A multi-type connection automatic modeling method for whole vehicle CAE multi-analysis target

Abstract

This application relates to an automatic modeling method for multiple connection types in CAE (Computer-Aided Engineering) for whole vehicles with multiple analysis objectives. The method involves scanning several candidate connection objects in the whole vehicle model and constructing feature vectors for each candidate connection object based on the whole vehicle model information. Based on the analysis objective and feature vectors, the connection type of the candidate connection for each candidate connection object is determined. Topological constraint matching is performed on the candidate connections based on the feature vectors, and a comprehensive matching score is calculated for each candidate connection. Based on the connection type of the candidate connection whose comprehensive matching score is not lower than a preset score threshold, the method generates the connection implementation method of the candidate connection under the corresponding analysis objective. The method automatically connects the parts in the corresponding candidate connection objects based on the connection implementation method until all candidate connection objects have been traversed, completing the connection of the whole vehicle's parts. When the connection quality of the whole vehicle is unqualified, adaptive repair is performed, thus realizing automatic modeling of multiple connection types in the whole vehicle model under the corresponding analysis objective.

Description

Technical Field

[0001] This application relates to the field of automated modeling technology for whole vehicle CAE, and in particular to an automated modeling method for multiple types of connections for multi-analysis objectives in whole vehicle CAE. Background Technology

[0002] In existing vehicle CAE modeling processes, existing technologies generate connection units based on connection information data files, extract weld point / adhesive / weld / bolt connection information from the original data of opening and closing components and generate connection files, and batch create weld point, adhesive, bolt, and weld connections in the pre-processing workflow and perform connection checks. However, existing methods have the following shortcomings: 1. Insufficient joint judgment of connection type and connection object: The existing process usually generates connections based on the weld point table, connection file, hole features or local geometric relationships. When the CAD, mesh, assembly level or part naming changes, it is easy to have problems such as incorrect connection, missing connection, duplicate connection or misjudging ordinary hole as bolt connection.

[0003] 2. Lack of automatic mapping for connection implementation methods under different analysis objectives: Existing connection tools can create connections in batches, but they usually do not uniformly associate analysis objectives such as stiffness, modal / NVH, durability and collision with connection implementation methods, stiffness levels and failure expressions, which can easily lead to local stiffness being artificially high, modal frequency shifts or fatigue hot spot distortions.

[0004] 3. Lack of executable cause diagnosis after connection failure: Existing processes can usually only give the status of connection not being implemented or connection failure, and it is difficult to automatically distinguish the causes such as missing objects, connection distance exceeding tolerance, holes not being on the same axis, shell element normal error, abnormal number of connection layers, poor local mesh quality, misconnected parts or abnormal connection stiffness.

[0005] 4. Lack of parameterized closed-loop connection repair: When a connection fails, engineers need to locate and manually modify each connection individually. There is a lack of mechanisms to call repair actions such as search radius adjustment, connection point reprojection, hole centerline reconstruction, normal correction, layer re-identification, and connection mode switching based on the failure type.

[0006] 5. Lack of a unified evaluation of connection quality for the whole vehicle model: The whole vehicle model contains a large number of connection objects. Relying solely on manual spot checks cannot systematically evaluate the connection success rate, misconnection rate, duplicate connection rate, abnormal connection stiffness rate, and risk level of critical connections. It is also difficult to trace the cause of failure and the repair process.

[0007] 6. Poor model version reusability and rule migration: After changes in CAD, mesh, part naming and assembly level of the whole vehicle model, the original connections often need to be repaired. Existing connection files or local scripts are difficult to migrate stably to new models, new configurations or new analysis targets. Summary of the Invention

[0008] Therefore, it is necessary to provide an automatic modeling method for multiple connection types in vehicle CAE with multiple analysis targets. This method aims to solve the problems still existing in current automatic modeling and connector batch creation technologies for vehicle CAE, such as unstable connection object judgment, mismatch between connection type and analysis target, undiagnostic reasons for connection failure, non-reusable repair actions, and untraceable connection quality. This method is based on a finite element preprocessing platform and includes: S1: Import the complete vehicle model from CAD or CAE and read the complete vehicle model information, including parts, materials, plate thickness, attributes, and analysis targets. S2: Scan out several candidate connection objects in the whole vehicle model, and construct the feature vector of each candidate connection object based on the whole vehicle model information; determine the candidate region to which the candidate connection object belongs based on the feature vector of the candidate connection object; S3: Within the candidate region, based on the analysis target and feature vector, determine the connection type of the candidate connection corresponding to the candidate connection object; perform topological constraint matching on the candidate connection based on the feature vector, and calculate the comprehensive matching score corresponding to the candidate connection according to the satisfaction of various topological constraints; S4: Based on the connection type, analysis target, material, plate thickness, and attributes of candidate connections whose comprehensive matching score is not lower than a preset score threshold, generate connection implementation methods for candidate connections under the corresponding analysis target; automatically connect the parts in the corresponding candidate connection objects based on the connection implementation methods; S5: Repeat S3-S4 until all candidate connection objects have been traversed to complete the connection of the vehicle parts. When the connection quality of the whole vehicle is unqualified, adaptive repair is performed to realize the automatic modeling of multiple types of connections of the whole vehicle model under the corresponding analysis target.

[0009] Preferably, the candidate connection objects include the set of parts in the plate overlap area, hole-to-hole coaxial area, hole-to-mounting surface area, adhesive surface area, continuous boundary fitting area, small gap area, and bushing mounting area.

[0010] Preferably, the feature vector of the candidate connection object includes: connection distance feature, hole center or geometric center matching feature, normal consistency feature, overlap area or overlap length feature, part semantic matching feature, number of connection layers feature, and local connection stiffness level.

[0011] Preferably, determining the connection type of the candidate connection object based on the analysis target and feature vector includes: ; in, Indicates the first Connection types of candidate connections, In the eigenvector and analysis objectives The following candidate connections belong to the connection type. The probability, This represents a set of connection types; the connection types in the set include spot welding, bolting, adhesive bonding, seam welding, bushing, and contact connection.

[0012] The preferred formula for calculating the comprehensive matching score is: ; in, Indicates the first A comprehensive matching score for each candidate connection. Indicates distance score, Indicates the score for matching the hole center or geometric center. Indicates the normal consistency score. The score indicates the overlap area or overlap length. This indicates the semantic matching score of the parts. Indicates the score based on the number of connection layers. , , , , , They are respectively , , , , , The weighting coefficients.

[0013] Preferably, the expression for the connection implementation method is: ; in, Indicates the first The connection implementation methods for each candidate connection. This indicates that the overall matching score is not lower than the preset score threshold. Connection types of candidate connections, Indicate the analysis objective, Indicates the first The local geometric features of the candidate connections include connection distance features, hole center or geometric center matching features, normal consistency features, overlap area or overlap length features, and number of connection layers features. Indicates material, plate thickness, and property information. Indicates the target connection stiffness level. This represents the mapping function for the connection implementation method driven by the analysis target.

[0014] Preferably, the analysis objectives include stiffness analysis, modal / NVH analysis, durability analysis, and collision analysis.

[0015] Preferably, the formula for calculating connection quality is: ; in, Indicates the overall vehicle connection quality. Indicates the connection success rate. Indicates the connection failure rate. Indicates the fault connection rate. Indicates the repeat connection rate. Indicates the connection stiffness anomaly rate. , , , , They represent , , , , The weighting coefficients.

[0016] Preferably, the adaptive repair when the overall vehicle connection quality is substandard includes: The causes of connection failures are attributed to the following: missing connection objects, connection distance exceeding tolerance, incorrect normal, misaligned holes, abnormal number of connection layers, incorrect connection, poor local mesh quality, and abnormal connection stiffness. Based on the reasons for failed connections, the feature vectors of the candidate connection objects corresponding to the failed connections are adaptively repaired, and S2-S5 are executed based on the repaired feature vectors until the connection quality of the whole vehicle is qualified.

[0017] Preferably, the adaptive repair of the feature vector of the candidate connection object corresponding to the failed connection based on the cause of the failed connection includes: ; in, Indicates the first During the nth iteration The feature vectors of the candidate connection objects corresponding to each connection. This represents an adaptive repair function. Indicates the first During the nth iteration The feature vectors of the candidate connection objects corresponding to each connection. Indicates the first Reasons for connection failure Indicates the reason for the failed connection. The corresponding repair parameters.

[0018] Beneficial effects: This method imports a complete vehicle model from CAD or CAE and reads the vehicle model information; scans out several candidate connection objects in the vehicle model and constructs feature vectors for each candidate connection object based on the vehicle model information; within the candidate region, based on the analysis target and feature vectors, determines the connection type of the candidate connection of the corresponding candidate connection object; performs topological constraint matching on the candidate connection based on the feature vectors, and calculates the comprehensive matching score corresponding to the candidate connection based on the satisfaction of various topological constraints; generates the connection implementation method of the candidate connection under the corresponding analysis target based on the connection type of the candidate connection with a comprehensive matching score not lower than a preset score threshold and the analysis target; automatically connects the parts in the corresponding candidate connection object based on the connection implementation method until all candidate connection objects have been traversed, completing the connection of the parts of the whole vehicle; when the connection quality of the whole vehicle is unqualified, adaptive repair is performed, realizing automatic modeling of multiple types of connections of the whole vehicle model under the corresponding analysis target. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart of an automatic modeling method for multiple types of connections for multi-analysis targets in vehicle CAE, as described in this application.

[0021] Figure 2 This is a schematic diagram illustrating the identification of candidate connection regions in an embodiment of this application.

[0022] Figure 3 The diagram shows the automatic identification effect of each connection type in the embodiments of this application.

[0023] Figure 4 This is a comparative diagram showing the automatic implementation of local connections before and after in the embodiments of this application.

[0024] Figure 5 This is a closed-loop flowchart of connection failure diagnosis and adaptive repair in the embodiments of this application.

[0025] Figure 6 This is a schematic diagram illustrating the process of re-pairing and repairing after connection failure caused by misalignment of holes in an embodiment of this application.

[0026] Figure 7 This is a diagram showing the output effect of the vehicle connectivity quality report in this embodiment. Detailed Implementation

[0027] To make the above-mentioned objectives, features, and advantages of this application more apparent and understandable, the specific embodiments of this application are described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of this application. However, this application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of this application. Therefore, this application is not limited to the specific embodiments disclosed below.

[0028] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0029] This embodiment provides an automated modeling method for multiple connection types in vehicle CAE with multiple analysis targets. It aims to address existing problems in automated vehicle CAE modeling and connector batch creation technologies, such as unstable connection object identification, mismatch between connection type and analysis target, undiagnostic reasons for connection failures, non-reusable repair actions, and untraceable connection quality. This method forms a complete technology chain from candidate feature construction to failure repair. It can be implemented in finite element preprocessing platforms such as HyperMesh via script interfaces or secondary development interfaces, and can also be extended to other preprocessing environments with connector or finite element element creation capabilities.

[0030] like Figure 1 As shown, this embodiment provides a multi-type connectivity automatic modeling method for multiple analysis objectives in whole-vehicle CAE, including: S1: Import the complete vehicle model from CAD or CAE and read the vehicle model information, which includes parts, nodes, units, materials, plate thickness, attributes, assembly level, local coordinate system, and analysis targets. The analysis targets include, but are not limited to, stiffness analysis, modal analysis, NVH analysis, durability analysis, and collision analysis.

[0031] S2: Scan out several candidate connection objects in the vehicle model, and construct the feature vector of each candidate connection object based on the vehicle model information; determine the candidate region to which the candidate connection object belongs based on the feature vector of the candidate connection object.

[0032] In this embodiment, a number of candidate connection objects are formed by performing geometric and topological scanning on the whole vehicle model, instead of directly converting existing connection information files into connection units. The candidate connection objects include a collection of parts in the plate overlap area, hole-to-hole coaxial area, hole-to-mounting surface area, adhesive surface area, continuous boundary fitting area, small gap area, and bushing mounting area.

[0033] Specifically, determining the candidate region to which a candidate connection object belongs based on its feature vector includes: For plate overlap areas, the system determines whether they belong to spot welding, adhesive bonding, or seam welding candidate areas based on plate thickness, overlap distance, overlap width, boundary continuity, and part semantics. For hole feature areas, the system determines whether they belong to bolt connection candidate areas based on hole diameter, hole center distance, hole normal, hole perimeter washer area, and the semantics of adjacent parts. For areas containing rubber parts, mounting sleeves, suspension attributes, or six-way stiffness parameters, the system determines whether they belong to bushing mounting areas. For areas with small gaps but no fixed connection semantics, the system determines whether they belong to small gap contact areas. A schematic diagram illustrating the identification of spot welding candidate areas, adhesive bonding candidate areas, seam welding candidate areas, bushing mounting areas, and small gap contact areas is shown below. Figure 2 As shown.

[0034] Furthermore, the feature vector of the candidate connection object includes: connection distance feature, hole center or geometric center matching feature, normal consistency feature, overlap area or overlap length feature, part semantic matching feature, number of connection layers feature, and local connection stiffness level; the feature vector of the candidate connection object is denoted as... ,in, Indicates connection distance characteristics, Indicates the matching feature of the hole center or geometric center. Represents the normal consistency feature, Indicates the characteristics of overlap area or overlap length. Represents semantic matching features of parts, Indicates the number of connection layers, This indicates the level of local connection stiffness.

[0035] It is worth noting that the feature vectors of candidate connection objects Constructed in the order of "geometric features—topological features—semantic features—stiffness features". For the ... iFor each candidate connection object, the set of parts involved, candidate center points, candidate connection directions, and local search domain are first determined. Then, within the local search domain, nodes, elements, hole boundaries, overlap boundaries, plate thickness, material properties, assembly levels, part names, and existing connection information are read. All features are then normalized to the range of 0 to 1, where 1 indicates that the constraints of the current connection type are fully satisfied, and 0 indicates that they are not satisfied, ensuring that different connection types can be compared within the same scoring framework. Wherein: Connection distance features The distance from the candidate connection point to the nearest projection point of the connected parts, the distance between the hole centers, or the gap between the lap surfaces is determined by the following formula: ; in, This indicates the actual connection distance of the current candidate connection object. Depending on the candidate area, it can correspond to the distance from the candidate connection point to the nearest projection point of the connected part, the distance between the centers of the two holes, or the gap of the overlapping surfaces. This indicates the maximum allowed search distance preset for the current connection type. The smaller the actual distance, the better. d i The closer it is to 1; when it reaches or exceeds the allowable limit, d i Take 0.

[0036] Hole center or geometric center matching features The calculation formula is composed of hole center deviation, hole diameter difference, hole shaft angle, and washer area overlap: ; in, Indicates the number of matching terms that are actually applicable to the current candidate connection object; Indicates the first The actual deviation of each applicable matching item This indicates the maximum allowable deviation for this item. Applicable items include hole center deviation, hole diameter difference, hole axis angle, and washer area overlap; for overlap, "1 - actual overlap" is used as the deviation. Items are disregarded when there are no holes or washer areas. The smaller the deviation for each item, the better. c i The closer it is to 1.

[0037] Normal consistency feature The angle between the normals of adjacent shell elements, the angle between the connection direction and the local normal, and the normal flip state are determined. When there is a shell element normal flip, Take 0 directly; when there is no flip. The formula for calculation can be: ; in, Indicates the included angle between the normals of adjacent shell elements. Indicates the angle between the connection direction and the local normal; the corresponding , This is a preset allowable angle. The more unfavorable of the two types of angular deviations determines normal consistency; the smaller the angle, the better. The closer it is to 1.

[0038] Features of overlap area or overlap length The overlap area is determined by the projected area of ​​the overlap region, the effective overlap width, the continuous length of the boundary, and the coverage of the adhesive path. The calculation formula can be: ; in, This indicates the number of overlap evaluation items actually used for the current connection type; Indicates the first Actual value of the item This indicates the minimum required value for this item. Evaluation items are selected from effective overlap area, effective overlap width, continuous boundary length, and adhesive coverage based on the connection type; items that meet or exceed the required value are counted as 1, and items that do not meet the requirement are counted as the ratio of the actual value to the required value.

[0039] Part semantic matching features Composed of part name keywords, assembly level, functional area, material / plate thickness combination, and exclusion rules, the calculation formula can be: ; in, This indicates the number of positive semantic rules that are matched in the part name, assembly level, functional area, and material / plate thickness combination. Indicates the number of exclusion rules that were matched; This indicates the total number of semantic rules applicable to the current connection type. A higher percentage of positive rules are matched, while fewer exclusion rules are applied. The closer to 1, the better. For example, when thin plate semantics such as "body, side, floor, rail" are combined with overlapping areas, they are given priority in spot welding, adhesive bonding, or seam welding candidates; when installation semantics such as "subframe, mount, bracket, seat, belt" are combined with coaxial holes, they are given priority in bolt candidates; when semantics such as "bushing, mount, rubber, sleeve" are combined with local coordinate systems or 6-axis stiffness parameters, they are given priority in bushing candidates.

[0040] Connection layer number features This is obtained by performing a penetration search along the candidate connection direction or the shell element normal. The number of valid plate layers, duplicate layers, and abnormal layers traversed along the search line are counted. When duplicate penetration or penetration to an incorrect part occurs, [further details are provided]. Take 0 directly; otherwise, calculate using the following formula. : ; in, This indicates the actual number of effective connection layers obtained by penetrating the candidate connection direction or shell element normal. Indicates the target number of tiers allowed for the current connection type. When the actual number of tiers matches the target number... h i =1, the larger the deviation. h i The smaller.

[0041] Local connection stiffness level The calculation formula is determined based on the connection type, material plate thickness, hole perimeter constraint range, target analysis type, and user-preset stiffness rules: ; ; in, The target connection stiffness level is indicated. In this embodiment, the three levels of low, medium and high are denoted as 1, 2 and 3, respectively. The stiffness is determined by the connection type, material and plate thickness, hole perimeter constraint range, analysis objective, and preset stiffness rules. It is its 0-1 normalized representation. Used for selecting subsequent connection implementation methods and target stiffness levels, not included in the comprehensive matching score. Q i Repeated weighting is used. For critical bearer connections... At the same time, record whether it belongs to the suspension point, subframe mounting point, shock absorber tower top, seat belt fixing point, seat mounting point, collision load path or fatigue hot spot area, so as to facilitate subsequent graded review.

[0042] S3: Within the candidate region, based on the analysis target and feature vector, determine the connection type of the candidate connection of the corresponding candidate connection object; perform topological constraint matching on the candidate connection based on the feature vector, and calculate the comprehensive matching score corresponding to the candidate connection according to the satisfaction of various topological constraints.

[0043] It is worth noting that connection type determination and topology constraint matching are performed only within the allowed connection types of the candidate regions; for example, plate overlap regions enter the spot welding, adhesive bonding or seam welding candidate regions, hole-to-hole coaxial regions and hole-to-mounting surface regions enter the bolt candidate regions, bushing mounting regions enter the bushing candidate regions, and small gap regions enter the contact candidate regions.

[0044] In this embodiment, the method also establishes a connection semantic rule base, which includes at least component name rules, assembly level rules, material thickness rules, hole feature rules, overlap relationship rules, connection position rules, analysis target rules, and exception exclusion rules.

[0045] It is worth noting that the set of connection types is defined as follows: ,in, Indicates spot welding connection, Indicates bolted connection, Indicates adhesive bonding, Indicates seam welded connection, Indicates bushing connection, This indicates a contact connection; schematic diagrams of various connection types are shown below. Figure 3 As shown.

[0046] Furthermore, determining the connection type of the candidate connection object based on the analysis target and feature vector includes: ; in, Indicates the first Connection types of candidate connections, In the eigenvector and analysis objectives The following candidate connections belong to the connection type. The probability, Represents a set of connection types.

[0047] This embodiment provides the following examples to further illustrate the principle of connection type determination: When the thickness of two thin plate parts is less than a set threshold, the overlap gap is small, and there are coordinates of weld points or point-like intervals, the system prioritizes spot welding connection; when two parts have coaxial holes, similar hole diameters, and mounting surfaces or washer areas around the holes, the system prioritizes bolt connection; when two thin plates have continuous surface overlap and the adhesive path width meets the rules, the system prioritizes adhesive connection; when two parts are continuously bonded along the boundary line, the system prioritizes seam welding connection; when the connection area includes bushings, suspensions, rubber parts, or 6-axis stiffness parameters, the system prioritizes bushing connection.

[0048] After initial determination of the connection type, topological constraint matching is performed on candidate connections to avoid connecting to incorrect parts or incorrect levels. The topological constraints include at least: distance constraints, hole coaxial constraints, normal consistency constraints, overlap area constraints, connection layer number constraints, part semantic constraints, and local mesh quality constraints.

[0049] In some optional embodiments, the comprehensive matching score is calculated as follows: ; ; ; in, Indicates the first A comprehensive matching score for each candidate connection. Indicates distance score, Indicates the score for matching the hole center or geometric center. Indicates the normal consistency score. The score indicates the overlap area or overlap length. This indicates the semantic matching score of the parts. Indicates the score based on the number of connection layers. , , , , , They are respectively , , , , , The weighting coefficients. It is worth noting that for a scoring item that is not applicable to a certain connection type, its corresponding weight is reset to 0, and the remaining existing weights are normalized according to their original proportions and then participate in the comprehensive matching score calculation; if the exclusion rules or hard topology constraints are not met, the corresponding score is directly set to 0 and enters the queue for repair or manual review.

[0050] Furthermore, each score is calculated using "constraint satisfaction". Distance score The score is obtained by comparing the candidate distance with the allowed search radius; the smaller the distance, the higher the score. The score is based on the matching of the hole center or geometric center. The score is determined by the hole center deviation, hole diameter difference, and hole axis included angle; normal consistency score. The score is determined by the angle between the normals of adjacent shell elements. When the normals are opposite or the angle exceeds a threshold, the score is reduced and a normal check is triggered; the score is also determined by the overlap area or overlap length. It is obtained by comparing the effective overlap area, overlap width, and continuous boundary length with the minimum requirements of the target connection type.

[0051] Part semantic matching score The semantic score is determined by semantic keyword matching, assembly level consistency, functional area consistency, and exclusion rules. When a candidate object crosses a system boundary that should not be connected, matches non-carrying semantics such as decorative parts / wiring harness clips, or does not match the analysis objective, the semantic score is reduced; the connection layer score... The number of layers is determined by comparing the actual number of layers identified with the number of layers allowed for the target connection type. If the number of layers is 0, there are too many layers, or there is repeated penetration or penetration to the wrong part, it is judged as an abnormal number of layers and enters the repair queue.

[0052] Topological constraints include, but are not limited to, the following rules: Distance constraints require that the distance between the candidate connection point and the surface of the connected part or the center of the hole does not exceed the search radius of the corresponding connection type; Hole coaxial constraints require that the angle between the centerlines of the two holes and the offset of the hole center are less than a threshold; Normal consistency constraints require that the normal of the connected shell element and the connection direction meet a preset angle; Overlap constraints require that the effective overlap area or overlap length reaches the minimum value; Layer constraints require that the number of identified layers is consistent with the allowed number of layers for spot welding, bolting, adhesive bonding, seam welding, bushing, or contact connection; Semantic constraints require that the part name, assembly level, and functional area are consistent with the connection type; Local mesh quality constraints require that there are no severely distorted elements around the holes, around the weld points, or in the adhesive path area.

[0053] When a single score falls below its threshold but the overall matching score remains close to a preset threshold, the system does not directly delete the candidate connection. Instead, it generates repair labels based on the low scores. For example, a low distance score corresponds to expanding the search radius or reprojection; a low hole center score corresponds to hole centerline reconstruction; a low normal score corresponds to normal unification; a low layer score corresponds to layer re-identification; and a low semantic score corresponds to re-filtering the part set. This ensures a traceable correspondence between score calculation, topological constraints, and adaptive repair actions.

[0054] It is worth noting that when When, the candidate connection is determined to meet the automatic fulfillment conditions; when At that time, candidate connections enter the queue for repair or manual review. This is a preset scoring threshold.

[0055] S4: Based on the connection type, analysis target, material, plate thickness, and attributes of candidate connections with a comprehensive matching score not lower than a preset score threshold, generate connection implementation methods for candidate connections under the corresponding analysis target; automatically connect the parts in the corresponding candidate connection objects based on the connection implementation methods.

[0056] To automatically select the appropriate finite element method (FEM) implementation based on different analysis objectives such as stiffness, modal / NVH, durability, and collision, and to match the physical properties of the connection, local stiffness level, and failure representation, this embodiment establishes an analysis objective-driven connection implementation matrix to address the differentiated modeling needs of the same candidate connection under different analysis objectives. Specifically, the FEM implementation method is automatically selected based on the connection type, analysis objective, local geometric features, material properties, and target stiffness level. In some optional embodiments, the expression for the connection implementation method is: ; in, Indicates the first The connection implementation methods for each candidate connection. This indicates that the overall matching score is not lower than the preset score threshold. Connection types of candidate connections, Indicate the analysis objective, Indicates the first The local geometric features of each candidate connection include connection distance features, hole center or geometric center matching features, normal consistency features, overlap area or overlap length features, and number of connection layers features, denoted as . ; Indicates material, plate thickness, and property information. Indicates the target connection stiffness level. This represents the mapping function for the connection implementation method driven by the analysis objective. It's worth noting that material, plate thickness, and attribute information are directly obtained from S1; the target connection stiffness level is determined based on the aforementioned stiffness constraint characteristics, analysis objective, and preset stiffness rules. Specifically, the mapping function for the connection implementation method driven by the analysis objective selects the corresponding finite element implementation method from the connection implementation matrix and outputs the connection implementation method based on the connection type, analysis objective, local geometric features, material and plate thickness attributes, and target connection stiffness level.

[0057] As a preferred embodiment, for spot welded connections, the system can select ACM, CWELD, equivalent weld point, or failable weld point according to the analysis objective; for bolted connections, the system can select RBE2+CBAR, RBE3+CBUSH, periphery distribution connection, or failable bolt; for adhesive connections, the system can select equivalent surface connection, adhesive layer equivalent element, or failable adhesive; for bushing connections, the system can select CBUSH, 6-axis stiffness CBUSH, or nonlinear bushing model. Examples of connection implementation methods for each connection type under each analysis objective are shown in Table 1.

[0058] Table 1. Examples of target-driven connection implementation method selection.

[0059] S5: Repeat S3-S4 until all candidate connection objects have been traversed to complete the connection of the vehicle parts. When the connection quality of the whole vehicle is unqualified, adaptive repair is performed to realize the automatic modeling of multiple types of connections of the whole vehicle model under the corresponding analysis target.

[0060] It is worth noting that by repeatedly executing S3-S4 until all candidate connection objects have been traversed, spot weld connections, bolted connections, adhesive connections, seam weld connections, bushing connections, and contact connections between parts of the entire vehicle are generated. Unlike simply generating connection units based on connection information files, this step simultaneously records the candidate source, judgment score, connection ID, connection type, connection coordinates, connected parts, number of connection layers, connection diameter, connection thickness, connection implementation method, analysis target, connection status, and generation time.

[0061] For spot welds, the system searches the plate layers along the weld point coordinates or candidate point locations and generates corresponding weld points according to the number of connection layers. For bolted connections, distributed or elastic connections are generated based on the hole centerline, hole perimeter nodes, and washer region. For adhesive connections, area connections or adhesive layer elements are generated based on the adhesive path centerline, adhesive width, and adhesive thickness. For seam welds, line connections are generated along continuous boundary lines. For bushing connections, the system establishes a local coordinate system and assigns 6-axis stiffness. For contact connections, the system generates contact pairs or tie relationships based on gap, normal, and relative motion constraints. Figure 4 As shown, before automatic implementation, no reliable connection was generated between the plates; after automatic implementation, the system automatically selects spot welding, bolting or adhesive bonding methods based on candidate features, topology scores and analysis objectives, and records the connection attributes.

[0062] Furthermore, after the connection is generated, a connection quality check is automatically performed. The check items include connection success rate, number of failed connections, number of unrealized connections, number of duplicate connections, missing connection areas, incorrectly connected components, abnormal connection distance, hole coaxial deviation, abnormal normal direction, abnormal number of connection layers, abnormal local mesh quality, missing CBUSH coordinate system, abnormal connection stiffness, and risk level of critical connections.

[0063] In some optional embodiments, the connection quality is calculated as follows: ; in, Indicates the overall vehicle connection quality. Indicates the connection success rate. Indicates the connection failure rate. Indicates the fault connection rate. Indicates the repeat connection rate. Indicates the connection stiffness anomaly rate. , , , , They represent , , , , The weighting coefficients.

[0064] Furthermore, such as Figure 5 As shown, the adaptive repair when the overall vehicle connection quality is substandard includes: The causes of connection quality failures are attributed to various factors, including: missing connection objects. Connection distance exceeds tolerance Error in normal direction Holes not coaxial Abnormal number of connection layers Misconnection Poor local mesh quality and abnormal connection stiffness ; Based on the reasons for failed connections, the feature vectors of the candidate connection objects are adaptively repaired. ; in, Indicates the first During the nth iteration The feature vectors of the candidate connection objects corresponding to each connection. This represents an adaptive repair function. Indicates the first During the nth iteration The feature vectors of the candidate connection objects corresponding to each connection. Indicates the first Reasons for connection failure Indicates the reason for the failed connection. The corresponding repair parameters.

[0065] like Figure 6 As shown, taking the failure of a bolt connection due to misalignment of holes as an example, the system first diagnoses the cause of the failed connection as follows: Then, the hole centerline was recalculated and the connection point was reprojected. Finally, the implementation was re-implemented and the repair log was recorded.

[0066] This embodiment provides the following examples to further illustrate the adaptive repair process: When the failure type is missing object or excessive connection distance, the system adaptively adjusts the search radius and reprojects the connection points. When the failure type is incorrect shell element normal, the system checks and unifies the shell element normal. When the failure type is misaligned holes, the system recalculates the hole centerline and re-pairs the hole features. When the failure type is an abnormal number of connection layers, the system re-identifies the number of penetration layers along the normal. When the failure type is incorrectly connected to other parts, the system re-filters the connection objects according to the part semantic rules. When the failure type is abnormal stiffness, the system switches the connection implementation method according to the analysis objective, for example, switching from RBE2 to RBE3+CBUSH.

[0067] S2-S5 are executed based on the repaired feature vectors until the overall vehicle connection quality is qualified.

[0068] It is worth noting that the search is performed only locally within the candidate region and its associated part set, and the re-search after a connection failure is also preferentially performed within the original candidate region. This transforms the global search of the entire vehicle into a locally constrained search, which can reduce cross-system misconnections, misjudgments of ordinary holes, and repeated searches, and form a continuous closed loop for candidate identification, connection type determination, connection implementation, and failure repair.

[0069] In this implementation, after completing the automatic modeling of multiple types of connections for the entire vehicle, a vehicle connection quality report is output to ensure the traceability of the modeling process. For example... Figure 7 As shown, the vehicle connection quality report includes the total number of connections, the number of spot welds, the number of bolts, the number of adhesives, the number of seam welds, the number of bushings, the number of contacts, the connection success rate, the number of failed connections, the number of automatically repaired connections, the number of manually reviewed connections, the statistics of failure reasons, the risk level of critical connections, connection visualization images, repair logs, and solver connection files.

[0070] The system can also classify connected objects into three categories based on risk level: pass, require review, and fail. For critical connections such as powertrain mounting points, subframe mounting points, shock absorber towers, seat belt anchor points, seat mounting points, collision load paths, and fatigue hotspots, the system includes them in Category A for key review. For ordinary spot welds, ordinary bracket bolts, and adhesive areas on the body-in-white, the system processes them in batches as standard connectors. For non-load-bearing accessories, wiring harness clips, and trim parts, the system processes them as simplified connections or quality points.

[0071] The multi-type connectivity automatic modeling method for multi-analysis objectives in whole-vehicle CAE provided in this embodiment has the following beneficial effects: 1. A multi-source candidate connection feature vector construction mechanism is proposed: This method unifies part name, assembly level, material plate thickness, hole center matching, normal consistency, overlap area or overlap length, number of connection layers, local stiffness level and analysis target into the candidate connection feature vector, so that the candidate connection is no longer determined by the weld point table, connection file or single hole feature.

[0072] 2. Propose a semantic-topology joint scoring and three-state flow separation method: This method unifies distance constraints, hole coaxial constraints, normal consistency constraints, overlap area constraints, connection layer constraints, part semantic constraints and local mesh quality constraints into the same judgment framework through comprehensive matching scoring, and divides candidate connections into three categories: automatically realized, to be repaired and manually verified.

[0073] 3. Propose an analysis target-driven connection implementation method: This method automatically selects different finite element implementation methods based on different analysis targets such as stiffness, modal / NVH, durability and collision, avoiding unifying all connections into rigid connections or the same connector implementation method, thereby reducing the risk of model stiffness distortion and stress distortion in key areas.

[0074] 4. Propose a classification of failed connections and an adaptive repair closed loop: This method classifies failed connections into types such as missing objects, excessive distance, incorrect normals, misaligned holes, abnormal layer counts, incorrectly connected components, poor mesh quality, and abnormal stiffness. Based on the failure type, the corresponding repair function is called to achieve re-search, reprojection, normal correction, hole centerline reconstruction, layer count re-identification, semantic filtering, and connection mode switching.

[0075] 5. Propose a vehicle connection quality index and a graded review mechanism for key connections: This method provides a unified evaluation of connection success rate, failure rate, incorrect connection rate, duplicate connection rate, and stiffness anomaly rate. It also includes suspension points, subframe mounting points, shock absorber tower tops, seat belt fixing points, seat mounting points, collision load paths, and fatigue hotspot areas as key review targets, enabling traceable output of connection results, failure reasons, repair logs, and manually reviewed targets.

[0076] 6. The above technical features are combined in sequence to form a closed loop: the candidate object first enters the implementation or repair queue through semantic-topological joint scoring, then the finite element expression is determined by the analysis target, and finally the closed loop verification is completed through quality index and hierarchical review.

[0077] In summary, this method imports a complete vehicle model from CAD or CAE and reads the model information; scans out several candidate connection objects in the model and constructs feature vectors for each object based on the model information; within the candidate region, based on the analysis objective and feature vectors, determines the connection type of the candidate connection for each object; performs topological constraint matching on the candidate connections based on the feature vectors and calculates the comprehensive matching score based on the satisfaction of various topological constraints; generates connection implementation methods for candidate connections under the corresponding analysis objective based on the connection type of the candidate connection with a comprehensive matching score not lower than a preset score threshold and the analysis objective; automatically connects the parts in the corresponding candidate connection objects based on the connection implementation methods until all candidate connection objects are traversed, completing the connection of the vehicle's parts; when the connection quality of the vehicle is unqualified, adaptive repair is performed, realizing automatic modeling of multiple connection types for the vehicle model under the corresponding analysis objective.

[0078] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0079] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these modifications and improvements all fall within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the appended claims.

Claims

1. A multi-type connection automatic modeling method for multiple analysis targets in whole vehicle CAE, based on a finite element preprocessing platform, characterized in that, include: S1: Import the complete vehicle model from CAD or CAE and read the complete vehicle model information, including parts, materials, plate thickness, attributes, and analysis targets. S2: Scan out several candidate connection objects in the whole vehicle model, and construct the feature vector of each candidate connection object based on the whole vehicle model information; determine the candidate region to which the candidate connection object belongs based on the feature vector of the candidate connection object; S3: Within the candidate region, based on the analysis target and feature vector, determine the connection type to which the candidate connection of the corresponding candidate connection object belongs; The candidate connections are matched for topological constraints based on feature vectors, and the comprehensive matching score corresponding to the candidate connections is calculated based on the satisfaction of various topological constraints. S4: Based on the connection type, analysis target, material, plate thickness, and attributes of candidate connections whose comprehensive matching score is not lower than a preset score threshold, generate connection implementation methods for candidate connections under the corresponding analysis target; automatically connect the parts in the corresponding candidate connection objects based on the connection implementation methods; S5: Repeat S3-S4 until all candidate connection objects have been traversed to complete the connection of the vehicle parts. When the connection quality of the whole vehicle is unqualified, adaptive repair is performed to realize the automatic modeling of multiple types of connections of the whole vehicle model under the corresponding analysis target.

2. The method according to claim 1, characterized in that, Candidate connection objects include parts in plate overlap areas, hole-to-hole coaxial areas, hole-to-mounting surface areas, adhesive surface areas, continuous boundary contact areas, small gap areas, and bushing mounting areas.

3. The method according to claim 1, characterized in that, The feature vectors of the candidate connection objects include: connection distance features, hole center or geometric center matching features, normal consistency features, overlap area or overlap length features, part semantic matching features, connection layer number features, and local connection stiffness level.

4. The method according to claim 1, characterized in that, The process of determining the connection type of a candidate connection object based on the analysis target and feature vectors includes: ； in, Indicates the first Connection types of candidate connections, In the eigenvector and analysis objectives The following candidate connections belong to the connection type. The probability, This represents a set of connection types; the connection types in the set include spot welding, bolting, adhesive bonding, seam welding, bushing, and contact connection.

5. The method according to claim 1, characterized in that, The formula for calculating the overall matching score is: ； in, Indicates the first A comprehensive matching score for each candidate connection. Indicates distance score, Indicates the score for matching the hole center or geometric center. Indicates the normal consistency score. The score indicates the overlap area or overlap length. This indicates the semantic matching score of the parts. Indicates the score based on the number of connection layers. , , , , , They are respectively , , , , , The weighting coefficients.

6. The method according to claim 1, characterized in that, The expression for the connection implementation method is: ； in, Indicates the first The connection implementation methods for each candidate connection. This indicates that the overall matching score is not lower than the preset score threshold. Connection types of candidate connections, Indicate the analysis objective, Indicates the first The local geometric features of the candidate connections include connection distance features, hole center or geometric center matching features, normal consistency features, overlap area or overlap length features, and number of connection layers features. Indicates material, plate thickness, and property information. Indicates the target connection stiffness level. This represents the mapping function for the connection implementation method driven by the analysis target.

7. The method according to claim 1, characterized in that, The analysis objectives include stiffness analysis, modal / NVH analysis, durability analysis, and collision analysis.

8. The method according to claim 1, characterized in that, The formula for calculating connection quality is: ； in, Indicates the overall vehicle connection quality. Indicates the connection success rate. Indicates the connection failure rate. Indicates the fault connection rate. Indicates the repeat connection rate. Indicates the connection stiffness anomaly rate. , , , , They represent , , , , The weighting coefficients.

9. The method according to claim 8, characterized in that, When the overall vehicle connection quality is substandard, adaptive repair is performed, including: The causes of connection failures are attributed to the following: missing connection objects, connection distance exceeding tolerance, incorrect normal, misaligned holes, abnormal number of connection layers, incorrect connection, poor local mesh quality, and abnormal connection stiffness. Based on the reasons for failed connections, the feature vectors of the candidate connection objects corresponding to the failed connections are adaptively repaired, and S2-S5 are executed based on the repaired feature vectors until the connection quality of the whole vehicle is qualified.

10. The method according to claim 9, characterized in that, The adaptive repair of the feature vectors of candidate connection objects corresponding to the failed connections based on the reasons for the failed connections includes: ； in, Indicates the first During the nth iteration The feature vectors of the candidate connection objects corresponding to each connection. This represents an adaptive repair function. Indicates the first During the nth iteration The feature vectors of the candidate connection objects corresponding to each connection. Indicates the first Reasons for connection failure Indicates the reason for the failed connection. The corresponding repair parameters.

Images