Simulation data management and collaborative review method and system

Abstract

The application relates to the field of CAE simulation whole-process management, and discloses a simulation data management and collaborative review method and system, which comprises the following steps: rendering a result view layer corresponding to a post-processing result node in a Web interface, receiving graphical annotation data and text annotation data containing a space relative coordinate system parameter input by an online client for the result view layer, serializing the graphical annotation data and the text annotation data into a message body in a preset format, and broadcasting the message body to other online clients in the same review group through a WebSocket protocol for position alignment superimposition rendering; classifying and summarizing the extracted annotation data to generate a review record node, and when a pre-processing input node or a solving task node with a new structured version identifier is generated based on the review record node in a to-be-modified state, a new directed association edge from the review record node to the corresponding new node is automatically established in a graph database. According to the scheme, the review cycle is reduced to one day, and the efficiency is improved by 85.7%.

Description

Technical Field

[0001] This invention relates to the field of CAE simulation end-to-end management, and in particular to a simulation data management and collaborative review method and system. Background Technology

[0002] CAE simulation processes involve multiple stages: "design model import → preprocessing (mesh / material) → solution calculation → post-processing analysis," generating massive amounts of data (design files, mesh files, solution logs, result contour plots, etc.). Existing data management and collaboration methods suffer from four core pain points: Version management is chaotic and difficult to trace: The traditional method of "file naming + shared folder" management has no structured version identification. The number of versions of the same simulation task often exceeds 10. Engineers need to manually filter the valid versions and trace the input parameters corresponding to a certain result, which takes more than 4 hours. Furthermore, version confusion can easily lead to simulation reproduction failure. Data correlation is lacking, and the data chain is broken: the design model (STEP / IGES), simulation inputs (mesh.inp, material.mat), and output results (stress contour plot.png, life report.pdf) are isolated from each other, lacking a unified correlation mechanism. For example, when the design model is updated to V2.1, it cannot automatically locate all simulation tasks that depend on the model, resulting in some simulations still using the old design model, leading to a result deviation exceeding 15%. The SPDM process is opaque and has poor reproducibility: the simulation process lacks standardized record templates, and steps, executors, and reasons for modifications rely solely on local engineer documentation, leaving no trace of missing processes or parameter changes. When reproducing simulation results from three months prior, issues such as "solver version not recorded" and "no explanation of boundary condition modifications" result in reproduction taking over 24 hours and yielding results with a deviation exceeding 20%. Collaborative review is inefficient and opinions are scattered: the review relies on "email attachments + Excel annotations", the annotations are scattered in different emails and documents, which need to be manually summarized, and it is impossible to see other people's annotations in real time. There is an "information gap" in multi-team collaboration, the review cycle is as long as 1 week, and the rework rate exceeds 30%.

[0003] The industry urgently needs a version-controllable, traceable, transparent, and highly efficient SDM and collaborative review system to support standardized management and multi-team collaboration throughout the CAE simulation process. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a simulation data management and collaborative review method, comprising the following steps: A structured simulation data version management model is constructed. According to the preset rules, the project identifier, model type, simulation task, version number and timestamp are sequentially concatenated to generate a unique structured version identifier for the target simulation data. When the target simulation data undergoes a version change, change log information including the person making the change, the content of the change, the reason for the change and the related documents is recorded. Construct a full-link simulation data association graph. In the graph database, establish design model nodes, preprocessing input nodes, solution task nodes, postprocessing result nodes, and review record nodes respectively. During the generation and flow of simulation data, automatically extract data nodes with pre- and post-dependent relationships, establish directed association edges between data nodes with pre- and post-dependent relationships, and output the upstream and downstream link node set of a specific data node by forward traversal and backward traversal of the directed association edges. The standardized simulation process is recorded and loaded. A workflow template containing multiple preset simulation execution steps is loaded. During the operation of each simulation execution step, the application programming interface of an external computer-aided engineering tool is called to extract the mesh parameters of the preprocessing stage, the configuration parameters of the solving stage, and the numerical characteristics of the results of the postprocessing stage as process execution parameters. The extracted process execution parameters are then bound to the executor and execution time and written into the corresponding preset simulation execution step node. The system performs web-based collaborative review, rendering the result view layer corresponding to the post-processing result node in the web interface. It receives graphic and text annotation data containing spatial relative coordinate system parameters from online clients for input to the result view layer. The graphic and text annotation data are serialized into a message body of a preset format and broadcast to other online clients in the same review group via the WebSocket protocol for overlay rendering with position alignment. The extracted annotation data is categorized and summarized to generate the review record node. When a pre-processing input node or solution task node with a new structured version identifier is generated based on the review record node containing a state to be modified, a new directed association edge pointing from the review record node to the corresponding new node is automatically established in the graph database.

[0005] This invention also provides a simulation data management and collaborative review system, comprising: The version management module is used to build a structured simulation data version management model. According to preset rules, it sequentially concatenates the project identifier, model type, simulation task, version number and timestamp to generate a unique structured version identifier for the target simulation data. When the target simulation data undergoes version changes, it records change log information including the person making the change, the content of the change, the reason for the change and the related documents. The data association module is configured with a graph database to construct a full-link simulation data association graph. Design model nodes, preprocessing input nodes, solution task nodes, postprocessing result nodes, and review record nodes are established in the graph database. During the generation and flow of simulation data, data nodes with pre- and post-dependent relationships are automatically extracted, directed association edges are established between data nodes with pre- and post-dependent relationships, and the upstream and downstream link node sets of specific data nodes are output by forward and backward traversal of the directed association edges. The SPDM process management module is used to execute standardized simulation process records, load workflow templates containing multiple preset simulation execution steps, and extract mesh parameters from the preprocessing stage, configuration parameters from the solving stage, and numerical characteristics of the results from the postprocessing stage as process execution parameters by calling the application programming interface of external computer-aided engineering tools. The extracted process execution parameters are then bound to the executor and execution time and written into the corresponding preset simulation execution step nodes. The Web Collaborative Review Module is used to perform web-based collaborative reviews. It renders the result view layer corresponding to the post-processing result node in the web interface, receives graphical and textual annotation data containing spatial relative coordinate system parameters input from online clients to the result view layer, serializes the graphical and textual annotation data into a message body of a preset format, and broadcasts it via the WebSocket protocol to other online clients in the same review group for position-aligned overlay rendering. It categorizes and summarizes the extracted annotation data to generate the review record node. When a pre-processing input node or solution task node with a new structured version identifier is generated based on the review record node containing a state to be modified, a new directed edge pointing from the review record node to the corresponding new node is automatically established in the graph database.

[0006] Compared with the prior art, the beneficial effects of the present invention are as follows: Version management efficiency has been significantly improved. Version traceability time: reduced from the traditional 4 hours to 5 minutes, improving efficiency by 95.8%; Version chaos rework rate: reduced from 30% to 5%, reducing the cost of rework per project by 83%; Version management automation rate: increased from 0% to 99%, reducing manual operation by engineers by 90%.

[0007] Breakthrough in data association and traceability capabilities Data link integrity: From isolated data to full-link association, the association accuracy is ≥99.9%; Simulation reproducibility rate: Increased from 60% to 100%, and the reproduction time was reduced from 24 hours to 1 hour, with an efficiency improvement of 95.8%; Identification of change impact scope: After the design model is updated, all dependent simulation tasks are located within 10 minutes to avoid missing updates (traditionally it takes 2 hours).

[0008] SPDM process transparency and standardization Completeness of process records: From "no records" to "automatic recording of all parameters", with a recording accuracy of ≥99%; Visualization of processes: From "verbal communication" to "real-time display of Gantt charts", improving progress tracking efficiency by 80%; Standardized template coverage: Supports more than 80% of CAE simulation scenarios, with enterprise-customized templates taking ≤30 minutes.

[0009] Collaborative review efficiency increased exponentially Review cycle: reduced from 1 week to 1 day, improving efficiency by 85.7%; time spent summarizing opinions: reduced from 2 days to 10 minutes, improving efficiency by 99.2%; cost of multi-team collaboration: cost of cross-regional team communication reduced by 70%, and the closure rate of review opinions increased to 100%. Attached Figure Description

[0010] Figure 1 A flowchart provided for an embodiment of the present invention. Detailed Implementation

[0011] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” or “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0012] like Figure 1 A simulation data management and collaborative review method includes the following steps: A structured simulation data version management model is constructed. According to the preset rules, the project identifier, model type, simulation task, version number and timestamp are sequentially concatenated to generate a unique structured version identifier for the target simulation data. When the target simulation data undergoes a version change, change log information including the person making the change, the content of the change, the reason for the change and the related documents is recorded. Construct a full-link simulation data association graph. In the graph database, establish design model nodes, preprocessing input nodes, solution task nodes, postprocessing result nodes, and review record nodes respectively. During the generation and flow of simulation data, automatically extract data nodes with pre- and post-dependent relationships, establish directed association edges between data nodes with pre- and post-dependent relationships, and output the upstream and downstream link node set of a specific data node by forward traversal and backward traversal of the directed association edges. The standardized simulation process is recorded and loaded. A workflow template containing multiple preset simulation execution steps is loaded. During the operation of each simulation execution step, the application programming interface of an external computer-aided engineering tool is called to extract the mesh parameters of the preprocessing stage, the configuration parameters of the solving stage, and the numerical characteristics of the results of the postprocessing stage as process execution parameters. The extracted process execution parameters are then bound to the executor and execution time and written into the corresponding preset simulation execution step node. The system performs web-based collaborative review, rendering the result view layer corresponding to the post-processing result node in the web interface. It receives graphic and text annotation data containing spatial relative coordinate system parameters from online clients for input to the result view layer. The graphic and text annotation data are serialized into a message body of a preset format and broadcast to other online clients in the same review group via the WebSocket protocol for overlay rendering with position alignment. The extracted annotation data is categorized and summarized to generate the review record node. When a pre-processing input node or solution task node with a new structured version identifier is generated based on the review record node containing a state to be modified, a new directed association edge pointing from the review record node to the corresponding new node is automatically established in the graph database.

[0013] The construction of the structured simulation data version management model also includes: Based on the base version data node with a locked state, a first branch version node corresponding to parameter optimization or a second branch version node corresponding to comparison and verification is generated in response to the branch creation instruction. An operation permission verification strategy based on role identification is configured. When the status of the review record node is changed to "review passed", the data attribute of the simulation data version on the corresponding data link is changed to read-only locked state. Data overwrite instructions for the simulation data version in the read-only locked state are intercepted. When a modification request is received, a new version branch is forced to be created based on the simulation data version in the read-only locked state.

[0014] The process of establishing directed edges between data nodes with pre- and post-dependent relationships, and outputting the upstream and downstream link node set of a specific data node by traversing the directed edges in both forward and reverse directions, specifically includes: When a design model is imported and preprocessing is performed to generate preprocessing input data, directed edges pointing from the corresponding design model node to the preprocessing input node are automatically established in the graph database. When a solution task is submitted, the preprocessing input data called by the solution task is parsed and extracted, and directed edges pointing from the corresponding preprocessing input node to the solution task node are automatically established. When the solver completes the calculation and outputs the result file, directed edges pointing from the corresponding solution task node to the postprocessing result node are automatically established. When a forward tracing request based on a certain design model node is received, the system traverses sequentially along the out-degree direction of the directed edges and outputs the version evolution chain of all preprocessing input nodes, solution task nodes, and postprocessing result nodes derived from that design model node. When a reverse tracing request based on a certain postprocessing result node is received, the system traverses sequentially along the in-degree direction of the directed edges and outputs the solution stage configuration parameters, preprocessing stage mesh parameters, and initial design model node version sequence on which the generation of the postprocessing result node depends.

[0015] The extraction of mesh parameters from the preprocessing stage, configuration parameters from the solution stage, and numerical features of the results from the postprocessing stage as process execution parameters specifically includes: In the preprocessing stage, the application programming interface (API) of the mesh generation tool is called to extract the mesh algorithm configuration, element type, size field parameters, and mesh quality indicators including distortion rate values. In the solving stage, the API of the solver is called to extract the solver version number, calculation step size, number of iterations, convergence criteria, and the calculation log file generated by the solver is captured and saved. In the post-processing stage, the API of the post-processing tool is called to extract the numerical values ​​of stress extrema, displacement extrema, or lifetime extrema of the target result type, as well as the position coordinate parameters of the corresponding extrema in the three-dimensional model space. The process execution parameters of each simulation execution step, the version information of the corresponding computer-aided engineering tools called, and the storage and download address of the generated data files are structured and encapsulated to generate a reproduction configuration file for process restoration.

[0016] The steps for performing web-based collaborative review also include: Configure color label attributes for the text annotation data to indicate rectification priority, and perform classification and aggregation on the extracted annotation data according to the reviewed object, review role group, and opinion type; configure a flow state machine for the review record node, and the state set of the flow state machine includes pending review, under review, summarized, rectified, and passed; when the system detects that a post-processing result node with a new structured version identifier has been generated due to the rectification operation, it automatically loads the review opinions in the pending review state in the original mapping relationship into the pending review list of the new post-processing result node, and sends a second review notification instruction to other online clients in the same review group until the flow state machine of the review record node changes to passed, triggering the extraction of annotation view coordinate data, text content, and rectification operation records from each flow in the graph database to generate a standardized review report file.

[0017] In embodiments of the present invention, a simulation data management and collaborative review system is also involved, comprising: The version management module is used to build a structured simulation data version management model. According to preset rules, it sequentially concatenates the project identifier, model type, simulation task, version number and timestamp to generate a unique structured version identifier for the target simulation data. When the target simulation data undergoes version changes, it records change log information including the person making the change, the content of the change, the reason for the change and the related documents. The data association module is configured with a graph database to construct a full-link simulation data association graph. Design model nodes, preprocessing input nodes, solution task nodes, postprocessing result nodes, and review record nodes are established in the graph database. During the generation and flow of simulation data, data nodes with pre- and post-dependent relationships are automatically extracted, directed association edges are established between data nodes with pre- and post-dependent relationships, and the upstream and downstream link node sets of specific data nodes are output by forward and backward traversal of the directed association edges. The SPDM process management module is used to execute standardized simulation process records, load workflow templates containing multiple preset simulation execution steps, and extract mesh parameters from the preprocessing stage, configuration parameters from the solving stage, and numerical characteristics of the results from the postprocessing stage as process execution parameters by calling the application programming interface of external computer-aided engineering tools. The extracted process execution parameters are then bound to the executor and execution time and written into the corresponding preset simulation execution step nodes. The Web Collaborative Review Module is used to perform web-based collaborative reviews. It renders the result view layer corresponding to the post-processing result node in the web interface, receives graphical and textual annotation data containing spatial relative coordinate system parameters input from online clients to the result view layer, serializes the graphical and textual annotation data into a message body of a preset format, and broadcasts it via the WebSocket protocol to other online clients in the same review group for position-aligned overlay rendering. It categorizes and summarizes the extracted annotation data to generate the review record node. When a pre-processing input node or solution task node with a new structured version identifier is generated based on the review record node containing a state to be modified, a new directed edge pointing from the review record node to the corresponding new node is automatically established in the graph database.

[0018] Specifically, this method achieves full lifecycle management and efficient collaboration of simulation data through a technical path of "structured version management model → end-to-end data association → standardized SPDM process recording → real-time collaborative review via Web". The specific steps are as follows: 1. Construction of a structured simulation data version management model Breaking away from the traditional "file naming" method, we designed a management model of "multi-level version identification + change tracking" to ensure that versions are unique and traceable. Version Identification Rules Definition 1.1 The structured identifier, "Project ID-Model Type-Simulation Task-Version Number-Timestamp," is used, with the meaning and format of each level as follows: Project ID Unique identifier for the simulation project; Car-Chassis-Fatigue-001; letters + numbers + hyphens, globally unique.

[0019] Model type Distinguish between data types (Design / Input / Output); Design / Input / Output; enumerated values, only three types are supported.

[0020] Simulation task Corresponding to specific simulation stages; Mesh / Material / Solver / Result; enumerated values ​​covering the entire SPDM process.

[0021] Version number Differentiate between different versions of the same task; V1.0 / V1.1 / V2.0; major version number (feature changes). minor version number (parameter fine-tuning).

[0022] Timestamp Version creation time: 202410151430 (October 15, 2024, 14:30); year, month, day, hour, minute, accurate to the minute.

[0023] Example: "Car-Chassis-Fatigue-001-Input-Mesh-V1.1-202410151430" means "Input-class mesh data for the Car Chassis Fatigue Simulation Project (ID: Car-Chassis-Fatigue-001), version 1.1, created on October 15, 2024 at 14:30".

[0024] Version 1.2 Branches and Change Log a) Branch Management: Supports multiple branches such as "Base → Optimize → Verify". For example, the "Optimize" branch can be created based on Mesh-V1.0 for mesh size optimization (V1.1). The branches support difference comparison (such as displaying the difference in the number of mesh cells between V1.0 and V1.1). b) Change Tracking: Each version automatically records "Change Person - Change Content - Reason for Change - Related Documents", for example: 1. Person making the change: Mr. Zhang (ID: ZG001); 2. Changes: The grid size has been adjusted from 2mm to 1.5mm (see attached comparison chart of the old and new sizes). 3. Reason for change: Insufficient accuracy in the stress concentration area (Review comment ID: Review-005); 4. Related document: Mesh Quality Report (ID: Report-Mesh-Quality-001).

[0025] Version 1.3 Operation Permission Control Role-based access control (RBAC): a) Administrator: Can create / delete versions and configure permissions; b) Simulation Engineer: Can create new versions and modify versions created by themselves; c) Reviewers: Can only view versions and add review comments, but cannot modify data; d) Version locking: Versions that have passed review (such as Result-V2.0) are automatically locked and cannot be modified. Only new branches (such as V2.1) can be created.

[0026] 2. End-to-end simulation data association mechanism Constructing a relationship graph of "design-simulation-results" to achieve bidirectional traceability between data, specifically including: 2.1 Association Dimensions and Identification Rules Five types of core data nodes and their relationships are defined, and a graph database (Neo4j) is used to store the relationship graph (data node type, unique identifier format, relationship, and example relationship link): Design Model Design-ID-Version Number; Associate → Preprocessing Input (Generate Mesh Based on Design Model); Design-001-V2.1 → Input-Mesh-001-V1.0.

[0027] Preprocessing input Input-ID-Task Type-Version Number; Associate → Solver Task (Input Parameter Driven Solver); Input-Mesh-001-V1.0 → Solver-001-V1.0.

[0028] Solve the task Solver-ID-Version Number; Associate → Post-processing Result (Solution Output); Solver-001-V1.0 → Output-Result-001-V1.0.

[0029] Post-processing results Output-ID-Type-Version Number; Associate → Review Record (Result is used for review); Output-Result-001-V1.0 → Review-001.

[0030] Review Record Review-ID; associated with → version change (review comments drive version update); Review-001 → Input-Mesh-001-V1.1.

[0031] 2.2 Combination of automatic and manual association Automatic association: The system automatically establishes associations when data is generated, for example: 1. After importing the design model (Design-001-V2.1), perform mesh generation to generate Input-Mesh-001-V1.0. The system will automatically add the "Design→Mesh" associated edge to the associated graph. 2. When submitting a solution task (Solver-001-V1.0), the selected Input (Mesh, Material) is automatically associated, generating an "Input→Solver" association; Manual association: Allows engineers to supplement the association, such as manually associating test data (Test-Data-001) with simulation results (Output-Result-001-V1.0) for comparative analysis.

[0032] 2.3 Two-way traceability function Forward tracing: Enter the design model ID to view all simulation tasks (Mesh, Solver, Result) and version evolution based on that model, such as Design-001-V2.1→Mesh-V1.0→Solver-V1.0→Result-V1.0→Mesh-V1.1 (updated after review). Reverse tracing: Input the result ID (Output-Result-001-V1.0) to view its corresponding solver parameters (Solver-V1.0: LS-DYNAR12.0, step size 1e-6s), mesh file (Mesh-V1.0: size 2mm, distortion rate 3.2%), and design model (Design-001-V2.1). Tracing takes ≤5 minutes.

[0033] 3. Standardized SPDM process recording and visualization Define an SPDM management approach that combines "template-based processes" with "automatic parameter recording" to ensure process transparency and reproducibility. 3.1 SPDM Process Template Definition Based on industry standards (such as ISO 10303-42), a standardized process template is designed, covering typical CAE scenarios (scenario type; standardized process steps; required parameters (automatically collected); executor / time (automatically recorded)): Structural strength simulation 1. Import geometry → 2. Mesh generation → 3. Material assignment → 4. Boundary condition setting → 5. Solve the problem → 6. Result analysis.

[0034] 1. Design file format / version; 2. Mesh algorithm (AdvancingFront) / size / pass rate; 3. Material model (*MAT_24) / parameters; 4. Load type (concentrated force 10kN) / constraints; 5. Solver version / step size; 6. Maximum stress / location.

[0035] Executor ID and start / end time for each step.

[0036] Fatigue life simulation 1. Import geometry → 2. Repair mesh → 3. Set fatigue material → 4. Import load spectrum → 5. Solve → 6. Calculate life.

[0037] 2. Repair defect type (crack / overlap); 3. Fatigue curve (SN curve); 4. Load spectrum format (.rlc); 6. Minimum lifetime / location.

[0038] Executor ID and start / end time for each step.

[0039] Template customization: Supports enterprises to extend the template according to their needs, such as adding a "multiphysics coupling" step (structure-thermal coupling settings) and defining the parameters to be recorded (heat transfer coefficient 50W / (m²)). 2 •K)).

[0040] 3.2 Automatic Parameter Acquisition and Recording Pre-processing stage: Connect to mesh generation tools (HyperMesh, ANSYSMeshing) to automatically collect mesh algorithm, element type, size field, and quality indicators (distortion rate, aspect ratio), eliminating the need for manual input; Solution phase: Connect to the solver (LS-DYNA, ABAQUS), automatically collect solver version, calculation step size, number of iterations, convergence criteria, and record solution log (.log file); Post-processing stage: Connect to post-processing tools (ParaView, ANSYSCFD-Post) to automatically collect result types (stress, displacement, life) and key values ​​(maximum, minimum, location), and associate them with result files (.odb, .d3plot).

[0041] 3.3 SPDM Process Visualization and Reproduction Process visualization: Display the SPDM progress in the form of Gantt charts / flowcharts, and mark the status of each step (pending execution, in progress, completed, rework required), such as "Mesh generation → completed (pass rate 98.2%)" and "Boundary condition setting → rework required (load direction error, review comment ID: Review-003)". Process Reproduction: Select a historical SPDM process (such as Solver-001-V1.0), and the system will automatically export a "Reproduction Manual" containing parameters, tool versions, and data file links for all steps. Clicking the link will allow you to directly download the corresponding version of the mesh and material files. The reproduction time is ≤1 hour (the traditional method requires 24 hours).

[0042] 4. Real-time collaborative review and annotation on the web platform Real-time collaboration among multiple teams based on web technology solves the problems of fragmented and inefficient reviews: 4.1 Review Targets and Permission Settings Review targets: simulation results (stress cloud diagram, life curve), SPDM process, data version, such as reviewing the rationality of stress distribution in "Output-Result-001-V1.0"; Review permissions: Set up review groups (design group, simulation group, test group), and assign reviewers to each group. For example, the design group reviews geometric consistency, the simulation group reviews mesh quality, and the test group reviews the deviation between the results and experimental data.

[0043] 4.2 Real-time annotation and feedback interaction Multi-format annotation: Supports annotation of result files (such as stress cloud diagrams.png, animations.mp4) on the web: 1. Graphical annotation: Circles (to mark stress concentration areas), straight lines (to indicate dimensions), and arrows (to indicate displacement direction); 2. Text annotation: Input comments ("The stress here exceeds the material yield strength of 350MPa, and the fillet needs to be optimized"). Rich text is supported (bold, color-coded priority: red = urgent, yellow = need optimization, green = approved). 3. Attachment annotation: Upload supplementary files (such as experimental data Excel) and link them to the annotation location; Real-time synchronization: Using WebSocket technology, once a reviewer adds a comment, other online reviewers can see it in real time (delay ≤100ms). It supports replies, likes (agreeing with opinions), and disagreements (marking "disagree, stress calculation is incorrect").

[0044] 4.3 Summary of Review Comments and Status Tracking Automatic summarization: The system automatically categorizes opinions by "Review Object - Review Group - Opinion Type", for example: 1. Review target: Output-Result-001-V1.0; 2. Simulation group comments: 2 items need modification (mesh accuracy, load direction), 3 items approved; 3. Test group comments: 1 point needs modification (the result deviates from the experimental result by 12%). Status tracking: The review status is divided into "Pending review → Under review → Summarized → Rectified → Passed". After rectification, the engineer uploads the new version (such as Mesh-V1.1), and the system automatically notifies the reviewer to review again until all comments are closed (no items to be modified). Review Report: Automatically generates a PDF review report, including information on the review object, all annotations (with screenshots), rectification records, and final conclusions. Generation time is ≤10 minutes (traditional manual compilation takes 2 days).

[0045] II. Simulation Data Management (SDM) and Collaborative Review System The system serves as the hardware and software implementation platform for the aforementioned methods. It adopts a hybrid architecture of "client-server (C / S) + Web (B / S)" and includes six core functional modules. Each module works collaboratively through standardized interfaces (module name - core hardware / software support - function description - performance indicators): Version Management Module Git 2.43 (version control), MySQL 8.0 (data storage); implements structured version identification, branch management, change logs, and access control; version rollback response time ≤ 1s; supports 1000+ parallel version management.

[0046] Data association module Neo4j 5.15 (graph database), Python graph algorithm library; constructs a "design-simulation-result" association graph, supports bidirectional traceability; traceability response ≤ 5 minutes; association accuracy ≥ 99.9%.

[0047] SPDM Process Management Module Activiti 7.1.0 (workflow engine), ECharts (visualization); provides SPDM templates, automatic parameter recording, process visualization and reproduction; process template creation ≤30 minutes; parameter collection accuracy ≥99%.

[0048] The Web-based collaborative review module, utilizing WebSocket (real-time communication), Canvas (annotation), and Vue3 (frontend), supports real-time annotation (graphics / text / attachments), synchronized comments, and summary annotation with a latency of ≤100ms; it also supports parallel review by 50+ users. Data integration module Third-party tool interfaces (HyperMesh / ABAQUS / ParaViewSDK); integrates with CAE tools to automatically collect mesh, solution, and result parameters; 100% interface adaptation success rate; parameter collection time ≤10s / step.

[0049] Results Output and Reporting Module PDFBox 2.0.32 (PDF generation), ExcelKit (data export); generates version reports, SPDM reproduction manuals, and review reports, and supports data export; report generation time ≤ 10 minutes; export formats supported: PDF / Excel / CSV. Hardware deployment architecture: Server-side: 2 physical servers (primary / backup architecture, Intel Xeon 8375C CPU, 256GB DDR5, 4TB NVMe SSD), deploying database (MySQL / Neo4j), application services (Tomcat 10), and file storage (distributed storage MinIO); Client: Engineer terminal (Windows 11 / macOS Ventura), supporting C / S client (interfacing with CAE tools) and B / S Web client (browser access to collaborative review function); Network environment: Supports LAN (10Gbps) and Internet (VPN encrypted) access to ensure collaboration among multi-regional teams (such as the Beijing design team and the Shanghai simulation team). Detailed Implementation Taking "fatigue simulation data management and collaborative review of automobile chassis lower control arm" as an example, the implementation process of this invention is explained in detail: 1. Input Data and Environment Preparation Project Information: Project ID=Car-Chassis-LowerArm-Fatigue-001, Task: Fatigue life simulation of lower control arm, Team: Design group (3 people), Simulation group (2 people), Testing group (2 people). Initial data: Design model (STEP format, Design-001-V2.1, lower control arm dimensions: 1200mm×150mm×3mm); Hardware environment: Server (Intel Xeon 8375C, 256GB DDR5), Client (Intel i7-13700K, 32GB DDR5); Software environment: This invention system V1.0, HyperMesh2023, LS-DYNAR12.0, ParaView5.11, Chrome browser 120.0.

[0051] 2. Simulation data version management and association Version 2.1 creation: When a simulation engineer imports a design model, the system automatically generates a version identifier: Car-Chassis-LowerArm-Fatigue-001-Design-Design-001-V2.1-202410150900; Perform mesh generation (HyperMesh, AdvancingFront algorithm, size 2mm), generate input data version: Car-Chassis-LowerArm-Fatigue-001-Input-Mesh-001-V1.0-202410151030, and the system automatically associates it with the design model (Design-001-V2.1). Assigned material (45 steel, *MAT_24, ρ=7850kg / m) 3 (E=206GPa), generated input data version: Input-Material-001-V1.0-202410151100, associated mesh version V1.0.

[0052] Version 2.2 Changes: Submit the solution task (LS-DYNA, step size 1e-6s), generate Solver-001-V1.0-202410151400, and associate Mesh-V1.0 with Material-V1.0; After the solution is completed, the generated result version is: Output-Result-001-V1.0-202410160900 (fatigue life 1 million cycles, maximum stress 420MPa), associated with Solver-V1.0.

[0053] 3. SPDM process recording and visualization 3.1 Process Template Selection: The system loads the "Fatigue Life Simulation" template and automatically generates the following steps: Geometry Import → Mesh Generation → Material Assignment → Load Spectrum Import → Solve → Result Analysis; 3.2 Automatic parameter recording: Mesh generation steps: Automatic acquisition algorithm = AdvancingFront, size = 2mm, pass rate = 98.2%, executor = Mr. Li (ID: LG002), time = 202410151000-1030; Solution steps: Automatic acquisition solver = LS-DYNAR12.0, step size = 1e-6s, number of iterations = 5000, convergence rate = 100%; 3.3 Process Visualization: View the Gantt chart on the system's web interface. It shows "Result Analysis → Completed". All other steps have been completed. The total process time is 24 hours (which cannot be statistically analyzed when there are no records in the traditional system).

[0054] 4. Web-based collaborative review 4.1 Review Initiation: The simulation engineer initiates a review, selects the review object = Output-Result-001-V1.0, and specifies the review group = Design Group (review geometry), Simulation Group (review mesh / results), Test Group (review experimental comparison). 4.2 Real-time annotation: Wang, from the simulation team: Circled "Stress around bolt hole 420MPa > 350MPa (yield strength)" on the stress cloud diagram, and commented "Mesh accuracy needs to be optimized to 1.5mm", with priority in red; Test team member Zhang: Upload experimental data (lifetime 1.2 million cycles), marked "simulation lifetime 1 million cycles, deviation 16%, load spectrum needs to be checked", priority = yellow; All annotations were synchronized to other reviewers in real time. After reviewing them, Engineer Li from the design team replied, "The geometry is consistent with the design model, no problem," and the priority was set to green. 4.3 Summary of Feedback and Rectification: The system automatically summarized: 3 comments (2 pending revision, 1 approved); Simulation engineer rectification: 1. Optimize the mesh to 1.5mm, generate Mesh-V1.1-202410161400, and associate it with review comment ID=Review-001; 2. Correct the load spectrum, generate Solver-V1.1-202410161600, and associate it with Mesh-V1.1; Resolve and generate Result-V1.1-202410171000 (life 1.18 million cycles, stress 345 MPa), and initiate a second review; 4.4 Review Passed: If no comments are required for modification during the second review, the system generates a review report and updates the status to "Passed".

[0055] 5. Results traceability and reproduction Reverse tracing: Enter the ID of Result-V1.1, and the system will display the associated link: Design-001-V2.1→Mesh-V1.1→Material-V1.0→Solver-V1.1→Result-V1.1. Clicking on Mesh-V1.1 will show the dimension of 1.5mm and the pass rate of 99.1%. Clicking on Solver-V1.1 will show the load spectrum correction record. Process Reproduction: Export the SPDM reproduction manual, which includes parameters for all steps, tool versions, and download links for data files. New engineers can reproduce the process by following the manual, which takes 45 minutes (the traditional method takes 24 hours).

[0056] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A simulation data management and collaborative review method, characterized in that, Includes the following steps: A structured simulation data version management model is constructed. According to the preset rules, the project identifier, model type, simulation task, version number and timestamp are sequentially concatenated to generate a unique structured version identifier for the target simulation data. When the target simulation data undergoes a version change, change log information including the person making the change, the content of the change, the reason for the change and the related documents is recorded. Construct a full-link simulation data association graph. In the graph database, establish design model nodes, preprocessing input nodes, solution task nodes, postprocessing result nodes, and review record nodes respectively. During the generation and flow of simulation data, automatically extract data nodes with pre- and post-dependent relationships, establish directed association edges between data nodes with pre- and post-dependent relationships, and output the upstream and downstream link node set of a specific data node by forward traversal and backward traversal of the directed association edges. The standardized simulation process is recorded and loaded. A workflow template containing multiple preset simulation execution steps is loaded. During the operation of each simulation execution step, the application programming interface of an external computer-aided engineering tool is called to extract the mesh parameters of the preprocessing stage, the configuration parameters of the solving stage, and the numerical characteristics of the results of the postprocessing stage as process execution parameters. The extracted process execution parameters are then bound to the executor and execution time and written into the corresponding preset simulation execution step node. The system performs web-based collaborative review, rendering the result view layer corresponding to the post-processing result node in the web interface. It receives graphic and text annotation data containing spatial relative coordinate system parameters from online clients for input to the result view layer. The graphic and text annotation data are serialized into a message body of a preset format and broadcast to other online clients in the same review group via the WebSocket protocol for overlay rendering with position alignment. The extracted annotation data is categorized and summarized to generate the review record node. When a pre-processing input node or solution task node with a new structured version identifier is generated based on the review record node containing a state to be modified, a new directed association edge pointing from the review record node to the corresponding new node is automatically established in the graph database.

2. The method according to claim 1, characterized in that, The construction of the structured simulation data version management model also includes: Based on the base version data node with a locked state, a first branch version node corresponding to parameter optimization or a second branch version node corresponding to comparison and verification is generated in response to the branch creation command. Configure an operation permission verification strategy based on role identifiers. When the status of the review record node changes to "review passed", change the data attribute of the simulation data version on the corresponding data link to read-only locked state, intercept the data overwrite instruction for the simulation data version in the read-only locked state, and when a modification request is received, force the creation of a new version branch based on the simulation data version in the read-only locked state.

3. The method according to claim 1, characterized in that, The step of establishing directed edges between data nodes with pre- and post-dependent relationships, and outputting the upstream and downstream link node set of a specific data node by traversing the directed edges in both forward and reverse directions, specifically includes: When a design model is imported and preprocessing is performed to generate preprocessing input data, directed edges pointing from the corresponding design model nodes to the preprocessing input nodes are automatically established in the graph database. When a solution task is submitted, the preprocessing input data called by the solution task is parsed and extracted, and a directed association edge is automatically established from the corresponding preprocessing input node to the solution task node. When the solver completes the calculation and outputs the result file, it automatically establishes a directed association edge from the corresponding solver task node to the post-processing result node; When a forward tracing request based on a certain design model node is received, the system traverses sequentially along the out-degree direction of the directed associated edge and outputs the version evolution link of all preprocessing input nodes, solution task nodes and postprocessing result nodes derived from the design model node. When a reverse tracing request based on a certain post-processing result node is received, the system traverses sequentially along the in-degree direction of the directed associated edges, and outputs the solution stage configuration parameters, pre-processing stage mesh parameters, and initial design model node version sequence on which the generation of the post-processing result node depends.

4. The method according to claim 1, characterized in that, The extraction of mesh parameters from the preprocessing stage, configuration parameters from the solution stage, and numerical features of the results from the postprocessing stage as process execution parameters specifically includes: In the preprocessing stage, the application interface of the mesh generation tool is called to extract the mesh algorithm configuration, element type, size field parameters, and mesh quality index including distortion rate values; During the solution phase, the solver's application programming interface is called to extract the solver version number, computation step size, number of iterations, convergence criteria, and the computation log file generated by the solver is captured and saved. In the post-processing stage, the application programming interface of the post-processing tool is called to extract the numerical values ​​of the stress extrema, displacement extrema, or lifetime extrema of the target result type, as well as the position coordinate parameters of the corresponding extrema in the three-dimensional model space. The process execution parameters of each simulation execution step, the version information of the corresponding computer-aided engineering tools called, and the storage and download address of the generated data files are structurally encapsulated to generate a reproduction configuration file for process restoration.

5. The method according to claim 1, characterized in that, The steps for performing web-based collaborative review also include: Configure color label attributes for the text annotation data to indicate rectification priority, and perform classification and aggregation on the extracted annotation data according to the reviewed object, review role group and opinion type; Configure a flow state machine for the review record node. The state set of the flow state machine includes pending review, under review, summarized, rectified, and passed. When the system detects that a post-processing result node with a new structured version identifier has been generated due to rectification operations, it automatically loads the review comments that are in the pending review state in the original mapping relationship into the pending review list of the new post-processing result node, and sends a second review notification instruction to other online clients in the same review group until the flow state machine of the review record node changes to "passed", triggering the extraction of the labeled view coordinate data, text content and rectification operation records of each flow in the graph database to aggregate and generate a standardized review report file.

6. A simulation data management and collaborative review system, characterized in that, include: The version management module is used to build a structured simulation data version management model. According to preset rules, it sequentially concatenates the project identifier, model type, simulation task, version number and timestamp to generate a unique structured version identifier for the target simulation data. When the target simulation data undergoes version changes, it records change log information including the person making the change, the content of the change, the reason for the change and the related documents. The data association module is configured with a graph database to construct a full-link simulation data association graph. Design model nodes, preprocessing input nodes, solution task nodes, postprocessing result nodes, and review record nodes are established in the graph database. During the generation and flow of simulation data, data nodes with pre- and post-dependent relationships are automatically extracted, directed association edges are established between data nodes with pre- and post-dependent relationships, and the upstream and downstream link node sets of specific data nodes are output by forward and backward traversal of the directed association edges. The SPDM process management module is used to execute standardized simulation process records, load workflow templates containing multiple preset simulation execution steps, and extract mesh parameters from the preprocessing stage, configuration parameters from the solving stage, and numerical characteristics of the results from the postprocessing stage as process execution parameters by calling the application programming interface of external computer-aided engineering tools. The extracted process execution parameters are then bound to the executor and execution time and written into the corresponding preset simulation execution step nodes. The Web Collaborative Review Module is used to perform web-based collaborative reviews. It renders the result view layer corresponding to the post-processing result node in the web interface, receives graphical and textual annotation data containing spatial relative coordinate system parameters input from online clients to the result view layer, serializes the graphical and textual annotation data into a message body of a preset format, and broadcasts it via the WebSocket protocol to other online clients in the same review group for position-aligned overlay rendering. It categorizes and summarizes the extracted annotation data to generate the review record node. When a pre-processing input node or solution task node with a new structured version identifier is generated based on the review record node containing a state to be modified, a new directed edge pointing from the review record node to the corresponding new node is automatically established in the graph database.

7. The system according to claim 6, characterized in that, The version management module is also used for: Based on the base version data node with a locked state, a first branch version node corresponding to parameter optimization or a second branch version node corresponding to comparison and verification is generated in response to the branch creation command. Configure an operation permission verification strategy based on role identifiers. When the status of the review record node changes to "review passed", change the data attribute of the simulation data version on the corresponding data link to read-only locked state, intercept the data overwrite instruction for the simulation data version in the read-only locked state, and when a modification request is received, force the creation of a new version branch based on the simulation data version in the read-only locked state.

8. The system according to claim 6, characterized in that, The data association module is specifically used for: When a design model is imported and preprocessing is performed to generate preprocessing input data, directed edges pointing from the corresponding design model nodes to the preprocessing input nodes are automatically established in the graph database. When a solution task is submitted, the preprocessing input data called by the solution task is parsed and extracted, and a directed association edge is automatically established from the corresponding preprocessing input node to the solution task node. When the solver completes the calculation and outputs the result file, it automatically establishes a directed association edge from the corresponding solver task node to the post-processing result node; When a forward tracing request based on a certain design model node is received, the system traverses sequentially along the out-degree direction of the directed associated edge and outputs the version evolution link of all preprocessing input nodes, solution task nodes and postprocessing result nodes derived from the design model node. When a reverse tracing request based on a certain post-processing result node is received, the system traverses sequentially along the in-degree direction of the directed associated edges, and outputs the solution stage configuration parameters, pre-processing stage mesh parameters, and initial design model node version sequence on which the generation of the post-processing result node depends.

9. The system according to claim 6, characterized in that, The SPDM workflow management module includes a pre-processing adaptation unit, a solution adaptation unit, and a post-processing adaptation unit: The preprocessing adaptation unit is used to call the application programming interface of the mesh generation tool to extract the mesh algorithm configuration, element type, size field parameters, and mesh quality index including distortion rate values. The solver adaptation unit is used to call the solver's application programming interface to extract the solver version number, computation step size, number of iterations, convergence criteria, and capture and save the computation log file generated by the solver. The post-processing adaptation unit is used to call the application programming interface of the post-processing tool to extract the numerical values ​​of the stress extrema, displacement extrema, or lifetime extrema of the target result type, as well as the position coordinate parameters of the corresponding extrema in the three-dimensional model space. The SPDM process management module is also used to structurally encapsulate the process execution parameters of each of the above simulation execution steps, the version information of the corresponding computer-aided engineering tools called, and the storage and download address of the generated data files, and generate a reproduction configuration file for process restoration.

10. The system according to claim 6, characterized in that, The Web collaborative review module is also used for: Configure color label attributes for the text annotation data to indicate rectification priority, and perform classification and aggregation on the extracted annotation data according to the reviewed object, review role group and opinion type; Configure a flow state machine for the review record node. The state set of the flow state machine includes pending review, under review, summarized, rectified, and passed. When the system detects that a post-processing result node with a new structured version identifier has been generated due to rectification operations, it automatically loads the review comments that are in the pending review state in the original mapping relationship into the pending review list of the new post-processing result node, and sends a second review notification instruction to other online clients in the same review group until the flow state machine of the review record node changes to "passed", triggering the extraction of the labeled view coordinate data, text content and rectification operation records of each flow in the graph database to aggregate and generate a standardized review report file.

Images