Human body biomechanical model generation method

Abstract

The invention belongs to the field of numerical modeling, and particularly relates to a human body biomechanical model generation method, human body geometric models of different age groups are obtained based on anatomy and a reverse modeling method, and rapid mapping generation of human body biomechanical models of corresponding age groups can be rapidly completed only through matching of a plurality of geometric feature points. The generation difficulty of human body biomechanical models of other age groups is greatly reduced, the modeling efficiency is improved, and meanwhile, the precision of generating the human body grid model can also be ensured. According to the method, high-fidelity biomechanical models of key impact parts such as the head, the neck and the chest are established on the basis of anatomy, then connection between the head and the neck and connection between the neck and the chest are achieved by referring to the connection mode of human organs / tissues, physical verification of the effectiveness of the models is carried out on the basis, and the precision and reliability of the models are guaranteed.

Description

Technical Field

[0001] The present application belongs to the field of numerical modeling, and in particular relates to a method for generating a human biomechanical model. Background Art

[0002] In recent years, the light and small drone industry has flourished and has been widely used in many fields, reshaping people's work and life paradigms. Drone manufacturers represented by DJI occupy an absolute dominant position in the global market, and the drone industry has become another beautiful business card in the industrial field. At the same time, drones overlap with ground personnel and others in time and space during operation, resulting in frequent collision accidents. There are thousands of collision cases every year around the world, which has caused widespread concern among the public. The introduction of flight restrictions and flight control requirements has become a major challenge restricting the sustainable and healthy development of the drone industry. For this reason, it is urgent to carry out drone collision safety assessments to evaluate the human body injuries during drone collisions. Although this method can more accurately obtain human impact responses and injuries, it has difficulties in conducting experiments, high experimental costs, incomplete parameter acquisition, and long test cycles.

[0003] With the development and breakthrough of computer simulation technology, digital modeling based on living data has gradually become the first choice for scientific research. The human biomechanical model is a method of simulating the tissue structure of the human body for research. Its purpose is to analyze the interaction between various parts of the human body. It can accurately evaluate the response of various parts of the human body to the load in a numerical simulation method that is relatively consistent with human ethical norms, thereby meeting the needs of various application fields for quantitative analysis of the human body system. It is used to carry out the analysis of injuries to personnel caused by drone collisions. In addition to improving the efficiency and economy of the research, it can also obtain physical quantities that cannot be measured through experiments, quantitatively reveal the force transmission path and deformation degree of the human body during the collision, and clarify the law and degree of injury to the human body caused by collisions.

[0004] Therefore, how to establish a human biomechanical model that can quantitatively reveal the force transmission path and deformation degree of the human body during collision is a problem that needs to be solved. Summary of the invention

[0005] The purpose of this application is to provide a method for generating a human biomechanical model to solve the problem in the prior art that it is difficult to quantitatively reveal the force transmission path and deformation degree of the human body during a collision.

[0006] The technical solution of the present application is: a method for generating a human biomechanical model, comprising: Determine the human body parts that need to be reconstructed, conduct demand analysis, and obtain the different human organs and tissues required for different human body parts; obtain existing medical image data for image segmentation and classification, and then combine the data after image segmentation of different categories to model different human organs and tissues separately, and assemble them to obtain the human upper body geometric model based on anatomy; the human upper body geometric model includes the geometric models of various parts of the human body, the human head geometric model, the human neck geometric model, and the human chest geometric model; The geometric model of the upper body of the human body is meshed. According to the anatomical characteristics, a hexahedral mesh model is established for each bone and tissue of the upper body of the human body, a quadrilateral shell unit is established for the skin, and a spring unit is established for the neck muscles. After splicing, a refined mesh model of the upper body of the human body is obtained; On the basis of the established human upper body geometric model, based on the radial basis theory, the correspondence between key landmarks in the human upper body geometric model of different age groups is established; firstly, the human body mesh transformation is performed to obtain the target human body mesh model of each part, and then the target human body mesh model of each part is spliced ​​to generate the human biomechanical models of different age groups; Based on the refined mesh model of the upper body and the biomechanical models of the human body at different ages, combined with the existing test data of volunteers, the collision analysis of the head, neck and chest was carried out to correct and verify the model, and the corrected biomechanical model of the human body was obtained. For the selected UAV products, the corresponding target finite element mesh model is established, and analysis conditions of maximum speed falling and hitting the top of the head, maximum speed horizontal impact on the forehead, and maximum speed impact on the chest are carried out. The sensor response data of the modified human biomechanical model is extracted, and the degree of human injury is evaluated according to the human injury criteria.

[0007] Preferably, the geometric model of each part of the human body is modeled by: Using image segmentation methods, the skin, bones, muscles, brain, and ligament tissues in different parts are separated to obtain segmented images; Based on the segmented image, the medical image data points are geometrically reconstructed through the threshold segmentation method. According to the shape characteristics of different parts, organs and tissues of the human body, geometric segmentation is performed against the anatomical atlas to obtain the geometric models of various parts and tissues of the human body.

[0008] Preferably, the geometric models of tissues in various parts of the human body are cleaned up until self-intersections of the curved surfaces, highly refractive edges and small holes in the geometry are eliminated.

[0009] Preferably, the specific steps of establishing the hexahedral mesh model are: Import the human upper body geometry model into the Truegrid software, open the human upper body geometry model in IGES format, and obtain the point, line, and surface data of the human upper body geometry model; Define blocks according to the point, line and surface data of the human upper body geometric model until the grid is evenly distributed in the three directions of X, Y and Z, adjust the shape of the block to be consistent with the human upper body geometric model, and move the block to the position of the center of mass of the human upper body geometric model to obtain the boundary points of the block on the grid; Project the boundary points of the block on the grid to the boundary on the geometry, project the line connecting any two points in the block to the geometric boundary line, and finally project the plane surrounded by the four edges of the block onto the grid surface to obtain a hexahedral grid model.

[0010] Preferably, the quadrilateral shell element is constructed according to the anatomical atlas, using a combination of shell elements that reflect the real geometric shape of the ligament and simplified discrete beam elements to construct the ligament model; the spring element is modeled by analyzing muscle fiber bundles to simplify the muscle into spring elements.

[0011] Preferably, after the hexahedral mesh model, quadrilateral shell unit and spring unit are established, the established human head geometry model, human neck geometry model and human chest geometry model are spliced ​​based on the anatomical structure to generate a refined mesh model of the human upper body.

[0012] Preferably, a mesh quality check is performed on the established refined mesh model of the upper body of the human body, and the Jacobian, warping, distortion and aspect ratio of the mesh are obtained by a geometric integrity check method. An evaluation index of mesh quality is set to determine whether there are mesh penetration and interference problems between the connected parts of the meshes and between the meshes of each part. If the evaluation index requirements of the mesh quality are met, it is determined that there are no mesh penetration and interference problems. The mesh quality evaluation index includes: for a hexahedral mesh model, the number of meshes with a Jacobian of 0.4 or higher is ≥99%, the number of meshes with a warping of 40 or lower is ≥99%, the number of meshes with a distortion of 50 or lower is ≥95%, and the number of meshes with an aspect ratio of 5 or lower is ≥95%; for quadrilateral shell elements, the number of meshes with a Jacobian of 0.3 or higher is ≥99%, the number of meshes with a warping of 40 or lower is ≥99%, the number of meshes with a distortion of 50 or lower is ≥99%, and the number of meshes with an aspect ratio of 3 or lower is ≥99%.

[0013] Preferably, the method for establishing the target finite element mesh model is: The upper body geometry model of the human body of other age groups other than the current age group is used as the target geometry model; a certain number of landmark points in the upper body geometry model of the human body are extracted as the reference landmark points ( x i , y i, z i ), and at the same time, the landmark points corresponding to the reference landmark points are extracted on the target geometric model as the target landmark points ( x j , y j , z j ); Assume that the number of extracted benchmark landmarks and target landmarks is n The number of all nodes in the refined mesh model of the upper body is N ; The coordinates of the transformed target geometric model are marked as ( M, P, Q ); Find the distance between two corresponding points using the Euclidean norm r Substitute the basis function and use the coordinate information of the basic geometric model to build the matrix , A is a parameter, is the transpose of P; According to the target landmark coordinate information of the target geometric model, calculate the column vector of the interpolation weight coefficient , and , the calculation formula of the nodes of the target finite element mesh model of each part is obtained as follows: ; In the formula, , , , are column vectors, is a fixed parameter; Different interpolation weight coefficients and the column vector composed of the coordinate parameters of all nodes of the refined mesh model of the upper body of the human body are used to perform the optimal solution mapping between the benchmark landmark points and the target landmark points, and the three column vectors M, P, and Q of the coordinates of the target finite element mesh model of each part after interpolation are obtained. The target finite element mesh model is constructed through the three column vectors M, P, and Q.

[0014] Preferably, based on the geometric model of the upper body and the target geometric model, the reference landmarks and target landmarks are selected and marked until the positions of the reference landmarks and the target landmarks correspond and their numbers are consistent; the optimal solution mapping method is specifically as follows: the one-to-one corresponding reference landmarks and target landmarks and the node coordinate information in the refined mesh model of the upper body are used as input conditions for the mesh transformation, and all the node coordinate information of the target finite element mesh model of each part is calculated, and the unit connection remains unchanged.

[0015] Preferably, the specific method for model correction and verification is: According to the working condition parameters of the volunteer head impact test obtained from the biomechanical research experiment of craniocerebral injury, the same analysis conditions were set for collision analysis. The accuracy of the human head geometric model was verified through the collision contact force curve, and the head test results were obtained. According to the existing volunteer trolley test condition parameters, the same analysis condition is set for collision analysis to obtain the cervical curve, and the accuracy of the human neck geometric model is verified through the cervical curve to obtain the neck test results; According to the existing volunteer chest impact test parameters, the same analysis conditions were set for collision analysis, the reliability verification of the human chest geometry model was carried out, and the chest test results were obtained; The human biomechanical model was modified and confirmed based on the test results of the head, neck and chest.

[0016] The human biomechanical model generation and transformation method of the present application has the following advantages: Based on the anatomy and inverse modeling methods, the geometric models of the human body of different age groups are obtained. Only by matching a number of geometric feature points, the rapid mapping and generation of the human body biomechanical model of the corresponding age group can be completed, which greatly reduces the difficulty of generating human body biomechanical models of other age groups and improves the modeling efficiency. At the same time, it can also ensure the accuracy of the generated human body mesh model.

[0017] Based on anatomy, a highly realistic biomechanical model of key impact parts such as the head, neck, and chest was established. Then, with reference to the connection methods of human organs / tissues, the connection between the head and neck, and the neck and chest was realized. On this basis, physical verification of the model's effectiveness was carried out to ensure the accuracy and reliability of the model.

[0018] The head, neck, and chest are modeled with high precision, and non-critical parts such as the lower body are modeled with simplified modeling. Mechanical response sensors are set up to quickly extract the collision response time curves of acceleration, displacement, or force of typical parts of concern, facilitating collision damage analysis and assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solution provided by the present application, the following is a brief introduction to the accompanying drawings. Obviously, the accompanying drawings described below are only some embodiments of the present application.

[0020] Figure 1 This is a schematic diagram of the overall process of this application; Figure 2 This is a schematic diagram of the geometric model of the human head of the applicant; Figure 3 This is a schematic diagram of the geometric model of the human neck of the applicant; Figure 4 This is a schematic diagram of the geometric model of the human chest of the applicant; Figure 5 This is a schematic diagram of the refined mesh model of the head of this application; Figure 6 Flowchart for establishing the finite element mesh model for the purpose of this application. DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0022] A method for generating a human biomechanical model. This application proposes a process for generating a human biomechanical model. Only the 50th percentile adult male is taken as an example. The method for generating human models of other age groups, genders and body shapes is the same and will not be repeated here.

[0023] like Figure 1 , including the following steps: Step S100, constructing a geometric model of the upper body of the human body: Specifically, the following steps are taken: determine the human body parts that need to be reconstructed, conduct demand analysis, and obtain the different human organs and tissues required for different human body parts; obtain existing medical image data for image segmentation and classification, and then combine the segmented data of different categories of images to model different human organs and tissues separately, and then assemble to obtain an anatomically based human upper body geometric model. The human upper body geometric model includes geometric models of tissues in various parts of the human body, geometric models of the human head, geometric models of the human neck, and geometric models of the human chest.

[0024] The human body parts include the upper body parts and the rest. The upper body parts are divided into the head, neck and chest, which can be subdivided into brain tissue, skull, scalp, cervical vertebrae, cervical intervertebral disc, cervical joints, thoracic vertebrae, ribs, sternum, scapula, clavicle, chest soft tissue, etc. The rest of the body parts include the upper limbs, abdomen, lower limbs, etc., which are not the parts of collision injury research. Anatomy-based human biomechanical modeling is not carried out, and simplified models are used instead.

[0025] Preferably, the means of acquiring the existing data is: for different human organs and tissues, medical images of the head, neck and chest are obtained by CT or MRI respectively.

[0026] Preferably, the geometric model of each part of the human body is modeled by: Using image segmentation methods, skin, bones, muscles, brain, ligaments and other tissues in different parts are separated to obtain segmented images; ensuring that the segmented images can fully reflect the structure and shape of the tissues and have sufficient biological fidelity; Based on the segmented image, the medical image data points are geometrically reconstructed through the threshold segmentation method. According to the shape characteristics of different parts, organs and tissues of the human body, geometric segmentation is performed in accordance with the anatomical atlas to obtain the geometric models of various parts and tissues of the human body. Since the biological geometric surface of the human body obtained by the threshold segmentation method has hole defects, the tissue model is further processed, specifically: the geometric model of the tissues in various parts of the human body is geometrically cleaned until the self-intersection of the surface, highly refractive edges, small holes in the geometry and other small structures are eliminated, the curvature of the geometric surface is improved, and the repaired geometric model of the tissues in various parts of the human body suitable for mesh modeling is obtained.

[0027] Preferably, the established human head geometric model includes skull, facial bones, head skin, brain tissue, cerebellum tissue, brain stem tissue, etc.; the established human neck geometric model includes 7 cervical vertebrae, intervertebral discs, ligaments, muscles, skin, etc.; the established human chest geometric model includes 12 thoracic vertebrae, 12 pairs of ribs, sternum, scapula, clavicle, lung tissue, heart tissue, muscle tissue, skin tissue, diaphragm, etc. Figure 2-Figure 4 .

[0028] The geometric models of various parts of the human body, the human head, the human neck and the human chest are assembled to complete the three-dimensional geometric construction of the head, neck and chest, forming an anatomically based geometric model of the human upper body.

[0029] Step S200, constructing a refined mesh model of the upper body of the human body: Specifically, the geometric model of the upper body of the human body is meshed, and according to the anatomical characteristics, hexahedral mesh models are established for the bones and tissues of the upper body of the human body, quadrilateral shell units are established for the skin, and spring units are established for the neck muscles. After splicing, a refined mesh model of the upper body of the human body is obtained.

[0030] Preferably, the Truegrid software is used to mesh the geometric model of the human upper body; the specific steps of establishing the hexahedral mesh model are: Import the human upper body geometry model into the Truegrid software, open the human upper body geometry model in IGES format, and obtain the point, line, and surface data of the human upper body geometry model; Define blocks according to the point, line and surface data of the human upper body geometric model until the grid is evenly distributed in the three directions of X, Y and Z, adjust the shape of the block to be consistent with the human upper body geometric model, and move the block to the position of the center of mass of the human upper body geometric model to obtain the boundary points of the block on the grid; Project the boundary points of the block on the grid to the geometric boundary, project the line connecting any two points in the block to the geometric boundary line, and finally project the plane surrounded by the four edges of the block onto the grid surface to obtain a hexahedral grid model; in the projection work of points, lines and surfaces, the use of lines and surfaces can solve related problems such as uneven grid distribution, and the internal grid can be smoothed by inserting nodes.

[0031] Preferably, for human ligament tissue, according to the anatomical atlas, a ligament model is constructed by combining shell elements that can truly reflect the true geometric shape of the ligament and simplified discrete beam elements to obtain a quadrilateral shell element. For human muscle tissue, the muscle is simplified into a spring element by analyzing the muscle fiber bundles to obtain each spring element.

[0032] Preferably, after the hexahedral mesh model, quadrilateral shell unit and spring unit are established, the established human head geometric model, human neck geometric model and human chest geometric model are spliced ​​based on the anatomical structure to generate a refined mesh model of the upper body of the human body. Among them, the neck is the part between the head and the chest, and its upper part is connected to the head model through ligaments, muscles, etc., and its lower part is connected to the chest model through intervertebral discs, ligaments, muscles, etc.

[0033] Preferably, for the mechanical response parameters that are the focus of collision analysis, such as head acceleration, six-degree-of-freedom cervical force, chest compression and chest acceleration, acceleration sensors, force sensors and displacement sensors are respectively set at designated positions of the refined mesh model of the upper body of the human body to extract the response time history signal during the collision analysis process.

[0034] Check the model quality. Check the mesh quality of the upper body refined mesh model. Manually adjust the mesh nodes to improve the mesh quality as much as possible. Use the geometric integrity check method to check whether there are mesh penetration and interference problems between the connected parts of the mesh and the internal meshes of each part. Set the mesh quality evaluation index. Use the geometric integrity check method to obtain the main parameters of the mesh, as shown in Table 1. The main parameters include the Jacobian, warping, distortion, aspect ratio, etc. of the mesh. Control the calculation results and calculation time by controlling the mesh quality evaluation index. If the mesh quality evaluation index requirements are met, it is judged that there is no mesh penetration and interference problem.

[0035] For the abdomen and lower limbs that are not of concern in collision analysis, the corresponding parts of the mature and reliable Hybrid III 50 percentile adult dummy model are used to replace them, and then spliced ​​with the established human upper body biomechanics to complete the final construction of the human body finite element model. The refined mesh models of the head, neck, chest, upper body and whole body are established, among which the refined mesh model of the head is as follows Figure 5 shown.

[0036] Table 1 Evaluation indexes of mesh quality of human biological tissue

[0037] Step S300, performing human body model transformation based on landmark point mapping theory: Specifically: On the basis of the established human upper body geometric model, based on the radial basis theory, the correspondence between key landmark points in the human upper body geometric model of different age groups is established, and the human body mesh transformation is first performed to obtain the target finite element mesh model of each part, and then the target finite element mesh model of each part is spliced ​​to generate the human biomechanical models of different age groups.

[0038] Preferably, if Figure 6 , the method of establishing the target finite element mesh model is: The upper body geometry model of the human body of other age groups other than the current age group is used as the target geometry model; a certain number of landmark points in the upper body geometry model of the human body are extracted as the reference landmark points ( x i , y i , z i ), and at the same time, the landmark points corresponding to the reference landmark points are extracted on the target geometric model as the target landmark points ( x j , y j , z j ). Assume that the number of extracted benchmark landmarks and target landmarks is n The number of all nodes in the refined mesh model of the upper body is N The coordinates of the transformed target geometric model are ( M, P, Q ).

[0039] , , .

[0040] Find the distance between two corresponding points using the Euclidean norm r Substitute the basis function and use the coordinate information of the basic geometric model to build the matrix , A is a parameter, is the transpose of P.

[0041] According to the target landmark coordinate information of the target geometric model, calculate the column vector of the interpolation weight coefficient , and , the calculation formula of the nodes of the target finite element mesh model of each part is obtained as follows: (1) In the formula, , , , are column vectors, is a fixed parameter.

[0042] By using different interpolation weight coefficients and the column vector composed of the coordinate parameters of all nodes of the refined mesh model of the upper body of the human body, the optimal solution mapping between the benchmark landmark points and the target landmark points is performed by calculating the minimum value of the surface energy equation. The three column vectors M, P, and Q of the coordinates of the target finite element mesh model of each part after interpolation can be obtained, and the target finite element mesh model is constructed through the three column vectors M, P, and Q.

[0043] Preferably, the target geometric model is modeled in the same manner as the method for constructing the human upper body geometric model in step S100, and is obtained by separately modeling different human organs and tissues and then splicing the models.

[0044] Preferably, based on the human upper body geometric model and the target geometric model, reference markers and target markers are selected and marked until the positions of the reference markers and the target markers correspond and the numbers are consistent.

[0045] Preferably, the optimal solution mapping method is specifically as follows: using the one-to-one corresponding reference landmark points and target landmark points and the node coordinate information in the refined mesh model of the upper body of the human body as the input conditions of the mesh transformation, calculating all the node coordinate information of the target finite element mesh model of each part, and keeping the unit connection unchanged.

[0046] The shapes and sizes of human body geometric models of different age groups vary greatly, which may cause mesh distortion after transformation. Therefore, mesh quality optimization is required to obtain a high-quality target mesh model. The mesh quality optimization method can be engineering simulation mesh optimization method, etc.

[0047] The method for splicing the target finite element mesh models of various parts is the same as the method for splicing in step S200. Taking a 6-year-old child as an example, a human biomechanical model of a 6-year-old child is obtained by the transformation method.

[0048] Step S400, verifying a highly realistic human biomechanical model: Based on the refined mesh model of the human upper body and the human biomechanical models of different age groups, combined with the experimental data obtained through volunteer experiments, collision analysis of the three parts of the head, neck and chest was carried out, and the model was modified and verified to obtain the revised human biomechanical model.

[0049] Preferably, the specific method for model correction and verification is: According to the working condition parameters of the volunteer head impact test obtained from the Nahum test (biomechanical research test of craniocerebral injury), the same analysis conditions were set for collision analysis. The accuracy of the human head geometric model was verified through the collision contact force curve, and the head test results were obtained; According to the existing volunteer trolley test condition parameters, the same analysis condition is set for collision analysis to obtain the cervical curve, and the accuracy of the human neck geometric model is verified through the cervical curve to obtain the neck test results; According to the existing volunteer chest impact test parameters, the same analysis conditions were set for collision analysis, the reliability verification of the human chest geometry model was carried out, and the chest test results were obtained; Based on the test results of the head, neck and chest, the human biomechanical model is modified and confirmed to ensure that the model analysis response curve falls within the envelope of the test curve.

[0050] Step S500, conducting simulation analysis and damage assessment of drone impact on human body: For the selected UAV products, the corresponding target finite element mesh model is established, and analysis conditions of maximum speed falling and hitting the top of the head, maximum speed horizontal impact on the forehead, and maximum speed impact on the chest are carried out. The sensor response data of the modified human biomechanical model is extracted, and the degree of human injury is evaluated according to the human injury criteria.

[0051] In summary, this application has the following advantages: Based on the anatomy and inverse modeling methods, the geometric models of the human body of different age groups are obtained. Only by matching a number of geometric feature points, the rapid mapping and generation of the human body biomechanical model of the corresponding age group can be completed, which greatly reduces the difficulty of generating human body biomechanical models of other age groups and improves the modeling efficiency. At the same time, it can also ensure the accuracy of the generated human body mesh model.

[0052] Based on anatomy, a highly realistic biomechanical model of key impact parts such as the head, neck, and chest was established. Then, with reference to the connection methods of human organs / tissues, the connection between the head and neck, and the neck and chest was realized. On this basis, physical verification of the model's effectiveness was carried out to ensure the accuracy and reliability of the model.

[0053] The head, neck, and chest are modeled with high precision, and non-critical parts such as the lower body are modeled with simplified modeling. Mechanical response sensors are set up to quickly extract the collision response time curves of acceleration, displacement, or force of typical parts of concern, facilitating collision damage analysis and assessment.

[0054] Based on the establishment of the adult model, the rapid mapping and generation of models of human bodies of different ages (different sizes and weights) can be realized, which greatly reduces the difficulty of generating biomechanical models of human bodies of other age groups and improves modeling efficiency.

[0055] Finally, it should be noted that: the drawings of the embodiments disclosed in the present invention only involve structures related to the embodiments disclosed in the present invention, and other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of the present invention can be combined with each other; Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for generating a human biomechanical model, characterized in that: include: Determine the body parts that need to be reconstructed, conduct demand analysis, and obtain the different human organs and tissues required for different body parts; Obtain existing medical image data for image segmentation and classification, and then combine the segmented data of different categories of images to model different human organs and tissues respectively, and assemble to obtain an anatomically based upper body geometric model of the human body; the upper body geometric model of the human body includes geometric models of various parts of the human body, a human head geometric model, a human neck geometric model, and a human chest geometric model; The geometric model of the upper body of the human body is meshed. According to the anatomical characteristics, a hexahedral mesh model is established for each bone and tissue of the upper body of the human body, a quadrilateral shell unit is established for the skin, and a spring unit is established for the neck muscles. After splicing, a refined mesh model of the upper body of the human body is obtained; On the basis of the established human upper body geometric model, based on the radial basis theory, the correspondence between key landmarks in the human upper body geometric model of different age groups is established; firstly, the human body mesh transformation is performed to obtain the target human body mesh model of each part, and then the target human body mesh model of each part is spliced ​​to generate the human biomechanical models of different age groups; Based on the refined mesh model of the upper body and the biomechanical models of the human body at different ages, combined with the existing test data of volunteers, the collision analysis of the head, neck and chest was carried out to correct and verify the model, and the corrected biomechanical model of the human body was obtained. For the selected UAV products, the corresponding target finite element mesh model is established, and analysis conditions of maximum speed falling and hitting the top of the head, maximum speed horizontal impact on the forehead, and maximum speed impact on the chest are carried out. The sensor response data of the modified human biomechanical model is extracted, and the degree of human injury is evaluated according to the human injury criteria.

2. The method for generating a human biomechanical model according to claim 1, wherein: The modeling method of the geometric model of various parts of the human body is: Using image segmentation methods, the skin, bones, muscles, brain, and ligament tissues in different parts are separated to obtain segmented images; Based on the segmented image, the medical image data points are geometrically reconstructed through the threshold segmentation method. According to the shape characteristics of different parts, organs and tissues of the human body, geometric segmentation is performed against the anatomical atlas to obtain the geometric models of various parts and tissues of the human body.

3. The method for generating a human biomechanical model according to claim 2, wherein: The geometric models of various parts of the human body are cleaned up until the self-intersections of the surfaces, highly refractive edges and small holes in the geometry are eliminated.

4. The method for generating a human biomechanical model according to claim 1, wherein: The specific steps to establish a hexahedral mesh model are: Import the human upper body geometry model into the Truegrid software, open the human upper body geometry model in IGES format, and obtain the point, line, and surface data of the human upper body geometry model; Define blocks according to the point, line and surface data of the human upper body geometric model until the grid is evenly distributed in the three directions of X, Y and Z, adjust the shape of the block to be consistent with the human upper body geometric model, and move the block to the position of the center of mass of the human upper body geometric model to obtain the boundary points of the block on the grid; Project the boundary points of the block on the grid to the boundary on the geometry, project the line connecting any two points in the block to the geometric boundary line, and finally project the plane surrounded by the four edges of the block onto the grid surface to obtain a hexahedral grid model.

5. The method for generating a human biomechanical model according to claim 4, wherein: The quadrilateral shell element is constructed according to the anatomical atlas, and the ligament model is obtained by combining the shell element that reflects the real geometric shape of the ligament and the simplified discrete beam element. The spring element is obtained by analyzing the muscle fiber bundles and simplifying the muscle into spring elements for modeling.

6. The method for generating a human biomechanical model according to claim 5, wherein: After the hexahedral mesh model, quadrilateral shell unit and spring unit are established, the established human head geometry model, human neck geometry model and human chest geometry model are spliced ​​based on the anatomical structure to generate a refined mesh model of the human upper body.

7. The method for generating a human biomechanical model according to claim 1, wherein: The mesh quality of the established refined mesh model of the upper body of the human body is checked. The Jacobian, warping, distortion and aspect ratio of the mesh are obtained by using the geometric integrity check method. The mesh quality evaluation index is set to determine whether there are mesh penetration and interference problems between the connected parts of the mesh and between the meshes of each part. If the mesh quality evaluation index requirements are met, it is determined that there are no mesh penetration and interference problems. The evaluation indicators of mesh quality include: for hexahedral mesh models, the number of meshes with Jacobi above 0.4 is ≥99%, the number of meshes with warpage below 40 is ≥99%, the number of meshes with distortion below 50 is ≥95%, and the number of meshes with aspect ratio below 5 is ≥95%; for quadrilateral shell elements, the number of meshes with Jacobi above 0.3 is ≥99%, the number of meshes with warpage below 40 is ≥99%, the number of meshes with distortion below 50 is ≥99%, and the number of meshes with aspect ratio below 3 is ≥99%.

8. The method for generating a human biomechanical model according to claim 1, wherein: The method for establishing the target finite element mesh model is: The upper body geometry model of the human body of other age groups other than the current age group is used as the target geometry model; a certain number of landmark points in the upper body geometry model of the human body are extracted as the reference landmark points ( x i , y i , z i ), and at the same time, the landmark points corresponding to the reference landmark points are extracted on the target geometric model as the target landmark points ( x j , y j , z j ); Assume that the number of extracted benchmark landmarks and target landmarks is n The number of all nodes in the refined mesh model of the upper body is N ; The coordinates of the transformed target geometric model are marked as ( M, P, Q ); Find the distance between two corresponding points using the Euclidean norm r Substitute the basis function and use the coordinate information of the basic geometric model to build the matrix , A is a parameter, is the transpose of P; According to the target landmark coordinate information of the target geometric model, calculate the column vector of the interpolation weight coefficient , and , the calculation formula of the nodes of the target finite element mesh model of each part is obtained as follows: ； In the formula, , , , are column vectors, is a fixed parameter; Using different interpolation weight coefficients and the column vector composed of the coordinate parameters of all nodes of the refined mesh model of the upper body of the human body, the optimal solution mapping between the benchmark landmark points and the target landmark points is performed by calculating the minimum value of the surface energy equation, and the three column vectors M, P, Q of the coordinates of the target finite element mesh model of each part after interpolation are obtained. The target finite element mesh model is constructed through the three column vectors M, P, Q.

9. The method for generating a human biomechanical model according to claim 8, wherein: Based on the upper body geometric model and the target geometric model, the reference landmarks and target landmarks are selected and marked until the positions of the reference landmarks and the target landmarks correspond and their numbers are consistent. The optimal solution mapping method is as follows: the one-to-one corresponding reference landmarks and target landmarks and the node coordinate information in the refined mesh model of the upper body are used as input conditions for mesh transformation, and all the node coordinate information of the target finite element mesh model of each part is calculated, while the unit connection remains unchanged.

10. The method for generating a human biomechanical model according to claim 1, wherein: The specific methods for model modification and verification are as follows: According to the working condition parameters of the volunteer head impact test obtained from the biomechanical research experiment of craniocerebral injury, the same analysis conditions were set for collision analysis. The accuracy of the human head geometric model was verified through the collision contact force curve, and the head test results were obtained. According to the existing volunteer trolley test condition parameters, the same analysis condition is set for collision analysis to obtain the cervical curve, and the accuracy of the human neck geometric model is verified through the cervical curve to obtain the neck test results; According to the existing volunteer chest impact test parameters, the same analysis conditions were set for collision analysis, the reliability verification of the human chest geometry model was carried out, and the chest test results were obtained; The human biomechanical model was modified and confirmed based on the test results of the head, neck and chest.

Images