A method of pedestrian head injury assessment for a vehicle and related apparatus

Abstract

The present disclosure relates to a pedestrian head injury evaluation method of a vehicle and related equipment. The method comprises: obtaining a simulation model and impact information of a vehicle to be evaluated; inputting the simulation model and the impact information into a pre-trained evaluation model to obtain a pedestrian head injury of the vehicle to be evaluated output by the evaluation model, wherein the evaluation model is used to extract metal component features and non-metal component features of the vehicle to be evaluated from the simulation model and the impact information, and obtain the pedestrian head injury according to the metal component features and the non-metal component features. The method provided by the present disclosure extracts the features of metal components and non-metal components respectively, differentiates the collision features according to the material categories, represents the material mechanics properties of the vehicle in collision which is closer to the actual situation, and then obtains more accurate pedestrian head injury results, thereby improving the efficiency of vehicle safety testing.

Description

Technical Field

[0001] The present disclosure relates to the field of data processing technology, and in particular to a method for assessing pedestrian head injuries in a vehicle and related equipment. Background Art

[0002] Global automotive safety tests have clear requirements for pedestrian protection. Head Injury Criterion (HIC) is usually used as an evaluation standard to measure the degree of head injury to pedestrians in a collision. During the automotive development process, the test requirements for pedestrian protection in automotive safety tests must be met. Summary of the Invention

[0003] In order to solve the above technical problems, the present disclosure provides a pedestrian head injury assessment method and related equipment for a vehicle.

[0004] The present disclosure provides a method for assessing pedestrian head injuries in a vehicle, comprising: obtaining a simulation model and collision information of a vehicle to be assessed; inputting the simulation model and the collision information into a pre-trained assessment model to obtain the pedestrian head injury of the vehicle to be assessed output by the assessment model, wherein the assessment model is used to extract metal component features and non-metal component features of the vehicle to be assessed from the simulation model and the collision information, and obtain the pedestrian head injury based on the metal component features and the non-metal component features; obtaining the pedestrian head injury based on the metal component features and the non-metal component features, comprising: constructing a three-dimensional matrix of the head of the vehicle to be assessed, wherein the three-dimensional matrix of the head includes a plurality of voxel units; mapping the metal component features and the non-metal component features to corresponding voxel units to obtain a structured representation matrix of the vehicle to be assessed; and obtaining the pedestrian head injury based on the structured representation matrix.

[0005] Optionally, the above-mentioned metal component characteristics include the yield strength of the metal component; the above-mentioned metal component characteristics of the vehicle to be evaluated are extracted from the above-mentioned simulation model and the above-mentioned impact information, including: extracting the yield strength of the metal component of the vehicle to be evaluated from the above-mentioned simulation model and the above-mentioned impact information by using the 0.2% offset method combined with the tangent modulus change rate; the above-mentioned non-metallic component characteristics include the peak strength and fracture toughness of the non-metallic component; the above-mentioned non-metallic component characteristics of the vehicle to be evaluated are extracted from the above-mentioned simulation model and the above-mentioned impact information, including: extracting the peak strength of the non-metallic component of the vehicle to be evaluated from the above-mentioned simulation model and the above-mentioned impact information by using the strain softening slope method, and extracting the fracture toughness of the non-metallic component of the vehicle to be evaluated from the above-mentioned simulation model and the above-mentioned impact information by using the energy absorption rate.

[0006] Optionally, extracting the above-mentioned metal component features of the above-mentioned vehicle to be evaluated from the above-mentioned simulation model and the above-mentioned collision information also includes: using logarithmic normalization to process the above-mentioned non-metallic component features of the above-mentioned vehicle to be evaluated from the above-mentioned simulation model and the above-mentioned collision information also includes: using maximum and minimum value normalization to process the above-mentioned non-metallic component features.

[0007] Optionally, the above-mentioned metal component characteristics include the elastic modulus and proportional limit of the metal components of the above-mentioned vehicle to be evaluated in the elastic stage, the yield strength, hardening index and work hardening rate of the above-mentioned metal components in the plastic stage, and the fracture strain of the above-mentioned metal components in the fracture stage; the above-mentioned non-metallic component characteristics include the elastic modulus and proportional limit of the non-metallic components of the above-mentioned vehicle to be evaluated in the elastic stage, the peak strength and work hardening rate of the above-mentioned non-metallic components in the plastic stage, and the fracture strain and fracture toughness of the above-mentioned non-metallic components in the fracture stage.

[0008] Optionally, the above-mentioned mapping of the above-mentioned metal component features and the above-mentioned non-metal component features into corresponding voxel units to obtain the above-mentioned structured representation matrix of the vehicle to be evaluated includes: mapping the above-mentioned metal component features and the above-mentioned non-metal component features into corresponding voxel units, and assigning different weights to at least two different voxel units to obtain the above-mentioned structured representation matrix of the above-mentioned vehicle to be evaluated.

[0009] Optionally, mapping the metal component features and the non-metal component features to corresponding voxel units to obtain a structured representation matrix of the vehicle to be evaluated also includes: in response to any voxel unit not being mapped to a metal component feature or a non-metal component feature, defining the voxel unit as an empty unit.

[0010] Based on the same inventive concept, the present disclosure also provides a vehicle pedestrian head injury assessment device, including: an information acquisition module, used to obtain a simulation model and collision information of the vehicle to be assessed; an assessment module, used to input the above-mentioned simulation model and the above-mentioned collision information into a pre-trained assessment model, and obtain the pedestrian head injury of the above-mentioned vehicle to be assessed output by the above-mentioned assessment model, wherein the above-mentioned assessment model is used to extract the metal component features and non-metal component features of the above-mentioned vehicle to be assessed from the above-mentioned simulation model and the above-mentioned collision information, and obtain the above-mentioned pedestrian head injury based on the above-mentioned metal component features and the above-mentioned non-metal component features; wherein, obtaining the pedestrian head injury based on the metal component features and the above-mentioned non-metal component features includes: constructing a three-dimensional head matrix of the vehicle to be assessed, wherein the three-dimensional head matrix includes multiple voxel units; mapping the metal component features and the non-metal component features to corresponding voxel units to obtain a structured representation matrix of the vehicle to be assessed; and obtaining the pedestrian head injury based on the structured representation matrix.

[0011] Based on the same inventive concept, the disclosure further provides an electronic device, comprising: a processor; a memory for storing executable instructions; wherein the processor is used to read the executable instructions from the memory and execute the executable instructions to realize any one of the above methods.

[0012] Based on the same inventive concept, the disclosure further provides a computer readable storage medium having a computer program stored thereon, characterized in that the storage medium stores a computer program, when the computer program is executed by a processor, the processor realizes any one of the above methods.

[0013] The technical solution provided by the disclosure has the following advantages compared with the prior art: the method provided by the disclosure extracts the features of the metal parts and the non-metal parts respectively, differentiates the collision features according to the material categories, represents the material mechanical properties of the vehicle in the collision which are closer to the actual situation, and then obtains more accurate pedestrian head injury results, thereby improving the efficiency of vehicle safety testing. Moreover, the method provided by the disclosure proposes a fusion material collision feature and three-dimensional voxel representation method, which can represent the coupling effect of the material mechanical properties and the spatial structure of the vehicle in the collision, makes the evaluation model more sensitive to the differences and changes of material and structure details, and solves the problem that the coupling of material mechanical properties and spatial structure is difficult to represent in related technologies. BRIEF DESCRIPTION OF DRAWINGS

[0014] The accompanying drawings, which are incorporated into and form part of the specification, illustrate embodiments consistent with the present disclosure and, together with the specification, serve to explain the principles of the present disclosure.

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure or the prior art, the drawings needed in the embodiment or prior art description will be briefly introduced as follows, and obviously, other drawings can also be obtained by those skilled in the art without creative labor.

[0016] Figure 1 A flowchart of a pedestrian head injury evaluation method provided by the embodiment of the present disclosure is shown in the figure;

[0017] Figure 2 A vehicle head impact point position diagram provided by the embodiment of the present disclosure is shown in the figure;

[0018] Figure 3 A pedestrian head injury data cloud diagram of the vehicle head provided by the embodiment of the present disclosure is shown in the figure;

[0019] Figure 4A schematic diagram of a process for obtaining pedestrian head injury based on metal component characteristics and non-metal component characteristics in a pedestrian head injury assessment method provided by an embodiment of the present disclosure;

[0020] Figure 5 A simplified three-dimensional matrix diagram of a vehicle provided in an embodiment of the present disclosure;

[0021] Figure 6 A schematic structural diagram of a pedestrian head injury assessment device provided in an embodiment of the present disclosure;

[0022] Figure 7 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0023] In order to more clearly understand the above-mentioned purposes, features and advantages of the embodiments of the present disclosure, the scheme of the embodiments of the present disclosure will be further described below. It should be noted that the embodiments of the present disclosure and the features therein can be combined with each other in the absence of conflict.

[0024] In the following description, many specific details are set forth to facilitate a full understanding of the embodiments of the present disclosure, but the embodiments of the present disclosure may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present disclosure, not all of the embodiments.

[0025] Related technologies use AI models to intelligently predict pedestrian protection head collision injuries in vehicles. The input values ​​of the AI ​​model include some spatial parameters and material parameters extracted from the vehicle's simulation model. However, in the vehicle's simulation model, the materials used for different structural parts are quite different, and the material properties during the collision involve complex collision relationships. Different material categories (metal / plastic / glass materials, etc.) have significant differences in dynamic properties during collision. The related technologies use the static properties of the material to define the mechanical properties of the material during the collision, which cannot be used as an effective input for the AI ​​model, resulting in poor prediction of vehicle head injuries by the AI ​​model and inability to be applied to predict new vehicle models.

[0026] In view of this, an embodiment of the present disclosure provides a method for assessing pedestrian head injuries in a vehicle, such as Figure 1 Shown, including:

[0027] S1. Obtain a simulation model and collision information of a vehicle to be evaluated.

[0028] Specifically, the simulation model of the vehicle may only include a simulation model of the vehicle head, and the collision information includes the position, angle, and speed of the collision of the vehicle.

[0029] S2. Input the simulation model and the impact information into a pre-trained evaluation model to obtain the pedestrian head injury of the vehicle to be evaluated as output by the evaluation model, wherein the evaluation model is used to extract the metal component features and non-metal component features of the vehicle to be evaluated from the simulation model and the impact information, and obtain the pedestrian head injury based on the metal component features and the non-metal component features.

[0030] Metallic and non-metallic materials exhibit different collision mechanical properties during a collision. Metallic materials have a distinct elastic phase and yield plateau, while non-metallic materials have a shorter elastic phase and may directly enter a strain softening or hardening phase after yielding, without a distinct yield plateau. The method provided in the disclosed embodiments extracts the characteristics of metallic and non-metallic components separately, differentiating the collision characteristics based on material type. This characterizes the material mechanical properties of vehicles during a collision that are more realistic, thereby obtaining more accurate pedestrian head injury results and improving the efficiency of vehicle safety testing.

[0031] Specifically, the pedestrian head injury obtained by the above-mentioned evaluation model is the pedestrian head injury corresponding to the impact information of the input impact point. During specific implementation, it is necessary to execute the method of the above-mentioned embodiment of the present disclosure multiple times to obtain the pedestrian head injury at multiple impact points of the vehicle to be evaluated. Figure 2 FIG2 shows multiple pedestrian protection head impact points of a vehicle that is required to be evaluated by a regulation in an embodiment. Figure 3 A cloud diagram of pedestrian head injury data for each impact point obtained using the method provided in the above embodiment of the present disclosure is shown, wherein the horizontal axis Pedestrian headforms (pedestrian head impact point), the vertical axis WAD on centerline (collision point grid line determined according to safety regulations), and multiple grids define multiple impact points. The score on the grid is the pedestrian head injury score for that impact point, and the darker the color, the higher the score. In specific implementation, the simulation model of the vehicle to be evaluated includes material information for each component of the vehicle, which can be obtained from a pre-defined standard material library. The material information includes the material of each component and the corresponding mechanical properties of the material.

[0032] In specific implementation, the simulation model of the vehicle to be evaluated that is input into the trained evaluation model is a normalized finite element simulation model. Specifically, the materials used for each component in the finite element simulation model are named normalized, and the format of the simulation model input into the evaluation model is unified to adapt to the evaluation model.

[0033] In some embodiments, as Figure 4 As shown, in the above S2, the process of obtaining the pedestrian head injury based on the characteristics of the metal parts and the non-metal parts includes:

[0034] S21. Construct a three-dimensional head matrix of the vehicle to be evaluated, wherein the three-dimensional head matrix includes a plurality of voxel units. Figure 5 A simplified three-dimensional matrix diagram including four voxel units is shown.

[0035] S22. Map the metal component features and the non-metal component features to corresponding voxel units to obtain a structured representation matrix for the vehicle to be evaluated. Specifically, the metal component features and the non-metal component features must first be normalized to obtain a standardized collision feature vector. The standardized collision feature vector is then mapped to the corresponding voxel units to obtain a structured representation matrix for the vehicle to be evaluated.

[0036] S23. Obtain pedestrian head injury based on the structured representation matrix.

[0037] The method provided in the embodiment of the present disclosure proposes a method for integrating material collision characteristics with three-dimensional voxel characterization, which can characterize the coupling effects of the material mechanical properties and spatial structure of a vehicle during a collision, making the evaluation model more sensitive to the differences in material and structural details, and solving the problem of difficulty in characterizing the coupling of material mechanical properties and spatial structure in related technologies.

[0038] In specific implementation, the above method includes: 1) first, defining an n×n×n three-dimensional matrix that can completely cover the entire vehicle head finite element simulation model to ensure that all parts of the model are included and no component details are lost; 2) setting a certain voxel resolution, where each element in the matrix represents a cube with a side length of a, i.e., a voxel unit. The size of the voxel resolution can be adjusted according to actual needs to adapt to different accuracy and computing resource requirements; 3) performing multi-directional geometric detection on each voxel unit, such as Figure 5 As shown, a detection line is emitted from the voxel unit along the x / y / z axis direction. When the detection line intersects with the finite element simulation model, the collision feature vector generated by the intersecting component in the previous step is extracted and mapped to the corresponding voxel unit. The classification code of the intersecting component is extracted and also mapped to the corresponding voxel unit.

[0039] In some embodiments, the metal component characteristics include the yield strength of the metal component. In S2, the process of extracting the metal component characteristics of the vehicle to be evaluated from the simulation model and the collision information by the evaluation model includes:

[0040] The yield strength of the metal parts of the vehicle to be evaluated is extracted from the simulation model and impact information using the 0.2% offset method combined with the tangent modulus change rate.

[0041] In specific implementation, the above process includes: extracting the impact stress-strain curve of the metal component and determining the linear segment in the elastic stage; drawing an offset line parallel to the linear segment at 0.2% of the strain axis; detecting the point where the tangent modulus change rate significantly changes; and taking the stress value corresponding to the intersection of the offset line and the curve near this point as the yield strength. Because metal materials have obvious elastic-plastic transition characteristics, their stress-strain curves obey Hooke's law in the elastic stage and undergo irreversible deformation after entering the plastic stage. This characteristic makes the 0.2% offset method suitable for determining the yield point. In addition, the microstructural characteristics of metal materials, such as the significant change in the tangent modulus change rate caused by dislocation movement, are also closely related to the yield point. In contrast, the stress-strain curves of non-metallic materials usually do not have a clear elastic-plastic transition stage, and their microstructure is complex, making it difficult to accurately find the yield point using similar methods.

[0042] In some embodiments, the non-metallic component characteristics include peak strength and fracture toughness of the non-metallic component. In S2, the process of extracting the non-metallic component characteristics of the vehicle to be evaluated from the simulation model and the collision information includes:

[0043] The peak strength of the non-metallic parts of the vehicle to be evaluated is extracted from the simulation model and the impact information using the strain softening slope method, and the fracture toughness of the non-metallic parts of the vehicle to be evaluated is extracted from the simulation model and the impact information using the energy absorption rate.

[0044] In specific implementation, the above process includes: first, extracting the impact stress-strain curve of a non-metallic component, locating the stress peak point and identifying the subsequent strain-softening region. The slope of the softening region is calculated (when the absolute value of the slope exceeds a preset threshold), and the stress value corresponding to the peak point is determined as the peak intensity. The area under the curve from the peak point to the fracture point is then calculated as the energy absorption rate. The ratio of this value to the fracture strain determines the fracture toughness, which is used to characterize the dynamic mechanical properties of the component during a collision. Non-metallic materials (such as glass, plastic, and rubber used in automotive development) typically exhibit different mechanical behavior from metals. During loading, they often exhibit a significant strain softening phenomenon: after reaching peak stress, the stress decreases with increasing strain, rather than entering a stable plastic deformation phase like metals. Furthermore, the fracture process of non-metallic materials is often accompanied by energy absorption, and the energy absorption rate can reflect the toughness or brittleness of the material during the fracture process. In contrast, metals generally do not have a distinct strain-softening phase, and their energy absorption is primarily concentrated in the plastic deformation phase, making their characteristics difficult to characterize using the strain-softening slope and energy absorption rate. Therefore, the strain softening slope method combined with the energy absorption rate is more suitable for extracting the characteristics of non-metallic materials and can more accurately reflect their mechanical behavior during the destruction process.

[0045] In some embodiments, the metal component features include the elastic modulus and proportional limit of the metal component of the vehicle to be evaluated in the elastic stage, the yield strength of the metal component in the plastic stage, the hardening index and work hardening rate, and the fracture strain of the metal component in the fracture stage; and the non-metal component features include the elastic modulus and proportional limit of the non-metal component of the vehicle to be evaluated in the elastic stage, the peak strength and work hardening rate of the non-metal component in the plastic stage, and the fracture strain and fracture toughness of the non-metal component in the fracture stage.

[0046] The elastic stage, the plastic stage and the fracture stage are three important stages in material mechanics for describing the process of material deformation under force. The elastic stage refers to the reversible elastic deformation of the material under the action of external force, and the stress and strain are linearly related. When the external force is removed, the material can completely recover to its original state. When the external force exceeds the elastic limit, the material enters the plastic stage, at which point irreversible plastic deformation occurs. When the stress reaches the yield strength, the material continues to deform but the stress hardly increases any more, and cold work hardening phenomenon appears with the increase of deformation degree. Finally, when the external force continues to increase beyond the bearing capacity of the material, the material enters the fracture stage, brittle fracture or ductile fracture occurs, and the material is completely destroyed. These three stages reflect the whole process from material deformation under force to final destruction, and are an important basis for engineering design and material selection.

[0047] It can be understood that although the method of the embodiments of the present disclosure needs to extract some same feature parameters for metal components and non-metal components, the processes and methods for extracting features are different, such as the method in the above embodiments. Moreover, for the feature parameters for which the specific extraction method is not described, the person skilled in the art can select the corresponding method in the related art based on the embodiments of the present disclosure for extraction, which is within the protection scope of the present disclosure.

[0048] In some embodiments, the process of extracting the metal component features of the vehicle to be evaluated from the simulation model and the impact information further includes:

[0049] The nonlinear metal component features are processed by logarithmic normalization to generate a standardized collision feature vector, which is used for subsequent generation of a structured characterization matrix.

[0050] Specifically, logarithmic normalization is a data processing method mainly used to convert data into a relatively unified range while preserving the logarithmic characteristics of the data. The logarithmic transformation conforms to the physical law of metal hardening "fast first and slow later", which makes the evaluation model more easily capture the early hardening characteristics of the metal component. In specific implementation, the metal component features can be processed by logarithmic normalization according to the following formula:

[0051]

[0052] wherein, is the original eigenvalue (i.e., the metal component feature that has not been normalized), is the normalized eigenvalue, The value range of the feature.

[0053] In some embodiments, the process of extracting features of non-metallic components of the vehicle to be evaluated from the simulation model and the collision information further includes:

[0054] The maximum and minimum value normalization is used to process the features of non-metallic parts and generate standardized collision feature vectors for the subsequent generation of structured representation matrix.

[0055] Min-Max Normalization is a data preprocessing method that aims to scale data to a fixed range. This method is implemented through linear transformation and can preserve the original distribution characteristics of the data, thereby preserving the strain softening characteristics of non-metallic component features. In specific implementation, the following formula can be used to refer to the maximum and minimum normalization processing of metal component features:

[0056]

[0057] in, is the original eigenvalue (i.e., the non-metallic component feature that has not been normalized), is the normalized eigenvalue, The value range of the feature.

[0058] In some embodiments, the above S22 specifically includes:

[0059] The metal component features and the non-metal component features are mapped to corresponding voxel units, and different weights are assigned to at least two different voxel units to obtain a structured representation matrix of the vehicle to be evaluated.

[0060] Since different vehicle components have different effects on pedestrian head injuries, distinguishing the degree of influence of key structural components and secondary structural components in collision deformation can make the final pedestrian head injury results closer to the actual value. Increasing the weight of key components can also make the evaluation model more sensitive to differences in materials and structural details.

[0061] Specifically, the above method includes: locating the position of each impact point in the three-dimensional matrix of the head according to the impact information, taking the local area of ​​the relevant area around the position, and the size of the local representation matrix p, which can be adjusted according to different actual conditions; among the components corresponding to all voxel units contained in the local representation matrix, applying enhancement weights to the voxel units belonging to the key components , the voxel units of other components are given a benchmark weight , Greater than .

[0062] In a specific embodiment, the local area of ​​the relevant area around the position includes: taking x-20 to x+20 in the x direction, taking y-20 to y+20 in the y direction, and taking z-30 to z+10 in the z direction, with a size of 40×40×40 local matrix; among the components corresponding to all voxel units included in the local representation matrix, the weight coefficient assigned to the voxel units belonging to the key components is: applying an enhanced weight to the hood outer panel =3.0, applies enhanced weight to the hood inner panel =2.5, to apply enhanced weight to the reinforcement plate =2.0, and the voxel units of other parts are given the base weight =1.0.

[0063] After performing a three-dimensional voxel representation of the features of various vehicle components, the matrices stored using traditional matrix storage methods will occupy a large amount of memory space. For example, when using npy files (NumPy binary format) to store the above structured representation matrix, if the matrix includes a large number of zero elements, the matrix will be too large, and the data storage will take up a lot of space, further resulting in excessive consumption of evaluation model computing resources and low efficiency. In view of this, the embodiments of the present disclosure provide a new matrix storage method. Specifically, the above S22 also includes:

[0064] In response to any voxel cell not being mapped to a metal component feature or a non-metal component feature, the voxel cell is defined as an empty cell.

[0065] The above embodiments of the present disclosure define empty cells, so that the structured representation matrix that is finally constructed and stored is very small, thereby reducing the consumption of data storage and computing resources and time.

[0066] It can be understood that the embodiments provided by the present disclosure have a high degree of freedom and scalability. For example, the information meaning and voxel resolution of the elements within the above-mentioned voxel units, as well as the weight coefficients of the above-mentioned parts, can be arbitrarily adjusted.

[0067] Specifically, before executing the above-mentioned vehicle pedestrian head injury assessment method, the embodiment of the present disclosure further provides a method for training an assessment model, including:

[0068] Acquire simulation models, collision information, and pedestrian head injuries obtained through experiments or simulation calculations of multiple vehicles to be evaluated; extract metal component features and non-metal component features of the vehicles to be evaluated from the simulation models and collision information; and train the simulation model using the simulation models, collision information, metal component features, non-metal component features, and pedestrian head injuries of the vehicles to be evaluated.

[0069] In a specific implementation, metal component features and non-metal component features (which may be the structured representation matrix in the above embodiment) are used as the main input data sets of the evaluation model, and pedestrian head injuries are used as the output data set of the evaluation model.

[0070] The method also includes: partitioning the input dataset into training, validation, and test sets proportionally; using a deep neural network algorithm to automatically extract features from the structured representation matrix of the impact points; incorporating an attention mechanism to further increase the importance of key features, allowing the model to more closely focus on them; and using a fully connected neural network for regression output. Finally, a deep evaluation model is trained to achieve accuracy on the validation set that exceeds a certain threshold. Finally, model hyperparameters are adjusted based on the evaluation model's performance on the test set.

[0071] In a specific embodiment, the above method includes:

[0072] Using finite element simulation models of approximately 5,000 points on the head of 25 vehicle models, impact point data, and simulation results, pedestrian head injury values ​​at the impact point were determined. Data from four vehicle models were randomly selected as the test set, with 70% of the remaining data used as the training set and 30% as the validation set, which was randomly shuffled. A deep convolutional neural network algorithm and an attention mechanism were used to automatically extract features from the local weighted representation matrix of the impact point, while a three-layer fully connected neural network was used for regression output. The entire deep learning model was trained, and the model was considered valid when the prediction accuracy on the test set exceeded 90%. Otherwise, the model was optimized and hyperparameters were adjusted.

[0073] During deep learning model training, accuracy is defined using the following formula:

[0074]

[0075] Where y is the predicted value, yt is the HIC value calculated by simulation, and n is the number of data sets used for prediction.

[0076] The HIC value is obtained through simulation calculation of the finite element model, and the calculation formula is shown as follows:

[0077]

[0078] It should be noted that the method of the embodiment of the present application can be performed by a single device, such as a computer or server. The method of this embodiment can also be applied in a distributed scenario, where multiple devices cooperate with each other to complete the method. In such a distributed scenario, one of the multiple devices may only perform one or more steps of the method of the embodiment of the present application, and the multiple devices will interact with each other to complete the above method.

[0079] It should be noted that the above description is limited to some embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in an order different from that described in the above embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0080] Based on the same inventive concept, corresponding to any of the above embodiments and methods, the present application also provides a vehicle pedestrian head injury assessment device, such as Figure 6 Shown, including:

[0081] An information acquisition module 10 is used to obtain a simulation model and collision information of a vehicle to be evaluated;

[0082] The evaluation module 20 is used to input the simulation model and the impact information into a pre-trained evaluation model to obtain the pedestrian head injury of the vehicle to be evaluated output by the evaluation model, wherein the evaluation model is used to extract the metal component features and non-metal component features of the vehicle to be evaluated from the simulation model and the impact information, and obtain the pedestrian head injury based on the metal component features and the non-metal component features; wherein the pedestrian head injury is obtained based on the metal component features and the non-metal component features, including: constructing a three-dimensional head matrix of the vehicle to be evaluated, wherein the three-dimensional head matrix includes multiple voxel units; mapping the metal component features and the non-metal component features to corresponding voxel units to obtain a structured representation matrix of the vehicle to be evaluated; and obtaining the pedestrian head injury based on the structured representation matrix.

[0083] The device provided by the embodiment of the present disclosure extracts the characteristics of metal parts and non-metal parts separately, differentiates the collision characteristics according to the material category, and characterizes the material mechanical properties of the vehicle during the collision that are closer to the actual situation, thereby obtaining more accurate pedestrian head injury results and improving the efficiency of vehicle safety testing. In addition, the method provided by the embodiment of the present disclosure proposes a method that integrates material collision characteristics with three-dimensional voxel characterization, which can characterize the coupling effect of the material mechanical properties and spatial structure of the vehicle during the collision, making the evaluation model more sensitive to the differences in material and structural details, and solving the problem of difficulty in characterizing the coupling of material mechanical properties and spatial structure in related technologies.

[0084] For the convenience of description, the above devices are described as being divided into various modules according to their functions. Of course, when implementing this application, the functions of each module can be implemented in the same or multiple software and / or hardware.

[0085] The device of the above embodiment is used to implement the corresponding pedestrian head injury assessment method in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be repeated here.

[0086] Figure 7 A schematic structural diagram of an electronic device provided by an embodiment of the present disclosure is shown.

[0087] like Figure 7 As shown, the electronic device may include a processor 1101 and a memory 1102 storing computer program instructions.

[0088] Specifically, the processor 1101 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0089] Memory 1102 may include a large-capacity memory for information or instructions. By way of example and not limitation, memory 1102 may include a hard disk drive (HDD), a floppy disk drive, flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 1102 may include removable or non-removable (or fixed) media. Where appropriate, memory 1102 may be internal or external to the integrated gateway device. In a specific embodiment, memory 1102 is a non-volatile solid-state memory. In a specific embodiment, memory 1102 includes a read-only memory (ROM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically alterable ROM (EAROM), or flash memory, or a combination of two or more of these.

[0090] The processor 1101 reads and executes the computer program instructions stored in the memory 1102 to perform the steps of the pedestrian head injury assessment method provided in the embodiment of the present disclosure.

[0091] In one example, the electronic device may further include a transceiver 1103 and a bus 1104. Figure 7 As shown, the processor 1101 , the memory 1102 and the transceiver 1103 are connected via a bus 1104 and communicate with each other.

[0092] The bus 1104 may include hardware, software, or both. By way of example, and not limitation, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a HyperTransport (HT) interconnect, an Industrial Standard Architecture (ISA) bus, an InfiniBand interconnect, a Low Pin Count (LPC) bus, a memory bus, a MicroChannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local Bus (VLB) bus, or other suitable buses, or a combination of two or more of these. Where appropriate, the bus 1104 may include one or more buses. Although embodiments herein describe and illustrate a particular bus, this application contemplates any suitable bus or interconnect.

[0093] The following is an embodiment of a computer-readable storage medium provided in an embodiment of the present disclosure. The computer-readable storage medium and the pedestrian head injury assessment method of the above-mentioned embodiments belong to the same inventive concept. For details not fully described in the embodiment of the computer-readable storage medium, please refer to the embodiment of the above-mentioned pedestrian head injury assessment method.

[0094] This embodiment provides a storage medium containing computer-executable instructions. When the computer-executable instructions are executed by a computer processor, they are used to perform a pedestrian head injury assessment method.

[0095] Of course, the storage medium containing computer-executable instructions provided in an embodiment of the present disclosure is not limited to the above method operations, and its computer-executable instructions can also execute related operations in the pedestrian head injury assessment method provided in any embodiment of the present disclosure.

[0096] Through the above description of the implementation methods, those skilled in the art can clearly understand that the present disclosure can be implemented with the help of software and necessary general-purpose hardware. Of course, it can also be implemented through hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present disclosure, or the part that contributes to the existing technology, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., and includes a number of instructions for enabling a computer cloud platform (which can be a personal computer, server, or network cloud platform, etc.) to execute the pedestrian head injury assessment method provided by each embodiment of the present disclosure.

[0097] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises", "includes" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprises a..." do not exclude the presence of other identical elements in the process, method, article or device that includes the above elements.

[0098] The foregoing are merely specific embodiments of the present disclosure, intended to enable those skilled in the art to understand and implement the present disclosure. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure is not limited to the foregoing embodiments, but is intended to be construed in the broadest manner consistent with the principles and novel features disclosed herein.

Claims

1. A method for assessing pedestrian head injuries in a vehicle, characterized in that: include: Obtain the simulation model and collision information of the vehicle to be evaluated; Inputting the simulation model and the collision information into a pre-trained evaluation model to obtain a pedestrian head injury of the vehicle to be evaluated as output by the evaluation model, wherein the evaluation model is used to extract metal component features and non-metal component features of the vehicle to be evaluated from the simulation model and the collision information, and obtain the pedestrian head injury based on the metal component features and the non-metal component features; Obtaining the pedestrian head injury according to the metal component characteristics and the non-metal component characteristics includes: Constructing a three-dimensional head matrix of the vehicle to be evaluated, wherein the three-dimensional head matrix includes a plurality of voxel units; Mapping the metal component features and the non-metal component features into corresponding voxel units to obtain a structured representation matrix of the vehicle to be evaluated; Obtaining the pedestrian head injury according to the structured characterization matrix; The metal component feature includes the yield strength of the metal component; extracting the metal component feature of the vehicle to be evaluated from the simulation model and the collision information includes: Extracting the yield strength of the metal parts of the vehicle to be evaluated from the simulation model and the impact information using a 0.2% offset method combined with a tangent modulus change rate; The non-metallic component characteristics include peak strength and fracture toughness of the non-metallic component; extracting the non-metallic component characteristics of the vehicle to be evaluated from the simulation model and the collision information includes: extracting the peak strength of the non-metallic component of the vehicle to be evaluated by using a strain softening slope method from the simulation model and the impact information, and extracting the fracture toughness of the non-metallic component of the vehicle to be evaluated by using an energy absorption rate from the simulation model and the impact information; Extracting the metal component features of the vehicle to be evaluated from the simulation model and the collision information further includes: Using logarithmic normalization to process nonlinear characteristics of the metal component; Extracting the non-metallic component features of the vehicle to be evaluated from the simulation model and the collision information further includes: The non-metallic component features are processed by using maximum and minimum value normalization; The metal component characteristics include the elastic modulus and proportional limit of the metal component of the vehicle to be evaluated in the elastic stage, the yield strength, hardening index and work hardening rate of the metal component in the plastic stage, and the fracture strain of the metal component in the fracture stage; The non-metallic component characteristics include the elastic modulus and proportional limit of the non-metallic component of the vehicle to be evaluated in the elastic stage, the peak strength and work hardening rate of the non-metallic component in the plastic stage, and the fracture strain and fracture toughness of the non-metallic component in the fracture stage.

2. The method according to claim 1, characterized in that Mapping the metal component features and the non-metal component features into corresponding voxel units to obtain a structured representation matrix of the vehicle to be evaluated includes: The metal component features and the non-metal component features are mapped into corresponding voxel units, and different weights are assigned to at least two different voxel units to obtain the structured representation matrix of the vehicle to be evaluated.

3. The method according to claim 1, characterized in that Mapping the metal component features and the non-metal component features into corresponding voxel units to obtain a structured representation matrix of the vehicle to be evaluated includes: In response to any of the voxel cells not being mapped to the metal component feature or the non-metal component feature, the voxel cell is defined as an empty cell.

4. A pedestrian head injury assessment device for a vehicle, using the method according to claim 1, characterized in that: include: An information acquisition module, used to obtain the simulation model and collision information of the vehicle to be evaluated; An evaluation module is used to input the simulation model and the collision information into a pre-trained evaluation model to obtain the pedestrian head injury of the vehicle to be evaluated output by the evaluation model, wherein the evaluation model is used to extract the metal component features and non-metal component features of the vehicle to be evaluated from the simulation model and the collision information, and obtain the pedestrian head injury based on the metal component features and the non-metal component features; wherein obtaining the pedestrian head injury based on the metal component features and the non-metal component features includes: constructing a three-dimensional head matrix of the vehicle to be evaluated, wherein the three-dimensional head matrix includes multiple voxel units; mapping the metal component features and the non-metal component features to corresponding voxel units to obtain a structured representation matrix of the vehicle to be evaluated; and obtaining the pedestrian head injury based on the structured representation matrix.

5. An electronic device, characterized in that: include: processor; A memory for storing executable instructions; wherein the processor is used to read the executable instructions from the memory and execute the executable instructions to implement the method according to any one of claims 1 to 3.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the processor implements the method according to any one of claims 1 to 3.

Images