Ground collision head injury evaluation method and system based on accident reconstruction

Abstract

The invention provides a ground collision head injury evaluation method and system based on accident reconstruction. The method comprises the following steps: acquiring accident information and an accident video of a traffic accident case; establishing a multi-rigid-body human body model, a multi-rigid-body vehicle model and a rigid-body ground model based on the accident information, and performing kinematics reconstruction; comparing and verifying the kinematics reconstruction result and the accident video, and constructing a human body finite element model at the moment before the head collides with the ground; and finally, performing head-ground collision simulation, calculating various head injury criteria based on head kinematics and brain tissues, and analyzing intracranial injury distribution of the VRU. According to the method, based on video information verification and two-stage reconstruction, the head injury caused by ground collision can be effectively evaluated; besides, the effectiveness of various head injury criteria in the aspect of predicting the serious head injury is systematically evaluated, a specific prediction capability level is given, and a scientific basis is provided for selection of the injury criteria.

Description

Technical Field

[0001] The present invention relates to the technical field of road traffic safety, and more specifically, to a method and system for evaluating head injuries in ground collisions based on accident reconstruction. Background Art

[0002] Road traffic accident investigations show that ground collision is one of the important causes of severe head injuries to vulnerable road users (VRUs). Existing research mainly focuses on the vehicle collision stage, while research on ground collision is relatively rare.

[0003] For example, an existing patent document discloses a method for determining the influence of head kinematic boundary conditions on head injury, comprising the following steps: obtaining accident information of multiple traffic accident cases that meet preset conditions, wherein the multiple traffic accident cases all involve collisions between vulnerable road users (VRUs) and vehicles; simulating the traffic accident process according to the accident information of the traffic accident cases, outputting head injury indicators and extracting head kinematic boundary conditions in the collision; using a univariate linear regression analysis method to perform correlation analysis on the extracted head kinematic boundary conditions and the output head injury indicators from the simulation of multiple traffic accident cases, calculating the correlation between each head kinematic boundary condition and each head injury indicator, and determining the influence of each head kinematic boundary condition on head injury according to the calculation results; the prior art mainly focuses on the influence of head kinematic boundary conditions on head injury, especially head injury in the process of collision between vehicles and pedestrians, but does not focus on the mechanism and risk of pedestrian head landing injury, and it is difficult to evaluate the influence of ground collision on head injury; in addition, the prior art determines the influence of each head kinematic boundary condition on head injury through correlation analysis, but does not specifically give the prediction ability level and prediction effectiveness evaluation of each criterion.

[0004] Due to the large number of uncertainties in ground collisions, existing head injury criteria and damage tolerance have great limitations in predicting such injuries. Therefore, a scientific method based on real accident cases is needed to evaluate the effectiveness of these criteria. Summary of the invention

[0005] In order to overcome the defects of the above-mentioned prior art that it is difficult to evaluate the impact of ground collision on head injury and difficult to determine the predictive ability level of various head injury criteria, the present invention provides a ground collision head injury evaluation method and system based on accident reconstruction, which combines video information for detailed accident reconstruction, can more realistically simulate ground collision accident scenes, and systematically evaluate the predictive ability of multiple head injury criteria, providing a scientific basis for the selection of injury criteria; finally, a head injury evaluation scheme suitable for ground collision is proposed, filling the gap in existing research.

[0006] In order to solve the above technical problems, the technical solution of the present invention is as follows:

[0007] A method for evaluating head injuries in ground collisions based on accident reconstruction comprises the following steps:

[0008] S1: Acquire accident information and an accident video of a traffic accident case, wherein the traffic accident case involves a collision between a VRU and a vehicle, and the head of the VRU only collides with the ground but does not directly contact the vehicle;

[0009] S2: establishing a multi-rigid human body model, a multi-rigid vehicle model and a rigid ground model based on the accident information, and performing a simulation of the entire traffic accident process to obtain a kinematic reconstruction result of the traffic accident case;

[0010] S3: Compare and verify the kinematic reconstruction result with the accident video, obtain the motion response parameters of various parts of the body at the moment before the VRU head and the ground collide, and construct a finite element model of the human body at that moment;

[0011] S4: performing a head-ground collision simulation based on the human finite element model, and calculating a plurality of head injury criteria based on head kinematics and brain tissue;

[0012] S5: Analyze the intracranial injury distribution of the VRU according to the calculated head injury criterion to complete the ground impact head injury evaluation.

[0013] Preferably, in step S1, the accident information includes: accident scene environment information, scene photos, vehicle information, VRU information and head injury information;

[0014] The video information records the entire process of the traffic accident case, including the movement patterns of the vehicle and the VRU before, during and after the collision, the collision moment of the vehicle and the VRU, and the kinematic posture of the VRU head when it collides with the ground.

[0015] Preferably, the accident scene environmental information includes: the specific time, location, road type, road surface material and weather conditions of the accident;

[0016] The vehicle information includes: the specific model, production year, front structure parameters, collision speed and braking deceleration of the vehicle; the front structure parameters include engine hood length BL, engine hood angle BA, engine hood front edge height BLEH, windshield angle WA and vehicle ground clearance GC;

[0017] The VRU information includes: basic physiological information of the VRU, and the motion state of the VRU before the collision with the vehicle;

[0018] The head injury information includes: the head injury medical record of the VRU, the medical record includes a head CT / MRI scan image, and the AIS injury grade of each part of the head, the specific injury type and its detailed description.

[0019] Preferably, in step S2, the multi-rigid body human body model includes any one of a CPM pedestrian model, a Madymo pedestrian model, a LS-DYNA pedestrian model and a GHBMC pedestrian model, and the most preferred is the CPM pedestrian model;

[0020] The multi-rigid body vehicle model includes a front structure, a side structure, a rear structure, a chassis structure, tires and an internal structure.

[0021] Preferably, in step S2, the entire traffic accident process is simulated using multi-rigid body dynamics analysis software MADYMO to obtain the kinematic reconstruction result of the traffic accident case;

[0022] In the MADYMO software, the multi-rigid body human model is obtained by performing personalized adjustment on the TNO 50 percentile standard male model; the personalized adjustment includes: scaling the geometric dimensions according to the gender and height of the VRU in the traffic accident case, modifying the stiffness characteristics of each joint according to the age and weight of the VRU, and adjusting the model posture to be consistent with the VRU body posture at the time of collision;

[0023] In the MADYMO software, constructing the multi-rigid body vehicle model includes:

[0024] Front structure: The front structure is constructed using multiple ellipsoids and the stiffness characteristics of each geometric body are defined based on pre-collected Euro-NCAP test data; the front structure includes the vehicle's hood, front windshield and front bumper;

[0025] Side structure: The side structure is constructed using multiple ellipsoids or cuboids or cylinders, and the stiffness characteristics of each geometric body are defined based on pre-collected Euro-NCAP test data; the side structure includes doors and side windows;

[0026] Rear structure: The rear structure is constructed using multiple ellipsoids or cuboids or cylinders, and the stiffness characteristics of each geometric body are defined according to the pre-collected Euro-NCAP test data; the rear structure includes the trunk and the rear bumper;

[0027] Chassis structure: Use multiple ellipsoids or cuboids to build the frame structure, and use spring or damper elements to simulate the dynamic characteristics of the suspension system to build the chassis structure;

[0028] Tires: Build tire models using cylinders or custom-shaped geometry and define the tire's stiffness and friction characteristics;

[0029] Internal structure: Use multiple cuboids or ellipsoids to build the internal structure, which includes seats, steering wheels and instrument panels; and define the stiffness characteristics of each geometric body based on pre-collected Euro-NCAP test data;

[0030] In MADYMO software, the ground is simplified into a rigid plane, and the friction coefficient between people and the ground is set to obtain a rigid ground model.

[0031] Preferably, the step S3 comprises:

[0032] Comparing and verifying the kinematic reconstruction result with the accident video, the contents of the comparison and verification include: the initial collision position between the vehicle and the VRU, the flight trajectory of the VRU in the air, the rotation angle, the first contact position between the body and the ground, the collision sequence between each body part and the ground, and the final position;

[0033] The kinematic reconstruction results after comparison and verification are used to obtain the motion response parameters of various parts of the body of the VRU at the moment before the head and the ground collide. The motion response parameters include: the three-dimensional position of the center of gravity of the head of the VRU before the head and the ground collide, the three-axis velocity and angular velocity information, and the motion status of other parts of the VRU body;

[0034] A THUMS finite element whole human model is constructed and adjusted to be consistent with the posture of the multi-rigid body human model when it is about to hit the ground, the motion response parameters are loaded into the corresponding parts of the THUMS finite element whole human model, and the human finite element model at that moment is constructed.

[0035] Preferably, in step S4, a finite element analysis software is used to adopt a surface-to-surface contact algorithm as a contact algorithm for the cranial brain tissue in the human finite element model, and a friction coefficient between the brain tissue and the dura mater is set to simulate the relative sliding between the brain tissue and the dura mater when the head collides with the ground;

[0036] Perform head-ground collision simulation based on the human finite element model after contact definition, and calculate multiple head injury criteria based on head kinematics and brain tissue;

[0037] The head injury criteria based on head kinematics include: any one or more of head angular velocity, angular acceleration, linear acceleration, HIC, HIP, BRIC and BrIC;

[0038] The brain tissue-based head injury criteria include: any one or more of von Mises stress, shear stress, positive pressure, negative pressure, MPS, CSDM0.15, CSDM0.25 and DDM.

[0039] Preferably, the method further comprises:

[0040] S6: Repeat steps S1 to S5 to evaluate ground collision head injuries in multiple traffic accident cases, and comprehensively evaluate the predictive ability of various head injury criteria for head injuries.

[0041] Preferably, in step S6, comprehensively evaluating the predictive ability of each head injury criterion for head injury comprises the following steps:

[0042] Compare the calculated value of each head injury criterion with the corresponding preset injury tolerance limit;

[0043] Re-rank the traffic accident cases according to the calculated values ​​to intuitively display the comparison between the calculated values ​​of each head injury criterion and the injury tolerance limit;

[0044] For each head injury criterion, if its calculated value is greater than or equal to the corresponding injury tolerance limit, it is an effective prediction; the number of cases in which each head injury criterion effectively predicts the head injury level is counted, and the prediction accuracy of each head injury criterion is calculated;

[0045] According to the calculated prediction accuracy, the predictive ability of each head injury criterion is divided into four levels:

[0046] A prediction accuracy between 75% and 100% is considered “excellent”; between 50% and 75% is considered “good”; between 25% and 50% is considered “fair”; and between 0% and 25% is considered “poor”.

[0047] The present invention also provides a ground collision head injury assessment system based on accident reconstruction, which uses the above-mentioned ground collision head injury assessment method based on accident reconstruction, including:

[0048] Case information collection unit: used to obtain accident information and accident video of a traffic accident case, wherein the traffic accident case involves a collision between a VRU and a vehicle, and the head of the VRU only collides with the ground without direct contact with the vehicle;

[0049] Kinematic reconstruction unit: used to establish a multi-rigid human body model, a multi-rigid vehicle model and a rigid ground model based on the accident information, and simulate the entire traffic accident process to obtain the kinematic reconstruction result of the traffic accident case;

[0050] Kinematic verification unit: used to compare and verify the kinematic reconstruction result with the accident video, obtain the motion response parameters of various parts of the body at the moment before the VRU head and the ground collide, and construct a finite element model of the human body at that moment;

[0051] A head injury reconstruction unit: used for performing head-ground collision simulation based on the human finite element model, and calculating a variety of head injury criteria based on head kinematics and brain tissue;

[0052] Head injury analysis unit: used to analyze the intracranial injury distribution of the VRU according to the calculated head injury criteria, and complete the ground collision head injury evaluation.

[0053] Compared with the prior art, the technical solution of the present invention has the following beneficial effects:

[0054] The present invention provides a ground collision head injury evaluation method based on accident reconstruction, firstly, the accident information and accident video of a traffic accident case are obtained, the traffic accident case involves a collision between a VRU and a vehicle, and the head of the VRU only collides with the ground but does not come into direct contact with the vehicle; then, a multi-rigid body human model, a multi-rigid body vehicle model and a rigid body ground model are established based on the accident information, and the whole traffic accident process is simulated to obtain the kinematic reconstruction result of the traffic accident case; then, the kinematic reconstruction result is compared and verified with the accident video, the motion response parameters of various parts of the body at a moment before the VRU head and the ground are obtained and a human finite element model at that moment is constructed; then, a head-ground collision simulation is performed based on the human finite element model, and a plurality of head injury criteria based on head kinematics and brain tissue are calculated; finally, the intracranial injury distribution of the VRU is analyzed according to the calculated head injury criteria, and the ground collision head injury evaluation is completed;

[0055] The present invention proposes a head injury assessment method suitable for ground collision, which fills the gap in existing research; firstly, the present invention combines accident video information to perform detailed accident reconstruction, which can load collision boundary conditions more accurately, ensure that the kinematic response of accident reconstruction is highly consistent with the actual situation, thereby improving the credibility of accident reconstruction; secondly, the present invention adopts a two-stage reconstruction method, which fully combines the advantages of multi-rigid body model and finite element model, and can simulate the accident process more realistically, especially in the analysis of intracranial dynamic response and injury mechanism in the head-ground collision stage; at the same time, the present invention not only focuses on the overall injury of the head, but also goes deep into the injury conditions of different areas in the skull. This analysis can provide more detailed information on injury mechanism and injury degree, so as to better understand the mechanism of injury and provide more specific data support for improving vehicle safety design and protective measures; in addition, the present invention also systematically evaluates the effectiveness of multiple head injury criteria in predicting severe head injuries, and gives specific prediction ability levels, providing a scientific basis for the selection of injury criteria. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 This is a flow chart of a ground collision head injury assessment method based on accident reconstruction provided in Example 1.

[0057] Figure 2 This is an overall flow chart of a ground collision head injury assessment method based on accident reconstruction provided in Example 2.

[0058] Figure 3 This is a schematic diagram of the multi-rigid body vehicle model provided in Example 2.

[0059] Figure 4 This is a schematic diagram of the multi-rigid body human body model and the human body finite element model provided in Example 2.

[0060] Figure 5 This is a schematic diagram of reordering the cases according to the calculated values ​​of each head injury criterion provided in Example 2.

[0061] Figure 6 This is a structural diagram of a ground collision head injury assessment system based on accident reconstruction provided in Example 3. DETAILED DESCRIPTION

[0062] The drawings are for illustrative purposes only and should not be construed as limiting the present patent;

[0063] In order to better illustrate the present embodiment, some parts in the drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product;

[0064] It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0065] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0066] Example 1

[0067] like Figure 1 As shown, this embodiment provides a ground collision head injury assessment method based on accident reconstruction, comprising the following steps:

[0068] S1: Acquire accident information and an accident video of a traffic accident case, wherein the traffic accident case involves a collision between a VRU and a vehicle, and the head of the VRU only collides with the ground but does not directly contact the vehicle;

[0069] S2: establishing a multi-rigid human body model, a multi-rigid vehicle model and a rigid ground model based on the accident information, and performing a simulation of the entire traffic accident process to obtain a kinematic reconstruction result of the traffic accident case;

[0070] S3: Compare and verify the kinematic reconstruction result with the accident video, obtain the motion response parameters of various parts of the body at the moment before the VRU head and the ground collide, and construct a finite element model of the human body at that moment;

[0071] S4: performing a head-ground collision simulation based on the human finite element model, and calculating a plurality of head injury criteria based on head kinematics and brain tissue;

[0072] S5: Analyze the intracranial injury distribution of the VRU according to the calculated head injury criterion to complete the ground impact head injury evaluation.

[0073] In the specific implementation process, firstly, the accident information and accident video of the traffic accident case are obtained. The traffic accident case involves the collision between the VRU and the vehicle, and the head of the VRU only collides with the ground but does not come into direct contact with the vehicle;

[0074] Then, based on the accident information, a multi-rigid human model, a multi-rigid vehicle model and a rigid ground model are established, and the entire traffic accident process is simulated to obtain the kinematic reconstruction results of the traffic accident case;

[0075] The kinematic reconstruction results were then compared and verified with the accident video to obtain the motion response parameters of various body parts at the moment before the VRU head and the ground collided, and a finite element model of the human body at that moment was constructed;

[0076] This embodiment uses the comparative verification of the accident video information and the kinematic reconstruction results to more accurately load the collision boundary conditions, ensure that the kinematic response of the accident reconstruction is highly consistent with the actual situation, and help improve the credibility of the accident reconstruction;

[0077] Then, a head-ground collision simulation was performed based on the human finite element model to calculate various head injury criteria based on head kinematics and brain tissue;

[0078] Finally, the distribution of VRU intracranial injuries was analyzed according to the calculated head injury criteria to complete the ground collision head injury evaluation;

[0079] This embodiment performs a detailed analysis of the distribution of intracranial injuries, which means that it not only focuses on the overall injury of the head, but also goes deep into the injury conditions of different regions of the brain; this analysis can provide more detailed information on the injury mechanism and injury degree; this detailed analysis is not common in traditional head injury research, because most studies mainly focus on the overall injury indicators of the head (such as HIC, GAMBIT, etc.); through the detailed analysis of the distribution of intracranial injuries, the mechanism of injury can be better understood, and more specific data support can be provided for improving vehicle safety design and protective measures;

[0080] This embodiment combines the accident video information to perform detailed accident reconstruction, which can load the collision boundary conditions more accurately and ensure that the kinematic response of the accident reconstruction is highly consistent with the actual situation, thereby improving the credibility of the accident reconstruction; in addition, this embodiment adopts a two-stage reconstruction method, which fully combines the advantages of the multi-rigid body model and the finite element model, and can more realistically simulate the accident process, especially in the intracranial dynamic response and injury mechanism analysis during the head-ground collision stage.

[0081] Example 2

[0082] This embodiment provides a method for evaluating head injuries in ground collisions based on accident reconstruction, comprising the following steps:

[0083] S1: Acquire accident information and an accident video of a traffic accident case, wherein the traffic accident case involves a collision between a VRU and a vehicle, and the head of the VRU only collides with the ground but does not directly contact the vehicle;

[0084] S2: establishing a multi-rigid human body model, a multi-rigid vehicle model and a rigid ground model based on the accident information, and performing a simulation of the entire traffic accident process to obtain a kinematic reconstruction result of the traffic accident case;

[0085] S3: Compare and verify the kinematic reconstruction result with the accident video, obtain the motion response parameters of various parts of the body at the moment before the VRU head and the ground collide, and construct a finite element model of the human body at that moment;

[0086] S4: performing a head-ground collision simulation based on the human finite element model, and calculating a plurality of head injury criteria based on head kinematics and brain tissue;

[0087] S5: analyzing the intracranial injury distribution of the VRU according to the calculated head injury criterion to complete the ground collision head injury evaluation;

[0088] S6: repeating steps S1 to S5, evaluating ground collision head injuries for multiple traffic accident cases, and comprehensively evaluating the predictive ability of various head injury criteria for head injuries;

[0089] In step S1, the accident information includes: accident scene environment information, scene photos, vehicle information, VRU information and head injury information;

[0090] The video information records the entire process of the traffic accident case, including the movement patterns of the vehicle and the VRU before, during and after the collision, the collision moment of the vehicle and the VRU, and the kinematic posture of the VRU head when it collides with the ground;

[0091] The accident scene environmental information includes: the specific time, location, road type, road surface material and weather conditions of the accident;

[0092] The vehicle information includes: the specific model, production year, front structure parameters, collision speed and braking deceleration of the vehicle; the front structure parameters include engine hood length BL, engine hood angle BA, engine hood front edge height BLEH, windshield angle WA and vehicle ground clearance GC;

[0093] The VRU information includes: basic physiological information of the VRU, and the motion state of the VRU before the collision with the vehicle;

[0094] The head injury information includes: the medical record of the head injury of the VRU, the medical record including the head CT / MRI scan image, and the AIS injury grade of each part of the head, the specific injury type and its detailed description;

[0095] In step S2, the multi-rigid body human body model includes any one of a CPM pedestrian model, a Madymo pedestrian model, a LS-DYNA pedestrian model and a GHBMC pedestrian model, and the CPM pedestrian model is most preferred;

[0096] The multi-rigid body vehicle model includes a front structure, a side structure, a rear structure, a chassis structure, a tire and an internal structure;

[0097] In the step S2, the entire traffic accident process is simulated using the multi-rigid body dynamics analysis software MADYMO to obtain the kinematic reconstruction result of the traffic accident case;

[0098] In the MADYMO software, the multi-rigid body human model is obtained by performing personalized adjustment on the TNO 50 percentile standard male model; the personalized adjustment includes: scaling the geometric dimensions according to the gender and height of the VRU in the traffic accident case, modifying the stiffness characteristics of each joint according to the age and weight of the VRU, and adjusting the model posture to be consistent with the VRU body posture at the time of collision;

[0099] In the MADYMO software, constructing the multi-rigid body vehicle model includes:

[0100] Front structure: The front structure is constructed using multiple ellipsoids and the stiffness characteristics of each geometric body are defined based on pre-collected Euro-NCAP test data; the front structure includes the vehicle's hood, front windshield and front bumper;

[0101] Side structure: The side structure is constructed using multiple ellipsoids or cuboids or cylinders, and the stiffness characteristics of each geometric body are defined based on pre-collected Euro-NCAP test data; the side structure includes doors and side windows;

[0102] Rear structure: The rear structure is constructed using multiple ellipsoids or cuboids or cylinders, and the stiffness characteristics of each geometric body are defined according to the pre-collected Euro-NCAP test data; the rear structure includes the trunk and the rear bumper;

[0103] Chassis structure: Use multiple ellipsoids or cuboids to build the frame structure, and use spring or damper elements to simulate the dynamic characteristics of the suspension system to build the chassis structure;

[0104] Tires: Build tire models using cylinders or custom-shaped geometry and define the tire's stiffness and friction characteristics;

[0105] Internal structure: Use multiple cuboids or ellipsoids to build the internal structure, which includes seats, steering wheels and instrument panels; and define the stiffness characteristics of each geometric body based on pre-collected Euro-NCAP test data;

[0106] In MADYMO software, the ground is simplified into a rigid plane, and the friction coefficient between the person and the ground is set to obtain a rigid ground model;

[0107] The step S3 comprises:

[0108] Comparing and verifying the kinematic reconstruction result with the accident video, the contents of the comparison and verification include: the initial collision position between the vehicle and the VRU, the flight trajectory of the VRU in the air, the rotation angle, the first contact position between the body and the ground, the collision sequence between each body part and the ground and the final position;

[0109] The kinematic reconstruction results after comparison and verification are used to obtain the motion response parameters of various parts of the body of the VRU at the moment before the head and the ground collide. The motion response parameters include: the three-dimensional position of the center of gravity of the head of the VRU before the head and the ground collide, the three-axis velocity and angular velocity information, and the motion status of other parts of the VRU body;

[0110] Constructing a THUMS finite element whole human model and adjusting it to be consistent with the posture of the multi-rigid body human model when it is about to hit the ground, loading the motion response parameters into the corresponding parts of the THUMS finite element whole human model, and constructing the human finite element model at that moment;

[0111] In step S4, using finite element analysis software, a surface-to-surface contact algorithm is used as a contact algorithm for the cranial brain tissue in the human finite element model, and a friction coefficient between the brain tissue and the dura mater is set to simulate the relative sliding between the brain tissue and the dura mater when the head collides with the ground;

[0112] Perform head-ground collision simulation based on the human finite element model after contact definition, and calculate multiple head injury criteria based on head kinematics and brain tissue;

[0113] The head injury criteria based on head kinematics include: any one or more of head angular velocity, angular acceleration, linear acceleration, HIC, HIP, BRIC and BrIC;

[0114] The brain tissue-based head injury criteria include: any one or more of von Mises stress, shear stress, positive pressure, negative pressure, MPS, CSDM0.15, CSDM0.25 and DDM;

[0115] In step S6, comprehensively evaluating the predictive ability of each head injury criterion for head injury includes the following steps:

[0116] Compare the calculated value of each head injury criterion with the corresponding preset injury tolerance limit;

[0117] Re-rank the traffic accident cases according to the calculated values ​​to intuitively display the comparison between the calculated values ​​of each head injury criterion and the injury tolerance limit;

[0118] For each head injury criterion, if its calculated value is greater than or equal to the corresponding injury tolerance limit, it is an effective prediction; the number of cases in which each head injury criterion effectively predicts the head injury level is counted, and the prediction accuracy of each head injury criterion is calculated;

[0119] According to the calculated prediction accuracy, the predictive ability of each head injury criterion is divided into four levels:

[0120] A prediction accuracy between 75% and 100% is considered “excellent”; between 50% and 75% is considered “good”; between 25% and 50% is considered “fair”; and between 0% and 25% is considered “poor”.

[0121] In the specific implementation process, Figure 2 As shown, firstly, accident cases with complete information are selected from the high-precision human injury dataset of traffic accidents; the selected cases must have detailed accident scene sketches, scene photos, vehicle information, VRU information, injury information and accident video information at the same time; the video information should clearly record the entire process of the accident, including the movement of the vehicle and VRU before, during and after the collision, especially the kinematic posture at the time of vehicle-VRU collision and VRU head-ground collision; to ensure the pertinence of this embodiment, the VRU head in the selected case should only collide with the ground without direct contact with the vehicle; in addition, the case should also contain a complete head injury record, including the AIS level and specific injury description information;

[0122] For the selection of cases, this embodiment sets strict screening criteria; in terms of accident scene information, it is necessary to include environmental parameters such as the specific time, location, road type, road surface material, and weather conditions of the accident; in terms of vehicle information, it is necessary to record the specific model, year, front structure parameters (including engine hood length BL, engine hood angle BA, engine hood front edge height BLEH, windshield angle WA, and vehicle ground clearance GC, etc.), collision speed, braking deceleration, etc. of the vehicle; in terms of VRU information, it is necessary to include basic information such as age, gender, height, weight, and movement status before the collision (such as walking speed, posture, etc.); in terms of injury information, detailed medical records are required, including head CT / MRI scan images, AIS injury grades of various parts, specific injury types (such as skull fracture SF, subdural hematoma SDH, subarachnoid hematoma SAH, etc.) and their detailed descriptions;

[0123] Specifically, this embodiment selected 10 accident cases with complete information for analysis, including 3 pedestrian accident cases (case numbers C1 to C3) and 7 two-wheeler accident cases (case numbers C4 to C10); the selected cases all have complete accident video information and detailed head injury records, and the VRU head only collides with the ground;

[0124] After completing the case selection, this embodiment uses a two-stage accident reconstruction method to conduct in-depth analysis;

[0125] The first stage is the VRU kinematic reconstruction. In this stage, the CPM (Chalmers Pedestrian Model) pedestrian model is used to reconstruct the entire collision process in the MADYMO software environment. The CPM model is selected based on its highly anthropomorphic knee joint characteristics, which can effectively predict the knee joint kinematic response in side collisions. The CPM pedestrian model has the following advantages:

[0126] 1) Accuracy and reliability of the model;

[0127] High accuracy: The CPM model has been verified many times and can accurately simulate the entire kinematic process of a pedestrian-vehicle collision; this high accuracy is crucial for accident reconstruction, especially when it is necessary to compare and verify with actual video information;

[0128] Well-validated: The CPM model performs well in multi-body dynamics simulations. Its geometric structure and dynamic parameters have been verified through a large number of experiments and simulations, and have high credibility.

[0129] 2) Model flexibility;

[0130] Scalability: The CPM model can easily adjust the gender and height of the pedestrian model through the GEBO (Gender and Body Shape) scaling module to adapt it to the basic information of different VRUs and the body posture at the time of collision. This flexibility enables the model to be applied to a variety of different accident scenarios;

[0131] Modularity: The CPM model can be easily coupled with other vehicle models and ground models to facilitate the construction of complex accident simulation environments;

[0132] 3) The computational efficiency of the model;

[0133] Efficiency: The CPM model has high computational efficiency in multi-body dynamics simulation and can complete complex kinematic reconstruction in a short time, which is very beneficial for large-scale accident reconstruction and multi-condition analysis;

[0134] Real-time: During the accident reconstruction process, it is necessary to quickly verify the consistency between the model's kinematic response and the actual video; the efficiency of the CPM model ensures the smooth progress of this process;

[0135] 4) Scope of application of the model;

[0136] Widely applicable: The CPM model can be used not only for the collision process between pedestrians and vehicles, but also for the kinematic reconstruction of pedestrians landing. This wide applicability enables the model to cover all stages of an accident and provide comprehensive kinematic information.

[0137] 5) Compatibility with finite element models;

[0138] Coupling analysis: When performing in-depth reconstruction of head-ground collisions, the kinematic parameters of the multi-body model need to be transferred to the finite element model; the coupling between the CPM model and the THUMS finite element model is relatively mature, which can ensure the accuracy and consistency of data transfer;

[0139] Although the CPM model has advantages in many aspects, in some cases, other models can also be considered, such as the Madymo pedestrian model, the LS-DYNA pedestrian model, or the GHBMC (Global Human Body Models Consortium) pedestrian model.

[0140] This embodiment uses the GEBOD scaling module to perform personalized adjustments to the 50th standard male model, including: scaling the geometric dimensions according to the gender and height of the VRU in the actual case, modifying the joint stiffness characteristics according to age and weight, and adjusting the model posture to be consistent with the VRU body posture at the time of collision; This embodiment specifically modifies the standard model by:

[0141] 1) Geometry scaling;

[0142] a) Gender and height adjustment: First, the geometric dimensions of the 50th standard male model are scaled using the GEBOD module according to the gender and height of the VRU in the actual case; this step ensures that the height, weight and other geometric parameters of the model are consistent with the actual VRU;

[0143] b) Height adjustment: by adjusting the model's bone length and joint position, the model's height is made consistent with the actual VRU's height;

[0144] c) Gender adjustment: By adjusting the geometry and proportions of the model, the model's morphology is made closer to the gender characteristics of the actual VRU;

[0145] 2) Modification of joint stiffness characteristics;

[0146] a) Age and weight adjustment: The joint stiffness characteristics will be affected by age and weight, so the joint stiffness of the model needs to be corrected according to the age and weight of the actual VRU;

[0147] b) Age correction: As people age, their joint stiffness changes. Generally, the joint stiffness of older people decreases, while that of younger people is relatively high. Therefore, the stiffness parameters of the model joints are adjusted according to the actual age of the VRU. For example, the joint stiffness can be adjusted according to the age range by referring to the existing biomechanical research data. A common practice is to use linear interpolation or empirical formulas to determine the joint stiffness values ​​for different age groups.

[0148] c) Body weight correction: Body weight affects the load and stiffness of joints. Generally, the joint stiffness of individuals with a larger body weight increases, while the joint stiffness of individuals with a smaller body weight decreases. Therefore, the stiffness parameters of the model joints are adjusted according to the actual body weight of the VRU. For example, the empirical relationship between body weight and joint stiffness, such as a linear relationship or a power function relationship, can be used to adjust the joint stiffness.

[0149] In terms of vehicle model construction, in a collision accident, it is necessary to focus on fine modeling of the front structure of the vehicle; first use multiple ellipsoids to construct the front structures such as the engine hood, windshield, and front bumper, and define the stiffness characteristics of each part according to the Euro-NCAP test data (downloaded from the Euro-NCAP official website, which provides a wealth of test data and reports, including vehicle collision test results, scores, and detailed technical reports); for the geometric parameters of each ellipsoid, the length of the major axis of the ellipsoid needs to be accurately set according to the size of the actual vehicle's front structure; the length of the minor axis of the ellipsoid also needs to be accurately set according to the size of the actual vehicle's front structure; the position of the ellipsoid in three-dimensional space needs to be accurately set to ensure that the model is consistent with the actual vehicle's front structure; the direction of the ellipsoid needs to be accurately set according to the orientation of the vehicle's front structure; the stiffness distribution of the ellipsoid needs to be set according to the stiffness distribution of the actual vehicle's front structure to ensure that the stiffness characteristics of the model are consistent with the actual vehicle;

[0150] After the front-end modeling is completed, other parts of the vehicle need to be modeled. In this embodiment, multiple ellipsoids or other geometric bodies are used to construct side structures such as doors and side windows, and the stiffness characteristics of each part are defined; multiple ellipsoids or other geometric bodies are used to construct rear structures such as trunks and rear bumpers, and the stiffness characteristics of each part are defined; multiple rectangular blocks or ellipsoids are used to construct the frame structure, and components such as springs and dampers are used to simulate the dynamic characteristics of the suspension system; cylinders or more complex geometric bodies are used to construct tire models, and the stiffness and friction characteristics of the tires are defined; multiple rectangular blocks or ellipsoids are used to construct internal structures such as seats and dashboards, and the stiffness characteristics of each part are defined; the final modeled vehicle model is as follows: Figure 3 As shown;

[0151] In this embodiment, for the two-wheeled vehicle case, the stiffness characteristics of each part of the two-wheeled vehicle are defined with reference to relevant existing literature; the ground model is simplified to a rigid body, the friction coefficient between the person and the ground is set to 0.58, and the contact friction coefficient between the person and the vehicle is set to 0.2; a constant braking deceleration of 0.8g is applied to the vehicle to simulate the braking effect between the wheel and the ground; the hysteresis slope of all vehicle contact parts is set to 10 8 ;

[0152] The kinematic reconstruction results need to be compared and verified in detail with the video information to ensure that the kinematic reconstruction results of all cases are highly consistent with the actual accident process; the verification content includes: the initial collision position between the vehicle and the VRU, the flight trajectory of the VRU in the air, the rotation angle, the first contact position between the body and the ground, the collision sequence between various parts of the body and the ground, and the final position; video verification has the following functions:

[0153] 1) It can improve the accuracy of reconstruction results;

[0154] Verify boundary conditions: Video information can provide detailed kinematic information before, during and after a collision, including the initial collision location of the vehicle and VRU, rotation angle, collision sequence of body parts with the ground and final position. This information helps verify whether the boundary conditions of the kinematic reconstruction are accurate;

[0155] Correct model parameters: By comparing with video information, potential errors in the model can be discovered, and then the model parameters can be adjusted to improve the accuracy of the reconstruction results;

[0156] 2) Ensure the authenticity and reliability of the model;

[0157] Verify motion response: Video information can intuitively display the motion response of the VRU during a collision, including rotation, translation, and attitude changes. By comparing with video information, it can be ensured that the motion response of the model is consistent with the actual situation;

[0158] Verify the injury mechanism: Video information can provide the posture changes of the VRU during the collision and the moment when the head hits the ground, which is very important for understanding the mechanism of head injury. By comparing, it can be verified whether the model can accurately reproduce these key moments;

[0159] 3) It can enhance the scientificity and credibility of the research;

[0160] Data support: Video information provides real-world data support, making the research results more scientific and credible; by comparing with video information, the effectiveness and applicability of the model can be demonstrated;

[0161] Multi-dimensional verification: Video information can provide multi-dimensional verification, including visual verification, kinematic verification, and dynamic verification, so as to comprehensively evaluate the performance of the model;

[0162] 4) Accident reconstruction methods can be optimized;

[0163] Method improvement: Video verification can reveal deficiencies in existing methods and thus optimize accident reconstruction methods; for example, the model geometry, stiffness properties, or the definition of boundary conditions can be improved;

[0164] Parameter optimization: Video information can provide detailed parameters of the actual collision process, helping to optimize the parameter settings in the model and improve the model's predictive ability;

[0165] In order to quantitatively describe the boundary conditions of head-ground collision, this embodiment counts the three-axis velocity and angular velocity information of the center of gravity of the head before the head-ground collision in 10 cases; wherein, the linear velocity range of the head in the direction perpendicular to the ground is -2.95 to 7.07 m / s, the linear velocity range in the horizontal direction is -3.81 to 2.25 m / s, and the angular velocity range of the head around each axis is -26.60 to 37.37 rad / s;

[0166] The second stage is the reconstruction of head-ground collision injury. In this stage, the THUMSV4.0.2 finite element whole person model is used for in-depth analysis. The relevant material properties of the head and intracranial soft tissues of this model have been fully verified. The shear modulus of the cranial tissue is G0=6kPa and G0=1.2kPa, which are similar to the material properties obtained in the tissue sample test. In this embodiment, the shear modulus "G0=6kPa and G0=1.2kPa" has the following meanings:

[0167] 1) G0 = 6 kPa refers to the shear modulus under static or low-frequency loading conditions. G0 usually represents the shear modulus under static or low-frequency loading conditions (i.e., slow loading rate); this value reflects the mechanical properties of the cranial tissue under slow loading; when the cranial tissue is subjected to slow shear stress, its ability to resist deformation is 6 kPa, which means that when a shear stress of 6 kPa is applied per unit area, the cranial tissue will produce a unit shear strain;

[0168] 2) G0 = 1.2 kPa refers to the shear modulus under dynamic or high-frequency loading conditions. G0 usually represents the shear modulus under dynamic or high-frequency loading conditions (i.e., fast loading rate); this value reflects the mechanical properties of the brain tissue under rapid loading; when the brain tissue is subjected to rapid shear stress, its ability to resist deformation is 1.2 kPa; this means that when a shear stress of 1.2 kPa is applied per unit area, the brain tissue will produce a unit shear strain;

[0169] Brain tissue is a complex biological material, and its shear modulus can change with the change of loading conditions. In addition, its mechanical behavior is time-dependent. Under different loading rates, the shear modulus of brain tissue will be different: 1) Under static or low-frequency loading conditions, the viscoelastic effect of brain tissue is not obvious, and its behavior is closer to elastic materials, so the shear modulus is higher (G0 = 6 kPa); under dynamic or high-frequency loading conditions, the viscoelastic effect of brain tissue is significant, and its behavior is closer to fluid, so the shear modulus is lower (G0 = 1.2 kPa);

[0170] In the finite element model, these shear modulus values ​​are used to define the material properties of the cranial brain tissue, thereby more accurately simulating the response of the cranial brain under different loading conditions; these parameters reflect the mechanical properties of the cranial brain tissue under different loading conditions, which is of great significance for accurately simulating the injury mechanism of the cranial brain during collision; in the finite element model, these parameters are used to define the material properties of the cranial brain tissue, thereby improving the accuracy and reliability of the model;

[0171] In the second stage, the motion response parameters of various parts of the body at the moment before the head-ground collision were first extracted from the kinematic reconstruction of the first stage, including the three-dimensional position, linear velocity and angular velocity of the head, as well as the motion state of the neck and other parts of the body; then the THUMSV4.0.2 finite element whole-person model was adjusted to be consistent with the posture of the CPM before the collision with the ground, and the motion response parameters extracted from the CPM model were loaded into the corresponding parts of the finite element model, thereby achieving accurate reconstruction of the intracranial dynamic response;

[0172] In terms of boundary condition setting, this embodiment takes into account the influence of road material on collision response; for dry asphalt pavement, the friction coefficient between the ground and the human body is set to 0.58; for wet pavement, the friction coefficient is reduced to 0.3; in addition, environmental parameters such as gravity field and contact algorithm need to be set; for the contact definition of cranial tissue, the surface-to-surface contact algorithm is used, and an appropriate friction coefficient is set to simulate the relative sliding between brain tissue and dura mater;

[0173] In the finite element model, the contact definition is very important because it directly affects the simulation results of the model; finite element software (such as LS-DYNA, ABAQUS, etc.) provides a variety of contact algorithms, and the surface-to-surface contact algorithm is one of them; the surface-to-surface contact algorithm is a commonly used finite element contact algorithm used to simulate the interaction between two or more surfaces; the surface-to-surface contact algorithm can accurately simulate the interaction between contact surfaces, especially in complex geometric structures; at the same time, in dynamic simulation, the surface-to-surface contact algorithm usually has better stability and convergence;

[0174] In addition, other contact algorithms can also be considered based on specific research requirements and model complexity. The following are some common alternative algorithms:

[0175] 1) Node-surface contact algorithm: The node-surface contact algorithm defines the contact between the nodes on the contact surface and another surface. It is suitable for scenes with relatively simple contact surfaces and has high computational efficiency. However, in complex geometric structures, problems such as contact penetration may occur, affecting the accuracy of the simulation results.

[0176] 2) Particle-surface contact algorithm: The particle-surface contact algorithm defines the interaction between discrete particles and contact surfaces. It is applicable to granular media or discrete element models and can simulate the contact between particles and surfaces. However, the computational complexity is high and it is only applicable to specific types of models.

[0177] 3) Adaptive contact algorithm: The adaptive contact algorithm automatically adjusts the definition of the contact surface according to the contact state during the simulation process. It is suitable for dynamic simulation with large changes in the contact surface. It can improve the accuracy and stability of the simulation, automatically handle changes in the contact surface, and reduce human intervention.

[0178] The comparison between the multi-rigid body human body model and the finite element model is shown below. Figure 4 As shown;

[0179] Based on the finite element reconstruction results, this embodiment simultaneously calculates multiple injury criteria based on head kinematics and brain tissue; in terms of injury criteria based on head kinematics, head angular velocity, angular acceleration, linear acceleration, HIC, HIP, BRIC and BrIC are calculated; among which, the calculated value range of angular acceleration is 11702~29500rad / s 2 , the calculated value range of linear acceleration is 216~404g, and the calculated value range of HIC value is 966~4808; in terms of injury criteria based on brain tissue, von Mises stress, shear stress, positive pressure, negative pressure, MPS, CSDM0.15, CSDM0.25 and DDM are calculated; among them, the calculated value range of positive pressure is 605~2402kPa, the calculated value range of MPS is 0.34~2.46, and the calculated value range of CSDM0.15 is 3.4%~99.4%;

[0180] This embodiment makes a detailed analysis of the distribution of intracranial injuries. Specifically, this embodiment makes a detailed analysis of the intracranial distribution of the maximum calculated values ​​of the brain tissue injury criterion obtained by finite element reconstruction in 10 cases. As for the distribution law of intracranial pressure, it can be clearly seen from the stress cloud map that positive pressure appears on the collision side and negative pressure appears on the opposite side of the collision. There is a certain similarity in the distribution positions of von Mises stress and shear stress, which is due to a certain conversion ratio relationship between the two stress indicators. In cases C1, C2, C4, C5, C8 and C10, the maximum intracranial shear stress appears in the area adjacent to the anterior lobe of the brain and the falx cerebri. The maximum values ​​of von Mises stress, shear stress and MPS caused by all cases appear in the brain. In addition, except for cases C1 and C6, the distribution positions of von Mises stress, shear stress and MPS in the brain are also the same. As for intracranial pressure, the maximum positive pressure is mainly concentrated in the brain, while the maximum negative pressure is mainly concentrated in the cerebellum and brainstem.

[0181] Finally, this embodiment systematically compares the calculated values ​​of the reconstructed injury criteria with the existing head AIS4+ injury tolerance limits in the existing literature to evaluate the predictive ability of each injury criterion for severe head injury. In the specific evaluation process, each case is first re-ranked according to the size of the calculated value (such as Figure 5 ) to visually display the comparison between the calculated value of each injury criterion and the injury tolerance limit; then count the number of cases in which each head injury criterion effectively predicts severe head injury, that is, the number of cases in which the calculated value exceeds the severe head injury tolerance limit; finally, the prediction ability of each injury criterion is divided into four levels: the prediction accuracy rate is between 75% and 100% as "excellent", between 50% and 75% as "good", between 25% and 50% as "average", and between 0% and 25% as "poor";

[0182] In this embodiment, the number of cases in which the kinematics-based injury criteria effectively predicted severe head injuries were: 4 cases of angular velocity, 10 cases of angular acceleration, 10 cases of linear acceleration, 8 cases of HIC, 4 cases of HIP, 4 cases of BRIC, and 1 case of BrIC; the number of cases in which the brain tissue-based injury criteria effectively predicted severe head injuries were: 0 cases of von Mises stress, 7 cases of shear stress, 10 cases of positive pressure, 6 cases of negative pressure, 9 cases of MPS, 0.158 cases of CSDM, 0.257 cases of CSDM, and 0 cases of DDM; thus, angular acceleration (threshold = 10000rad / s 2 ), linear acceleration (threshold = 200g), HIC (threshold = 1440), normal pressure (threshold = 234kPa), MPS (threshold = 0.9) and CSDM0.15 (threshold = 55%) have an “excellent” ability to predict severe head injury, with prediction accuracy rates above 75%;

[0183] This embodiment proposes a head injury assessment method suitable for ground collision, which fills the gap in existing research; first, this embodiment combines the accident video information to perform detailed accident reconstruction, which can load the collision boundary conditions more accurately, ensure that the kinematic response of the accident reconstruction is highly consistent with the actual situation, and thus improve the credibility of the accident reconstruction; second, this embodiment adopts a two-stage reconstruction method, which fully combines the advantages of the multi-rigid body model and the finite element model, and can simulate the accident process more realistically, especially in the analysis of intracranial dynamic response and injury mechanism in the head-ground collision stage; at the same time, this embodiment not only focuses on the overall injury of the head, but also goes deep into the injury conditions of different areas of the skull. This analysis can provide more detailed information on the injury mechanism and degree of injury, so as to better understand the mechanism of injury and provide more specific data support for improving vehicle safety design and protective measures; in addition, this embodiment also systematically evaluates the effectiveness of multiple head injury criteria in predicting severe head injuries, and gives specific prediction ability levels, providing a scientific basis for the selection of injury criteria.

[0184] Example 3

[0185] like Figure 6 As shown, this embodiment provides that the present invention also provides a ground collision head injury assessment system based on accident reconstruction, which applies the above-mentioned ground collision head injury assessment method based on accident reconstruction, including:

[0186] The case information collection unit 301 is used to obtain accident information and accident video of a traffic accident case, wherein the traffic accident case involves a collision between a VRU and a vehicle, and the head of the VRU only collides with the ground but does not directly contact the vehicle;

[0187] Kinematic reconstruction unit 302: used to establish a multi-rigid human body model, a multi-rigid vehicle model and a rigid ground model based on the accident information, and simulate the entire traffic accident process to obtain the kinematic reconstruction result of the traffic accident case;

[0188] Kinematic verification unit 303: used to compare and verify the kinematic reconstruction result with the accident video, obtain the motion response parameters of various parts of the body at the moment before the VRU head and the ground collide, and construct a finite element model of the human body at that moment;

[0189] A head injury reconstruction unit 304 is used for performing a head-ground collision simulation based on the human body finite element model, and calculating a plurality of head injury criteria based on head kinematics and brain tissue;

[0190] The head injury analysis unit 305 is used to analyze the intracranial injury distribution of the VRU according to the calculated head injury criterion to complete the ground collision head injury evaluation.

[0191] In the specific implementation process, first, the case information collection unit 301 obtains the accident information and accident video of the traffic accident case, where the traffic accident case involves a collision between a VRU and a vehicle, and the head of the VRU only collides with the ground but does not come into direct contact with the vehicle;

[0192] Then the kinematic reconstruction unit 302 establishes a multi-rigid human model, a multi-rigid vehicle model and a rigid ground model based on the accident information, and simulates the entire traffic accident process to obtain the kinematic reconstruction result of the traffic accident case;

[0193] Afterwards, the kinematic verification unit 303 compares and verifies the kinematic reconstruction result with the accident video, obtains the motion response parameters of each body part of the VRU head and the ground before the collision, and constructs a finite element model of the human body at that moment;

[0194] Then, the head injury reconstruction unit 304 performs a head-ground collision simulation based on the human body finite element model, and calculates various head injury criteria based on head kinematics and brain tissue;

[0195] Finally, the head injury analysis unit 305 analyzes the intracranial injury distribution of the VRU according to the calculated head injury criterion to complete the ground collision head injury evaluation;

[0196] This system combines accident video information to perform detailed accident reconstruction, which can load collision boundary conditions more accurately and ensure that the kinematic response of accident reconstruction is highly consistent with the actual situation, thereby improving the credibility of accident reconstruction; in addition, this system adopts a two-stage reconstruction method, which fully combines the advantages of multi-rigid body model and finite element model, and can simulate the accident process more realistically, especially in the analysis of intracranial dynamic response and injury mechanism in the head-ground collision stage.

[0197] The same or similar reference numerals correspond to the same or similar components;

[0198] The terms used in the drawings to describe positional relationships are only used for illustrative purposes and should not be construed as limiting this patent;

[0199] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A ground collision head injury assessment method based on accident reconstruction, characterized in that: The following steps are involved: S1: Acquire accident information and an accident video of a traffic accident case, wherein the traffic accident case involves a collision between a VRU and a vehicle, and the head of the VRU only collides with the ground but does not directly contact the vehicle; S2: establishing a multi-rigid human body model, a multi-rigid vehicle model and a rigid ground model based on the accident information, and performing a simulation of the entire traffic accident process to obtain a kinematic reconstruction result of the traffic accident case; S3: Compare and verify the kinematic reconstruction result with the accident video, obtain the motion response parameters of various parts of the body at the moment before the VRU head and the ground collide, and construct a finite element model of the human body at that moment; S4: performing a head-ground collision simulation based on the human finite element model, and calculating a plurality of head injury criteria based on head kinematics and brain tissue; S5: Analyze the intracranial injury distribution of the VRU according to the calculated head injury criterion to complete the ground impact head injury evaluation.

2. The method for evaluating head injuries in ground collisions based on accident reconstruction according to claim 1, characterized in that: In step S1, the accident information includes: accident scene environment information, scene photos, vehicle information, VRU information and head injury information; The video information records the entire process of the traffic accident case, including the movement patterns of the vehicle and the VRU before, during and after the collision, the collision moment of the vehicle and the VRU, and the kinematic posture of the VRU head when it collides with the ground.

3. The method for evaluating head injuries in ground collisions based on accident reconstruction according to claim 2, characterized in that: The accident scene environmental information includes: the specific time, location, road type, road surface material and weather conditions of the accident; The vehicle information includes: the specific model, production year, front structure parameters, collision speed and braking deceleration of the vehicle; the front structure parameters include engine hood length BL, engine hood angle BA, engine hood front edge height BLEH, windshield angle WA and vehicle ground clearance GC; The VRU information includes: basic physiological information of the VRU, and the motion state of the VRU before the collision with the vehicle; The head injury information includes: the head injury medical record of the VRU, the medical record includes a head CT / MRI scan image, and the AIS injury grade of each part of the head, the specific injury type and its detailed description.

4. The method for evaluating head injuries in ground collisions based on accident reconstruction according to claim 1, characterized in that: In step S2, the multi-rigid body human body model includes any one of a CPM pedestrian model, a Madymo pedestrian model, a LS-DYNA pedestrian model and a GHBMC pedestrian model, and the CPM pedestrian model is most preferred; The multi-rigid body vehicle model includes a front structure, a side structure, a rear structure, a chassis structure, tires and an internal structure.

5. The method for evaluating head injuries in ground collisions based on accident reconstruction according to claim 4, characterized in that: In the step S2, the entire traffic accident process is simulated using the multi-rigid body dynamics analysis software MADYMO to obtain the kinematic reconstruction result of the traffic accident case; In the MADYMO software, the multi-rigid body human model is obtained by performing personalized adjustment on the TNO 50 percentile standard male model; the personalized adjustment includes: scaling the geometric dimensions according to the gender and height of the VRU in the traffic accident case, modifying the stiffness characteristics of each joint according to the age and weight of the VRU, and adjusting the model posture to be consistent with the VRU body posture at the time of collision; In the MADYMO software, constructing the multi-rigid body vehicle model includes: Front structure: The front structure is constructed using multiple ellipsoids and the stiffness characteristics of each geometric body are defined based on pre-collected Euro-NCAP test data; the front structure includes the vehicle's hood, front windshield and front bumper; Side structure: The side structure is constructed using multiple ellipsoids or cuboids or cylinders, and the stiffness characteristics of each geometric body are defined based on pre-collected Euro-NCAP test data; the side structure includes doors and side windows; Rear structure: The rear structure is constructed using multiple ellipsoids or cuboids or cylinders, and the stiffness characteristics of each geometric body are defined according to the pre-collected Euro-NCAP test data; the rear structure includes the trunk and the rear bumper; Chassis structure: Use multiple ellipsoids or cuboids to build the frame structure, and use spring or damper elements to simulate the dynamic characteristics of the suspension system to build the chassis structure; Tires: Build tire models using cylinders or custom-shaped geometry and define the tire's stiffness and friction characteristics; Internal structure: Use multiple cuboids or ellipsoids to build the internal structure, which includes seats, steering wheels and instrument panels; and define the stiffness characteristics of each geometric body based on pre-collected Euro-NCAP test data; In MADYMO software, the ground is simplified into a rigid plane, and the friction coefficient between people and the ground is set to obtain a rigid ground model.

6. The method for evaluating head injuries in ground collisions based on accident reconstruction according to claim 1, characterized in that: The step S3 comprises: Comparing and verifying the kinematic reconstruction result with the accident video, the contents of the comparison and verification include: the initial collision position between the vehicle and the VRU, the flight trajectory of the VRU in the air, the rotation angle, the first contact position between the body and the ground, the collision sequence between each body part and the ground, and the final position; The kinematic reconstruction results after comparison and verification are used to obtain the motion response parameters of various parts of the body of the VRU at the moment before the head and the ground collide. The motion response parameters include: the three-dimensional position of the center of gravity of the head of the VRU before the head and the ground collide, the three-axis velocity and angular velocity information, and the motion status of other parts of the VRU body; A THUMS finite element whole human model is constructed and adjusted to be consistent with the posture of the multi-rigid body human model when it is about to hit the ground, the motion response parameters are loaded into the corresponding parts of the THUMS finite element whole human model, and the human finite element model at that moment is constructed.

7. The method for evaluating head injuries in ground collisions based on accident reconstruction according to claim 1, characterized in that: In step S4, using finite element analysis software, a surface-to-surface contact algorithm is used as a contact algorithm for the cranial brain tissue in the human finite element model, and a friction coefficient between the brain tissue and the dura mater is set to simulate the relative sliding between the brain tissue and the dura mater when the head collides with the ground; Perform head-ground collision simulation based on the human finite element model after contact definition, and calculate multiple head injury criteria based on head kinematics and brain tissue; The head injury criteria based on head kinematics include: any one or more of head angular velocity, angular acceleration, linear acceleration, HIC, HIP, BRIC and BrIC; The brain tissue-based head injury criteria include: any one or more of von Mises stress, shear stress, positive pressure, negative pressure, MPS, CSDM0.15, CSDM0.25 and DDM.

8. A method for evaluating head injuries in ground collisions based on accident reconstruction according to any one of claims 1 to 7, characterized in that: The method further comprises: S6: Repeat steps S1 to S5 to evaluate ground collision head injuries in multiple traffic accident cases, and comprehensively evaluate the predictive ability of various head injury criteria for head injuries.

9. The method for evaluating head injuries in ground collisions based on accident reconstruction according to claim 8, characterized in that: In step S6, comprehensively evaluating the predictive ability of each head injury criterion for head injury includes the following steps: Compare the calculated value of each head injury criterion with the corresponding preset injury tolerance limit; Re-rank the traffic accident cases according to the calculated values ​​to intuitively display the comparison between the calculated values ​​of each head injury criterion and the injury tolerance limit; For each head injury criterion, if its calculated value is greater than or equal to the corresponding injury tolerance limit, it is an effective prediction; the number of cases in which each head injury criterion effectively predicts the head injury level is counted, and the prediction accuracy of each head injury criterion is calculated; According to the calculated prediction accuracy, the predictive ability of each head injury criterion is divided into four levels: A prediction accuracy between 75% and 100% is considered "excellent"; between 50% and 75% is considered "good"; between 25% and 50% is considered "fair"; and between 0% and 25% is considered "poor".

10. A ground collision head injury assessment system based on accident reconstruction, using a ground collision head injury assessment method based on accident reconstruction as claimed in any one of claims 1 to 9, characterized in that: include: Case information collection unit: used to obtain accident information and accident video of a traffic accident case, wherein the traffic accident case involves a collision between a VRU and a vehicle, and the head of the VRU only collides with the ground without direct contact with the vehicle; Kinematic reconstruction unit: used to establish a multi-rigid human body model, a multi-rigid vehicle model and a rigid ground model based on the accident information, and simulate the entire traffic accident process to obtain the kinematic reconstruction result of the traffic accident case; Kinematic verification unit: used to compare and verify the kinematic reconstruction result with the accident video, obtain the motion response parameters of various parts of the body at the moment before the VRU head and the ground collide, and construct a finite element model of the human body at that moment; A head injury reconstruction unit: used for performing head-ground collision simulation based on the human finite element model, and calculating a variety of head injury criteria based on head kinematics and brain tissue; Head injury analysis unit: used to analyze the intracranial injury distribution of the VRU according to the calculated head injury criteria, and complete the ground collision head injury evaluation.

Images