Human motion simulation and CFD simulation method and device under complex actions

Abstract

The invention discloses a human motion simulation and CFD simulation method and device under complex actions. The method comprises the following steps: constructing a multi-degree-of-freedom mechanical human body model containing simplified entity parts and preset-degree-of-freedom joints; based on the model, making a motion control strategy, decomposing a target action into multiple stages, applying a driving force to perform kinematics simulation, and exporting speed-time data of each part; based on the exported data, human motion is defined in CFD software and simulation is performed. According to the method, motion simulation and CFD are integrated, and optimized data processing (cubic polynomial fitting) and motion definition (smooth transition switching logic) are adopted, so that an effective scheme for economically, stably and accurately simulating complex human motion fluid dynamic behaviors is provided.

Description

Technical Field

[0001] The present application belongs to the technical field of human motion simulation, and in particular to a method and device for human motion simulation and CFD simulation under complex movements. Background Art

[0002] Computational fluid dynamics (CFD), a powerful numerical simulation tool, plays an increasingly important role in studying the interaction between the human body and its surrounding physical environment. For example, in sports engineering, analyzing airflow around athletes to optimize performance; in the built environment, evaluating the impact of indoor air flow and pollutant diffusion on the human body; simulating heat transfer on the human body surface in thermal comfort research; and studying respiratory airflow in the medical field, all require accurate simulation of fluid flow around the human body and related physical phenomena.

[0003] In many real-world scenarios, the human body is not static but rather in various motions, such as walking, running, squatting, and turning. This motion significantly alters the surrounding flow field structure, temperature distribution, and pollutant diffusion paths. Therefore, to obtain realistic simulation results, CFD simulations must accurately reproduce the dynamics of the human body.

[0004] Currently, the mainstream technologies for handling moving boundaries in CFD simulations include dynamic mesh and overlapping / overset mesh. These techniques allow the computational mesh to deform, increase, decrease, or slide relative to the object's boundaries as they move, thereby simulating the flow field around the moving object. However, the effective application of these technologies requires the precise definition of the motion patterns of the object's boundaries (i.e., the surfaces of various parts of the human body).

[0005] In the existing technology, obtaining accurate human motion data usually relies on technologies such as motion capture and three-dimensional scanning. These technologies are usually based on complex sensor systems or machine vision technologies. Although these methods can provide motion data to a certain extent, they generally have problems such as complex processes, high costs, and difficulty in data processing. In the field of fluid mechanics calculations, the movement of objects is mainly achieved by technologies such as dynamic grids and overlapping grids. However, complex human motion is difficult to define directly in fluid computing software, and the above-mentioned technologies for accurate simulation of human motion cannot directly provide the parameters of dynamic grids and overlapping grids. Therefore, it is impossible to implement CFD simulation calculations of complex human motion in computational fluid software.

[0006] In addition, even if the user manages to obtain motion data and input it into the CFD software, complex or non-smooth boundary motion can easily lead to a sharp decline in the quality of the dynamic mesh (such as mesh distortion, stretching, or even negative volume), which in turn affects the stability and accuracy of the calculation and may even cause the calculation to diverge.

[0007] In summary, existing technologies for CFD simulation of complex human motion face the following challenges: acquiring and processing human motion data suitable for CFD is costly and complex; directly defining complex human motion in CFD is impractical; and there remain challenges in smoothly integrating motion data with CFD dynamic mesh technology and ensuring simulation stability. Therefore, a more efficient, cost-effective, and easy-to-implement method is urgently needed to easily generate human motion descriptions suitable for CFD simulation and stably reproduce complex human motion in CFD software for accurate fluid dynamics analysis. Summary of the Invention

[0008] This application proposes a human motion simulation and CFD simulation method under complex movements, which can efficiently, accurately and economically generate human motion descriptions suitable for CFD simulation to perform accurate fluid dynamics analysis.

[0009] According to one embodiment of the present application, a method for simulating human motion and CFD simulation under complex movements is proposed, the method comprising:

[0010] Constructing a mechanical human body structure model with multiple degrees of freedom, the mechanical human body structure model including simplified representations of human body part solid models and human joint models with preset degrees of freedom connected to the solid models;

[0011] Constructing a motion control strategy for simulating a target motion based on the mechanical human body structure model, wherein the motion control strategy is used to decompose the target motion into multiple motion stages and select a rotational driving force or a linear driving force to apply to one or more human body parts to control their motion;

[0012] Performing kinematic simulation based on the motion control strategy to simulate the motion of various parts of the human body at various motion stages, and deriving parameters representing the motion state of each part, including rotational velocity and linear velocity that vary over time, including speed-time data of the translational velocity of each part in each direction and the rotational velocity around each axis in a preset coordinate system;

[0013] Based on the derived rotational speed and linear speed parameters, the motion of each part of the human body is defined in computational fluid dynamics (CFD) software, and CFD simulation calculations are performed.

[0014] In some embodiments, when constructing a mechanical human body structure model with multiple degrees of freedom, a human joint model is selected based on the simulated motion requirements and the degree of freedom characteristics of the real human body joints, and the connection relationship between the human joint model and the corresponding human body part entity model is constrained by setting a coincident fit or a concentric fit mode.

[0015] In some embodiments, when the target motion is walking, the motion control strategy simulates the motion state of the lower body of the human body and keeps the arms, torso, and head consistent or relatively still, simulating the motion state of the lower body of the human body.

[0016] In some embodiments, based on the derived rotational speed and linear speed parameters, the motion of various parts of the human body is defined in computational fluid dynamics (CFD) software, and CFD simulation calculations are performed, including:

[0017] For each motion phase included in the target action, the derived velocity-time data is fitted to obtain a continuous velocity function that describes the translational and rotational trajectories of each part and is defined in stages;

[0018] The obtained continuous velocity functions for each part in different stages and the switching logic for determining the corresponding motion stage and calling the corresponding velocity function according to the current simulation time are compiled into a motion program that can be called by CFD software;

[0019] In the dynamic mesh function of the CFD software, the motion program is assigned to the wall surface or the computational domain corresponding to the human body part to define the motion of the wall surface or the computational domain corresponding to the human body part;

[0020] Under given flow field boundary conditions, CFD simulation calculations are performed to obtain the fluid dynamics results around the moving human body.

[0021] In some embodiments, fitting the derived velocity-time data comprises:

[0022] A cubic polynomial function is selected to fit the velocity-time data. The cubic polynomial is selected based on its ability to describe the nonlinear characteristics of human motion and to balance the computational efficiency of frequent calls to motion functions by the dynamic mesh in subsequent CFD simulations.

[0023] In some implementations, the switching logic further includes:

[0024] In a preset transition time interval near a boundary time point between two consecutive motion phases, a smooth transition process is implemented to ensure that the speed curve output by the motion program maintains continuity in the transition time interval.

[0025] In some embodiments, the smooth transition process uses weighted averaging to achieve smooth blending of speed functions, including:

[0026] In the transition time interval, based on a weight function that changes smoothly with time, the speed value calculated by the speed function of the previous stage is mixed with the speed value calculated by the speed function of the next stage.

[0027] In some embodiments, before assigning the motion program to the wall or computational domain corresponding to the human body part in the dynamic mesh function of the CFD software, the method further includes:

[0028] In the pre-processing stage of the CFD software, the imported overall human body geometry model is cut or grouped according to the part division of the mechanical human body structure model, so that each human body part in the CFD model can independently load the corresponding motion program.

[0029] In some embodiments, in the dynamic mesh function of the CFD software, when assigning the motion program to the wall or calculation domain of the corresponding human body part, it also includes specifying the center point coordinates for each human body part to represent the translational trajectory of the corresponding human body part.

[0030] According to one embodiment of the present application, a human motion simulation and CFD simulation device for complex movements is provided, the device comprising:

[0031] A first construction unit is configured to construct a mechanical human body structure model having multiple degrees of freedom, wherein the mechanical human body structure model includes simplified representations of human body part solid models and human joint models having preset degrees of freedom connected to the solid models;

[0032] a second construction unit, configured to construct a motion control strategy for simulating a target action based on the mechanical human body structure model, wherein the motion control strategy is configured to decompose the target action into multiple motion stages and select a rotational driving force or a linear driving force to apply to one or more human body parts to control their motion;

[0033] a first simulation unit, configured to perform kinematic simulation based on the motion control strategy, simulate the motion of various parts of the human body at various motion stages, and derive parameters representing the motion state of each part, namely, rotational velocity and linear velocity that vary with time, the parameters including velocity-time data of the translational velocity of each part in each direction and the rotational velocity around each axis in a preset coordinate system;

[0034] The second simulation is used to define the motion of various parts of the human body in computational fluid dynamics (CFD) software based on the derived rotational speed and linear speed parameters, and perform CFD simulation calculations.

[0035] Compared with the prior art, the present application provides a human motion simulation and CFD simulation method under complex actions, which has significant beneficial effects. First, the present application greatly reduces the difficulty and cost of obtaining human motion data suitable for CFD simulation by constructing a simplified multi-degree-of-freedom mechanical human model and utilizing the drive and constraint mechanism in the kinematic simulation software to simulate complex actions. Users do not need to rely on expensive motion capture equipment or complex biomechanical models. Only by reasonably decomposing the target action and setting simple drive parameters, a realistic and demand-compliant motion process can be generated, which significantly improves the accessibility and efficiency of human motion CFD simulation. Moreover, the present application proposes fitting the discrete velocity data obtained by kinematic simulation into a continuous velocity function (such as a cubic polynomial function with higher computational efficiency), and compiling it into a motion program (such as UDF) that can be called by CFD software, which effectively solves the interface problem between the kinematic simulation results and the CFD dynamic mesh technology, and realizes the seamless integration of the two.

[0036] In addition, the method provided by the present application further improves the stability and accuracy of CFD simulation of complex human motion. By incorporating time-based phase switching logic into the motion program and introducing a smooth transition processing mechanism (e.g., using weighted average mixing), it is possible to effectively eliminate or weaken the speed or acceleration mutations that may result from switching motion phases, thereby ensuring that the motion instructions given to the moving mesh are continuous and smooth. This avoids the problems of severe deformation and negative volume of the mesh caused thereby, and significantly improves the stability and robustness of the CFD calculation process. At the same time, by cutting and grouping the human body model corresponding to the motion model in the CFD pre-processing, and accurately specifying the motion program and center point coordinates of each part, it is ensured that the complex motion generated by the simulation can be accurately reproduced in the CFD environment.

[0037] In summary, this application provides a complete, efficient and reliable technical process that can realize refined CFD analysis of the interaction between the human body and the surrounding fluid environment under complex motion states, and has important application value.

[0038] Further details and advantages of this application are detailed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the specification and, together with the description, serve to explain the principles of the specification.

[0040] Figure 1 A flow chart of a human motion simulation and CFD simulation method under complex actions provided in accordance with an embodiment of the present application is shown.

[0041] Figure 2An overall schematic diagram of a mechanical human body structure model according to an exemplary embodiment of the present application is shown.

[0042] Figure 3 A structural schematic diagram of a joint connector (joint model) according to an exemplary embodiment of the present application is shown.

[0043] Figure 4 A schematic diagram showing a process of human squatting movement simulated according to an exemplary embodiment of the present application is shown.

[0044] Figure 5 A schematic diagram showing a human walking motion process simulated according to an exemplary embodiment of the present application is shown.

[0045] Figure 6 A schematic diagram of the skin surface flow velocity distribution results obtained by CFD simulation calculation of the walking motion of a moving human body according to an exemplary embodiment of the present application is shown.

[0046] Figure 7 A schematic diagram of the skin surface temperature distribution results of a CFD simulation of a moving human walking motion obtained according to an exemplary embodiment of the present application is shown. DETAILED DESCRIPTION

[0047] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

[0048] Figure 1 The flowchart of the human motion simulation and CFD simulation method under complex motion provided by an embodiment of the present application is shown. The method is used to simulate the interaction between the human body and the surrounding fluid environment under complex motion state. Figure 1 As shown, the method includes steps 1 to 4.

[0049] Step 1: construct a mechanical human body structure model with multiple degrees of freedom. The mechanical human body structure model includes a simplified representation of a human body part entity model and a human body joint model with preset degrees of freedom connected to the entity model.

[0050] Step 1 creates a digital mechanical model of the human body. Rather than pursuing detailed anatomical structures, this model focuses on kinematic expression, simplifying the main body parts (such as the head, torso, upper arms, lower arms, hands, thighs, lower legs, and feet) into geometric shapes (for example, a spherical segment for the head, a frustum for the limbs, and polyhedrons for the torso, hands, and feet). This facilitates subsequent motion simulation and computational fluid dynamics (CFD) mesh processing.

[0051] At the same time, based on the actual human motion capabilities, joint models with specific degrees of freedom (for example, one rotational degree of freedom for the elbow and knee, and two or three rotational degrees of freedom for the shoulder and hip) are defined between these solid models. These joints are connected to the corresponding solid models through standard mechanical coordination methods (such as coincidence and concentric constraints), forming a multi-degree-of-freedom system that can simulate the basic posture and range of motion of the human body. This creates a simplified basic mechanical model that not only reflects the main motion characteristics of the human body but is also suitable for efficient kinematic simulation.

[0052] In some embodiments, when constructing a mechanical human body structure model with multiple degrees of freedom, a human joint model is selected based on the simulated motion requirements and the degree of freedom characteristics of the real human body joints, and the connection relationship between the human joint model and the corresponding human body part entity model is constrained by setting a coincident fit or a concentric fit mode.

[0053] According to this embodiment, when constructing the multi-degree-of-freedom mechanical human body structure model, the human joint model is selected based on the specific needs of simulating the target action and the physiological degrees of freedom of the human body's real joints, thereby ensuring that the model meets both the simulation needs and the principles of biomechanics in terms of motion function, greatly improving the accuracy and fidelity of subsequent kinematic simulation results, and providing more reliable motion boundary conditions for CFD analysis. Moreover, the connection between the joint model and the corresponding body part physical model is achieved through a set mechanical matching method (such as overlapping matching or concentric matching), ensuring the structural stability of the constructed model and the certainty of the motion path, avoiding the model errors or instability problems that may occur during the simulation process, thereby improving the reliability of the entire simulation process.

[0054] Step 2: Based on the mechanical human body structure model, a motion control strategy for simulating the target action is constructed. The motion control strategy is used to decompose the target action into multiple motion stages, and select a rotational driving force or a linear driving force to be applied to one or more parts of the human body to control its movement.

[0055] Through step 2, it is possible to plan how to drive the model to move in order to reproduce the target action. For complex human body movements (such as walking, squatting, etc.), according to this embodiment, they can usually be decomposed into a series of relatively simple, continuous motion stages. Then, for each stage, the key body parts that need to be actively driven (such as thighs, calves, etc.) are determined, and a virtual driving force or torque is applied to the corresponding joints or parts (as if a motor is installed), and its mode of action (rotational or linear) and control law (such as constant speed, specific acceleration, etc.) are set to form a complete motion control strategy to guide how subsequent kinematic simulations are performed.

[0056] In some embodiments, when the target motion is walking, the motion control strategy simulates the motion state of the lower body of the human body and keeps the arms, torso, and head consistent or relatively still, simulating the motion state of the lower body of the human body. This simplifies the motion control strategy when the target motion to be simulated is walking. The simplified strategy focuses on simulating the motion state of the lower body of the human body (legs and feet), while processing the motion of the arms, torso, and head to remain consistent (for example, translation and limited rotation as a whole with the movement of the lower body) or relatively still, thereby simulating the main motion characteristics of the lower body of the human body during walking.

[0057] According to the simplified strategy for walking movements of this embodiment, the complexity of the motion control strategy and the computational burden of subsequent kinematic simulations are significantly reduced. By focusing the simulation on the lower body movements that have the greatest impact on walking posture, the simulation parameters can be set more quickly and the required motion data can be obtained. For many CFD studies that focus on the flow field around the human body during walking (especially the lower limb area or overall resistance), the precise independent swinging of upper limbs such as arms is usually a secondary factor. This simplified strategy effectively balances the authenticity and computational efficiency of the simulation while ensuring the capture of key motion characteristics, making the simulation analysis more targeted and practical, and significantly improving the efficiency of walking movement simulation.

[0058] Step 3: Perform kinematic simulation based on the motion control strategy to simulate the motion of various parts of the human body in various motion stages, and derive parameters of the rotational speed and linear speed that characterize the motion state of each part and change with time. The parameters include the speed-time data of the translational speed of each part in each direction in the preset coordinate system and the rotational speed around each axis.

[0059] After step 3, professional kinematics or multi-body dynamics simulation software can be used to simulate the constructed mechanical human body model according to the control strategy set in the previous step. The software can solve the precise motion trajectory of each body part in the model throughout time under the driving force and joint constraints. The key result of the simulation is the export of a series of detailed quantitative data, such as the translational velocity of each body part in the x, y, and z directions in the global coordinate system, as well as the rotational velocity around the x, y, and z axes. These velocity values are a sequence of discrete data points that change over time, converting abstract motion instructions into specific, quantified six-degree-of-freedom velocity information that can be used for subsequent CFD calculations. This velocity-time data can accurately record the simulated motion state of each body part, serving as the raw data foundation for connecting motion simulation and CFD simulation.

[0060] Step 4: Based on the derived rotational speed and linear speed parameters, define the motion of various parts of the human body in the CFD software and perform CFD simulation calculations.

[0061] Through step 4, the kinematic results generated in the previous steps are applied to fluid simulation, including processing the exported discrete velocity-time data, and then writing it into a motion program (such as UDF) that the CFD software can understand and execute. Then, in the CFD software (usually using the dynamic mesh function), these motion programs are loaded onto the surfaces of the corresponding human body model parts, thereby accurately defining the movement of these surfaces during the simulation process. When the motion of all relevant parts has been defined and the boundary conditions of the flow field have been set, the CFD solver can be started to perform fluid dynamics calculations, and finally obtain detailed information about the flow field around the human body in a complex motion state (such as air flow velocity, pressure, temperature distribution, etc.). This step completes the transformation from motion description to final flow field analysis.

[0062] In some embodiments, step 4 may specifically include:

[0063] For each motion phase included in the target action, the derived velocity-time data is fitted to obtain a continuous velocity function that describes the translational and rotational trajectories of each part and is defined in stages;

[0064] The obtained continuous velocity functions for each part in different stages and the switching logic for determining the corresponding motion stage and calling the corresponding velocity function according to the current simulation time are compiled into a motion program that can be called by CFD software;

[0065] In the dynamic mesh function of the CFD software, the motion program is assigned to the wall surface or the computational domain corresponding to the human body part to define the motion of the wall surface or the computational domain corresponding to the human body part;

[0066] Under given flow field boundary conditions, CFD simulation calculations are performed to obtain the fluid dynamics results around the moving human body.

[0067] According to the above embodiment, for each motion stage (for example, the early and late stages of the left leg swing, the early and late stages of the right leg swing, etc. in the walking action), a corresponding velocity function can be fitted (for example, a function that describes the change in the translational or rotational velocity of a certain part over time during the stage), so that the most suitable mathematical description for the different parts of the complex action is adopted, which significantly improves the flexibility and accuracy of the fitting. The switching logic is used to automatically determine which motion stage should be in according to the current simulation time, and call the velocity function corresponding to the stage. A series of velocity functions and corresponding switching logic can be written into a motion program (for example, a user-defined function UDF) that can be called and executed by CFD software (such as Fluent).

[0068] Then, within the standard dynamic mesh module of the CFD software, the programmed motion program is loaded onto the wall or computational domain corresponding to the human body part, driving these motion units to move according to the program's defined patterns. Finally, after setting the necessary flow field boundary conditions (such as inlet velocity and outlet pressure), the CFD solution is executed, ultimately obtaining detailed fluid dynamics simulation results around the human body in motion.

[0069] This embodiment adopts a staged definition of velocity functions, making the description of complex, multi-mode motion more accurate and flexible. By combining switching logic to write motion programs, complex, staged discrete motion data is converted into standardized instructions that are easy to process and call by CFD software, effectively solving the problem of defining complex motion directly in CFD. And using mature dynamic mesh technology, the motion program is assigned to the motion units of the corresponding human body parts, ensuring that the simulated motion can be accurately reproduced in the CFD environment. On this basis, the final CFD calculation can accurately capture the impact of human motion on the surrounding flow field, thereby realizing CFD simulation of complex human motion, which is highly operational and reliable.

[0070] In some embodiments, a cubic polynomial function is selected to fit the velocity-time data. The cubic polynomial is selected because it can describe the nonlinear characteristics of human motion and balance the computational efficiency of frequent calls to motion functions by the dynamic mesh in subsequent CFD simulations.

[0071] The speed change of human body parts during movement is often not a simple linear relationship, but rather presents certain nonlinear characteristics. For example, the acceleration and deceleration process may be similar to a parabola or an S-shaped curve. The inventors have conducted extensive research and believe that a cubic polynomial function (e.g., a function of the form at 3 +bt 2+ct+d) can capture and describe this common low-order nonlinear variation trend more flexibly and more accurately than lower-order polynomials (such as linear or quadratic), thereby making the speed curve fitted closer to real motion situation. On the other hand, in CFD simulation, moving grid technology needs to frequently call motion function to update boundary position at each time step (or iteration step). If the fitting function selected is too complicated (such as higher-order polynomial or complex transcendental function), although precision may be higher in data fitting, the computational load of motion program (UDF) may be sharply increased, significantly slowing down the speed of whole CFD simulation, even making large-scale, long-time simulation become infeasible. Therefore, the inventor, after in-depth research and comprehensive evaluation, preferably selects cubic polynomial to be fitted to obtain continuous velocity function. Cubic polynomial form is relatively concise, and its evaluation calculation (only involving several multiplications and additions) is very fast, can when guaranteeing basic nonlinear feature description ability, greatly reduce the computational cost when moving grid calls, significantly improve the overall computational efficiency of CFD simulation.

[0072] In some implementations, the switching logic further includes:

[0073] In a preset transition time interval near a boundary time point between two consecutive motion phases, a smooth transition process is implemented to ensure that the speed curve output by the motion program maintains continuity in the transition time interval.

[0074] According to this embodiment, when the simulation time falls within the transition time interval between two motion stages, the motion program no longer simply switches from the previous stage function to the next stage function, but instead performs a smooth transition process, thereby further improving the stability of the CFD simulation and the quality of the mesh. Because sudden changes in speed can have an impact on the calculation of the dynamic mesh, it is easy to cause the mesh to deform and twist violently in a short period of time, or even to have a negative volume, resulting in calculation failure. By implementing a smooth transition, the speed change is made more moderate and natural, significantly reducing the difficulty and risk of dynamic mesh calculations, and improving the robustness and success rate of the entire CFD simulation.

[0075] In some embodiments, the smooth transition process uses weighted averaging to achieve smooth blending of speed functions, including:

[0076] In the transition time interval, based on a weight function that changes smoothly with time, the speed value calculated by the speed function of the previous stage is mixed with the speed value calculated by the speed function of the next stage.

[0077] According to this embodiment, when the simulation time is within a preset transition time interval, the motion program will simultaneously calculate the value of the speed function of the previous stage at that time point and the value of the speed function of the next stage at that time point, and then, based on a weight function that changes smoothly over time, mix these two speed values to obtain the final output speed value. For example, the weight function can be a smooth normalized function, such as a normalized function based on a cosine function or a polynomial function. At the beginning of the transition interval, the output speed can be mainly determined by the speed function of the previous stage; as time goes by, the weight gradually changes, and the influence of the speed function of the next stage gradually increases; by the end of the transition interval, the output speed is completely determined by the speed function of the next stage, thereby naturally achieving a smooth transition of the speed curve between the two motion stages through this progressive mixing process.

[0078] In some embodiments, before assigning the motion program to the wall or computational domain corresponding to the human body part in the dynamic mesh function of the CFD software, the method further includes:

[0079] In the pre-processing stage of the CFD software, the imported overall human body geometry model is cut or grouped according to the part division of the mechanical human body structure model, so that each human body part in the CFD model can independently load the corresponding motion program.

[0080] According to this embodiment, in the pre-processing stage of the CFD software, the part division of the previously constructed mechanical human body structure model (i.e., which parts are defined as independent motion units, such as thigh, calf, torso, etc.) can be referred to, and the overall human body geometric model imported into the CFD software can be cut or grouped accordingly to ensure that each part in the CFD model (e.g., the geometric body representing the thigh, the geometric body representing the calf) can be independently identified and selected, so that the corresponding motion programs can be loaded independently. That is, when the dynamic mesh is subsequently set, the program describing the thigh motion can be assigned only to the thigh model, the program describing the calf motion can be assigned only to the calf model, and so on, thereby realizing CFD simulation of complex coordinated motion of multiple parts.

[0081] In some embodiments, in the dynamic mesh function of the CFD software, when assigning the motion program to the wall or calculation domain of the corresponding human body part, it also includes specifying the center point coordinates for each human body part to represent the translational trajectory of the corresponding human body part.

[0082] The coordinates of the center point of the translational trajectory can be recorded and exported in step 2 above. According to this embodiment, these coordinates are entered into the CFD dynamic mesh settings, so that the dynamic mesh module knows the point around which to apply the translational velocity component of the corresponding part. This ensures that the translational trajectory simulated in CFD is completely consistent with the trajectory obtained by kinematic simulation, avoiding motion deviations that may be caused by inconsistent reference points.

[0083] Figure 1 The illustrated embodiment provides an efficient, economical, and reliable method for simulating human motion and CFD simulation under complex movements. By using simplified mechanical modeling and kinematic simulation to generate motion data, and selecting a cubic polynomial that balances nonlinear description with computational efficiency for fitting, the cost and difficulty of data acquisition are significantly reduced, and CFD computational efficiency is optimized. In particular, the staged motion function and switching logic including smooth transition processing are written as a motion program and integrated into the CFD dynamic mesh, effectively resolving the interface issues between motion data and CFD simulation, and greatly improving the stability and robustness of the simulation process, ultimately achieving accurate and efficient simulation of the fluid dynamics behavior of complex human motion.

[0084] In order to make the purpose, technical solutions and advantages of this application more clear, the following Figures 2 to 7 The human motion simulation and CFD simulation method under complex actions proposed in this application are further described in detail with reference to specific application examples. It should be understood that the specific application examples described here are only used to explain this application and are not used to limit this application.

[0085] The core process of this application example can be roughly divided into three main stages: building a mechanical human body structure model, performing kinematic simulation and exporting motion data, and integrating the motion data into CFD software for simulation calculations.

[0086] Phase 1: Construct a mechanical human body structure model.

[0087] First, a multi-degree-of-freedom mechanical human structure model for kinematic simulation can be constructed, which can be referred to Figure 2 and Figure 3 .

[0088] like Figure 2 As shown, the model includes multiple simplified representations of human body parts, such as the head 1-1 (which can be simplified as a spherical segment), the torso 1-14, and the limbs (including the left upper arm 1-2, left forearm 1-3, left hand 1-12; the left thigh 1-4, left calf 1-5, left foot 1-6; the right upper arm 1-7, right forearm 1-8, right hand 1-13; the right thigh 1-9, right calf 1-10, and right foot 1-11). These limbs can be simplified as frustums or similar geometric solids of varying sizes, while the hands and feet can be simplified as polyhedrons. The dimensions of the simplified models of each part can be referenced to relevant standards, such as the dimensional parameters for specific populations (e.g., the P50 percentile of adult males aged 18-25) in GB-T 10000-2023, "Chinese Adult Human Body Dimensions."

[0089] The model also includes human joint models that connect these solid models. These joint models have preset degrees of freedom. Figure 3 As shown, the joint model may include, for example, a single degree of freedom (1-DOF) joint model 3-1 ( Figure 3 Middle left) and two-degree-of-freedom (2-DOF) joint model 3-2 ( Figure 3 The choice of joint model should be based on the simulated motion requirements and the degree of freedom characteristics of real human joints. In this application example, the specific joint settings can be:

[0090] The neck joint between the head 1-1 and the torso 1-14: adopts a 1-DOF joint model (e.g., allowing rotation around a specific axis);

[0091] The shoulder joint between the trunk 1-14 and the upper arm (1-2, 1-7) uses a 2-DOF joint model;

[0092] Elbow joint between upper arm (1-2, 1-7) and lower arm (1-3, 1-8): adopts 1-DOF joint model;

[0093] The wrist joint between the forearm (1-3, 1-8) and the palm (1-12, 1-13) uses a 1-DOF joint model;

[0094] Hip joint between torso 1-14 and thigh (1-4, 1-9): adopts 2-DOF joint model;

[0095] The knee joint between the thigh (1-4, 1-9) and the calf (1-5, 1-10) uses a 1-DOF joint model.

[0096] Ankle joint between the calf (1-5, 1-10) and the foot (1-6, 1-11): a 1-DOF joint model is used.

[0097] The connection relationship between the joint model and the corresponding human body part physical model is constrained by setting a precise matching method. The matching methods according to the embodiments of the present application include "coincidence matching" and "concentric matching". For example, a "coincidence" matching can be set between the bottom surface of the head model 1-1 and the connection surface of the neck joint model 3-4 to ensure that the two are in the same plane; at the same time, a "concentric" matching can be set (and "locked rotation" may also be required to limit rotation in non-physiological directions) to ensure that the head rotates correctly around the neck joint.

[0098] Similarly, similar coincidence and concentricity are used to establish stable connections with clear kinematic relationships between other joints and body parts. The trunk and shoulder joints, the shoulder joints and upper arms, the elbow joints and lower arms, the wrist joints and hands, the trunk and hip joints, the hip joints and thighs, the knee joints and lower legs, and the ankle joints and feet are connected using "coincidence" and "concentricity (locked rotation)" cooperation methods. The connections between the remaining part models and joint models are fully constrained using the outline of the 3-4 connection surface of the 1-DOF joint model drawn in the part model and the "coincidence" cooperation method of the 3-4 connection surface of the 1-DOF joint model.

[0099] With the above settings, this application example builds a simple mechanical human structure model with 17 degrees of freedom.

[0100] Phase 2: Motion control and simulation.

[0101] Based on the constructed mechanical human body structure model, a motion control strategy can be constructed to simulate the target motion. This strategy first involves decomposing the complex target motion into multiple simple motion stages. Then, special constraints are set for specific stages or parts as needed. Next, a rotational driving force or a linear driving force (which can be achieved by setting up a "virtual motor" that can be applied to different parts of the human body to control its motion posture) is applied to one or more parts of the human body, and the appropriate driving rate (rotational rate or translational rate) and action time are set. Finally, a kinematic simulation is performed to transmit motion through joint constraints, simulate the overall motion, and export the required motion data.

[0102] Example 1: Squatting simulation, such as Figure 4 shown.

[0103] Movement decomposition: The squat movement is broken down into the squatting phase and the rising phase. During the squatting phase, the calf rotates positively around the x-axis, while the thigh rotates negatively around the x-axis, both of which can be described by simple functional relationships.

[0104] Special constraints: Set the feet (1-6, 1-11) as fixed constraints to prevent them from moving during simulation; set the head 1-1 and torso 1-14 as "parallel" constraints to ensure that the head rotates around the X-axis with the torso.

[0105] Drive control: specifically including:

[0106] During the squat phase, rotary motors are set at the ankle joints to control the positive rotation of the calves (1-5, 1-10) around the x-axis; rotary motors are set at the knee joints to control the negative rotation of the thighs (1-4, 1-9) around the x-axis, and appropriate rotation rates and execution times are set.

[0107] During the standing-up phase, a rotary motor is set at the same position, the control position remains unchanged, and the rotation direction is changed. That is, the rotary motor at the ankle joint is controlled to rotate in the negative direction around the x-axis, and the rotary motor at the knee joint is controlled to rotate in the positive direction around the x-axis. The motor speed and execution time are the same as those set in the squatting phase to ensure that the human body returns to the initial posture after standing up.

[0108] Simulation and Adjustment: After completing the motor settings for each stage, perform full-process motion simulation and perform kinematic simulation. The thigh and calf movements will be transmitted to other parts through the constraint relationship to form an overall motion effect. If the simulation effect is not good, you can return to the previous step to adjust the motor speed or time. The final motion simulation result is as follows Figure 4 shown.

[0109] Example 2: Walking action simulation, such as Figure 5 shown.

[0110] Movement decomposition: The walking movement (one gait cycle) is decomposed into 4 main phases: left leg early swing (flexion), left leg late swing (extension), right leg early swing (flexion), right leg late swing (extension).

[0111] Special constraints (simplification strategy): To simplify the model and focus on the movement of the lower body, set the arms (upper arms, forearms, hands) and torso 1-14 to "parallel" constraints, so that the arms, torso and head 1-1 remain consistent or relatively stationary (only with translation and slight rotation with the whole body), mainly simulating the movement state of the lower body.

[0112] Drive control: At each stage, drive is applied to key parts. The active moving parts of the walking behavior are the thighs, calves, and torso. The thigh and calf movements simulate the posture of the legs at each stage of the human walking process, and the torso movement realizes the movement of the entire human model in the positive direction of the z-axis. Specific settings include:

[0113] Early swing (flexion) phase of the left leg:

[0114] A rotary motor is set at the left hip joint to control the left thigh to rotate in the opposite direction around the x-axis.

[0115] A rotary motor is set at the left knee joint to control the left calf to rotate around the positive direction of the x-axis.

[0116] A rotary motor is set at the ankle joint of the right leg to control the rotation of the right calf around the positive direction of the x-axis.

[0117] A rotary motor is set at the right knee joint to control the right thigh to rotate around the positive direction of the x-axis.

[0118] Set appropriate motor speeds and execution times.

[0119] Left leg late swing (extension) phase:

[0120] A rotary motor is set at the left knee joint to control the left thigh to rotate around the positive direction of the x-axis.

[0121] Set appropriate motor speeds and execution times.

[0122] Right leg early swing (flexion) phase:

[0123] A rotary motor is set at the right leg hip joint to control the right thigh to rotate in the opposite direction around the x-axis.

[0124] A rotary motor is set at the right knee joint to control the left calf to rotate around the positive direction of the x-axis.

[0125] A rotary motor is set at the left ankle joint to control the right calf to rotate around the positive direction of the x-axis.

[0126] A rotary motor is set at the left knee joint to control the right thigh to rotate around the positive direction of the x-axis.

[0127] Set appropriate motor speeds and execution times.

[0128] Right leg late swing (extension) phase:

[0129] A rotary motor is set at the right knee joint to control the right thigh to rotate around the positive direction of the x-axis.

[0130] Set appropriate motor speeds and execution times.

[0131] Trunk Movement:

[0132] A linear motor is set on the surface of the torso to control the movement of the torso along the positive direction of the z-axis, simulating the movement of the entire model in the positive direction of the z-axis.

[0133] Set the appropriate drive rate and execution time.

[0134] Simulation and Adjustment: After completing the motor settings for each stage, perform full-process motion simulation. The movements of the thigh, calf, and torso will be transmitted to other parts through the constraint relationship to form an overall motion effect. If the simulation effect is poor, return to the previous step to adjust the motor speed. The final motion simulation result is as follows Figure 5 shown.

[0135] After the simulation is completed, the parameters of the rotational speed and linear speed that characterize the motion state of each part and change with time can be derived. Specifically, the parameters of the rotational speed and linear speed of each part (for example Figure 2Velocity-time data series for the translational velocity in each direction (x, y, z) and the rotational velocity around each axis (x, y, z) for the head 1-1, torso 1-14, and limb segments 1-2 through 1-11 (labeled in the figure) in a preset coordinate system (e.g., the xyz global coordinate system). Because the body parts in this application example only translate along the y and z axes, only translational data in the y and z directions needs to be derived. The coordinates of the center point representing the translational trajectory of each body part are also recorded.

[0136] Phase 3: CFD Integration and Calculation: In this phase, the kinematic simulation results are applied to CFD calculations.

[0137] Fitting velocity function: Fit the derived velocity-time data for each part and each motion stage to obtain a continuous velocity function defined in stages. This application example uses the following cubic polynomial function for fitting:

[0138] V=a1t 3 +b1t 2 +c1t+d1,

[0139] Where V is the translational velocity or rotational velocity, t is the time, and a1, b1, c1, and d1 are fitting coefficients.

[0140] For the walking example, due to the simplification of the upper body, it is actually only necessary to fit the movements of seven independent moving parts (torso / head / arm as a whole + left and right thighs + left and right calves + left and right feet) in four stages.

[0141] Write a motion program: Write the continuous velocity function in the form of a cubic polynomial for each part in stages and the switching logic for determining the corresponding motion stage according to the current simulation time and calling the corresponding velocity function into a motion program (such as a user-defined function UDF) that can be called by CFD software (such as Fluent). Different parts in fluid mechanics calculations are controlled by separate motion function programs, so the translational equation and rotational equation for the same part are written in the same function program. The switching logic of this application example also includes a smooth transition processing mechanism, that is, in the preset transition time interval near the boundary time point between two consecutive motion stages, a weighted average is used to achieve smooth mixing of the velocity function. For example, in the transition interval, the velocity values of the previous stage and the next stage are mixed based on a weight function that changes smoothly over time to ensure the continuity of the output velocity curve and improve the stability of the CFD calculation.

[0142] CFD pre-processing: In the pre-processing stage of CFD software, the imported overall human body geometry model is cut or grouped according to the part division of the mechanical human body structure model, so that each human body part in the CFD model (such as thigh and calf) can independently load the corresponding motion program.

[0143] Compilation and Dynamic Mesh Setup: Compile the pre-written motion program (e.g., a UDF) in the CFD software environment. Then, in the Dynamic Mesh function module, select the wall mesh and computational domain mesh corresponding to the body part to load the compiled motion program. When loading the motion program, specify the center point coordinates for each body part to represent its translational trajectory. For example, select the wall of the thigh model, specify the call to the UDF describing the thigh's motion, and enter the center point coordinates of the thigh. Repeat this process for all parts that need to be moved.

[0144] Perform a CFD simulation: Set the necessary flow field boundary conditions (such as incoming flow velocity, ambient temperature, and pressure) and initial conditions, then start the CFD solver. The dynamic mesh module updates the position and posture of the human body model in real time based on the loaded motion program, and the CFD solver calculates the resulting flow field changes.

[0145] Results acquisition: After the simulation is completed, the fluid dynamics results around the moving human body can be obtained, such as Figure 6 The velocity distribution cloud of the human body surface during walking is shown, or Figure 7 The temperature distribution cloud diagram of the human body surface during walking is shown.

[0146] Through the detailed steps described above, this application example implements the complete process from building a simplified human body model, simulating complex motion, processing motion data, and ultimately accurately reproducing motion in CFD and performing fluid simulation calculations.

[0147] Although this specification includes many specific implementation details, these should not be interpreted as limiting the scope of any invention or the scope of protection claimed, but are mainly used to describe the features of specific embodiments of specific inventions. Certain features described in multiple embodiments within this specification may also be implemented in combination in a single embodiment. On the other hand, the various features described in a single embodiment may also be implemented separately in multiple embodiments or in any suitable sub-combination. In addition, although features may work in certain combinations as described above and even initially claimed as such, one or more features from the claimed combination may be removed from the combination in some cases, and the claimed combination may point to a sub-combination or a variation of the sub-combination.

[0148] The above description is merely a preferred embodiment of one or more embodiments of this specification and is not intended to limit one or more embodiments of this specification. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of one or more embodiments of this specification shall be included in the scope of protection of one or more embodiments of this specification.

Claims

1. A human motion simulation and CFD simulation method under complex movements, characterized by: The method comprises: Constructing a mechanical human body structure model with multiple degrees of freedom, the mechanical human body structure model including simplified representations of human body part solid models and human joint models with preset degrees of freedom connected to the solid models; Constructing a motion control strategy for simulating a target motion based on the mechanical human body structure model, wherein the motion control strategy is used to decompose the target motion into multiple motion stages and select a rotational driving force or a linear driving force to apply to one or more human body parts to control their motion; Performing kinematic simulation based on the motion control strategy to simulate the motion of various parts of the human body at various motion stages, and deriving parameters representing the motion state of each part, including rotational velocity and linear velocity that vary over time, including speed-time data of the translational velocity of each part in each direction and the rotational velocity around each axis in a preset coordinate system; Based on the derived rotational speed and linear speed parameters, the motion of each part of the human body is defined in computational fluid dynamics (CFD) software, and CFD simulation calculations are performed.

2. The method according to claim 1, characterized in that In constructing a mechanical human body structure model with multiple degrees of freedom, a human joint model is selected according to the simulated motion requirements and the degree of freedom characteristics of the real human joints, and the connection relationship between the human joint model and the corresponding human body part entity model is constrained by setting a coincident fit or a concentric fit mode.

3. The method according to claim 1, characterized in that When the target action is walking, the motion control strategy simulates the motion state of the lower body of the human body and keeps the arms, torso and head consistent or relatively still, simulating the motion state of the lower body of the human body.

4. The method according to claim 1, wherein Based on the derived rotational and linear velocity parameters, the motion of each part of the human body is defined in computational fluid dynamics (CFD) software, and CFD simulation calculations are performed, including: For each motion phase included in the target action, the derived velocity-time data is fitted to obtain a continuous velocity function that describes the translational and rotational trajectories of each part and is defined in stages; The obtained continuous velocity functions for each part in different stages and the switching logic for determining the corresponding motion stage and calling the corresponding velocity function according to the current simulation time are compiled into a motion program that can be called by CFD software; In the dynamic mesh function of the CFD software, the motion program is assigned to the wall surface or the computational domain corresponding to the human body part to define the motion of the wall surface or the computational domain corresponding to the human body part; Under given flow field boundary conditions, CFD simulation calculations are performed to obtain the fluid dynamics results around the moving human body.

5. The method according to claim 4, characterized in that The velocity-time data derived from the fit include: A cubic polynomial function is selected to fit the velocity-time data. The cubic polynomial is selected based on its ability to describe the nonlinear characteristics of human motion and to balance the computational efficiency of frequent calls to motion functions by the dynamic mesh in subsequent CFD simulations.

6. The method according to claim 4, characterized in that The switching logic further includes: In a preset transition time interval near a boundary time point between two consecutive motion phases, a smooth transition process is implemented to ensure that the speed curve output by the motion program maintains continuity in the transition time interval.

7. The method according to claim 6, characterized in that The smooth transition process uses weighted averaging to achieve smooth blending of speed functions, including: In the transition time interval, based on a weight function that changes smoothly with time, the speed value calculated by the speed function of the previous stage is mixed with the speed value calculated by the speed function of the next stage.

8. The method according to claim 4, characterized in that Before assigning the motion program to the wall or computational domain of the corresponding human body part in the dynamic mesh function of the CFD software, the method further includes: In the pre-processing stage of the CFD software, the imported overall human body geometry model is cut or grouped according to the part division of the mechanical human body structure model, so that each human body part in the CFD model can independently load the corresponding motion program.

9. The method according to claim 4, characterized in that In the dynamic mesh function of the CFD software, when assigning the motion program to the wall or calculation domain of the corresponding human body part, it also includes specifying the center point coordinates for each human body part to represent the translation trajectory of the corresponding human body part.

10. A human motion simulation and CFD simulation device under complex movements, characterized in that: The device comprises: A first construction unit is configured to construct a mechanical human body structure model having multiple degrees of freedom, wherein the mechanical human body structure model includes simplified representations of human body part solid models and human joint models having preset degrees of freedom connected to the solid models; a second construction unit, configured to construct a motion control strategy for simulating a target action based on the mechanical human body structure model, wherein the motion control strategy is configured to decompose the target action into multiple motion stages and select a rotational driving force or a linear driving force to apply to one or more human body parts to control their motion; a first simulation unit, configured to perform kinematic simulation based on the motion control strategy, simulate the motion of various parts of the human body at various motion stages, and derive parameters representing the motion state of each part, namely, rotational velocity and linear velocity that vary with time, the parameters including velocity-time data of the translational velocity of each part in each direction and the rotational velocity around each axis in a preset coordinate system; The second simulation is used to define the motion of various parts of the human body in computational fluid dynamics (CFD) software based on the derived rotational speed and linear speed parameters, and perform CFD simulation calculations.

Images