Pedestrian head collision injury value prediction method and device, equipment and storage medium
By extracting the structural feature values of the vehicle model and using neural network models to predict the head collision damage value of pedestrians, the problem of time-consuming and difficult traditional simulation methods is solved, and fast and accurate design evaluation is achieved, and the automotive design process is simplified.
Patent Information
- Application Number
- CN202411571484.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-06
- Publication Date
- 2024-12-03
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Traditional pedestrian protection performance evaluation methods rely on complex simulation software and high technical thresholds, and have a long calculation time, hindering the rapid iteration of automotive design.
By obtaining the model data of the vehicle to be tested, the structural feature values near the impact point are extracted, the structural feature images are formed, and the neural network model is used for training to predict the head collision damage value of pedestrians.
Fast and accurate prediction of pedestrian head collision damage value is achieved, reducing calculation costs and design iteration time, and simplifying the design process.
Smart Images

Figure CN119066997A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the field of artificial intelligence, and in particular to a method, device, equipment and storage medium for predicting pedestrian head collision injury values. Background Art
[0002] Pedestrian protection is an important part of automobile safety performance and is also an important part of testing standards such as C-NCAP (China-New Car Assessment Program), US NCAP (US New Car Assessment Program), and Euro NCAP (European New Car Assessment Program). When a vehicle collides with a pedestrian, it is necessary to ensure the safety of the pedestrian as much as possible, especially the head safety. In the development process of the front part of the car (including the bumper and headlights, etc.) and the hood, not only the modal and stiffness performance should be considered, but also the consideration of pedestrian protection.
[0003] Traditional pedestrian protection performance assessments are mostly conducted using CAE (Computer Aided Engineering) technology. CAE is an approximate numerical analysis method that uses computers to assist in solving complex engineering and product structural strength, stiffness, buckling stability, dynamic response, heat conduction, three-dimensional multi-body contact, elastic-plastic analysis and other mechanical properties, as well as the optimization design of structural performance. However, the use of this method requires reliance on professional simulation software tools and requires certain simulation knowledge and technical thresholds. In addition, a single vehicle model often needs to calculate and process the simulation and results of hundreds of impact points, which takes a long time to calculate and requires a large amount of work, which seriously hinders the rapid iteration of automobile design. Summary of the invention
[0004] The present disclosure provides a method, device, equipment and storage medium for predicting pedestrian head collision injury value, so as to at least solve the above technical problems existing in the prior art.
[0005] According to a first aspect of the present disclosure, a method for predicting pedestrian head collision injury value is provided, the method comprising: Acquiring model data of a vehicle to be tested, wherein the vehicle to be tested includes a plurality of impact points; Taking each of the impact points as the center, obtaining a structural characteristic value within a preset range; According to the structural feature values, forming a structural feature image corresponding to each of the impact points; Train the neural network model to obtain a prediction model; The structural feature image is input into the prediction model to determine the pedestrian head collision injury value corresponding to each impact point.
[0006] In one possible implementation manner, taking each of the impact points as the center and obtaining a structural characteristic value within a preset range includes: Respectively obtain the structural characteristic values of the hood inner panel, hood outer panel and connecting parts of the vehicle to be tested; The structural feature values of the hair cover inner plate, the hair cover outer plate and the connecting member are respectively matched with a channel in the RGB image to form a first image; For the first image, taking each of the impact points as the center, a structural feature value within a preset range is obtained.
[0007] In one possible implementation manner, the training of the neural network model to obtain the prediction model includes: Acquire model data of a plurality of sample vehicles, each of the sample vehicles comprising a plurality of sample impact points; Taking each sample impact point as the center, obtaining a structural characteristic value within a preset range; According to the structural feature values, a sample structural feature image corresponding to each of the sample impact points is formed; Obtaining a pedestrian head collision injury value corresponding to each of the sample impact points; The neural network model is trained according to the sample structural feature image corresponding to each sample impact point and the pedestrian head collision injury value to obtain a prediction model.
[0008] In one possible implementation manner, the training of the neural network model to obtain the prediction model includes: Dividing the model data of the plurality of sample vehicles into a training data set and a verification data set, wherein the training data set is used to train the neural network model, and the verification data set is used to verify the accuracy of the trained neural network model; According to the training data set and the verification data set, a cross entropy loss function is used to determine a neural network model with the highest verification accuracy with the highest accuracy as the goal; The neural network model with the highest validation accuracy is determined as the trained prediction model.
[0009] In one embodiment, the structural characteristic value includes at least one of a maximum value, a minimum value, a median value, a mean value, a variance or a standard deviation.
[0010] According to a second aspect of the present disclosure, a pedestrian head collision injury value prediction device is provided, the device comprising: A first acquisition unit is configured to acquire model data of a vehicle to be tested, wherein the vehicle to be tested includes a plurality of impact points; A second acquisition unit is configured to acquire a structural characteristic value within a preset range with each of the impact points as the center; an image forming unit configured to form a structural feature image corresponding to each of the impact points according to the structural feature value; A training unit, configured to train the neural network model to obtain a prediction model; The determination unit is configured to input the structural feature image into the prediction model to determine the pedestrian head collision injury value corresponding to each of the impact points.
[0011] In one possible implementation, the second acquisition unit includes: A first sub-acquisition unit is configured to respectively acquire structural characteristic values of a hood inner panel, a hood outer panel, and a connecting member of the vehicle to be tested; A first sub-image forming unit is configured to correspond the structural feature values of the hair cover inner plate, the hair cover outer plate and the connecting member to a channel in the RGB image respectively to form a first image; The second sub-acquisition unit is configured to acquire a structural feature value within a preset range with each of the impact points as the center of the first image.
[0012] In one possible implementation, the training unit includes: A third sub-acquisition unit is configured to acquire model data of a plurality of sample vehicles, each of the sample vehicles comprising a plurality of sample impact points; A fourth sub-acquisition unit is configured to acquire a structural characteristic value within a preset range with each of the sample impact points as the center; A second sub-image forming unit, configured to form a sample structure characteristic image corresponding to each of the sample impact points according to the structure characteristic value; a fifth sub-acquisition unit, configured to acquire a pedestrian head collision injury value corresponding to each of the sample collision points; The first sub-training unit is configured to train the neural network model according to the sample structure feature image corresponding to each of the sample impact points and the pedestrian head collision injury value to obtain a prediction model.
[0013] According to a third aspect of the present disclosure, there is provided an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method described in the present disclosure.
[0014] According to a fourth aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to cause the computer to execute the method described in the present disclosure.
[0015] The disclosed pedestrian head collision injury value prediction method, device, equipment and storage medium obtain the model data of the vehicle to be tested and extract the structural feature values near the collision point to form a structural feature image, and input the structural feature image into the prediction model to obtain the pedestrian head collision injury value corresponding to each collision point. The disclosed embodiment has a good accuracy rate and extremely short prediction time, effectively solving the problem of high time cost of traditional simulation. The method only needs to input the relevant model of the vehicle to be tested, without complex configuration, and effectively solves the problem of high difficulty of traditional simulation.
[0016] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] The above and other objects, features and advantages of the exemplary embodiments of the present disclosure will become readily understood by reading the detailed description below with reference to the accompanying drawings. In the accompanying drawings, several embodiments of the present disclosure are shown in an exemplary and non-limiting manner, in which: In the drawings, the same or corresponding reference numerals represent the same or corresponding parts.
[0018] Figure 1 A flowchart of a method for predicting pedestrian head collision injury value provided by an embodiment of the present disclosure; Figure 2 It is a structural schematic diagram of the vehicle to be tested and the impact point; Figure 3 is the first image of the vehicle to be tested; Figure 4 for Figure 3 Schematic diagram of each impact point in the figure; Figure 5 The structural feature image corresponding to each impact point; Figure 6 It is a real distribution diagram of the danger level of pedestrian head collision injury value corresponding to the impact point; Figure 7 Loss curve graph for model training; Figure 8 A predicted distribution diagram of the danger level of pedestrian head collision injury value corresponding to the impact point; Fig. 9 A schematic diagram of the structure of a pedestrian head collision injury value prediction device provided by an embodiment of the present disclosure; Fig.10 A schematic diagram of the structure of an electronic device according to an embodiment of the present disclosure is shown. DETAILED DESCRIPTION
[0019] In order to make the purpose, features, and advantages of the present disclosure more obvious and easy to understand, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present disclosure.
[0020] The present disclosure provides a method for predicting pedestrian head collision injury value. Figure 1 A flowchart of a method for predicting pedestrian head collision injury value provided by an embodiment of the present disclosure, such as Figure 1 As shown, the method comprises the following steps: Step 101 : obtaining model data of a vehicle to be tested, where the vehicle to be tested includes a plurality of impact points.
[0021] Specifically, the car model mesh vertices can be obtained from the .key file containing the car model mesh and vertices generated by the LS_Prepost software and converted into point cloud data, that is, the model data of the vehicle to be tested, and the impact point position and collision damage value of the corresponding vehicle to be tested can be obtained from the .csv file containing the simulation results of the LS_Run software.
[0022] Figure 2 It is a schematic diagram of the structure of the vehicle to be tested and the impact point. Figure 2 As shown, the vehicle to be tested includes multiple impact points, and the different colors of each impact point also represent different pedestrian head collision injury values corresponding to the impact point.
[0023] Step 102, taking each impact point as the center, obtaining a structural characteristic value within a preset range.
[0024] In one embodiment, taking each impact point as the center, obtaining a structural characteristic value within a preset range includes: Respectively obtain the structural characteristic values of the hood inner panel, hood outer panel and connecting parts of the vehicle to be tested; The structural feature values of the hair cover inner plate, the hair cover outer plate and the connecting member are respectively matched with a channel in the RGB image to form a first image; For the first image, taking each impact point as the center, a structural feature value within a preset range is obtained.
[0025] The hood inner panel, hood outer panel and connecting parts are important components of the hood, and the image of the hood is constructed through the structural feature values of the hood inner panel, hood outer panel and connecting parts.
[0026] Specifically, the median of the structural height of each component within a certain range in the hood inner panel, the hood outer panel and the connecting member can be extracted as the structural feature value, and then the structural feature values of the hood inner panel, the hood outer panel and the connecting member are respectively corresponded to a channel in the RGB image. For example, the structural feature value of the hood inner panel is corresponded to the red channel in the RGB image, the structural feature value of the connecting member is corresponded to the green channel in the RGB image, and the structural feature value of the hood outer panel is corresponded to the blue channel in the RGB image, thereby forming a first image of the vehicle to be tested (see Figure 3 ), then in the first image, taking each collision point as the center, obtain the structural feature value within a preset range, for example, the first image is segmented within a range of 350×350 pixels with the collision point as the center to form multiple Figure 4 The image shown, then Figure 4 The structural feature value is extracted once every 7×7 pixels.
[0027] In one embodiment, the structural characteristic value includes at least one of a maximum value, a minimum value, a median value, a mean value, a variance, or a standard deviation. For example, the median value of vertices within a certain range can be extracted as the structural characteristic value.
[0028] Step 103: forming a structural feature image corresponding to each impact point according to the structural feature value.
[0029] Figure 5 The structural feature image corresponding to each impact point. Figure 5 As shown, for example, in the previous step, after extracting the structural feature value every 7×7 pixels within a range of 350×350 pixels with the impact point as the center, a structural feature image with an image size of 50×50 pixels can be formed.
[0030] In the disclosed embodiment, the structural feature values are stored as images. In other embodiments, the structural feature values may also be stored as videos, numerical values, arrays, matrices, lists, or dictionaries.
[0031] Step 104: train the neural network model to obtain a prediction model.
[0032] In one embodiment, training a neural network model to obtain a prediction model includes: Acquire model data of a plurality of sample vehicles, each sample vehicle including a plurality of sample impact points; Taking each sample impact point as the center, obtain the structural characteristic value within the preset range; According to the structural characteristic values, a sample structural characteristic image corresponding to each sample impact point is formed; Obtain the pedestrian head collision injury value corresponding to each sample impact point; According to the sample structural feature images corresponding to each sample impact point and the pedestrian head collision injury value, a neural network model is trained to obtain a prediction model.
[0033] Specifically, the method is the same as that of step 101 to step 103. First, the model data of multiple sample vehicles and the positions of sample impact points are obtained, and a sample structural feature image corresponding to each sample impact point is formed; then, the pedestrian head collision injury value corresponding to each impact point is obtained from the .csv file. The injury value can be calculated according to the C-NCAP management rules, and the injury value is converted into a danger level score. Table 1 is a correspondence table of pedestrian head collision injury values and danger level scores.
[0034] Table 1
[0035] The damage value in the disclosed embodiment may also be calculated according to the management rules of US NCAP (United States), Euro NCAP (Europe), JNCAP (Japan), Australia (ANCAP), South Korea (KNCAP), South America (Latin NCAP), and ASEAN (ASEAN NCAP).
[0036] Figure 6 is the real distribution diagram of the danger level of pedestrian head collision injury value corresponding to the impact point, compared with Figure 6 With Table 1, the danger level score of each impact point can be obtained.
[0037] It should be explained that after obtaining the sample structure characteristic image corresponding to each sample impact point, the sample structure characteristic image can be stored in a folder corresponding to the score.
[0038] Next, the neural network model is trained based on the sample structural feature image and pedestrian head collision injury value corresponding to each sample impact point to obtain a prediction model. Specifically, the sample structural feature image and the danger level score are input into the neural network model to train the neural network model.
[0039] In one embodiment, training a neural network model to obtain a prediction model includes: The model data of multiple sample vehicles are divided into a training data set and a verification data set, the training data set is used to train the neural network model, and the verification data set is used to verify the accuracy of the trained neural network model; Based on the training data set and the validation data set, the cross entropy loss function is used to determine the neural network model with the highest validation accuracy with the highest accuracy as the goal; The neural network model with the highest validation accuracy is determined as the trained prediction model.
[0040] Specifically, 90% of the data is set as the training data set, 10% of the data is set as the validation data set, the number of iterations is set to 125, the batch size is set to 5, and the cross entropy loss function is used as the loss function. The sample structure feature image and the danger level score corresponding to each sample impact point are read and passed into the AlexNet network model for training. The training loss is as follows: Figure 7 As shown. The model with the highest accuracy is saved during the training process. In this embodiment, after five experiments, the results are obtained, and the accuracy of the best model on the validation data set is 83.3%, and the average accuracy is 80.0%.
[0041] The embodiments of the present disclosure may also adopt other neural network models, such as GoogLeNet, UNet, FNO, VGG or ResNet network models.
[0042] Step 105 , input the structural feature image into the prediction model to determine the pedestrian head collision injury value corresponding to each impact point.
[0043] After the prediction model is obtained, the structural feature image corresponding to the impact point of the vehicle to be tested is input into the prediction model to obtain the danger level score of the pedestrian head collision injury value corresponding to the input impact point, and finally drawn as follows Figure 8 The predicted distribution diagram of the dangerous degree of pedestrian head collision injury value corresponding to the impact point is shown.
[0044] In the disclosed embodiment, it takes about 0.1 seconds to predict the danger level score corresponding to a single impact point. With the time for extracting structural features, it takes about 2 seconds for a single impact point, which is much lower than the time required for traditional CAE to solve a single impact point. Using the method of the present invention instead of the traditional CAE method can save the calculation cost of pedestrian head injuries, shorten the iteration time and R&D cycle of automobile product design, and improve the competitiveness of products.
[0045] The disclosed method is based on the theory of machine learning algorithms of two-dimensional images, and quickly predicts the pedestrian head collision injury value of the input car model. Compared with the traditional finite element simulation method, it greatly reduces the use threshold and improves the calculation efficiency. Compared with other machine learning algorithms based on three-dimensional geometry, the disclosed method is more concise and efficient in data processing and model use. The method disclosed in the present invention is based on the input vehicle model data to be tested and the impact point to be evaluated, and takes each of the impact points as the center, obtains the three-dimensional structural feature value within a preset range and converts it into a corresponding structural feature image. Through the prediction model disclosed in the present invention, the pedestrian head collision injury value corresponding to each of the impact points can be quickly predicted. The disclosed method greatly improves the prediction efficiency of pedestrian head collision injury value, effectively reduces the calculation threshold of pedestrian head collision injury value, accelerates the rapid iteration of automobile design, and shortens the product development cycle.
[0046] The disclosed embodiment also provides a pedestrian head collision injury value prediction device, Fig. 9 The schematic diagram of the structure of the pedestrian head collision injury value prediction device provided by the embodiment of the present disclosure is as follows: Fig. 9 As shown, the pedestrian head collision injury value prediction device includes: A first acquisition unit 901 is configured to acquire model data of a vehicle to be tested, where the vehicle to be tested includes a plurality of impact points; The second acquisition unit 902 is configured to acquire a structural characteristic value within a preset range with each impact point as the center; An image forming unit 903, configured to form a structural feature image corresponding to each impact point according to the structural feature value; A training unit 904 is configured to train the neural network model to obtain a prediction model; The determination unit 905 is configured to input the structural feature image into the prediction model to determine the pedestrian head collision injury value corresponding to each impact point.
[0047] In one embodiment, the second obtaining unit 902 includes: The first sub-acquisition unit 9021 is configured to respectively acquire structural characteristic values of the hood inner panel, hood outer panel and connecting parts of the vehicle to be tested; A first sub-image forming unit 9022 is configured to correspond the structural feature values of the hair cover inner plate, the hair cover outer plate and the connecting member to a channel in the RGB image respectively to form a first image; The second sub-acquisition unit 9023 is configured to acquire a structural feature value within a preset range with each impact point as the center for the first image.
[0048] In one embodiment, the training unit 904 includes: The third sub-acquisition unit 9041 is configured to acquire model data of a plurality of sample vehicles, each sample vehicle including a plurality of sample impact points; The fourth sub-acquisition unit 9042 is configured to acquire a structural characteristic value within a preset range with each sample impact point as the center; A second sub-image forming unit 9043 is configured to form a sample structure feature image corresponding to each sample impact point according to the structure feature value; A fifth sub-acquisition unit 9044 is configured to acquire a pedestrian head collision injury value corresponding to each sample collision point; The first sub-training unit 9045 is configured to train a neural network model according to the sample structure feature image corresponding to each sample impact point and the pedestrian head collision injury value to obtain a prediction model.
[0049] In one embodiment, the training unit 904 is specifically configured to divide the model data of the plurality of sample vehicles into a training data set and a verification data set, the training data set is used to train the neural network model, and the verification data set is used to verify the accuracy of the trained neural network model; Based on the training data set and the validation data set, the cross entropy loss function is used to determine the neural network model with the highest validation accuracy with the highest accuracy as the goal; The neural network model with the highest validation accuracy is determined as the trained prediction model.
[0050] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device and a readable storage medium.
[0051] Fig.10 A schematic block diagram of an example electronic device 1000 that can be used to implement an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.
[0052] like Fig.10 As shown, the device 1000 includes a computing unit 1001, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 1002 or a computer program loaded from a storage unit 1008 into a random access memory (RAM) 1003. In the RAM 1003, various programs and data required for the operation of the device 1000 can also be stored. The computing unit 1001, the ROM 1002, and the RAM 1003 are connected to each other via a bus 1004. An input / output (I / O) interface 1005 is also connected to the bus 1004.
[0053] A number of components in the device 1000 are connected to the I / O interface 1005, including: an input unit 1006, such as a keyboard, a mouse, etc.; an output unit 1007, such as various types of displays, speakers, etc.; a storage unit 1008, such as a disk, an optical disk, etc.; and a communication unit 1009, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 1009 allows the device 1000 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.
[0054] The computing unit 1001 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 1001 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 1001 performs the various methods and processes described above, such as a pedestrian head collision injury value prediction method. For example, in some embodiments, the pedestrian head collision injury value prediction method may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 1008. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 1000 via the ROM 1002 and / or the communication unit 1009. When the computer program is loaded into the RAM 1003 and executed by the computing unit 1001, one or more steps of the pedestrian head collision injury value prediction method described above may be performed. Alternatively, in other embodiments, the computing unit 1001 may be configured to execute the pedestrian head impact injury value prediction method in any other appropriate manner (for example, by means of firmware).
[0055] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.
[0056] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.
[0057] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or device, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
[0058] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).
[0059] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.
[0060] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.
[0061] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.
[0062] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of the present disclosure, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.
[0063] The above is only a specific embodiment of the present disclosure, but the protection scope of the present disclosure is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present disclosure, which should be included in the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure should be based on the protection scope of the claims.
Claims
1. A method for predicting pedestrian head collision injury value, characterized in that: The method comprises: Acquiring model data of a vehicle to be tested, wherein the vehicle to be tested includes a plurality of impact points; Taking each of the impact points as the center, obtaining a structural characteristic value within a preset range; According to the structural feature values, forming a structural feature image corresponding to each of the impact points; Train the neural network model to obtain a prediction model; The structural feature image is input into the prediction model to determine the pedestrian head collision injury value corresponding to each impact point.
2. The method according to claim 1, characterized in that The obtaining of structural characteristic values within a preset range with each of the impact points as the center includes: Respectively obtain the structural characteristic values of the hood inner panel, hood outer panel and connecting parts of the vehicle to be tested; The structural feature values of the hair cover inner plate, the hair cover outer plate and the connecting member are respectively matched with a channel in the RGB image to form a first image; For the first image, taking each of the impact points as the center, a structural feature value within a preset range is obtained.
3. The method according to claim 1, characterized in that: The neural network model is trained to obtain a prediction model, including: Acquire model data of a plurality of sample vehicles, each of the sample vehicles comprising a plurality of sample impact points; Taking each sample impact point as the center, obtaining a structural characteristic value within a preset range; According to the structural feature values, a sample structural feature image corresponding to each of the sample impact points is formed; Obtaining a pedestrian head collision injury value corresponding to each of the sample impact points; The neural network model is trained according to the sample structural feature image corresponding to each sample impact point and the pedestrian head collision injury value to obtain a prediction model.
4. The method according to claim 3, characterized in that The neural network model is trained to obtain a prediction model, including: Dividing the model data of the plurality of sample vehicles into a training data set and a verification data set, wherein the training data set is used to train the neural network model, and the verification data set is used to verify the accuracy of the trained neural network model; According to the training data set and the verification data set, a cross entropy loss function is used to determine a neural network model with the highest verification accuracy with the highest accuracy as the goal; The neural network model with the highest validation accuracy is determined as the trained prediction model.
5. The method according to claim 1, characterized in that The structural characteristic value includes at least one of a maximum value, a minimum value, a median value, a mean value, a variance or a standard deviation.
6. A pedestrian head collision injury value prediction device, characterized in that: The device comprises: A first acquisition unit is configured to acquire model data of a vehicle to be tested, wherein the vehicle to be tested includes a plurality of impact points; A second acquisition unit is configured to acquire a structural characteristic value within a preset range with each of the impact points as the center; an image forming unit configured to form a structural feature image corresponding to each of the impact points according to the structural feature value; A training unit, configured to train the neural network model to obtain a prediction model; The determination unit is configured to input the structural feature image into the prediction model to determine the pedestrian head collision injury value corresponding to each of the impact points.
7. The device according to claim 6, characterized in that The second acquisition unit includes: A first sub-acquisition unit is configured to respectively acquire structural characteristic values of a hood inner panel, a hood outer panel, and a connecting member of the vehicle to be tested; A first sub-image forming unit is configured to correspond the structural feature values of the hair cover inner plate, the hair cover outer plate and the connecting member to a channel in the RGB image respectively to form a first image; The second sub-acquisition unit is configured to acquire a structural feature value within a preset range with each of the impact points as the center of the first image.
8. The device according to claim 6, characterized in that The training unit comprises: A third sub-acquisition unit is configured to acquire model data of a plurality of sample vehicles, each of the sample vehicles comprising a plurality of sample impact points; A fourth sub-acquisition unit is configured to acquire a structural characteristic value within a preset range with each of the sample impact points as the center; A second sub-image forming unit, configured to form a sample structure characteristic image corresponding to each of the sample impact points according to the structure characteristic value; a fifth sub-acquisition unit, configured to acquire a pedestrian head collision injury value corresponding to each of the sample collision points; The first sub-training unit is configured to train the neural network model according to the sample structure feature image corresponding to each of the sample impact points and the pedestrian head collision injury value to obtain a prediction model.
9. An electronic device, characterized in that: include: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 5.
10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to make a computer execute the method according to any one of claims 1-5.
Citation Information
Patent Citations
Pedestrian protection system under autonomous traffic
CN115601955A
Method and system for analyzing head injury of pedestrian in pedestrian-vehicle collision
CN116229543A
Pedestrian protection-oriented head collision severity prediction method
CN117113826A
Automobile active energy absorption device based on pedestrian classification and recognition and control method thereof
CN117125020A
Pedestrian head collision injury prediction method based on neural network
CN117131731A
Cited By
Pedestrian head injury value prediction method and system based on multi-modal data interaction
WO2026145709A1
