Bending machine numerical control system processing piece pattern real-time drawing verification method

Abstract

The present application relates to the technical field of real-time verification of bending machines, and in particular to a method for real-time drawing and verification of workpiece patterns for bending machine numerical control systems. The method comprises the following steps: obtaining an original pattern and identifying geometric element features, generating feature area data, and calculating load safety margins to obtain safety threshold boundaries, decomposing principal stress directions and constructing a continuous stress flow field, constraining and reconstructing the geometric element features, identifying variable contour data, obtaining target bending data and planning bending trajectories, predicting pre-processing shapes, generating real-time processing gaps by comparing differences with target data, detecting the elastic rebound of the bent part, mapping compensation parameters based on rebound data and simulating secondary processing, and finally performing real-time comparison and dynamic processing parameter fine-tuning. The present application optimizes the processing process and improves the processing efficiency and product quality of the bending machine numerical control system as a whole.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of real-time verification of bending machines, and in particular to a method for real-time drawing and verification of a workpiece pattern of a numerical control system of a bending machine. BACKGROUND

[0002] In actual production, due to the influence of material properties, equipment precision, environmental factors and other factors, the actual forming of the bent part often cannot achieve the expected design effect, resulting in problems such as reduced production efficiency and material waste. The existing bending processing verification method relies mainly on experience and experiments. This subjectivity and uncertainty lead to unstable processing precision, making it difficult to meet the needs of modern manufacturing for high precision and high consistency. In addition, many traditional technologies lack corresponding analysis tools when faced with complex geometric shapes, and cannot effectively handle nonlinear deformation and multi-axis interaction. Secondly, existing technologies often ignore the dynamic characteristics of materials, especially under high frequency and high stress, the influence of material behavior changes on bending precision is not fully considered, resulting in difficulty in predicting the shape and size precision of the bent part in actual application, and the existence of springback phenomenon further increases the complexity of process adjustment. The existing springback compensation method relies on fixed parameters and lacks flexibility and adaptability, and cannot achieve the best compensation effect. SUMMARY

[0003] Therefore, it is necessary to provide a method for real-time drawing and verification of a workpiece pattern of a numerical control system of a bending machine to solve at least one of the above technical problems.

[0004] To achieve the above purpose, the method for real-time drawing and verification of a workpiece pattern of a numerical control system of a bending machine comprises the following steps:

[0005] Step S1: Obtain the original pattern of the bending workpiece; identify the geometric element features of the original pattern of the bending workpiece, and perform boundary feature cutting based on the geometric element features to generate feature region data; calculate the load safety margin of the feature region data based on the preset workpiece material data to obtain a safety threshold boundary;

[0006] Step S2: Decompose the safety threshold boundary in the principal stress direction, and construct a continuous stress flow field; perform geometric constraint reconstruction on the geometric element features based on the continuous stress flow field, and identify variable contour data;

[0007] Step S3: Obtain target bending data; perform bending trajectory planning on the variable contour data based on the target bending data to obtain a planned bending sequence; and predict the pre-processing bending shape according to the planned bending sequence;

[0008] Step S4: corresponding position projection is performed on the target bending data and the pre-processing bending shape, and difference comparison analysis is performed to generate real-time processing difference; based on the real-time processing difference, elastic springback detection is performed on the variable contour data to obtain bending part elastic springback data;

[0009] Step S5: springback compensation is performed based on the bending part elastic springback data to generate bending springback compensation data; compensation parameters are mapped according to the bending springback compensation data to obtain difference compensation parameters;

[0010] Step S6: secondary processing simulation is performed on the variable contour data according to the difference compensation parameters to generate secondary processing simulation data; real-time comparison is performed based on the secondary processing simulation data and the target bending data, and dynamic processing parameter fine-tuning is performed according to the real-time comparison data.

[0011] The present application provides a basic basis for the processing process by obtaining the original pattern of the bending processing part, and the feature region data generated by the recognition and cutting of geometric element features lays a foundation for subsequent analysis, and the calculation of the load safety margin ensures the safety of the processing part in use, the construction of the safety threshold boundary provides a reference for the optimization of the processing parameters, the implementation of the principal stress direction decomposition makes the stress analysis more accurate, the establishment of the continuous stress flow field promotes the reconstruction of the geometric constraint, the identification of the variable contour data lays a solid foundation for the subsequent bending trajectory planning, the acquisition of the target bending data provides an actual operation guide for the bending process, the bending trajectory planning enhances the flexibility and accuracy of the bending, the prediction of the pre-processing bending shape provides an estimate of the actual processing result, the generation of the real-time processing difference reflects the deviation of the processing precision and the target, the elastic springback detection ensures a comprehensive understanding of the springback characteristics of the bending part, the generation of the bending springback compensation data provides a necessary basis for optimizing the finished product precision, the mapping of the difference compensation parameters realizes the accurate adjustment of the original pattern, the data generation of the secondary processing simulation provides dynamic feedback for the verification of the processing process, the analysis of the real-time comparison data realizes the fine-tuning of the processing parameters, and the overall bending machine numerical control system processing efficiency and finished product quality are improved, which provides effective protection and support for realizing high-precision bending. BRIEF DESCRIPTION OF DRAWINGS

[0012] Fig. 1 The step flowchart of the real-time drawing verification method for the pattern of the processing part of the bending machine numerical control system is shown in the figure.

[0013] Fig. 2 The detailed implementation step flowchart of step S2 is shown in the figure.

[0014] The implementation of the present application, functional characteristics and advantages will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0015] The technical method of the present application will be described clearly and completely below in conjunction with the drawings. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0016] In addition, the drawings are only schematic illustrations of the present application and are not necessarily drawn to scale. Identical reference numerals in the drawings represent identical or similar parts, and thus repeated descriptions thereof will be omitted. Some of the block diagrams shown in the drawings are functional entities, which do not necessarily have to correspond to physically or logically independent entities. The functional entities can be implemented in the form of software, or in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0017] It should be understood that although the terms "first", "second" and the like can be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, without departing from the scope of the example embodiments, a first element can be called a second element, and similarly a second element can be called a first element. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0018] To achieve the above-mentioned purpose, please refer to Figs. 1-2 The method for real-time drawing and verifying a workpiece pattern of a numerical control system of a bending machine comprises the following steps:

[0019] Step S1: Obtain a bending workpiece original pattern; identify geometric element features of the bending workpiece original pattern, and perform boundary feature cutting based on the geometric element features to generate feature region data; perform load safety margin calculation on the feature region data based on preset workpiece material data to obtain a safety threshold boundary;

[0020] Step S2: Perform main stress direction decomposition on the safety threshold boundary, and construct a continuous stress flow field; perform geometric constraint reconstruction on the geometric element features based on the continuous stress flow field, and identify variable contour data;

[0021] Step S3: Obtain target bending data; perform bending trajectory planning on the variable contour data based on the target bending data to obtain a planned bending sequence; and predict a pre-processing bending shape according to the planned bending sequence;

[0022] Step S4: corresponding position projection is performed on the target bending data and the pre-processing bending shape, and difference comparison analysis is performed to generate real-time processing difference; based on the real-time processing difference, elastic springback detection is performed on the variable contour data to obtain bending part elastic springback data;

[0023] Step S5: springback compensation is performed based on the bending part elastic springback data to generate bending springback compensation data; and compensation parameters are mapped to the original pattern of the bending processing part according to the bending springback compensation data to obtain difference compensation parameters;

[0024] Step S6: secondary processing simulation is performed on the variable contour data according to the difference compensation parameters to generate secondary processing simulation data; real-time comparison is performed based on the secondary processing simulation data and the target bending data, and dynamic processing parameter fine-tuning is performed according to the real-time comparison data.

[0025] The present application provides a basic basis for the processing process by obtaining the original pattern of the bending processing part, and the feature region data generated by the recognition and cutting of the geometric element features lays a foundation for subsequent analysis, and the calculation of the load safety margin ensures the safety of the processing part in the use process, the construction of the safety threshold boundary provides a reference for the optimization of the processing parameters, the implementation of the principal stress direction decomposition makes the stress analysis more accurate, the establishment of the continuous stress flow field promotes the reconstruction of the geometric constraint, the identification of the variable contour data lays a solid foundation for the subsequent bending trajectory planning, the acquisition of the target bending data provides an actual operation guide for the bending process, the bending trajectory planning enhances the flexibility and accuracy of the bending, the prediction of the pre-processing bending shape provides an estimate of the actual processing result, the generation of the real-time processing difference reflects the deviation of the processing precision and the target, the elastic springback detection ensures the comprehensive understanding of the springback characteristics of the bending part, the generation of the bending springback compensation data provides a necessary basis for optimizing the finished product precision, the mapping of the difference compensation parameters realizes the accurate adjustment of the original pattern, the data generation of the secondary processing simulation provides dynamic feedback for the verification of the processing process, the analysis of the real-time comparison data realizes the fine-tuning of the processing parameters, and the overall processing efficiency and finished product quality of the bending machine numerical control system are improved, which provides effective protection and support for realizing high-precision bending.

[0026] In the embodiment of the present application, the processing part pattern real-time drawing verification method of the bending machine numerical control system comprises the following steps:

[0027] Step S1: obtaining the original pattern of the bending processing part; recognizing the geometric element features of the original pattern of the bending processing part, and performing boundary feature cutting based on the geometric element features to generate feature region data; and performing load safety margin calculation on the feature region data based on the preset processing part material data to obtain a safety threshold boundary;

[0028] In this embodiment, when obtaining the original drawing of the bending workpiece, a high-precision two-dimensional CAD drawing input module is used to access AutoCAD format (.dwg) or STEP format (.stp) drawing file. During the drawing reading process, a boundary recognition function is called and a boundary scan segmentation algorithm is used to convert the drawing contour data into a line segment set. All line segment sets need to meet the constraint conditions that the length threshold is greater than 1.0 mm and the angle tolerance is less than 1.5°. When identifying geometric element features, RANSAC (Random Sample Consensus Algorithm) is applied to straight line fitting and circular arc extraction of boundary line segments. Line segment combination structure establishes corner point markers at adjacent angle changes greater than 3°. Then, the region between the marked nodes calls the contour closure judgment module to complete the region division. Based on the identified geometric element structure, the primitive-level partitioning algorithm is executed to complete the boundary feature cutting. The generated feature region data includes the closed contour path of each region, internal structure element number and attribute label. The boundary feature cutting precision error needs to be less than ±0.2mm. Then, the material information corresponding to the workpiece drawing is called from the database. The material number is GB-T700 Q235, the yield strength is set to 235MPa, the elastic modulus is 210GPa, and the Poisson's ratio is 0.3. Based on the above material parameters, the finite element micro-unit distribution function is called to perform grid cell load distribution on each feature region. The equivalent stress method is used to calculate the Von Mises stress of each cell. The loading method uses standard constant pressure loading, and the pressure direction is perpendicular to the region boundary normal direction. The cell distribution range is controlled within 0.5mm grid spacing. According to the principle that the maximum cell stress does not exceed the yield limit, the critical load boundary line of each feature region is extracted, which is defined as the safety threshold boundary.

[0029] Step S2: decompose the safety threshold boundary into principal stress direction, and construct a continuous stress flow field; based on the continuous stress flow field, reconstruct the geometric constraint of the geometric element feature, and identify the variable contour data;

[0030] In this embodiment, when the safety threshold boundary is decomposed in the principal stress direction, the equivalent stress tensor matrix is used to perform eigenvalue decomposition operation on each boundary element node, the maximum principal stress direction (Major Principal Stress Direction) is extracted, the principal direction range is limited to 0° to 180°, and the direction vector field is established. The discrete vector field is extended into a continuous stress flow field by using the structural interpolation algorithm (Structural Interpolation Method), and the vector field distribution density is set to not less than 4 direction vector nodes per square millimeter. Then, the continuous stress flow field is superimposed on the original geometric element characteristic coordinate system, the direction offset fitting operation is performed, the geometric constraint reconstruction is carried out, the offset deformation angle is limited to not more than ±1.8°, and the reconstructed boundary curve is fitted and smoothed. The fitting error is controlled to be within 0.15 mm. Finally, the deformation sensitive segment is identified according to the reconstructed curve under the curvature continuity and length constraint, and is marked as variable contour data. The data structure includes curvature change rate, paragraph length, average principal stress direction and number index.

[0031] Step S3: obtaining target bending data; based on the target bending data, the variable contour data is subjected to bending trajectory planning to obtain a planned bending sequence; and predicting a pre-processing bending shape according to the planned bending sequence;

[0032] In this embodiment, when the target bending data is obtained, the standard bending process flow table is imported through the man-machine interaction terminal, including process number, bending sequence, angle setting, linear displacement and tool parameters. The target bending angle range is set to 30° to 150°, and the minimum bending radius is limited to 1.5 mm. According to the data, the variable contour data is subjected to trajectory control algorithm based on path optimization for bending trajectory planning. In the control algorithm, the variable step path evolution module is enabled, the trajectory node spacing is not greater than 2 mm, the bending path needs to construct a continuous and smooth path within the allowable minimum curvature radius range, and the control point sequence is recorded. Finally, the planned bending sequence is obtained. The bending sequence storage format is a structured data set containing curvature control points, bending angles and loading positions. Then, based on the planned bending sequence, the intermediate part deformation simulation module is constructed, and the nonlinear large deformation simulation is performed in the ANSYS Workbench environment. The material properties are consistent with the previous material parameters. The simulation loading speed is 5 mm / s, the simulation step number is 200 steps, the intermediate configuration stability is taken as the stop judgment condition, and the predicted pre-processing bending shape is generated. The data is a deformation node set with a time stamp, including the spatial position, principal strain direction and curvature vector information of each point.

[0033] Step S4: Project the target bending data and the pre-processing bending shape on corresponding positions, and perform difference comparison analysis to generate real-time processing difference; based on the real-time processing difference, perform elastic springback detection on the variable contour data to obtain bending part elastic springback data;

[0034] In this embodiment, when performing the corresponding position projection operation on the target bending data and the pre-processing bending shape, a bidirectional projection comparison function is called to perform full contour mapping on the two shape models, an error matching function based on point cloud density weighting is used to calculate the local normal distance difference at each mapping node, the error tolerance is set to ±0.25 mm, the real-time processing difference data is generated, then a nonlinear deformation response model is called based on the difference data to establish an elastic springback displacement path model between corresponding nodes, the reaction displacement threshold is set to 0.15 mm, the maximum springback angle deviation is 2°, the elastic springback data of the bending part is calculated by reverse displacement of the springback path, and the output result is a matrix structure data containing displacement vector direction, springback angle and residual stress mapping.

[0035] Step S5: Perform springback compensation based on the bending part elastic springback data to generate bending springback compensation data; perform compensation parameter mapping on the original pattern of the bending processing part according to the bending springback compensation data to obtain difference compensation parameters;

[0036] In this embodiment, based on the bending part elastic springback data, the high-sensitive springback area is extracted by the springback absorption area recognition algorithm, the local contour data of the area is subjected to stress inverse solution reconstruction operation, and a correction configuration function is generated, the area deformation gain parameters constructed by the correction function are superimposed with the original springback data to form the bending springback compensation data, the compensation data contains angle increment, line segment extension ratio and local curvature adjustment vector, under the guidance of the data, inverse geometric mapping is performed on the original pattern of the bending processing part to map the compensation amount to the boundary curve of the original pattern, the process needs to keep the boundary closure error within 0.05 mm, and the difference compensation parameters output include the deformation function expression of each boundary curve and the angle correction vector matrix.

[0037] Step S6: Perform secondary processing simulation on the variable contour data according to the difference compensation parameters to generate secondary processing simulation data; perform real-time comparison based on the secondary processing simulation data and the target bending data, and perform dynamic processing parameter fine tuning according to the real-time comparison data.

[0038] In this embodiment, the machining simulation subsystem is called according to the gap compensation parameter, the compensated contour model is constructed in the ABAQUS simulation platform, the grid element density is set to 0.25 mm, the material attribute is called from the original data model, the loading condition is set to the constant speed pressurization of the servo hydraulic system, the pressure increment is 50 N / step, the secondary machining simulation is executed, the simulation step is 0.1 s, the secondary machining simulation data is output, the data structure includes the configuration state of each time node in the bending process, the stress distribution atlas and the node position transformation matrix, then the data is spatially mapped and reconstructed with the target bending data, three-dimensional geometric alignment is executed, the least square error matching function is used to fit the residual error distribution graph, when the residual error extreme value is not more than 0.2 mm, the dynamic machining parameter fine adjustment module is started, the feed speed, bending angle setting and tool stroke output by the bending control system are finely adjusted, the feed speed adjustment range is ±1.2 mm / s, the bending angle adjustment range is ±1.5°, the tool stroke adjustment accuracy is ±0.1 mm, and finally the machining feedback correction parameter set is output for secondary closed loop correction of the control system.

[0039] Preferably, the geometric element feature of the original pattern of the bending workpiece is identified in step S1, and the boundary feature cutting is performed based on the geometric element feature.

[0040] The geometric element feature of the original pattern of the bending workpiece is identified.

[0041] The topological relationship is constructed according to the geometric element feature, and the connection topological graph is drawn based on the topological relationship, wherein the minimum spacing between nodes in the connection topological graph should not be less than 0.1 mm, and the maximum connectivity error in the topological path should be kept not more than 0.05 mm.

[0042] The pattern boundary contour is identified based on the connection topological graph.

[0043] The pattern boundary contour is subjected to solid feature cutting, the region is cut according to the minimum unit segmentation size of 1 mm², the aspect ratio of the cut region is controlled to be 1:5~5:1, and the feature region data is generated.

[0044] In the embodiment, in the processing of the processing piece pattern real-time drawing verification method of the numerical control system of the bending machine, the geometric element feature recognition operation of the original pattern of the processing piece is processed by using the vector graphics analysis based mode. First, the original pattern with the input format of DXF (Drawing Exchange Format) or SVG (Scalable Vector Graphics) type is introduced into the CAD core engine for vector boundary extraction processing. The ShapeAnalysis_ShapeTolerance class based on the OpenCascade geometric modeling kernel is called to realize the unified transformation and classification of all straight lines, circular arcs, broken lines and B-spline curves into a geometric element feature set. According to the grouping number of the entity primitive type, the boundary contact points between the endpoints of the line segments, the control points of the curves and the adjacent surfaces are marked. All point data is kept to 6 decimal places of accuracy, so that the error is not more than 0.0001 mm in the boundary connection process. After the initial feature point set and the line segment set are generated, the topology relationship construction is entered. The topology relationship construction is executed by using the graph traversal and the minimum connection optimization strategy. First, the directed graph structure in the Boost Graph Library (graph library) is called. The starting point and the ending point of the geometric element feature are respectively taken as the nodes of the graph. The line segment and the curve are defined as the edges of the graph. The edge weight is set as the actual length of the line segment. After the initial connection graph is constructed, the Dijkstra shortest path algorithm is used to identify the minimum path set between the geometric element features. After the non-closed structure or the local isolated features are screened out, the topology correction stage is entered. The connection distance between the nodes in the graph must be not less than 0.1 mm, otherwise it is regarded as a misconnection or a redundant point. The neighbor point merging strategy based on the Euclidean distance calculation is executed. The node pairs less than 0.1 mm are merged into a single node. At the same time, the boundary direction vector and the connectivity angle are recalculated. The maximum connectivity error of all topology paths must be not more than 0.05mm, check all path error cumulative values by floating point error cumulative check method, ensure the closure of the topology graph, and thus complete the connection topology graph. In the identification of the pattern boundary contour stage, call the BRepBuilderAPI_MakeWire class in OpenCascade to construct the boundary set in the topology graph into a closed contour structure, use the vector clockwise direction judgment method to filter out all geometric element feature combinations that constitute the outermost contour line, and correct the boundary normal vector direction by judging its consistency, complete the reconstruction of the pattern outer contour. On this basis, call the BRepOffsetAPI_Sewing interface to build the closed contour area into a two-dimensional solid structure, enter the solid feature cutting stage, and the solid feature cutting is discretely processed by setting the cutting grid parameters. Call the BRepTools_Substitution class in OpenCascade to partition the contour area according to the uniform grid method, set the minimum unit area cutting size to 1mm2, and use the horizontal and vertical layer-by-layer scanning cutting mode. The aspect ratio parameter of each cutting block is set to 1:5 to 5:1, and the boundary area is automatically adjusted to the nearest range of legal proportion. For example, when processing a rectangular contour with a width of 3mm and a height of 15mm, it is divided into 5 vertical regions, each with a unit size of 1mm x 3mm, totaling 15 sub-regions and automatically numbered as R1~R15. The geometric information of each sub-region includes corner point coordinates, boundary normal vector, corresponding relationship with geometric elements in the original pattern, and topological association number. Finally, all region data is saved as a feature region data structure.

[0045] Preferably, the step S1 includes:

[0046] Based on the preset machining part material data, the feature region data is associated with the material type to generate material distribution data;

[0047] Project anisotropy indexes on the material distribution data, and construct a directional intensity matrix based on the projected data;

[0048] Map material attributes on the feature region data according to the directional intensity matrix to obtain a material parameter set;

[0049] Reconstruct the material microstructure according to the material parameter set, and perform constitutive relationship conversion based on the material microstructure to generate a material stress tensor;

[0050] Calculate the critical load distribution based on the material stress tensor, and limit the safety margin based on the critical load distribution to obtain a safety threshold boundary.

[0051] In this embodiment, when the feature region data is associated with the material type based on the preset workpiece material data, first, the workpiece pattern is imported into the pattern analysis module, which performs voxel segmentation processing on the CAD three-dimensional model, divides the overall model into a plurality of micro-unit regions containing feature of machining behavior, each micro-unit contains a unique spatial coordinate index, on this basis, the material basic data of the preset workpiece is loaded, which includes material code, crystal structure type, heat treatment state, yield strength, elastic modulus, Poisson's ratio and hardening function surface, the material database is matched and selected according to the machining process version number, the material data is mapped to each feature region micro-unit to form a spatial distribution material atlas, each micro-unit is accurately bound with its corresponding material type in the spatial dimension to form a material distribution data containing three groups of primary key fields of three-dimensional coordinates, material code and material parameters, when the material distribution data is projected with anisotropy index, a machining direction coordinate system is constructed, which is established according to the centering reference of the upper and lower dies of the bending machine, the X direction is set as the workpiece axial direction, the Y direction is set as the width direction, and the Z direction is set as the normal loading direction, then the single crystal elastic modulus and shear modulus values in each direction are extracted according to the material crystal structure, the directional remapping is performed in the coordinate system, the rigidity response ability of each material in the current spatial direction is numerically projected to generate a directional index set, after the projection is completed, a directional intensity vector is established for each micro-unit, which contains the elastic response intensity of three orthogonal directions and the intensity level of three shear directions, a spatial intensity matrix is constructed for subsequent loading analysis, when the feature region data is mapped with material properties according to the directional intensity matrix, the material parameters of the micro-unit are adjusted relying on the anisotropy information, the intensity distribution information in different directions is coupled with the basic material parameters, the directional related elastic modulus variation curve, yield strength correction value and strain hardening interval are recorded in each micro-unit to generate a material parameter collection, which is a composite material feature table under three-dimensional coordinate index, the structure is a multi-field data table, wherein the main fields include spatial index, elastic modulus, Poisson's ratio, yield limit and numerical interval of yield strength in each direction, when the material microstructure is reconstructed according to the material parameter collection, a three-dimensional reconstruction engine is used for micro-grain modeling, the average grain size is set as 10 microns, the grain orientation is generated by ODF function distribution according to the machining heat treatment history, the grain boundary position is drawn by using the method of spatial minimum energy distribution, the grain boundary thickness is set as 0.2 microns, the critical shear model defined in GB2312-3 metal hot working data is used to construct the grain boundary slip behavior, the intracrystalline slip direction is defined by using the main slip system of FCC crystal structure, these microstructures are input into the constitutive model generator, the stress tensor field is generated by using the elastic-plastic constitutive relationship, the tensor expression is σ ijwhere i and j are coordinate direction indexes, and the numerical values represent the stress response of each direction per unit area, with the unit of mega-pascal (MPa). In the process of calculating the critical load distribution based on the material stress tensor, the maximum principal stress of each micro-unit is obtained by analyzing the maximum stress in the principal axis direction of the stress tensor, and compared with the yield strength of the material in the current processing direction. The von Mises yield criterion is adopted The yield state is judged, and a yield region map is established, where s ij are the components of the deviatoric stress tensor, representing the pure shear stress part after removing the average stress in each direction from the total stress tensor, describing the stress component leading to deformation in the material stress state, f(σ) is the von Mises yield function, used to judge whether the material enters the yield state, when f(σ)≥0, it means that the material is in the yield or plastic state, when f(σ)<0, the material is still in the elastic state, f(σ) is the yield strength, representing the equivalent stress value of the material when the stress reaches the yield, which is one of the material property parameters, set according to the material model, heat treatment state, grain structure, etc. The region in the map that exceeds the yield limit is defined as the critical load point set, based on the aggregation density and position distribution of the point set in space, a three-dimensional safety boundary model is constructed, the load deviation degree of all micro-units with yield risk is calculated and labeled, and finally a safety threshold boundary map is generated. This map is used as the input for dynamic interference judgment in real-time processing path optimization, and can be loaded into the numerical control system of the bending machine for pattern drawing and processing behavior control.

[0052] Preferably, step S2 comprises the following steps:

[0053] Step S21: identifying the surface features of the safety threshold boundary, and identifying the feature control area based on the surface feature identification; based on the feature control area, the finite element grid is obtained by grid division;

[0054] Step S22: decomposing the principal stress direction based on the finite element network to obtain the principal stress direction vector; constructing a continuous stress flow field according to the principal stress direction vector;

[0055] Step S23: mapping the density gradient based on the continuous stress flow field, and extracting stress hot zone data based on the density gradient; according to the stress hot zone data, the material density distribution is analyzed to obtain the material density distribution;

[0056] Step S24: identifying the unit stress concentration point based on the material density distribution; the local shape of the unit stress concentration point is corrected to obtain a corrected stress concentration point;

[0057] Step S25: Adjust the stress distribution of the geometric element feature based on the stress concentration point correction, and reconstruct the geometric constraint based on the stress distribution adjustment data to generate adjustment constraints;

[0058] Step S26: According to the adjustment constraints, the geometric element feature is not limited to the contour recognition, so as to obtain the variable contour data.

[0059] In this embodiment, the safety threshold boundary data generated in the previous stage is input to the Gaussian curvature calculation module in the form of a three-dimensional point cloud. For each boundary surface point, a Gaussian curvature calculation operation is performed. A spherical neighborhood with a local window radius of 2.5 mm is used to estimate the curvature of each boundary point. A least squares method is used to construct a local fitting surface function z = f(x, y) for the neighborhood point set, and then the principal curvatures k1 and k2 are calculated, and their product is taken as the Gaussian curvature value of the point. All regions with a Gaussian curvature value greater than 0.15 are marked as concave regions to form a feature control region. Subsequently, a three-dimensional hexahedral automatic mesh generation tool is called to perform a volume mesh generation operation on the volume within the control region. The initial mesh density is set to 0.8 mm, and a Poisson mesh optimization algorithm is used to fine-tune the boundary mesh, thereby generating a finite element mesh with continuous boundaries. After the finite element mesh is generated, the principal stress solving module is input based on the boundary conditions. Fixed boundary conditions are applied to each hexahedral element, and the initial external load direction is defined as the Y-axis positive direction. The loading amplitude is set to 250 MPa. The three-dimensional elastic finite element solver is used to solve the stress tensor components σxx, σyy, σzz, σxy, σyz, σxz for each element. The stress principal value decomposition method is used to perform eigenvalue decomposition on each tensor to calculate the principal stresses σ1, σ2, σ3 and their corresponding eigenvector directions v1, v2, v3. The first principal stress direction v1 is extracted as the principal stress direction vector set. The principal stress direction vectors are interpolated on a spherical surface based on their continuity. The interpolated principal stress direction vectors are used to construct a continuous stress flow field, which is defined as the vector field T(x, y, z). The direction components of the vector field are smoothed using a Bezier curve interpolation method to obtain a continuous vector field dataset. Based on the continuous stress flow field T(x, y, z), a density gradient projection mapping operation is performed. The initial element density is set to 7.85 g / cm³. The direction integral of all paths in the flow field is calculated, and the density change rate ∇ρ(x, y, z) of each unit path is calculated. The density change rate is normalized in the global domain to obtain the heat extraction weight. The weighted heat mapping method is used to extract the high-density change region and set it as the stress heat zone. The heat zone threshold is set to a density gradient ∇ρ greater than 1.2g / cm³ / mm, the corresponding material density data is extracted according to the grid cell coordinate index covered by the heat zone, the density distribution mean value ρ_i in each cell is calculated by volume average method, and a three-dimensional density distribution function ρ(x, y, z) is constructed, the material density distribution function ρ(x, y, z) is input into the stress concentration point identification module, the maximum value search of the density gradient modulus ∥∇ρ∥ is performed on each grid cell using the local peak detection algorithm, the window size is set to 3*3*3 cells, if the density of the center cell is greater than that of all its neighborhood cells and the density gradient modulus is greater than 3.5g / cm³ / mm² in the window, it is marked as a cell stress concentration point, the five-order Bezier surface around the point is fitted as an initial local geometric model, and when performing shape correction processing, the Bezier control points are projected and moved along the stress main direction by a distance d=(σ / σ_max)*Δx, wherein σ is the first principal stress of the cell, σ_max is the maximum principal stress of the whole field, and Δx is the cell edge length. The stress concentration point position data set obtained after adjustment is input into the stress distribution uniform adjustment of geometric element features, the structure topology optimization module is called, the optimization objective function is set as the maximum stress σ_minimization, and the geometric element boundary line is re-parameterized through the variable shape function under the constraint condition of the stress field σ(x, y, z). The control point spacing of each shape parameter vector is limited to be within 0.2mm, the control point positions are adjusted so that the stress difference Δσ_i=|σ_i-σ_mean| of each cell after correction is less than 15MPa, the control point coordinate set obtained after uniform adjustment is defined as a geometric constraint reconstruction condition, and is output to the next module for contour data identification. The non-restricted contour identification processing is performed on the geometric element features, the above control point coordinate set is input as a constraint control, the boundary point cloud graph structure of the three-dimensional geometric model is identified through the graph convolution network (GCN, Graph Convolutional Network) algorithm, the graph topology structure G(V, E) is constructed, the boundary continuity of each node V is judged, the non-restricted contour identification is performed on the area where the boundary curvature change rate is greater than 0.25 / mm, the graph traversal operation is performed with the curved boundary starting point as the root node in the identification process, all non-restricted contour boundary paths are recorded, and finally the contour path is output in the form of ordered coordinates to generate a variable contour data set.

[0060] Especially important is that step S23 comprises:

[0061] The continuous stress flow field is quantified in gradient direction to obtain stress gradient distribution data;

[0062] The stress gradient distribution data is graded in gradient intensity to generate gradient intensity classification data;

[0063] Performing hot zone threshold fitting on the gradient intensity classification data to generate stress hot zone threshold;

[0064] Extracting hot zone range based on the stress hot zone threshold and the gradient intensity classification data to obtain stress hot zone data;

[0065] Calculating the material density corresponding to the hot zone according to the stress hot zone data;

[0066] Performing material distribution mapping based on the material density corresponding to the hot zone to obtain material density distribution.

[0067] In this embodiment, the stress tensor matrix based on each unit node is constructed, the principal stress direction vector is coordinate mapped, and the Gaussian filter is selected to denoise each principal stress direction vector in the three-dimensional continuous stress field. The filter kernel width is set to σ = 1.5, and the step length is Δx = 1.0 unit length. The spatial derivative is solved for each node vector after denoising using the Sobol gradient operator, and the stress change rate in the X, Y and Z directions is calculated respectively. The direction derivative at each node is unified into a three-dimensional polar coordinate expression form through three-dimensional Euler angle (Euler angles) transformation, and finally a gradient direction angle distribution matrix between nodes is formed, with a unit of radian and a range set in [0, 2π]. In the process of classifying the stress gradient distribution data, the maximum-minimum normalization method is used to scale the gradient modulus value corresponding to each node to between 0 and 1. The gradient modulus value is calculated from the Euclidean norm of the vector derivative, and its expression form is G = √(∂σ / ∂x)² + (∂σ / ∂y)² + (∂σ / ∂z)². Then the normalized G value is divided into five levels, and the level threshold is set to [0, 0.2), [0.2, 0.4), [0.4, 0.6), [0.6, 0.8), [0.8, 1.0], respectively marked as L1 to L5. Each node is assigned to the corresponding category according to the gradient level it belongs to, and a hierarchical mapping grid in three-dimensional space is constructed, with each grid length set to 2.5 mm. In the process of fitting the hot zone threshold value to the gradient intensity classification data, the Logistic regression surface fitting method is used to fit the L4 and L5 level gradient node coordinate set in the spatial distribution, and the Logistic function is set to f(x, y, z) = 1 / (1 + e^-(a·x + b·y + c·z + d)). The fitting coefficients a = 0.12, b = -0.08, c = 0.1 and d = -1.7 are obtained by least squares fitting. In the three-dimensional space, the critical fitting threshold T = 0.5 is set according to the function output value, and the nodes corresponding to f(x, y, z) ≥ T are marked as potential stress hot zone nodes, forming the hot zone threshold space boundary. In the process of extracting the hot zone range based on the stress hot zone threshold and the gradient intensity classification data, the nodes that meet f(x, y, z) ≥ 0 are first screened out from the node set classified as L4 and L5.The node group of 5 is clustered and identified by using a voxel growth algorithm based on eight-neighbor connectivity, and a set of thermal zones with more than 50 voxels connected continuously is defined as an effective thermal zone. A boundary volume of the thermal zone is constructed, and a minimum containing volume space of each thermal zone is recorded in the form of a bounding box. In the process of calculating the material density corresponding to the stress thermal zone data, there is a relationship between the local structural stiffness K=E·A / L (where E is the Young's modulus of the material, A is the cross-sectional area, and L is the unit length) of the material and the local stress σ, which is σ=K·ε (strain). Assuming that the workpiece is in a static uniform state under stress, the average stress σ_avg and the average strain ε_avg within the boundary of each thermal zone are used to back-calculate the local structural stiffness K_avg=σ_avg / ε_avg, and the relative density ρ_local of the current thermal zone is further derived by the ratio of K_avg to the structural stiffness K_ref of the standard density material, which is 45 GPa. The density value distribution of each voxel in the thermal zone is obtained by the above method, with a unit of g / cm³, and a local material density matrix of each thermal zone is finally constructed. In the process of mapping the material distribution based on the material density corresponding to the thermal zone, the density values in the thermal zone are propagated to the outside of the thermal zone boundary by using an inverse distance weighted interpolation method (IDW, Inverse Distance Weighting), with an influence radius of 10 mm and an interpolation power index of 2. The calculation formula is ρ(x, y, z)=∑(ρ_i / d_i. 2 ) / ∑(1 / d_i 2 ), where ρ_i is the known density in the thermal zone, and d_i is the distance between the current point and ρ_i. Finally, a complete material density distribution map is constructed in the entire workpiece three-dimensional grid, with a density resolution of 0.1 g / cm³. The map is stored in the Voxel format, with a resolution of 0.5 mm³. Each Voxel contains the corresponding spatial position and its density value, and the output is a three-dimensional density volume data set.

[0068] Preferably, the step S3 of performing the bending trajectory planning on the variable profile data based on the target bending data comprises:

[0069] mapping the target bending data based on boundary constraints to obtain boundary constraint data;

[0070] performing bending angle serialization processing based on the boundary constraint data to obtain angle serialization data;

[0071] performing multi-axis collaborative path mapping on the variable profile data according to the angle serialization data to generate a candidate bending trajectory;

[0072] performing bending simulation based on the candidate bending trajectory, and generating deformation topology data based on the simulated bending data;

[0073] detecting a bending conflict structure in the deformed topological data;

[0074] performing conflict trajectory avoidance on the candidate bending trajectory based on the bending conflict structure, and performing bending action planning according to the conflict avoidance trajectory to obtain a planned bending sequence.

[0075] In this embodiment, B-spline fitting operation is performed on the two-dimensional bending profile during the CAD drawing input stage, the number of control points is set to be not less than 30, and the fitting error is kept to be less than 0.05 mm, after the fitting is completed, the curvature continuity of the bending starting point, the ending point and the turning point is extracted, the initial tangent vector and the terminal normal vector of each bending segment are calculated respectively, the initial boundary condition set is constructed, then the static configuration constraint framework is introduced, the boundary perpendicular constraint value of Z-axis direction is set to be ±0.03 mm under the reference workpiece coordinate system, the linear spread range constraint of X-axis direction is ±1.2 mm, the boundary value is converted into the local constraint matrix of each B-spline control point through the three-dimensional space boundary projection algorithm, finally the boundary constraint control set is formed, the boundary set corresponding to each bending segment contains six rigid parameters: three-dimensional coordinate offset, curvature offset and direction rotation angle, all boundary data are stored in three-dimensional tensor format, the size is [bending segment number, 6], during the bending angle serialization process based on the boundary constraint data, the workpiece geometric center is used as the coordinate origin to construct the polar coordinate expression, the θ angle and r radius of each bending control point are projected, the tangent angle between adjacent bending segments is calculated, the angle is recorded in degrees, the value range is in [0, 180], the angle accuracy is controlled to be 0.1°, the sorting algorithm is used to sort all angles in ascending order, to ensure that the bending path is arranged according to the angle evolution direction, if there are adjacent bending segments with an angle difference less than 2°, they are merged into one segment and the angle center point is reconstructed, finally the bending angle sequence is output, each sequence unit contains starting point, ending point, angle value and direction vector, the total length of the angle sequence is consistent with the number of bending segments, the sequence is saved in matrix form, the dimension is [n, 4], during the process of multi-axis collaborative path mapping of variable profile data according to the angle serialization data, the five-axis motion control algorithm is used, the rotation accuracy of each control axis is set to be ±0.01°, the five-axis posture mapping function M(θ, x, y, z) is introduced in the path generation, wherein θ is the bending angle, (x, y, z) is the bending starting point coordinate, each bending angle is mapped to five-axis coordinate parameters through the forward kinematics model, forming the space bending path of the variable profile, the node spacing in the path is 0.5 mm, the control axis angle and the tool contact point pose are recorded at the node, the preliminary candidate bending trajectory path set is generated, the trajectory set is recorded in the form of discrete path group, each path data contains more than 1000 discrete points, the path format is five-dimensional vector group: [x, y, z, θ1, θ2], θ1 and θ2 are the tool angles of Y-axis and Z-axis respectively, during the process of bending simulation based on the candidate bending trajectory and deformation field mapping based on the simulated bending data, the finite element simulation engine ABAQUS is used to load and simulate each bending trajectory segment, the material model is set to be elastic-plastic constitutive relation, the Young's modulus in the material parameter is 210 GPa, the Poisson's ratio is 0.3, the yield strength is set to 280 MPa, the loading process of each bending section is solved by 100 steps of nonlinear incremental solution, and the equivalent plastic strain (PEEQ), equivalent stress (von Mises) and principal strain direction information of the bending region nodes are extracted after the solution is completed. The deformation information on the nodes is mapped back to the CAD space using a three-dimensional grid interpolation function to form a three-dimensional deformation field. The deformation field is expressed in the form of a topological grid, with each grid element having a size of 0.25 mm³. The generated deformation topological data volume is 512×512×64 grid elements. Each element in the interior contains a stress-strain pair, a node displacement vector, and a plastic deformation index value. In the process of detecting the bending conflict structure in the deformation topological data, the principal strain direction extracted from the deformation field tensor is used for vector angle analysis with the tool motion direction. The angle threshold is set to 45°. If the angle between the local deformation principal strain direction and the current bending motion direction exceeds the threshold, and the equivalent plastic strain value of the region exceeds 0.12, it is determined to be a bending conflict region. The voxel clustering algorithm is used to partition the continuous conflict region, and the minimum boundary volume of each conflict structure must exceed 50 grid elements. After extracting the bounding box of the conflict region, the conflict voxel set is output, and the conflict segment number is marked in the three-dimensional trajectory space to generate a conflict structure identification map. The map is represented as a binary volume, with 1 representing conflict voxels and 0 representing non-conflict. In the process of avoiding conflict trajectories based on bending conflict structures and planning bending actions according to the conflict avoidance trajectory, the path obstacle avoidance algorithm RRT* (Rapidly-Exploring Random Tree Star) is called for each conflict segment to adjust the bending trajectory. The maximum step size is set to 1.5 mm, and the maximum iteration number is set to 3000 times. In the path generation, the non-conflict region is used as the passable space, and the conflict identification map is used as the obstacle. A new trajectory path segment is dynamically generated, and the trajectory node interpolation uses five B-spline smoothing processing. The path curvature is controlled within 0.1, and the corner smoothing error is not more than 0.2°. The merged and globally optimized path segments are output as the final bending action sequence. Each action contains five-axis control parameters, tool contact point posture, bending angle instruction, and execution timestamp. The bending action sequence is in the form of time-ordered structured data, and the precision requirement is one frame of control data per 0.01s.

[0076] Preferably, the step S3 of predicting the pre-machining bending shape according to the planned bending sequence comprises:

[0077] reconstructing the variable bending part structure according to the variable profile data;

[0078] mapping the path torque of the variable bending part structure based on the planned bending sequence to obtain bending force guide data;

[0079] folding the bending force guide data by displacement curvature to generate configuration trend data;

[0080] Based on the configuration trend data, a geometric change of the variable bending structure is speculated to obtain a pre-processing bending shape.

[0081] In this embodiment, in the process of reconstructing the variable bending structure according to the variable profile data, first, all closed profile boundaries are extracted by using two-dimensional CAD drawing data through the Open CASCADE Technology modeling engine, the boundaries are expressed in vector format, each boundary is composed of multiple line segments and circular arc segments, the line segments and circular arcs are uniformly converted into a set of midpoint coordinates, start and end coordinates, direction angles and normal angles, after generating complete profile topology information, a custom geometry segmentation function is called to perform profile segmentation with 1mm precision for each segment, a discrete point set sequence is extracted as an input coordinate set, after the coordinate set is input into the boundary structure reconstruction module, the variable bending structure units are generated in sequence according to the topological continuity of the profile, each structure unit is numbered in a fixed order, each boundary structure is recorded in the structure body model according to its connection angle and length, the plate thickness is uniformly set to 3mm during modeling, the bending inner radius is set to 4mm, the bending angle is inferred from the included angle of the boundary segment and mapped into the two-dimensional vector model, after all structure units are generated, they are integrated into a standard structure library data, which is uniformly output as a STEP format file for subsequent bending physical mapping calculation, the connection mode between the boundary segments is set as non-elastic rotational connection, there are no more than two structure segments that coincide at the intersection node in the structure model, in the process of path torque mapping of the variable bending structure based on the planned bending sequence, first, the generated STEP structure file is imported into the bending path simulator, the path simulator is composed of analysis tools based on finite segment loading models, the tool supports the definition of one-to-one mapping relationship between structure segments and loading nodes, during torque mapping, the number of each structure unit in the planned path is corresponding to the structure segment index, virtual loading nodes are inserted at equal intervals in each bending segment, the bending simulation displacement is applied in sequence at the loading nodes, after each displacement is applied, the reverse internal force distribution of the segment is extracted and mapped as an equivalent static torque value, the data of all loading nodes is stored as a multi-dimensional matrix, each element of the matrix represents the local equivalent torque of the structure segment under a certain loading state, the matrix finally forms the bending force guide data, which is stored as a TXT format file with row and column markers in ASCII encoding, the structure segment numbers not included in the bending sequence are skipped and no data is generated, in the process of displacement curvature folding of the bending force guide data, the TXT format bending torque data is loaded into the deformation folding reasoning module, the module performs local affine transformation operation on each structure segment, when processing each structure segment, the corresponding torque value is first read according to the loading node sequence, and then the torque value is mapped into an angle increment and a linear displacement vector in sequence, all displacement changes are combined and superimposed in a two-dimensional plane, the superimposition method uses the main direction of the structure segment as the reference axis, and cumulative rotation and offset are performed according to the bending loading sequence, the rotation operation is realized by constructing a two-dimensional rotation matrix, and the offset operation is realized by constructing a translation vector, after all structure segments are processed, a set of configuration trend data is generated, which records the endpoint displacement and angle change information of each structure segment under different loading states,The transformation information of all structure segments is uniformly coded as binary block structure, and each data block of the structure segment contains the start point coordinate, the end point coordinate, the cumulative angle change and the cumulative translation vector. All data blocks are spliced in sequence to form a set of configuration trend trajectory files. In the process of geometric change prediction of the variable bending structure based on the configuration trend data, the configuration trend trajectory files are loaded into the structure feedforward deformation module. The module reconstructs the complete structure connection graph according to the structure segment number when inputting, and then reads the configuration transformation data of each segment in sequence from the first structure segment. When applying each segment transformation data, the start point coordinate of the current segment is first corrected according to the cumulative translation vector, and then a local rotation transformation matrix is constructed according to the cumulative rotation angle. The end point coordinate of the current segment is recalculated after the direction vector of the current segment is rotated and transformed. Then, the current end point coordinate is taken as the start point coordinate of the next segment to continue the geometric change prediction operation of the next segment. After each segment operation is completed, the new start and end point coordinates are written into the output list. Finally, a complete set of structure segment updated coordinate sequences is output.

[0082] Preferably, the corresponding position projection of the target bending data and the pre-processing bending shape and the difference comparison analysis in step S4 include:

[0083] extracting target feature control points of the target bending data;

[0084] extracting actual feature control points of the pre-processing bending shape;

[0085] constructing a bidirectional projection matrix based on the target feature control points and the actual feature control points;

[0086] performing corresponding position projection according to the bidirectional projection matrix to obtain corresponding position projection data;

[0087] performing distance measurement calculation based on the corresponding position projection data to generate a shape deviation vector;

[0088] performing directional analysis on the shape deviation vector to obtain a shape deviation trend;

[0089] identifying a critical deviation point based on the form deviation trend;

[0090] determining a regionalized gap according to the critical deviation point; and integrating and statistically analyzing the regionalized gap to determine a real-time processing gap.

[0091] In the process of extracting the target feature control points of the target bending data in this embodiment, first, the DXF file of the standard target pattern is loaded and all vector elements are read by using the custom pattern analysis module, wherein only the polyline structure composed of LINE elements and ARC elements is retained, the connection nodes between the consecutive line segments in the pattern are used as the preliminary control point candidates, the node included angle θ is constructed for each connection node before and after the two line segments by calling the node included angle calculation module, the nodes with an absolute value of the included angle greater than 3 degrees are marked as high-angle feature points, the angle threshold is set to 3 degrees according to the actual detection accuracy requirement, the marked nodes are sequentially numbered and the two-dimensional coordinate values, the corresponding line segment numbers and the local angle attributes are recorded, all the extracted control points are stored in the form of a sequential array, each control point in the array includes four basic parameters of X coordinate, Y coordinate, angle of the bend and line segment number, the array output format is a CSV file, all data uniformly uses three decimal places to represent the accuracy, the angle value is accurate to one decimal place, and the coordinate unit is unified as millimeter, in the process of extracting the actual feature control points of the pre-processing bending shape, first, the pre-processing bending pattern DXF file generated by the bending configuration prediction module is loaded, all vector data is uniformly converted into a continuous vector sequence of structure segments, the start and end point coordinates of each structure segment are extracted by calling the structure segment analysis module, and then the included angles between adjacent structure segments are calculated, the connection points with an included angle value greater than 3 degrees are extracted as the actual feature control points, each control point includes the structure segment number, the connection coordinates, the angle information and the sequential number in the whole structure, all control point information generates a two-dimensional control point sequence matrix, the number of rows of the matrix is equal to the total number of control points, and the number of columns is 4, which are X coordinate, Y coordinate, included angle and segment number in turn, in the process of constructing the bidirectional projection matrix based on the target feature control points and the actual feature control points, first, the coordinate vector sets are extracted from the target control point CSV file and the actual control point JSON file respectively, the target coordinate set is denoted as G=[g1, g2,...,gn], and the actual coordinate set is denoted as A=[a1, a2,...,an], when the number of control points is not equal, the shorter set is interpolated to fill the number, the interpolation method is linear interpolation, and the interval is controlled within 2 millimeters, then the projection matrix T and the inverse matrix T' are constructed based on the least squares registration algorithm, the projection matrix T is used to map G to the space of A, the inverse matrix T' is used to map A back to the space of G, the projection matrix T=[R|t] is composed of a rotation matrix R and a translation vector t, R is obtained by singular value decomposition (SVD) calculation, and the convergence threshold of SVD is set to 10 -8, the entire matrix combination forms a 4x4 homogeneous matrix form, in the process of corresponding position projection according to the bidirectional projection matrix to obtain the corresponding position projection data, each target control point gi is expressed as [gi_x, gi_y, 1] using homogeneous coordinates, and is sequentially left multiplied by the projection matrix T to obtain the corresponding projection point Ti(gi) of each gi in the actual coordinate system. Similarly, each actual control point ai is also inversely projected through the T' matrix to obtain Ti'(ai), and finally a set of bidirectional mapping data point groups is formed, each group of data containing target points, actual points, target mapping points and actual mapping points. All data point groups are written into a unified projection data structure body, each row of which contains original point coordinates, mapping point coordinates, point distance difference values and mapping offset angles. The structure body is exported in CSV format for use in the subsequent shape deviation calculation stage. The coordinate error during each pair of point projection operation execution process shall not exceed 0.1 millimeter. In the process of generating the shape deviation vector based on the corresponding position projection data, the coordinate information of each point pair in the projection data structure body is read, and the Euclidean distance between the target control point and the actual control point is calculated by calling the coordinate distance calculation module. Each distance is recorded as , wherein di is the Euclidean distance (i.e. the absolute value of the shape deviation) between the target control point and the actual control point, in millimeters, representing the bending offset at that position, gi x is the coordinate value of the i-th target feature control point in the X-axis direction, gi y is the coordinate value of the i-th target feature control point in the Y-axis direction, ai x is the coordinate value of the i-th actual feature control point in the X-axis direction, ai y is the coordinate value of the i-th actual feature control point in the Y-axis direction. The structure segment number and point number corresponding to each di are also recorded. All distance values are output in list form and constructed as a deviation vector D=[d1, d2,..., dn], each element being the spatial offset of the current point pair. The offset value unit is millimeter, and the vector data precision is set to 0.001 millimeter. The deviation vector is output as a CSV format file, wherein each row contains point pair number, corresponding segment number, target point coordinates, actual point coordinates and Euclidean distance value. In the process of directional analysis of the shape deviation vector to obtain the shape deviation trend, the target point gi and the actual point ai corresponding to each deviation value di are constructed into a directional vector , wherein vi is the i-th offset directional vector, ai x is a two-dimensional vector representing the direction and size of the offset, gi x is the coordinate value of the i-th actual control point in the X-axis, gi y is the coordinate value of the i-th actual control point in the Y-axis, gi yThe coordinate value of the i-th target control point on the Y-axis is denoted by a vector array v, each element of which is a two-dimensional vector. Then, an angle clustering analysis operation is performed on all direction vectors, using the K-Means clustering algorithm to divide all deviation directions into three direction classes. The maximum number of iterations is set to 300, and the convergence condition is that the angle change is less than 0.5 degrees. The clustering center direction is output in vector form. Finally, each deviation direction is labeled as a trend direction ID, which corresponds to the deviation value to form a deviation trend structure set. Each element of the structure set contains the control point number, the offset direction vector, the offset distance value, and the direction classification label. In the process of identifying critical deviation points based on shape deviation trends, the deviation trend structure set is traversed, and the points with offset values greater than the average value plus two standard deviations in each direction class are counted as critical deviation points. The average value μ and the standard deviation σ are calculated by standard statistical methods, and the threshold is set to μ + 2σ. All points exceeding this threshold are recorded as critical deviation points, which are represented by a list of point numbers. The list is output in TXT format, and the segment number of the critical deviation points is used to identify the high error area segment. The regional deviation gap is determined based on the critical deviation points, and the regional deviation gap is integrated to determine the real-time processing gap. In the process of reading all critical deviation point numbers and classifying them by structure segment number, the number of critical points and the average offset in each structure segment are counted. If the number of critical points in a segment is greater than 3 and the average offset is greater than 1.5 mm, the segment is marked as a gap area segment. All gap area segments are constructed into a gap segment array, and each segment in the gap segment array contains the segment number, the number of critical points, the average offset value, the maximum offset direction, and the corresponding trend direction label. Finally, all gap segment information is integrated into a real-time gap report, and the report output format is an XML document containing the total number of gap segments, the maximum offset segment number, the most concentrated feature point segment number, the average offset statistical table, and the corresponding coordinate index table.

[0092] Preferably, the step S4 of performing elastic rebound detection on the variable profile data based on the real-time processing gap comprises:

[0093] Determining a variation gap direction based on the real-time processing gap;

[0094] Performing material characteristic parameterization processing on the variable profile data to obtain material characteristic parameters;

[0095] Performing collision simulation based on the material characteristic parameters, and performing elastic response analysis based on the simulated collision data to obtain material elastic response data;

[0096] Quantifying the rebound influence value according to the real-time processing gap and the variation gap direction;

[0097] Based on the rebound influence value, the response strength of the material elastic response data is derived to obtain the elastic rebound strength;

[0098] Based on the change gap direction, the elastic rebound direction is determined.

[0099] The elastic rebound strength and the elastic rebound direction are fused to generate the elastic rebound data of the bending part.

[0100] In this embodiment, based on the real-time processing gap determines the change gap direction in the process, first from the deviation calculation module to extract each corresponding control point of the offset vector, offset vector is the target point coordinates and actual point coordinates between the space difference, all the extracted offset vector composition two-dimensional vector array, then all the vectors in the array are unitized direction extraction processing, that is, all the offset vectors are normalized and mapped to the two-dimensional direction circle to form an angle sequence, the angle sequence is divided into direction intervals from 0 to 360 degrees, using 5 degrees as the minimum statistical unit, the number distribution of all offset directions in each angle interval is counted and the direction histogram is constructed, the peak value extraction operation is performed on the histogram to determine the concentrated area of the maximum offset direction, the direction interval with a direction concentration higher than 75% is defined as the dominant gap direction, the direction is represented in the form of unit vector and reserved to the subsequent rebound direction derivation module, in the process of material characteristic parameterization processing of variable profile data, the profile structure of the bending part is imported into the material mapping module, each bending segment is numbered in the structure topology, each numbered structure segment corresponds to a set of material parameter set, the material parameters include elastic modulus, yield strength, strain hardening index, Poisson's ratio and density, each material parameter is selected by the user in the database and injected into the model, the parameter setting range is elastic modulus 195-210GPa, yield strength 200-350MPa, strain hardening index between 0.2 and 0.3, Poisson's ratio is 0.3, and density is set to 7.85g / cm³, all parameters establish mapping relationship according to structure segment index, the mapping result is a group of parameter matrix in the form of multi-dimensional tensor, the parameters of each structure segment are combined into material characteristic parameter file in the form of structure body, in the process of collision simulation based on material characteristic parameters and elastic response analysis based on simulation collision data, the JSON format material parameter file is imported into the finite element solving environment, the grid is divided with each structure segment as the simulation sub-domain, the grid accuracy is set to 0.1mm, the material model adopts elastic-plastic constitutive relation, the boundary condition is set to one-way compression in the bending loading direction, the simulation adopts explicit integration method, each structure segment is subjected to equal interval step displacement loading, the total number of loading steps is 100 steps, and each step increment is 0.02mm, the maximum deformation and stress response values of each node on each structural segment are recorded during the simulation process, and the maximum node displacement and average springback recovery distance of each segment are output after the simulation is completed to construct an elastic response dataset, each element of the dataset containing the structural segment number, maximum stress value, maximum displacement value, and springback recovery ratio. In the process of quantifying the springback influence value according to the real-time processing gap and the change gap direction, first, the maximum offset value and direction vector of each structural segment are read, and then the offset direction vector is projected with the dominant gap direction. The segment data with an included angle less than 15 degrees is retained as the effective springback area. Then, the product of the springback direction vector and the offset value of each structural segment is taken as the springback influence base value, which is divided by the average length of the corresponding segment and multiplied by the offset value to obtain the springback influence intensity scalar of each segment. All intensity scalar values form a springback influence vector, which is indexed by the structural segment number to construct a table data, each row containing the segment number, influence intensity value, offset direction, and unit direction vector. In the process of deriving the response strength based on the springback influence value and the material elastic response data, the springback influence vector and the previously generated elastic response dataset are aligned by segment number. After alignment, point multiplication is performed on each structural segment to combine the springback influence intensity value with the corresponding segment elastic recovery ratio to obtain the comprehensive springback intensity index. The comprehensive springback intensity value represents the geometric change intensity caused by actual springback under the current material condition and loading state. The intensity index is sorted by structural segment number, and the maximum value is taken as the upper limit of the structural overall springback intensity. The springback intensity of all segments is stored in a list form, with each data including the segment number, direction vector, intensity value, and corresponding coordinate index position. The final output is a springback intensity distribution map in the structural segment dimension. In the process of determining the elastic springback direction based on the change gap direction, the initial direction vector of each structural segment is compared with its dominant offset direction vector in terms of vector angle. If the angle is less than 20 degrees, the direction is recorded as the effective springback direction. If the angle is greater than 20 degrees, the symmetric direction is replaced. After normalization, all direction vectors are converted into unit vector format to generate a structure segment number and direction vector table. Each record in the table contains the segment number, unit direction coordinates, and main offset angle. In the process of data fusion of elastic springback intensity and elastic springback direction to generate the bending part elastic springback data, the springback intensity distribution map and the unit direction vector table are merged into a unified compensation data structure. In the data structure, each structural segment is matched with its direction and intensity value. The final springback vector is obtained by multiplying the direction vector by the intensity value. Each springback vector represents the elastic recovery direction and recovery amplitude of the structural segment during the bending process. All springback vectors form a bending part elastic springback dataset, which is indexed by the segment number. Each segment contains a two-dimensional recovery vector and a vector length value. The final output is a standard DXF layer file for overlay plotting comparison.

[0101] Preferably, step S5 comprises the following steps:

[0102] Step S51: Principal component decomposition is performed on the elastic springback data of the bent part to obtain a set of main springback directions; a springback vector field is constructed based on the set of main springback directions;

[0103] Step S52: Direction transformation calculation is performed on the elastic springback data of the bent part according to the springback vector field to generate initial compensation vectors;

[0104] Step S53: Critical point compensation simulation is performed based on the initial compensation vectors to generate local compensation data; global coordinated compensation is performed according to the local compensation data to generate the bending springback compensation data;

[0105] Step S54: Feature point mapping is performed on the original drawing of the bent part according to the bending springback compensation data to obtain feature correspondence data;

[0106] Step S55: The bending springback compensation data is converted into processing parameter correction values based on the feature correspondence data to obtain gap compensation parameters.

[0107] In this embodiment, when the elastic springback data of the bent part is subjected to principal component decomposition, the principal component analysis (PCA) is used to process the springback displacement matrix composed of three-dimensional coordinates, the data matrix size is N x 3, N is the number of sampling points, each sampling point includes the springback displacement values in X, Y and Z directions, the singular value decomposition (SVD) algorithm is used to process the matrix, the first three principal component direction vectors are selected as the set of main springback directions, the principal component contribution rate threshold is set to 95%, the function svd(A) is called in the MATLAB platform to obtain three matrices U, S and V, the first three vectors in the V matrix are used as the main direction vectors V1, V2 and V3, the three groups of vectors are used to construct a three-dimensional springback vector field model, the scatter interpolation method is used to interpolate the main direction vectors at each sampling point to generate a complete vector distribution field, the distribution field is visualized and mapped in a uniform grid manner in the three-dimensional space, the grid density is set to 1 mm interval to ensure sufficient direction density sampling, when the direction transformation calculation is performed according to the springback vector field, the angle between the springback displacement vector of each sampling point and the direction vector in the springback vector field is used for orthogonal projection, the original springback vector is decomposed into a main direction component and a residual direction component, the main direction component is used to generate the initial compensation vector, the original springback vector is R, the main direction vector is V, and the initial compensation vector is , dot function represents dot product operation, all points of B vector set constitute initial compensation field, all vectors are normalized and multiplied by compensation weight factor λ, λ is set to 0.85, to avoid overcompensation or miscompensation phenomenon, finally the initial compensation vector set of all sampling points is obtained, the matrix calculation process is completed in Python environment using NumPy library, the obtained result is saved as initial compensation field coordinate point cloud, format is XYZ+B_x+B_y+B_z, when carrying out critical point compensation simulation based on initial compensation vector, first, the structure finite element modeling method is used to divide the processed bending part into refined grid model, the grid size is set to 1mm*1mm, Abaqus platform is used to carry out static mechanics analysis on the bending area, key monitoring points are established for groove corner, edge bending starting point and material thickness mutation area, initial compensation vector is introduced into the model as displacement boundary condition, different compensation directions and amplitudes are simulated for different areas, after simulation, key point coordinate change data is exported, called local compensation data, moving least square method is used to carry out interpolation expansion of these local compensation points to the whole part surface, Gaussian kernel function is selected as interpolation kernel function, kernel width is set to 3mm, according to the full coverage compensation vector field, global coordinated compensation is carried out on the whole bending part surface through weight weighted average method, so that the compensation vectors of all points remain relative consistency with the local compensation field, thereby generating final bending springback compensation data, when mapping feature points according to the bending springback compensation data, first, CAD vector data in the original pattern is subjected to surface parameterization modeling, B-spline (B-Spline) is used to represent curve contour, all control points are selected as feature point set, let the original control point coordinates be , the corresponding coordinates in the bending springback compensation data be , the mapping matrix T from to is constructed, wherein T is a nonlinear affine transformation matrix, least square matching method (Least-Square Matching) is used to carry out fitting optimization in Matlab through function lsqnonlin(), the target is to minimize The mapping set of the feature points of the whole pattern is obtained after solving, and the mapping result is a mapping data file containing the original and compensation coordinates of the control points. When the bending springback compensation data is converted into the machining parameter correction value based on the corresponding data of the features, the reverse modeling method is adopted to convert the coordinate difference ΔP=P'-P into the bending blade angle adjustment value Δθ and the mold stroke correction amount Δd. The material yield limit is set to 245 MPa, the elastic modulus is set to 210 GPa, and the bending radius R is set to 1.5 times the material thickness t. The functional relationship between ΔP and Δθ and Δd is established under the simulation condition, and the empirical relationship is obtained through the experimental calibration curve Δθ=k1×ΔP_total and Δd=k2×ΔP_total, wherein ΔP_total is the modulus of the compensation vector in each direction, k1=8.6° / mm, and k2=2.3 mm / mm. All compensation data are converted into the machining blade adjustment parameters and the mold height adjustment amount according to the above relationship, and a gap compensation parameter table is generated, which is in the format of workpiece number, control point number, Δθ, and Δd. The table is imported into the numerical control system control module of the bending machine, and is read by the real-time drawing module and dynamically refreshed to obtain the machining part pattern contour coordinate set.

[0108] Especially important is that step S52 comprises:

[0109] According to the feature points of the springback vector field, a feature weight matrix is constructed;

[0110] The feature weight matrix is subjected to nonlinear coordinate conversion to obtain a coordinate transformation field;

[0111] The coordinate transformation field is subjected to shortest path planning to generate transformation path data;

[0112] Based on the transformation path data, a compensation vector set is determined; based on the compensation vector set, clustering analysis is performed to obtain a compensation feature dimension;

[0113] Based on the compensation feature dimension, the compensation vector is reconstructed to generate an initial compensation vector.

[0114] In this embodiment, the feature points in the rebound vector field are indexed. First, the spatial coordinates of each feature point are classified according to the main rebound direction. Utilizing the differences in direction and position of the rebound data, the position and rebound vector of each feature point in three-dimensional space are calculated. Using the rebound vector length and direction of each feature point as input, a feature weight matrix is ​​constructed using a local weighting method. Specifically, each weight in the weight matrix is ​​calculated as a function of the Euclidean distance between two points. Assuming an adjacency threshold of 2.5 mm, the local density weighting coefficient is defined as w = exp(-d² / (2*1.2²)), where d is the Euclidean distance between two points, and σ is set to 1.2 mm, resulting in an n×n matrix where n is the number of feature points. After obtaining the feature weight matrix, a nonlinear coordinate transformation is performed, using the isomap method from manifold learning for dimensionality reduction. By selecting a neighborhood radius of 3 mm for each point and setting the adjacency boundary to 8, the shortest path distance between points is calculated, thereby mapping the feature points from three-dimensional space to a two-dimensional plane. This method simplifies the coordinates of feature points, preserving the original local structure of each feature point's positional relationship in the output two-dimensional coordinate space. A coordinate transformation field is generated, resulting in an M×2 coordinate matrix, where M is the number of mapped feature points. In the transformed coordinate space, a path planning method based on the A* algorithm is used to determine the shortest path. The heuristic function is set as the Manhattan distance h(n) = |xx|. t |+|yy t |, where (x) t ,y t Let be the coordinates of the target point, and g(n) be the path cost function, representing the cumulative cost from the starting point to the current node. During path planning, to prevent excessive backtracking, a minimum weight threshold of 0.6 is set for each path node, ultimately generating a transformed path dataset. Each path point contains its two-dimensional coordinates and corresponding three-dimensional compensation offset. Based on the compensation offset values ​​in the path data, a compensation vector set is constructed. Each compensation vector vᵢ=(Δxᵢ,Δyᵢ,Δzᵢ) is calculated from the difference between the original position and the target position of the transformed path point. Subsequently, these compensation vectors undergo cluster analysis using the k-means clustering algorithm, classifying them according to their magnitude and direction. The initial number of cluster centers is set to 6, and Euclidean distance is used to measure the similarity between vectors. The clustering results will serve as the basis for compensation feature dimension vectors in subsequent steps. After the clustering analysis is completed, the centroid of each cluster is extracted as the set of compensation feature dimension vectors. These feature dimension vectors will be used to reconstruct the compensation vector. The reconstruction of the compensation vector is accomplished through linear interpolation, and a weighted average is performed on the compensation vector of each class to obtain the standard initial compensation vector V0={v 01 ,v 02 ,...,v 0m} and ensure that the length of all vectors is between 0 and 5 mm and the direction error is not more than 2°.

[0115] Preferably, step S6 comprises the following steps:

[0116] Step S61: Geometric transformation of variable profile data according to gap compensation parameters to generate compensated profile parameters; based on the compensated profile parameters, a machining simulation environment is constructed, wherein the environment simulation accuracy is controlled at the level of 0.01 mm grid, the temperature field fluctuation range is kept within ±3℃, and the simulation step is 0.1s;

[0117] Step S62: Secondary machining simulation of variable profile data according to the machining simulation environment to generate secondary machining simulation data, wherein the loading speed range in the simulation process is controlled at 1-10mm / s, the material elastic modulus is set to the actual material value ±5% floating range, and the friction coefficient is limited to 0.1-0.3;

[0118] Step S63: Determining the secondary machining evolution sequence based on the secondary machining simulation data; extracting the machining part evolution result shape according to the secondary machining evolution sequence;

[0119] Step S64: Real-time shape gap comparison of the machining part evolution result shape and the target bending data, and dynamic machining parameter fine-tuning according to the real-time comparison data.

[0120] In this embodiment, according to the position and shape of each point in the contour data, affine transformation processing is performed through the set compensation parameters, including rotation angle, scaling factor and offset, which are applied through matrix operation. The specific steps include: for each data point, first perform position offset (translation), then adjust the size of the contour according to the compensation parameters (scaling), and finally adjust the direction of the contour according to the set rotation angle. After the compensation, the data is subjected to geometric transformation to generate the compensated contour parameters, which are used as input for subsequent simulation environment construction. The precision control of the simulation environment is set to 0.01 mm grid level, meaning that the length of each grid in the simulation process is 0.01 mm, accurate to the micron level. In addition, the temperature field fluctuation range in the simulation is controlled within ±3°C to ensure that the temperature change in the simulation environment conforms to the actual processing situation. The simulation step is set to 0.1s, and the time interval between each calculation step is 0.1s. The compensated contour data is imported into the processing simulation environment, and the simulated loading speed range is set between 1mm / s and 10mm / s, which is suitable for deformation characteristics in different processing scenarios. The material elastic modulus is set to a floating range of ±5% of the actual material value to ensure that the physical properties of the material conform to the actual situation. For example, if the material is aluminum alloy, its elastic modulus can be set to 66.5GPa to 73.5GPa. The friction coefficient range is set to 0.1 to 0.3 to adapt to the material deformation behavior under different surface friction conditions. Under these conditions, secondary processing simulation is performed to obtain simulation data and record the deformation characteristics in each processing step. By analyzing the simulation data of each processing step, the deformation and stress distribution at each time are extracted. Using time series analysis technology, the sequence of shape evolution at each time is established. Each item in the evolution sequence reflects the changes the material undergoes during processing, including shape change, stress distribution and material state evolution. Through processing of these sequences, evolution data at each time during processing is obtained, forming a complete processing evolution sequence. The evolution result shape of the workpiece is extracted from the secondary processing evolution sequence. Through curve fitting and interpolation algorithm, the shape data at each time is reconstructed to obtain the actual shape of the workpiece at each time during the entire processing process. Interpolation technology is used to fill in the discontinuous points to ensure the smoothness of the evolution result shape. The final workpiece shape data is formed by extracting the data points in the evolution sequence, which is used for comparison with the target bending data. When comparing the evolution result shape of the workpiece and the target bending data in real time, point cloud matching algorithm is used to compare the evolution shape of the workpiece with the target shape point by point. The Euclidean distance between the two is calculated to determine the difference value of each point. If the shape difference of a certain part exceeds the set threshold (e.g. 0.05mm), the processing parameters are dynamically adjusted according to the difference size, adjusting the loading speed, processing force or other key parameters.

[0121] Thus, the embodiments should be considered in all respects as illustrative and not restrictive, the scope of the application being indicated by the appended claims rather than by the description given above, and all changes which come within the meaning and range of equivalency of the claims are intended to be embraced therein.

[0122] The foregoing is considered as illustrative only of the principles of the application. Numerous modifications and changes will readily occur to those skilled in the art, and the generic principles defined herein can be applied to other embodiments without departing from the spirit or scope of the application. Therefore, the application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for verifying real-time drawing of a workpiece pattern of a numerical control system of a bending machine, characterized in that, The method comprises the following steps: Step S1: obtaining an original drawing of a bending workpiece; identifying geometric element features of the original drawing of the bending workpiece, and performing boundary feature cutting based on the geometric element features to generate feature region data; Based on the preset workpiece material data, the load safety margin of the feature region data is calculated to obtain a safety threshold boundary; Step S2: decomposing the safety threshold boundary in the principal stress direction, and constructing a continuous stress flow field; Based on the continuous stress flow field, the geometric element features are reconstructed by geometric constraint, and variable profile data is identified; Step S3: obtaining target bending data; Based on the target bending data, the variable profile data is bent trajectory planning to obtain a planned bending sequence; and the pre-processing bending shape is predicted according to the planned bending sequence; Step S4: projecting the target bending data and the pre-processing bending shape to the corresponding positions, and performing difference comparison analysis to generate real-time processing difference; based on the real-time processing difference, the variable profile data is detected for elastic rebound to obtain bending workpiece elastic rebound data; Step S5: based on the bending workpiece elastic rebound data, the rebound compensation is performed to generate bending rebound compensation data; and according to the bending rebound compensation data, the compensation parameter mapping of the original drawing of the bending workpiece is performed to obtain the difference compensation parameter; Step S5 comprises the following steps: Step S51: decomposing the bending workpiece elastic rebound data into principal component to obtain a set of principal rebound directions; and constructing a rebound vector field based on the set of principal rebound directions; Step S52: calculating the direction transformation of the bending workpiece elastic rebound data according to the rebound vector field to generate an initial compensation vector; Step S53: based on the initial compensation vector, a critical point compensation simulation is performed to generate local compensation data; and a global coordinated compensation is performed according to the local compensation data to generate the bending rebound compensation data; Step S54: according to the bending rebound compensation data, the feature point mapping of the original drawing of the bending workpiece is performed to obtain feature corresponding data; Step S55: based on the feature corresponding data, the bending rebound compensation data is converted into a processing parameter correction value to obtain the difference compensation parameter; Step S6: according to the difference compensation parameter, a secondary processing simulation is performed on the variable profile data to generate secondary processing simulation data; and a real-time comparison is performed based on the secondary processing simulation data and the target bending data, and a dynamic processing parameter fine tuning is performed according to the real-time comparison data.

2. The method according to claim 1, wherein the method is characterized by: The identification of the geometric element features of the original drawing of the bending workpiece in step S1 and the boundary feature cutting based on the geometric element features comprise: identifying the geometric element features of the original drawing of the bending workpiece; constructing a topological relationship according to the geometric element features, and drawing a connection topological graph based on the topological relationship, wherein the minimum distance between nodes in the connection topological graph should not be less than 0.1 mm, and the maximum connectivity error in the topological path should be kept within 0.05 mm; identifying the drawing boundary contour based on the connection topological graph; performing solid feature cutting on the drawing boundary contour, and performing region cutting according to a minimum unit segmentation size of 1 mm², the aspect ratio of the cutting region is controlled within 1:5~5:1, and the feature region data is generated.

3. The method according to claim 1, wherein the method is characterized by: The load safety margin calculation of the feature region data based on the preset workpiece material data in step S1 comprises: The feature region data is associated with the material type based on the preset machining part material data, and material distribution data is generated; Anisotropy index projection is performed on the material distribution data, and a directional intensity matrix is constructed based on the projected data; Material attribute mapping is performed on the feature region data according to the directional intensity matrix to obtain a material parameter set; The material microstructure is reconstructed according to the material parameter set, and the constitutive relationship is converted based on the material microstructure to generate a material stress tensor; The critical load distribution is calculated based on the material stress tensor, and the safety margin is limited based on the critical load distribution to obtain a safety threshold boundary.

4. The method according to claim 1, wherein the method is characterized by: Step S2 includes the following steps: Step S21: identifying the surface features of the safety threshold boundary, and identifying the feature control region based on the surface feature recognition; based on the feature control region, a finite element grid is obtained by grid division; Step S22: based on the finite element network, the principal stress direction is decomposed to obtain a principal stress direction vector; and a continuous stress flow field is constructed according to the principal stress direction vector; Step S23: based on the continuous stress flow field, density gradient mapping is performed, and stress hot zone data is extracted based on the density gradient; material density distribution analysis is performed according to the stress hot zone data to obtain the material density distribution; Step S24: identifying the unit stress concentration point based on the material density distribution; and performing local shape correction on the unit stress concentration point to obtain a corrected stress concentration point; Step S25: based on the corrected stress concentration point, stress distribution equalization adjustment is performed on the geometric element feature, and geometric constraint reconstruction is performed based on the stress distribution equalization adjustment data to generate adjustment limitation conditions; Step S26: identifying the unrestricted contour of the geometric element feature according to the adjustment limitation conditions to obtain variable contour data.

5. The method of claim 1, wherein the method further comprises: determining a bending angle of the bending machine; and determining a bending length of the bending machine. The bending trajectory planning of the variable contour data based on the target bending data in step S3 includes: boundary constraint mapping is performed on the target bending data to obtain boundary constraint data; bending angle serialization processing is performed based on the boundary constraint data to obtain angle serialization data; According to the angle serialization data, multi-axis collaborative path mapping is performed on the variable contour data to generate a candidate bending trajectory; bending simulation is performed based on the candidate bending trajectory, and deformation field mapping is performed based on the simulated bending data to generate deformation topology data; detecting the bending conflict structure in the deformation topology data; collision trajectory avoidance is performed on the candidate bending trajectory based on the bending conflict structure, and bending action planning is performed according to the collision avoidance trajectory to obtain a planned bending sequence.

6. The method of claim 1, wherein the method further comprises: The prediction of the pre-machining bending shape according to the planned bending sequence in step S3 includes: reconstructing the variable bending part structure according to the variable contour data; path torque mapping is performed on the variable bending part structure based on the planned bending sequence to obtain bending force guidance data; displacement curvature folding is performed on the bending force guidance data to generate configuration trend data; geometric change speculation is performed on the variable bending part structure based on the configuration trend data to obtain the pre-machining bending shape.

7. The method of claim 1, wherein the method further comprises: determining a bending angle of the bending machine; and determining a bending length of the bending machine. The corresponding position projection of the target bending data and the pre-machining bending shape in step S4 includes: extracting the target feature control point of the target bending data; extracting the actual feature control point of the pre-machining bending shape; The bidirectional projection matrix is constructed based on the target feature control points and the actual feature control points; The corresponding position projection data is obtained by performing corresponding position projection according to the bidirectional projection matrix; The shape deviation vector is generated by performing distance metric calculation based on the corresponding position projection data; The shape deviation trend is obtained by performing directionality analysis on the shape deviation vector; The critical deviation point is identified based on the shape deviation trend; The regionalized deviation gap is determined according to the critical deviation point; and the real-time processing gap is determined by integrating the regionalized deviation gap.

8. The method of claim 1, wherein the method further comprises: determining a bending angle of the bending machine; and determining a bending length of the bending machine. The elastic rebound detection of the variable profile data based on the real-time processing gap in step S4 includes: The change gap direction is determined based on the real-time processing gap; The material characteristic parameter is obtained by performing material characteristic parameterization processing on the variable profile data; The material elastic response data is obtained by performing collision simulation based on the material characteristic parameter and performing elastic response analysis based on the simulated collision data; The rebound influence numerical value is quantified according to the real-time processing gap and the change gap direction; The elastic rebound strength is derived based on the rebound influence numerical value and the material elastic response data, so as to obtain the elastic rebound strength; The elastic rebound direction is determined based on the change gap direction; The elastic rebound data of the bent part is generated by data fusion of the elastic rebound strength and the elastic rebound direction.

9. The method of claim 1, wherein the method further comprises: determining a bending angle of the bending machine; and determining a bending length of the bending machine. Step S6 includes the following steps: Step S61: Geometric transformation is performed on the variable profile data according to the gap compensation parameter to generate the compensated profile parameter; the processing simulation environment is constructed based on the compensated profile parameter, wherein the environment simulation accuracy is controlled at the 0.01mm grid level, the temperature field fluctuation range is kept within ±3℃, and the simulation step is 0.1s; Step S62: Secondary processing simulation is performed on the variable profile data according to the processing simulation environment to generate secondary processing simulation data, wherein the loading speed range is controlled at 1-10mm / s during the simulation process, the material elastic modulus is set to the actual material value±5% floating range, and the friction coefficient is limited to 0.1-0.3; Step S63: The secondary processing evolution sequence is determined based on the secondary processing simulation data; and the processing part evolution result shape is extracted according to the secondary processing evolution sequence; Step S64: Real-time shape gap comparison is performed on the processing part evolution result shape and the target bending data, and dynamic processing parameter fine-tuning is performed according to the real-time comparison data.

Images