Doll mechanism automatic adjustment method and system based on finite element pre-processing software

Abstract

The present application relates to the field of finite element pre-processing software, and discloses a dummy mechanism automatic adjustment method and system based on finite element pre-processing software, which comprises the following steps: obtaining full-field physical field distribution data of a target dummy model based on a preset simulation working condition, calculating stress gradient data of a grid unit, extracting a grid area with a numerical value greater than a preset threshold value in the stress gradient data, and performing encryption processing of local grid size reduction and node increase on the grid unit in the grid area; and outputting the processed target dummy model. The system focuses on solving the problems of complicated operation, non-uniform parameter configuration, and grid division depending on experience in the existing software during dummy mechanism modeling, and realizes efficient cooperation of dummy model posture adjustment, contact relationship definition, boundary condition setting, and grid quality optimization through the construction of a parameterized, automated, and intelligent adjustment framework.

Description

Technical Field

[0001] This invention relates to the field of finite element preprocessing software, and more particularly to an automatic adjustment method and system for dummy mechanisms based on finite element preprocessing software. Background Technology

[0002] Finite element preprocessing software (such as Abaqus, ANSYS, and LS-DYNA preprocessors) is a core tool for simulation modeling and is widely used to construct simulation models of whole vehicles or partial structures that include dummy models. However, in practical engineering applications, the modeling and adjustment of dummy mechanisms still face many challenges: 1. Manual operation is complex and inefficient: Current mainstream CAE software lacks dedicated adjustment tools for dummy models. Engineers must manually adjust each joint degree of freedom (such as hip, knee, shoulder, spine, etc.) one by one in the graphical interface, and set contact pairs, constraints, and material parameters. Taking the Hybrid III 50th male dummy model as an example, completing a full sitting posture modeling typically involves the coordinated configuration of more than 200 key parameters, with a single adjustment generally taking 2 to 3 hours, severely restricting project progress. In addition, there are strong coupling relationships between the parameters (such as changes in the knee joint angle affecting the contact position between the foot and the pedal), and manual adjustment is very likely to cause geometric interference or boundary condition distortion, leading to simulation failure or result deviation.

[0003] 2. Lack of standardization and reusability: When different projects or engineers handle similar scenarios (such as frontal collisions and side impacts), they often repeatedly perform the same parameter configuration work, lacking a unified parameter template and adjustment logic accumulation mechanism. This not only wastes human resources but also leads to inconsistent simulation model quality, affecting the comparability and credibility of the results. For example, in traditional whole-vehicle collision simulation, different teams can set boundary conditions for the same dummy model differently by up to 20%, reducing the consistency of results.

[0004] 3. The setting of boundary conditions is highly subjective, affecting the reliability of the simulation: The boundary conditions, such as the force distribution and constraint methods between the dummy and components like the seatbelt, seat, and steering wheel, are typically set based on the engineer's experience and judgment. For example, the location and magnitude of the seatbelt pretension force, and even minor deviations in the dummy's initial posture, can significantly affect the simulation results. This subjectivity introduces human error, reducing the repeatability and scientific rigor of the simulation.

[0005] 4. Mesh generation quality depends on human experience: In finite element analysis, mesh quality directly affects computational accuracy and convergence. High-stress areas in dummy models (such as the pelvis and thoracic vertebrae) require local mesh refinement, but current software does not provide intelligent recognition and automatic optimization functions. Engineers must manually mesh based on experience, which is labor-intensive and difficult to ensure consistency.

[0006] 5. Grid penetration risk: Achieving posture changes solely through mechanical joint rotation without considering the compression and deformation of soft components (such as bionic skin and lining) can easily lead to mesh cross-penetration, causing calculation errors or distorted results. 6. Compatibility limitations: Some adjustment methods rely on specific software or models, lack cross-platform adaptability, and do not integrate a collaborative mechanism between weight adjustment and attitude adjustment.

[0007] While some research has attempted to simplify certain operations through batch processing using scripts, such as Python scripts based on Abaqus / CAE, these methods remain limited to linear processes and cannot handle the dynamic linkages and conditional judgments of complex mechanisms. For example, published patent CN202511657889.5 proposes a cross-platform code portability scheme, which improves script portability but does not solve the problem of adaptive adjustment of the dummy mechanism under different postures, lacking intelligent response capabilities to physical constraints and simulation objectives. In the automotive industry, a car company optimized its collision safety design using Abaqus simulation, reducing physical testing costs by 40%, but dummy adjustment still relies on manual operation, with a single posture adjustment taking over 2 hours. Summary of the Invention

[0008] To address the aforementioned technical problems, this invention provides an automatic adjustment method for a dummy mechanism based on finite element preprocessing software, comprising: In response to the model loading command, the parameterized template corresponding to the target dummy model is loaded from the parameterized template library. The parameterized template includes the dummy's joint angle range, mass distribution parameters, contact pair definitions, and joint linkage rules. Acquire target posture data and counterweight parameters, calculate the linkage angle and displacement of each joint of the target dummy model based on the target posture data and the joint linkage rules, and adjust the posture of the target dummy model and the counterweight distribution of each part according to the linkage angle and displacement of each joint. Perform multi-step iterative mesh interference check on the target dummy model after posture adjustment, extract the set of interference mesh nodes that have spatial overlap, calculate the penetration depth of the set of interference mesh nodes, and perform mesh Morph deformation operation to drive the set of interference mesh nodes to move the distance value corresponding to the penetration depth in the opposite direction to the penetration direction. Based on the preset simulation conditions, the full-field physical field distribution data of the target dummy model is obtained, the stress gradient data of the grid cells is calculated, the grid regions with values ​​greater than a preset threshold in the stress gradient data are extracted, and the grid cells in the grid regions are subjected to local mesh size reduction and node addition densification processing. Output the processed target dummy model.

[0009] This invention also provides an automatic adjustment system for a dummy mechanism based on finite element preprocessing software. The system is integrated into the finite element preprocessing software as a plug-in, comprising: The template loading module is used to load the parameterized template corresponding to the target dummy model from the parameterized template library in response to the model loading command. The parameterized template includes the dummy joint angle range, mass distribution parameters, contact pair definition and joint linkage rules. The posture and weight adjustment module is used to acquire target posture data and weight parameters, calculate the linkage angle and displacement of each joint of the target dummy model based on the target posture data and the joint linkage rules, and adjust the posture of the target dummy model and the weight distribution of each part according to the linkage angle and displacement of each joint. The interference processing module is used to perform multi-step iterative mesh interference checks on the target dummy model after the posture is adjusted, extract the set of interference mesh nodes that have spatial overlap, calculate the penetration depth of the set of interference mesh nodes, and perform a mesh Morph deformation operation to drive the set of interference mesh nodes to move the distance value corresponding to the penetration depth in the opposite direction to the penetration direction. The mesh optimization module is used to obtain the full-field physical field distribution data of the target dummy model based on the preset simulation conditions, calculate the stress gradient data of the mesh cells, extract the mesh regions in the stress gradient data whose values ​​are greater than a preset threshold, and perform local mesh size reduction and node addition densification processing on the mesh cells in the mesh regions. The model output module is used to output the processed target dummy model.

[0010] Compared with the prior art, the beneficial effects of the present invention are as follows: The adjustment time of the dummy mechanism of the present invention is reduced from the traditional 2-3 hours to 10-20 minutes, the overall pre-processing efficiency is improved by 3-5 times, and the research and development cycle is significantly shortened.

[0011] This invention reduces the simulation error rate of key parts by more than 15% through a standardized parameter library and intelligent mesh optimization, and significantly improves the repeatability of the results.

[0012] This invention completely eliminates mesh penetration by performing multi-step iterative interference checks and eliminating interference or Morph deformation, keeping the attitude deviation within ±1.0mm and significantly improving simulation accuracy. This invention supports coordinated adjustment of posture and weight, is compatible with mainstream dummy models, and adapts to testing needs in various scenarios such as collision and vibration. Parameter templates and adjustment logic can be stored and shared long-term, forming enterprise-level knowledge assets and reducing reliance on the experience of individual engineers.

[0013] The system of this invention adopts a plug-in design, is compatible with mainstream software such as Ls-dyna and Pamcrash, and can be expanded to fields such as medical simulation and human factors engineering in the future. Visual wizards and automated processes enable even junior engineers to quickly get started, improving the overall modeling capabilities of the team. Attached Figure Description

[0014] Figure 1 A flowchart provided for an embodiment of the present invention. Detailed Implementation

[0015] 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.

[0016] This invention belongs to the field of Computer-Aided Engineering (CAE) technology, specifically involving the secondary development and intelligent functional expansion of finite element preprocessing software. It is particularly suitable for simulation scenarios requiring high-precision human models, such as automotive collision safety simulation, human biomechanical analysis, and verification of the mechanical performance of medical devices. The system focuses on solving problems existing in current software during dummy modeling, such as cumbersome operation, inconsistent parameter configuration, and reliance on experience for mesh generation. By constructing a parameterized, automated, and intelligent adjustment framework, it achieves efficient coordination in dummy model posture adjustment, contact relationship definition, boundary condition setting, and mesh quality optimization.

[0017] This invention aims to overcome the shortcomings of existing technologies and provide an efficient, accurate, and reusable automatic adjustment system and method for dummy mechanisms, specifically addressing the following technical problems: 1. Significantly reduces the degree of human intervention in the dummy modeling process and improves preprocessing efficiency; 2. Establish a standardized and configurable parameter template system to realize the knowledge accumulation and reuse of adjustment logic; 3. Optimize mesh generation or mesh Morph deformation and boundary condition settings through intelligent algorithms to reduce human error and improve the reliability and consistency of simulation results; 4. Build a user-friendly interface to lower the barrier to entry and enable non-expert users to quickly complete high-quality modeling.

[0018] The present invention is applicable to the following scenarios: I. Automotive Industry: Core Scenarios for Passive Safety Simulation This is the core application area of ​​the system, directly matching the basic design of collision and vibration tests in the patent, covering the entire process from R&D to certification: 1. Collision safety simulation: including legally mandated test scenarios such as frontal collision, side collision, rear-end collision, and rollover collision (e.g., C-NCAP, E-NCAP certification), compatible with bionic dummies such as AC-HUMs and THUMS, automatically adjusting the seating posture and weight distribution of the driver and passengers (including child dummies) to ensure the accuracy of force simulation for the head, chest, and legs during a collision, supporting the optimized design of airbags, seat belts, and vehicle body structure; 2. Vibration and Comfort Test: For road vibrations during vehicle operation (such as bumpy roads and high-speed cruising), the dummy's posture is adjusted to a natural sitting state. By optimizing the weight distribution of passengers of different body types, the vibration transmission characteristics of the seat and suspension system to the human body are simulated to improve riding comfort. 3. Special scenario simulation: such as the escape posture simulation after a new energy vehicle battery catches fire, the dummy posture response during emergency braking of an autonomous vehicle, and the seating posture adaptation simulation of commercial vehicle (truck, bus) drivers (requires adjustment of a large-size dummy model).

[0019] II. Aerospace: Crew Safety and Cabin Layout Scenarios Adapted to the aerospace industry's need for precise attitude control of dummies in extreme environments, and compatible with lightweight dummy models: 1. Aircraft Collision / Emergency Landing Simulation: Adjust the seating posture of pilot and passenger dummies in the cockpit / cabin, match seat belt restraints and seat layout, simulate the impact load transmission during emergency landing, and optimize seat strength and seat belt anchor point design; 2. Spacecraft (spacecraft, space station) crew attitude simulation: In response to the overload environment during microgravity and launch / return phases, the dummy attitude is adjusted to the working or restraint state, and the human body mass distribution is simulated by coordinating counterweight optimization to verify the reliability of the cabin operating space and restraint system. 3. Aircraft seat comfort test: Simulate the sitting posture of a dummy under different flight attitudes (takeoff, cruise, landing), and test the effect of seat back angle and headrest position on human fatigue by adjusting the posture and counterweight in coordination.

[0020] III. Rail Transit: Safety Scenarios for Train Passengers Adaptable to the collision and vibration testing requirements of high-speed rail, subway, and urban rail transit; compatible with multi-body dummy models. 1. Train collision / derailment simulation: Adjust passenger dummies to different postures such as sitting and standing, match the layout of handrails and seats in the carriage, simulate the risk of secondary collisions during a collision (such as collisions between the human body and the inner wall and seats of the carriage), and optimize the carriage structure and safety facilities. 2. Track vibration transmission test: To address the vibration of the carriage caused by track irregularities, the dummy was adjusted to a natural sitting / standing posture, and the mass characteristics of different groups (adults, children, and the elderly) were matched by adjusting the counterweights to verify the impact of vibration on passenger comfort and safety.

[0021] IV. Medical Devices and Protective Equipment: Human Adaptive Simulation Scenarios Leveraging the advantages of precise system attitude adjustment and penetration-resistant soft components, it is suitable for research and development testing of medical and protective equipment. 1. Performance testing of protective equipment (helmets, bulletproof vests): Adjust the head and torso posture of the dummy to the standard test angle (such as the 15° / 30° impact angle for helmet testing), optimize the human body mass distribution by using counterweights, and simulate the energy absorption effect and protective function of the protective equipment under impact load. 2. Rehabilitation equipment adaptability simulation: such as the development of wheelchairs, prostheses, and rehabilitation braces; adjusting the dummy posture to the typical posture of disabled people (such as the sitting posture of hemiplegic patients and the standing posture of lower limb amputees); simulating the mass distribution of human residual limbs with synergistic weights; and verifying the adaptability and safety of the equipment. 3. Medical collision scenario simulation: such as simulating the fixed posture of patients during ambulance transfer, adjusting the dummy to a lying / semi-lying posture, simulating the displacement risk of patients during sudden braking and turning, and optimizing the design of ambulance stretchers and restraint belts.

[0022] V. General Industry and Scientific Research: Customized Simulation Scenarios Adaptable to the customized dummy adjustment needs of research institutions and specialized industrial fields: 1. Industrial collision safety testing: such as collision protection simulation for operators of construction machinery (excavators, cranes), adjusting the dummy to the operating posture, matching the cab layout, and simulating the operator's safety status when the equipment tips over or collides; 2. Human Dynamics Research: Research conducted by universities and research institutions on human movement postures and impact responses, using systems to quickly adjust dummies to complex postures (such as jumping and falling), and precisely control the weights and grid shape to support the verification of dynamic models; Military equipment testing: such as safety simulation of the crew compartment of armored vehicles, paratrooper landing posture simulation, adjusting dummies to combat / landing posture, simulating the mass distribution of individual soldier equipment (weapons, backpacks) with coordinated counterweights, and optimizing equipment design and protection schemes.

[0023] This invention proposes an "intelligent adjustment framework" based on secondary development of finite element preprocessing software. By integrating three core modules—a parametric template library, an adaptive mesh optimization engine, and a real-time feedback control interface—it achieves fully automated adjustment of the dummy mechanism throughout the entire process. The system is embedded as a plug-in into the mainstream domestic CAE preprocessing software, DataGeology G5, without requiring modification of the underlying source code, and possesses excellent compatibility and scalability.

[0024] 1. Parametric Template Library Preset template integration: The system has built-in standard parameter templates for various internationally recognized dummy models (such as Hybrid Ⅲ 5th / 50th / 95th, THUMS v4.02, GHBMC), covering key parameters such as joint angle range, mass distribution, material properties, and contact pair definitions. Users can load the corresponding template with one click according to simulation needs, avoiding repeated configuration.

[0025] Custom template support: Users can customize adjustment logic through graphical configuration tools or XML files. For example, they can define conditional rules such as "automatically adjust the foot contact surface position when the knee flexion angle is greater than 90°", and the system will automatically execute the logical judgment and parameter linkage during runtime.

[0026] Version Management and Sharing: Templates support version control and team sharing, making it easy for enterprises to establish a unified modeling standard library and improve team collaboration efficiency.

[0027] 2. Adaptive Mesh Optimization Engine Mesh sensitivity analysis based on physical fields: Based on preliminary simulation or pre-calculation, the system analyzes the distribution of physical fields such as stress gradient and strain energy density, and identifies high deformation and high stress areas (such as the pelvis and head in a collision).

[0028] Dynamic grid encryption algorithm: Calculate the stress gradient across the entire field Automatically refine the mesh in regions where the gradient exceeds a threshold.

[0029] Machine learning-assisted prediction: The system can access historical simulation data to train a lightweight neural network model, predicting high-risk areas under typical working conditions and achieving "predictive" mesh optimization, further reducing computational resource waste. Experiments show that this engine can reduce manual mesh generation by more than 80% while improving the computational accuracy of critical areas.

[0030] 3. Real-time feedback control interface Visual wizard interface: The system provides a step-by-step graphical wizard, and users only need to complete the following 5 steps to complete the adjustment: A [Select Dummy Model] --> B [Load Parameter Template] B --> C [Set target attitude (angle / displacement)] C --> D [Activate Adaptive Mesh Optimization] D --> E [Generate Simulation Readiness Report] Multi-source data fusion capability: The system supports real-time docking with external devices (such as optical motion capture systems and force sensors) to directly map measured human posture data to the simulation model, realize "measurement-simulation" closed-loop calibration, and significantly improve the model's realism.

[0031] API Open Interface: Provides a Python API to support advanced users in script invocation and process integration, meeting the needs of automated simulation pipelines.

[0032] Specifically, such as Figure 1 The present invention provides an automatic adjustment method for a dummy mechanism based on finite element preprocessing software, comprising: In response to the model loading command, the parameterized template corresponding to the target dummy model is loaded from the parameterized template library. The parameterized template includes the dummy's joint angle range, mass distribution parameters, contact pair definitions, and joint linkage rules. Acquire target posture data and weight parameters, calculate the linkage angle and displacement of each joint of the target dummy model based on the target posture data and joint linkage rules, and adjust the posture of the target dummy model and the weight distribution of each part according to the linkage angle and displacement of each joint. Perform multi-step iterative mesh interference check on the target dummy model after attitude adjustment, extract the set of interference mesh nodes that have spatial overlap, calculate the penetration depth of the interference mesh node set, and perform mesh Morph deformation operation to drive the interference mesh node set to move along the direction opposite to the penetration direction by the distance value corresponding to the penetration depth. Based on the preset simulation conditions, the full-field physical field distribution data of the target dummy model is obtained, the stress gradient data of the grid cells is calculated, the grid regions with stress gradient values ​​greater than the preset threshold are extracted, and the grid cells within the grid regions are subjected to local mesh size reduction and node addition densification processing. Output the processed target dummy model.

[0033] In this invention, before the step of loading the parameterized template corresponding to the target dummy model from the parameterized template library, the method further includes: The process includes: reading a custom configuration file in Extensible Markup Language (XML) format; parsing the custom configuration file, extracting the conditional judgment logic for the flexion angle of the dummy joints, and configuring the conditional judgment logic into the joint linkage rules; and calculating the linkage angle and displacement of each joint of the target dummy model based on the target posture data and the joint linkage rules. This includes: obtaining the flexion angle data of the first adjustable joint; when the flexion angle data of the first adjustable joint is greater than the angle threshold set by the conditional judgment logic, extracting the coordinate compensation amount of the bound second adjustable joint contact surface based on the flexion angle data of the first adjustable joint, and modifying the spatial relative position of the second adjustable joint contact surface according to the coordinate compensation amount.

[0034] In this invention, the step of performing a Morph deformation operation on the mesh to drive the interferometric mesh node set to move along a direction opposite to the penetration direction by a distance corresponding to the penetration depth includes: The total attitude adjustment displacement is discretized into multiple adjustment steps for iteration. Within each adjustment step, the set of interferometric mesh nodes on the main contact surface and the secondary contact surface that have spatial overlap is extracted, and the penetration depth value of the set of interferometric mesh nodes in the normal direction is calculated. A spatial interpolation function is constructed using the penetration depth value as the spatial displacement boundary condition. Based on the spatial interpolation function, the set of interferometric mesh nodes and the surrounding mesh nodes are driven to perform position offset, and a multi-step iterative mesh interference check is triggered cyclically until the recalculated penetration depth value is zero.

[0035] In this invention, the steps of obtaining the full-field physical field distribution data of the target dummy model based on preset simulation conditions and calculating the stress gradient data of the mesh elements include: Extract the geometric parameters of the target dummy model and the load boundary parameters of the preset simulation conditions; input the geometric parameters and load boundary parameters into the pre-trained neural network model, and output the strain energy density matrix of the target dummy model under the preset simulation conditions; calculate the stress gradient values ​​of the grid elements in the whole field based on the strain energy density matrix, and generate the physical field distribution data of the whole field.

[0036] In this invention, the steps of obtaining target attitude data and counterweight parameters include: Establish a real-time data communication link with the external optical motion capture system and force sensor; receive the measured spatial coordinate sequence and force distribution data of the human body collected by the external optical motion capture system and force sensor; map the measured spatial coordinate sequence of the human body to the corresponding control node of the target dummy model through the coordinate transformation matrix to generate target posture data; generate counterweight parameters according to the force distribution data; when adjusting the posture of the target dummy model and the counterweight distribution of each part, allocate and bind the counterweight parameters to the corresponding counterweight grid node inside the target dummy model in the form of additional mass points.

[0037] This invention also relates to an automatic adjustment system for a dummy mechanism based on finite element preprocessing software. The system is integrated into the finite element preprocessing software as a plug-in, comprising: The template loading module is used to load the parametric template corresponding to the target dummy model from the parametric template library in response to the model loading command. The parametric template includes the dummy's joint angle range, mass distribution parameters, contact pair definitions, and joint linkage rules. The posture and weight adjustment module is used to acquire target posture data and weight parameters, calculate the linkage angle and displacement of each joint of the target dummy model based on the target posture data and joint linkage rules, and adjust the posture of the target dummy model and the weight distribution of each part according to the linkage angle and displacement of each joint. The interference processing module is used to perform multi-step iterative mesh interference checks on the target dummy model after the posture is adjusted, extract the set of interference mesh nodes that have spatial overlap, calculate the penetration depth of the set of interference mesh nodes, and perform mesh Morph deformation operation to drive the set of interference mesh nodes to move in the opposite direction to the penetration direction by a distance value corresponding to the penetration depth. The mesh optimization module is used to obtain the full-field physical field distribution data of the target dummy model based on the preset simulation conditions, calculate the stress gradient data of the mesh cells, extract the mesh regions in the stress gradient data whose values ​​are greater than the preset threshold, and perform local mesh size reduction and node addition densification processing on the mesh cells in the mesh regions. The model output module is used to output the processed target dummy model.

[0038] In this invention, the template loading module is configured with a custom configuration unit; The custom configuration unit is used to: read custom configuration files in Extensible Markup Language (EXPLAIN) format; parse the custom configuration files, extract the conditional judgment logic for the dummy joint flexion angle, and configure the conditional judgment logic into the joint linkage rules; The posture and weight adjustment module is specifically used to: obtain the flexion angle data of the first adjustment joint; when the flexion angle data of the first adjustment joint is greater than the angle threshold set by the condition judgment logic, extract the coordinate compensation amount of the bound second adjustment joint contact surface based on the flexion angle data of the first adjustment joint, and modify the spatial relative position of the second adjustment joint contact surface according to the coordinate compensation amount.

[0039] In this invention, the interference processing module includes a discrete iteration unit and a morphology repair unit; Discrete iterative units are used to discretize the total attitude adjustment displacement into multiple adjustment steps for iteration; The morphological repair unit is used to extract the set of interference mesh nodes on the main contact surface and the secondary contact surface where spatial overlap occurs within each adjustment step, and calculate the penetration depth value of the set of interference mesh nodes in the normal direction; construct a spatial interpolation function with the penetration depth value as the spatial displacement boundary condition, drive the position offset of the set of interference mesh nodes and the surrounding mesh nodes based on the spatial interpolation function, and repeatedly trigger multi-step iterative mesh interference checks until the recalculated penetration depth value is zero.

[0040] In this invention, the mesh optimization module has an embedded neural network prediction model pre-trained using historical simulation data; The mesh optimization module is specifically used for: extracting the geometric parameters of the target dummy model and the load boundary parameters of the preset simulation conditions; inputting the geometric parameters and load boundary parameters into a pre-trained neural network prediction model, and outputting the strain energy density matrix of the target dummy model under the preset simulation conditions; calculating the stress gradient values ​​of the grid elements in the entire field based on the strain energy density matrix, and generating the physical field distribution data of the entire field.

[0041] This invention also includes a data fusion interface module; The data fusion interface module is used to: establish a real-time data communication link with the external optical motion capture system and force sensor; receive the measured spatial coordinate sequence and force distribution data of the human body collected by the external optical motion capture system and force sensor; map the measured spatial coordinate sequence of the human body to the corresponding control node of the target dummy model through the coordinate transformation matrix to generate target posture data; generate counterweight parameters according to the force distribution data; and when adjusting the posture of the target dummy model and the counterweight distribution of each part, allocate and bind the counterweight parameters to the corresponding counterweight grid node inside the target dummy model in the form of additional mass points.

[0042] In existing technologies, dummy posture adjustment heavily relies on engineers manually configuring hundreds of tightly coupled joint parameters one by one in a graphical interface. This process is time-consuming and prone to model distortion or inconsistent standards due to subjective operation. This invention introduces parameterized templates and joint linkage rules (and supports custom conditional logic parsing XML format), achieving a leap from isolated manual adjustment to global linkage calculation. The system can automatically calculate the linkage angles and coordinate compensation of each joint based on the target posture, significantly reducing the time for a single complete dummy modeling from the traditional 2-3 hours to 10-20 minutes, improving overall preprocessing efficiency by 3-5 times. Simultaneously, this mechanism solidifies expert adjustment experience into executable underlying code rules, completely eliminating deviations caused by subjective configurations between different personnel and projects, achieving knowledge accumulation and high-fidelity reuse of enterprise-level simulation adjustment logic.

[0043] To address the problem that relying solely on rigid rotation of mechanical joints can easily lead to mesh compression and penetration in soft-covered components (such as bionic skin and linings), resulting in solver errors or distorted results, this invention innovatively employs a multi-step iterative mesh interference check combined with Morph mesh deformation processing. The system accurately extracts the penetration depth of spatially overlapping mesh nodes and uses this as a boundary condition to construct a spatial interpolation function, driving the interference mesh to precisely shift and yield along the penetration direction until the penetration depth value is strictly zero. This technique, without destroying the original mesh topology, completely eliminates mesh intersection and overlap dead angles from a physical mechanism, strictly constraining attitude geometry deviations within ±1.0 mm, greatly reducing the probability of solver errors under extreme deformation conditions and improving the convergence stability of the simulation model.

[0044] To address the issues of poor consistency and uneven computational power distribution caused by the heavy reliance on manual experience in mesh generation for high-stress regions, this invention proposes an adaptive mesh optimization scheme **based on the overall physical field distribution and stress gradient data (combined with predictions from a pre-trained neural network model)**. The system can automatically and predictively identify high-risk areas where stress gradients exceed limits under realistic simulation conditions (such as the pelvis and compressed intervertebral discs in collision scenarios), and automatically perform local mesh size reduction and node addition operations only for these high-deformation areas. This dynamic densification strategy, which allocates resources as needed, reduces the workload of manual mesh generation by more than 80% while effectively avoiding the massive waste of computational power caused by a globally high-density mesh. Experimental results show that the simulation calculation error rate for key parts is significantly reduced by more than 15% (e.g., the rib strain prediction error in side-impact simulations decreased from 18% to 5%), perfectly balancing computational accuracy and the economy of hardware and software resources.

[0045] To address the limitations of subjective boundary conditions and weight settings that are difficult to accurately match real-world physical scenarios, this invention establishes a real-time data fusion communication link with an external optical motion capture system and force sensors. Through a coordinate transformation matrix, the system directly maps the measured spatial coordinate sequence of a real human body to the control node posture of the dummy model, and converts the measured force distribution into additional mass points that are precisely bound to the internal weight mesh of the model. This mechanism synchronously imports the real human posture and mass weights in a three-dimensional dynamic manner, eliminating the blindness of human pre-setting at the source and constructing a high-fidelity "measurement-simulation" closed-loop calibration system. This effect enables the generated simulation model to not only accurately meet the requirements of automotive crash testing, but also seamlessly adapt to complex simulation scenarios with extremely high biomechanical realism requirements, such as aerospace microgravity occupant safety and medical rehabilitation equipment adaptation verification for disabled individuals.

[0046] Taking a C-NCAP frontal 50% offset crash simulation conducted by an OEM as an example, the application process of this system is explained: 1. Initialization: Load the plugin for this invention in Lsdyna / CAE and start the "Dummy Adjustment Wizard"; select the "HybridⅢ 50th Male" model from the template library, and the system will automatically load the standard parameter set.

[0047] 2. Attitude settings: The user drags the arm control point in the 3D view and sets the arm bending angle to 15°; the system automatically calculates the linkage angle of the shoulder and wrist joints and checks for geometric interference.

[0048] 3. Mesh optimization: When the "Adaptive Mesh Optimization" function is activated, the system predicts that the pelvis and chest are high-stress areas based on preset collision conditions, and automatically refines the mesh in these areas (reducing the element size from 5mm to 1mm), while keeping the mesh coarse in other areas to save computational resources; when the "Morph Mesh Optimization" function is activated, the system deforms the mesh with "distortion" or "interference" based on preset dummy pose to meet the solver's mesh quality requirements.

[0049] 4. Output and Verification: Generate a report containing a parameter list, mesh quality report (such as Jacobian value, aspect ratio), contact status and simulation readiness flags for engineers to review; the model can be directly submitted to the solver for subsequent simulation calculations.

[0050] Case 1: Simulation Optimization of Car Side Collision A European automaker used this system for side pole impact simulation when developing a new electric vehicle. Traditional methods required manual adjustment of the dummy's pelvic tilt and head position, taking 2.5 hours, and often resulted in inaccurate calculations of thoracic loads due to parameter coupling. After using this system: Load Template: Select the "Euro NCAP Side Impact" preset template to automatically set the dummy's initial posture and seat belt restraint; Intelligent adjustment: Import real vehicle sensor data through a real-time feedback interface to calibrate the dummy's hip displacement (error <2mm). Results: The adjustment time was reduced to 12 minutes. The simulation results and physical tests showed that the rib strain prediction error was reduced from 18% to 5%, and the project cycle was shortened by 40%.

[0051] Case 2: Biomechanical Medical Applications In the field of medical rehabilitation, a research institution used this system to build a simulation model of cervical spine injury. The traditional process relied on manual mesh generation using HyperMesh, requiring three days to build a single model and making it difficult to reproduce clinical data. After integrating this system: Custom template: Based on patient CT scan data, create an XML template to define the intervertebral disc stress rules; Dynamic optimization: The mesh engine automatically identifies high-strain zones (such as C5-C6 intervertebral discs) and locally refines the mesh to 0.5mm elements.

[0052] 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. An automatic adjustment method for a dummy mechanism based on finite element preprocessing software, characterized in that, include: In response to the model loading command, the parameterized template corresponding to the target dummy model is loaded from the parameterized template library. The parameterized template includes the dummy's joint angle range, mass distribution parameters, contact pair definitions, and joint linkage rules. Acquire target posture data and counterweight parameters, calculate the linkage angle and displacement of each joint of the target dummy model based on the target posture data and the joint linkage rules, and adjust the posture of the target dummy model and the counterweight distribution of each part according to the linkage angle and displacement of each joint. Perform multi-step iterative mesh interference check on the target dummy model after posture adjustment, extract the set of interference mesh nodes that have spatial overlap, calculate the penetration depth of the set of interference mesh nodes, and perform mesh Morph deformation operation to drive the set of interference mesh nodes to move the distance value corresponding to the penetration depth in the opposite direction to the penetration direction. Based on the preset simulation conditions, the full-field physical field distribution data of the target dummy model is obtained, the stress gradient data of the grid cells is calculated, the grid regions with values ​​greater than a preset threshold in the stress gradient data are extracted, and the grid cells in the grid regions are subjected to local mesh size reduction and node addition densification processing. Output the processed target dummy model.

2. The automatic adjustment method for a dummy mechanism based on finite element preprocessing software according to claim 1, characterized in that, Before the step of loading the parameterized template corresponding to the target dummy model from the parameterized template library, the method further includes: Read custom configuration files in Extensible Markup Language (XML) format; The custom configuration file is parsed to extract the conditional judgment logic for the flexion angle of the dummy joints, and the conditional judgment logic is configured into the joint linkage rules. The step of calculating the linkage angle and displacement of each joint of the target dummy model based on the target posture data and the joint linkage rules includes: Obtain the flexion angle data of the first adjustment joint. When the flexion angle data of the first adjustment joint is greater than the angle threshold set by the condition judgment logic, extract the coordinate compensation amount of the bound second adjustment joint contact surface based on the flexion angle data of the first adjustment joint, and modify the spatial relative position of the second adjustment joint contact surface according to the coordinate compensation amount.

3. The automatic adjustment method for a dummy mechanism based on finite element preprocessing software according to claim 1, characterized in that, The step of performing the Morph deformation operation to drive the interference mesh node set to move along a direction opposite to the penetration direction by a distance value corresponding to the penetration depth includes: The total attitude adjustment displacement is discretized into multiple adjustment steps and iterated; Within each adjustment step, the set of interference grid nodes on the main contact surface and the secondary contact surface that have spatial overlap is extracted, and the penetration depth of the set of interference grid nodes in the normal direction is calculated. A spatial interpolation function is constructed using the penetration depth value as the spatial displacement boundary condition. Based on the spatial interpolation function, the interferometric mesh node set and surrounding mesh nodes are driven to perform position offset, and the multi-step iterative mesh interference check is triggered cyclically until the recalculated penetration depth value is zero.

4. The automatic adjustment method for a dummy mechanism based on finite element preprocessing software according to claim 1, characterized in that, The steps of obtaining the full-field physical field distribution data of the target dummy model based on preset simulation conditions and calculating the stress gradient data of the mesh elements include: Extract the geometric parameters of the target dummy model and the load boundary parameters of the preset simulation condition; The geometric parameters and the load boundary parameters are input into a pre-trained neural network model, and the strain energy density matrix of the target dummy model under the preset simulation conditions is output. The stress gradient values ​​of the entire field grid cells are calculated based on the strain energy density matrix to generate the entire field physical field distribution data.

5. The automatic adjustment method for a dummy mechanism based on finite element preprocessing software according to claim 1, characterized in that, The steps of acquiring target posture data and weight parameters include: Establish a real-time data communication link with external optical motion capture systems and force sensors; Receives the measured spatial coordinate sequence and force distribution data of the human body collected by the external optical motion capture system and the force sensor; The measured spatial coordinate sequence of the human body is mapped to the corresponding control node of the target dummy model through a coordinate transformation matrix to generate the target posture data; The counterweight parameters are generated based on the force distribution data. When adjusting the posture of the target dummy model and the counterweight distribution of each part, the counterweight parameters are allocated and bound to the corresponding counterweight grid nodes inside the target dummy model in the form of additional mass points.

6. An automatic adjustment system for a dummy mechanism based on finite element preprocessing software, wherein the system is integrated into the finite element preprocessing software as a plug-in, characterized in that, include: The template loading module is used to load the parameterized template corresponding to the target dummy model from the parameterized template library in response to the model loading command. The parameterized template includes the dummy joint angle range, mass distribution parameters, contact pair definition and joint linkage rules. The posture and weight adjustment module is used to acquire target posture data and weight parameters, calculate the linkage angle and displacement of each joint of the target dummy model based on the target posture data and the joint linkage rules, and adjust the posture of the target dummy model and the weight distribution of each part according to the linkage angle and displacement of each joint. The interference processing module is used to perform multi-step iterative mesh interference checks on the target dummy model after the posture is adjusted, extract the set of interference mesh nodes that have spatial overlap, calculate the penetration depth of the set of interference mesh nodes, and perform a mesh Morph deformation operation to drive the set of interference mesh nodes to move the distance value corresponding to the penetration depth in the opposite direction to the penetration direction. The mesh optimization module is used to obtain the full-field physical field distribution data of the target dummy model based on the preset simulation conditions, calculate the stress gradient data of the mesh cells, extract the mesh regions in the stress gradient data whose values ​​are greater than a preset threshold, and perform local mesh size reduction and node addition densification processing on the mesh cells in the mesh regions. The model output module is used to output the processed target dummy model.

7. The automatic adjustment system for a dummy mechanism based on finite element preprocessing software according to claim 6, characterized in that, The template loading module is configured with a custom configuration unit; The custom configuration unit is used to: read a custom configuration file in Extensible Markup Language format; parse the custom configuration file, extract the condition judgment logic of the dummy joint flexion angle, and configure the condition judgment logic into the joint linkage rules; The posture and weight adjustment module is specifically used to: obtain the flexion angle data of the first adjustment joint; when the flexion angle data of the first adjustment joint is greater than the angle threshold set by the condition judgment logic, extract the coordinate compensation amount of the bound second adjustment joint contact surface based on the flexion angle data of the first adjustment joint, and modify the spatial relative position of the second adjustment joint contact surface according to the coordinate compensation amount.

8. The automatic adjustment system for a dummy mechanism based on finite element preprocessing software according to claim 6, characterized in that, The interference processing module includes a discrete iteration unit and a morphology repair unit; The discrete iteration unit is used to discretize the total attitude adjustment displacement into multiple adjustment steps for iteration; The morphology repair unit is used to extract the set of interference mesh nodes on the main contact surface and the secondary contact surface where spatial overlap occurs within each adjustment step, and calculate the penetration depth value of the set of interference mesh nodes in the normal direction; construct a spatial interpolation function using the penetration depth value as the spatial displacement boundary condition, drive the set of interference mesh nodes and surrounding mesh nodes to perform position offset based on the spatial interpolation function, and repeatedly trigger the multi-step iterative mesh interference check until the recalculated penetration depth value is zero.

9. The automatic adjustment system for a dummy mechanism based on finite element preprocessing software according to claim 6, characterized in that, The grid optimization module has an embedded neural network prediction model that has been trained in advance using historical simulation data; The mesh optimization module is specifically used for: extracting the geometric parameters of the target dummy model and the load boundary parameters of the preset simulation condition; inputting the geometric parameters and the load boundary parameters into the pre-trained neural network prediction model, and outputting the strain energy density matrix of the target dummy model under the preset simulation condition; calculating the stress gradient values ​​of the full-field mesh elements based on the strain energy density matrix, and generating the full-field physical field distribution data.

10. The automatic adjustment system for a dummy mechanism based on finite element preprocessing software according to claim 6, characterized in that, It also includes a data fusion interface module; The data fusion interface module is used to: establish a real-time data communication link with an external optical motion capture system and a force sensor; receive the measured spatial coordinate sequence and force distribution data of the human body collected by the external optical motion capture system and the force sensor; map the measured spatial coordinate sequence of the human body to the corresponding control node of the target dummy model through a coordinate transformation matrix to generate the target posture data; generate the counterweight parameters according to the force distribution data; and when adjusting the posture of the target dummy model and the counterweight distribution of each part, allocate and bind the counterweight parameters to the corresponding counterweight grid node inside the target dummy model in the form of additional mass points.

Images