding degree of individuals within each level is calculated to avoid the concentration of excellent individuals in local areas and to ensure population diversity. Then, a new generation of population is generated through genetic operations such as selection, crossover, and mutation. The above sorting and crowding degree calculation process is repeated, and the population is iterated and evolved until it converges to a stable state, obtaining one or more Pareto optimal parameter solutions.

[0116] From the Pareto optimal parameter solution, and combined with the priority requirements of actual production, the final optimal combination of process parameters and structural parameters is selected. Based on this combination, the process parameters and structural parameters of the initial material support core structure are updated accordingly. For example, if the optimization requires a reduction in the cutting edge width, the cutting edge size of the profile generated in step 2 is adjusted; if the thickness of the conformal body is required to increase, the stretching parameters of the conformal body in step 3 are modified, and the geometry of the working part and the conformal body are updated simultaneously.

[0117] After updating the process and structural parameters, the working part and the conformal body structure are re-integrated to generate the target material support core main structure. The updated structural parameters are then substituted into the IWOA-LSSVM model for performance prediction to confirm that the springback and surface roughness indicators meet the preset process requirements. If the requirements are met, the design is completed; otherwise, the process returns to the parameter optimization stage, adjusts the optimization strategy, and iterates again until the target material support core main structure that meets all performance requirements is obtained.

[0118] Based on the method provided in this application, its beneficial effects are as follows: by using explicit parameter constraints and formula definitions to replace the traditional experience-based design mode, the fit error between the working part surface and the product is controlled within 0.03mm, and the contour deviation is ≤0.01mm, significantly reducing human design errors and improving the forming accuracy of the initial structure of the support core. By optimizing the LSSVM model parameters through the improved IWOA whale optimization algorithm, the relative prediction error of the constructed IWOA-LSSVM regression model is reduced, allowing for accurate prediction of the springback and surface roughness of the initial structure in advance. This avoids repeated rework during the trial molding stage in traditional design, reducing mold trial production time and material costs. The NSGA-II algorithm is used to achieve dual-objective optimization of springback and surface roughness. Combined with the linear weighted method to screen the optimal engineering solution, this ensures that the springback of the support core support area meets the acceptance threshold, while also considering the surface quality of the product contact area. This solves the shortcomings of traditional single-objective optimization, which often overlooks certain aspects, thus improving the pass rate of stamped products. It can automatically, efficiently, and accurately create the main structure of the support core, saving a significant amount of manual operation and significantly improving design efficiency.

[0119] Please see Figure 2 , Figure 2 This is a flowchart illustrating an embodiment of the third stamping die support core body design method provided in this application, including:

[0120] 201. Based on the local coordinate system, determine the construction orientation of the working part, identify and extract the boundary line SL1 of the original product sheet, determine the cutting edge width of the working part, discretize the parting line into several discrete points, project them onto the reference plane D along the retraction direction, and fit and construct the parting line projection line SL1.

[0121] 202. Search for the two points on the original product sheet boundary line that are closest to the starting point of the parting line. Cut off the boundary line to obtain a boundary cut-off line that is close to the length of the parting line. Discretize the boundary cut-off line into several discrete points. First, offset it by 20mm along the negative Y direction, and then project it onto the reference plane D along the back direction. Fit and construct the boundary line projection line SL3.

[0122] 203. Identify the starting and ending points of the parting line projection line SL1 and the boundary line projection line SL3 to form a closed region abcd;

[0123] 204. Stretch the closed area abcd along the back direction to a certain height to generate the outline body. Cut and retain the process supplement surface along the negative Z direction as the surface body. Offset the parting line along the local negative Z direction to obtain curve S1. Offset curve S1 to the local negative Y direction by the cutting edge height to obtain curve S2.

[0124] 205. Construct a mesh surface, and based on the mesh surface, stretch it along the local negative Z direction to generate a back empty body. Perform Boolean difference between the empty body and the surface body to obtain the working part.

[0125] Based on the method provided in this application, the orientation of the working part is determined based on a local coordinate system. Combined with discrete projection fitting of the parting line, orientation deviations in the global coordinate system are avoided, ensuring consistency between the parting line projection and the process datum, thus laying a solid positioning foundation for subsequent structural forming. A suitable closed region is formed by truncating the boundary line and directional offset projection, and then the profile is obtained through stretching and cutting. Simultaneously, the cutting edge size is clearly defined through bidirectional offset of the parting line, which not only meets the requirements of the stamping process for the cutting edge width but also avoids workpiece forming defects caused by boundary deviations. By using mesh surface stretching to generate a hollow back and completing Boolean subtraction, the core forming function of the working part is retained while reserving space for assembly and movement, eliminating structural interference during mold operation, significantly improving the structural rationality and production adaptability of the working part, and laying a reliable structural foundation for the overall performance of the subsequent material support core.

[0126] The above describes the design method of the material support core body of the stamping die in the embodiments of this application. The following describes the design device of the material support core body of the stamping die in the embodiments of this application. Please refer to [link / reference]. Figure 3 , Figure 4 A schematic diagram of a material support core design device for a stamping die provided in this application embodiment includes:

[0127] 301. Local coordinate system establishment module, used to acquire process input data, calculate the global envelope size of the parting line, construct a reference plane in the vertical retraction direction, and establish a local coordinate system;

[0128] 302. Working part generation module, used to determine the orientation of the working part based on the local coordinate system, extract the boundary line to calculate the cutting edge width, fit the projection of the parting line to form a closed area, stretch the closed area to generate the contour body, cut the surface body through the process supplement surface, and obtain the working part after Boolean subtraction.

[0129] 303. Material support core body generation module, used to extract the surface edge line from the working part, construct the extruded conformal surface to generate the conformal body, and sum the working part and the conformal body through Boolean summation to obtain the initial material support core body structure;

[0130] 304. Regression Model Building Module: This module is used to build an LSSVM nonlinear regression model using LSSVM least squares support vector machine. It uses the IWOA improved whale optimization algorithm to optimize the regularization parameters and kernel function parameters of the LSSVM nonlinear regression model, thus obtaining the IWOA-LSSVM nonlinear regression model.

[0131] 305. Core support performance prediction module, used to predict the springback amount and surface roughness index of the initial core support body structure under the current process parameters and structural parameters using the IWOA-LSSVM nonlinear regression model, and obtain the prediction results;

[0132] 306. The material support core body optimization module is used to optimize the process parameters and structural parameters in a coordinated manner based on the prediction results and the NSGA-II algorithm. The process parameters and structural parameters of the initial material support core body structure are updated according to the optimization results to obtain the target material support core body structure.

[0133] Based on the device provided in this application, a local coordinate system establishment module replaces the traditional experience-based design mode, controlling the fit error between the working part surface and the product within 0.03mm and the contour deviation ≤0.01mm, significantly reducing human design errors and improving the forming accuracy of the initial structure of the support core. The regression model establishment module optimizes the LSSVM model parameters using the IWOA-improved whale optimization algorithm, reducing the relative prediction error of the constructed IWOA-LSSVM regression model. This allows for accurate prediction of the springback and surface roughness of the initial structure, avoiding repeated rework during the trial molding stage in traditional design and reducing mold trial production time and material costs. The support core body optimization module uses the NSGA-II algorithm to achieve dual-objective optimization of springback and surface roughness, combined with a linear weighted method to select the optimal engineering solution. This ensures that the springback in the support core support area meets the acceptance threshold, while also considering the surface quality of the product contact area, solving the shortcomings of traditional single-objective optimization that overlooks certain aspects, thus improving the pass rate of stamped products. It can automatically, efficiently, and accurately create the support core body structure, saving a significant amount of manual operation and significantly improving design efficiency.

[0134] Please see Figure 4 , Figure 4 This is a schematic diagram of a stamping die core support body design device 400 provided in an embodiment of this application. The stamping die core support body design device 400 can vary significantly due to different configurations or performance characteristics. It may include one or more processors 410, for example, one or more processors and a memory 420, and one or more storage media 430 storing application programs 433 or data 432. The memory 420 and storage media 430 can be temporary or persistent storage. The program stored in the storage media 430 may include one or more modules (not shown in the diagram), each module may include a series of instruction operations on the stamping die core support body design device 400. Furthermore, the processor 410 may be configured to communicate with the storage media 430 and execute the series of instruction operations in the storage media 430 on the stamping die core support body design device 400.

[0135] The stamping die support core body design device 400 may also include one or more power supplies 440, one or more wired or wireless network interfaces 450, one or more input / output interfaces 460, and / or one or more operating systems 431, such as Windows Server, MacOSX, Unix, Linux, FreeBSD, etc. Those skilled in the art will understand that... Figure 4 The illustrated design of the core support device for a stamping die does not constitute a limitation on the core support device design for a stamping die. It may include more or fewer components than illustrated, or combine certain components, or have different component arrangements.

[0136] The computer device includes a memory and a processor. The memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the processor performs each step of the design method for the core support body of the stamping die in the above embodiments.

[0137] This application also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, storing instructions that, when executed on a computer, cause the computer to perform various steps of the design method for the core body of a stamping die.

[0138] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0139] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device, such as a personal computer, server, or network device, to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory, ROM, random access memory, RAM, magnetic disks, or optical disks.

[0140] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0141] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method of designing a material supporting core body of a press die, characterized by, The method comprises the following steps: Obtain process input data, calculate the global envelope size of the parting line, construct a reference plane at the vertical retreat direction, and establish a local coordinate system; Determine the working part orientation based on the local coordinate system, extract the boundary line to calculate the blade width, project and fit the parting line to form a closed area, stretch the closed area to generate a contour body, and cut the surface body through the process supplement surface to obtain the working part; Determine the construction orientation of the working part based on the local coordinate system, identify and extract the boundary line SL1 of the original product sheet body, determine the blade width of the working part, and discretize the parting line into a plurality of discrete points, project them to the reference plane D along the retreat direction, and fit to construct the parting line projection line SL1; Search for the two points on the original product sheet body boundary line that are closest to the starting point of the parting line, obtain the boundary cut-off line on the cut-off boundary line, and discretize the boundary cut-off line into a plurality of discrete points, first offset by 20 mm in the Y negative direction, then project to the reference plane D along the retreat direction, and fit to construct the boundary line projection line SL3; Identify the starting point and ending point of the parting line projection line SL1 and the boundary line projection line SL3 to form a closed area abcd; Stretch the closed area abcd along the retreat direction to a certain height to generate a contour body, cut the surface body through the process supplement surface to retain the process supplement surface along the Z negative direction as a surface body, offset the parting line along the local Z negative direction to obtain a curve S1, and offset the curve S1 along the local Y negative direction by the blade height to obtain a curve S2; Construct a grid surface, generate a back empty body based on the grid surface along the local Z negative direction, and perform Boolean difference between the empty body and the surface body to obtain the working part; Extract the surface edge line from the working part, construct a stretch conformal surface to generate a conformal body, and perform Boolean sum between the working part and the conformal body to obtain the initial material core body structure; Extract the surface edge line from the working part, discretize the boundary surface edge line into a plurality of discrete points, offset the discrete points along the local Z negative direction by a drop height, fit the offset discrete points to construct a smooth and smooth conformal drop line, and construct a stretch conformal surface along the local Y negative direction; Construct a conformal body based on the conformal surface, automatically perform Boolean sum between the working part of the material core body and the conformal body to obtain the initial material core body structure; Establish an LSSVM nonlinear regression model using LSSVM least squares support vector machine, and use IWOA improved whale optimization algorithm to optimize the regularization parameter and kernel function parameter of the LSSVM nonlinear regression model to obtain an IWOA-LSSVM nonlinear regression model; Use the IWOA-LSSVM nonlinear regression model to predict the springback amount and surface roughness index of the initial material core body structure under the current process parameters and structure parameters to obtain a prediction result; Based on the prediction result, use NSGA-II algorithm to collaboratively optimize the process parameters and structure parameters, and update the process parameters and structure parameters of the initial material core body structure according to the optimization result to obtain a target material core body structure.

2. A method of designing a stripper core body of a stamping die according to claim 1, wherein The acquisition process input data, calculate the parting line global envelope size, construct a reference plane at the vertical retreat direction, establish a local coordinate system, comprising: Acquiring process input data, at least including original product sheet, process supplementary surface, parting line and retreat direction line; Identifying the parting line starting point, calculating the maximum profile envelope size of the parting line in the global coordinate system, the envelope width is recorded as W, and the length is recorded as L, L represents the working range of the flanging, and the envelope of the parting line in the global coordinate system is obtained: - - wherein, , represents the minimum / maximum coordinate value of the parting line in the X direction, defining the spatial range of the parting line on the X axis; , represents the minimum / maximum coordinate value of the parting line in the Y direction, defining the spatial range of the parting line on the Y axis; , represents the minimum / maximum coordinate value of the parting line in the Z direction, defining the spatial range of the parting line on the Z axis; A reference plane D is constructed at a certain distance perpendicular to the retreat direction line, a local coordinate system is constructed on the reference plane, the envelope center is set as the origin of the local coordinate system, the x-axis of the local coordinate system is parallel to the L long side of the envelope, the y-axis points to the original product sheet side, and the Z direction is along the retreat direction according to the right-hand rule.

3. The method of designing a stripper core body of a stamping die according to claim 1, wherein The LSSVM nonlinear regression model is established by using the LSSVM least square support vector machine, the regularization parameter and the kernel function parameter of the LSSVM nonlinear regression model are optimized by using the IWOA improved whale optimization algorithm, and the IWOA-LSSVM nonlinear regression model is obtained, comprising: Obtaining structural parameters, process parameters and corresponding actual springback of the core body structure in the database and surface roughness , the structural parameters at least including blade width , profile curvature , and contour thickness , the process parameters at least including punching speed and blank holder force ; The input variables of the model are determined as the structure parameters and process parameters of the supporting material core, and the output variables are the springback and surface roughness, the LSSVM nonlinear regression model is built based on the LSSVM least square support vector machine, the RBF radial basis kernel function is set as the kernel function of the model, and the calculation formula is as follows: wherein, represents a kernel function value used to measure the similarity between the current sample vector , the greater the kernel function value, the higher the similarity; represents a vector composed of structure parameters and process parameters; represents the input vector of the th sample; represents the kernel function width parameter, represents the Euclidean distance between the input vector and the sample vector , the range of the preliminary set regularization parameter is [0.01, 1000], and the range of the kernel function is [0.01, 100]; The mathematical expression of the LSSVM nonlinear regression model is: wherein, denotes a weight vector of a high-dimensional feature space, denotes a norm of the weight vector denotes a measure of model complexity, denotes a regularization parameter for balancing model complexity and fitting error, denotes a fitting error of the -th sample, denotes a total number of samples, denotes a fitting error of the -th sample, denotes an actual output value of the -th sample, denotes a transpose of the weight vector denotes a high-dimensional mapping function for mapping a low-dimensional sample input vector to a high-dimensional feature space.​ 4. The method of designing a stripper core body of a stamping die according to claim 1, wherein The IWOA-LSSVM nonlinear regression model is used to predict the springback and surface roughness indexes of the initial supporting material core main structure under the current process parameters and structure parameters, and the prediction result is obtained, comprising: The Logistic chaotic mapping initialization population is introduced on the basis of the traditional WOA whale optimization algorithm, the chaotic variable is converted into the parameter solution of the algorithm, and a linearly decreasing inertia weight is added, so that the improved IWOA whale optimization algorithm is obtained; Taking the RMSE root mean square error of the LSSVM nonlinear regression model as the objective function, the calculation formula is as follows: wherein, denotes the root mean square error, which is an evaluation index of model prediction accuracy, and the smaller the value, the more accurate the model prediction; denotes the number of test set samples, denotes the serial number of the test sample, denotes the model prediction value, denotes the actual measured value, and the regularization parameter and the kernel function parameter are taken as optimization variables, the iteration number is set to 200, the population size is set to 30, and the improved IWOA whale optimization algorithm is used for iterative optimization until the RMSE root mean square error reaches the minimum value, and the optimal parameter combination is obtained; The optimal parameter combination is substituted into the LSSVM nonlinear regression model to obtain the IWOA-LSSVM nonlinear regression model; Extracting process parameters and structure parameters of the initial core body structure, the process parameters including blade width , profile curvature , and profile thickness , the structure parameters including punching speed and blank holder force ; arranging the process parameters and structure parameters into a model recognizable feature vector , and performing normalization processing on the feature vector ; The normalized feature vector In the input IWOA-LSSVM nonlinear regression model, the actual rebound amount of the initial structure is respectively predicted And surface roughness , to get the prediction result.

5. A method of designing a stripper core body of a stamping die according to claim 4, wherein, Based on the prediction result, the process parameters and structure parameters are cooperatively optimized by using the NSGA-II algorithm, and the process parameters and structure parameters of the initial supporting material core main structure are updated according to the optimization result, so that the target supporting material core main structure is obtained, comprising: Based on the prediction result, the process parameters and structure parameters are cooperatively optimized by using the NSGA-II algorithm, and a double-objective optimization mathematical model is determined, and the expression is as follows: Wherein, is the abbreviation of the constraint condition, indicating the range limit that the variable needs to meet in the optimization process, the minimum rebound amount objective function , represents the basic rebound amount, which is the rebound reference value of the material supporting core under the reference condition; , , represents the fitting coefficient of the IWOA-LSSVM nonlinear regression model; the minimum surface roughness objective function , , , represents the fitting coefficient of the IWOA-LSSVM nonlinear regression model; the rebound and surface quality are cooperatively controlled through the double objective function; represents the lower limit of the th optimization variable, represents the upper limit of the th optimization variable; the number of optimization variables; Determination of optimization variables And set the value range of each variable, build the optimization model of NSGA-II algorithm, take the prediction result of IWOA-LSSVM model as the fitness function of the algorithm, and use linear weighting method to screen the engineering optimal solution from the Pareto optimal solution set; update the process parameters and structure parameters of the initial supporting core main structure according to the engineering optimal solution, and obtain the target supporting core main structure.

6. A design device for a core support body of a stamping die, characterized in that, The supporting material core main structure design device of the stamping die comprises the following modules: The local coordinate system establishment module is used for acquiring process input data, calculating the parting line global envelope size, constructing a reference plane at the vertical retreat direction, and establishing a local coordinate system. The working part generation module is configured to determine a working part orientation based on the local coordinate system, extract a boundary line to calculate a blade width, project and fit a parting line to form a closed area, stretch the closed area to generate a contour body, cut the contour body through a process supplement surface to obtain a workpiece, and obtain the working part through Boolean difference. The workpiece core main body generation module is configured to extract a surface edge line from the working part, construct a stretch conformal surface to generate a conformal body, and obtain an initial workpiece core main body structure by summing the working part and the conformal body through Boolean sum. The regression model establishment module is configured to establish an LSSVM nonlinear regression model by using an LSSVM least square support vector machine, and optimize regularization parameters and kernel function parameters of the LSSVM nonlinear regression model by using an IWOA improved whale optimization algorithm to obtain an IWOA-LSSVM nonlinear regression model. The workpiece core performance prediction module is configured to predict a springback amount and a surface roughness index of the initial workpiece core main body structure under current process parameters and structure parameters by using the IWOA-LSSVM nonlinear regression model, and obtain a prediction result. The workpiece core main body optimization module is configured to collaboratively optimize the process parameters and the structure parameters by using an NSGA-II algorithm based on the prediction result, update the process parameters and the structure parameters of the initial workpiece core main body structure according to an optimization result, and obtain a target workpiece core main body structure.

7. A device for designing the core support body of a stamping die, characterized in that, The blank holder core body design device of the stamping die comprises a memory and at least one processor, and the memory stores instructions; The at least one processor invokes the instructions in the memory to enable the blank holder core body design device of the stamping die to perform the steps of the blank holder core body design method of the stamping die according to any one of claims 1-5.

8. A computer-readable storage medium having stored thereon instructions, the computer-readable storage medium comprising: The instructions are executed by the processor to implement the steps of the blank holder core body design method of the stamping die according to any one of claims 1-5.

Images