Passenger injury prediction method and device, electronic equipment and storage medium

Abstract

The invention provides a passenger injury prediction method and device, electronic equipment and a storage medium, and the method comprises the steps: determining the discontinuous positions of two stages in a pulse curve according to an obtained original working condition collision pulse, so as to generate a reconstructed collision pulse curve corresponding to a plurality of collision working conditions; performing second-order half-sine wave fitting processing on the reconstructed collision pulse curve to obtain a second-order half-sine simplified pulse curve, and extracting to obtain interpretable waveform features; the interpretable waveform features and the obtained constraint system parameters are input into a pre-constructed passenger injury prediction model to output and obtain a passenger injury result, and the passenger injury prediction model is obtained through training of the interpretable waveform features, the constraint system parameters and the passenger injury result. According to the method, high-precision and strong-generalization passenger injury prediction can be realized, so that the engineering application bottleneck of dependence on small sample data and a black box model in the prior art is solved.

Description

Technical Field

[0001] The present application belongs to the field of automobile passive safety technology, and in particular relates to a method, device, electronic device and storage medium for predicting occupant injury. Background Art

[0002] With increasingly stringent automotive safety regulations and growing demand for intelligent driving, the design of passive safety systems urgently needs to shift from an experience-driven approach to a data-driven one. Traditional occupant injury prediction relies on real-vehicle crash tests and finite element simulations, but these methods' high costs and long lead times (a single simulation cycle can take dozens of hours) severely restrict design iteration efficiency. Furthermore, existing methods capture a limited amount of crash pulse data (typically covering only a few standard operating conditions required by regulations), resulting in insufficient generalization of trained occupant injury prediction models. This can lead to significant errors, particularly in the development of new vehicle models.

[0003] The current prediction methods have the following main problems:

[0004] 1) Lack of pulse data and inadequate reconstruction technology. Currently, crash pulses are primarily derived from finite element simulations and real-vehicle testing (such as C-NCAP). However, due to cost and time constraints, sample coverage is insufficient. Existing data augmentation methods (such as noise injection and time-domain scaling) can only generate simple variants and struggle to realistically reconstruct complex operating conditions (such as multi-directional coupled collisions). Consequently, prediction models trained on small samples (such as support vector machines and random forests) generally have R² values ​​below 0.7 in new data scenarios, failing to meet engineering requirements.

[0005] 2) The "black box" problem of pulse feature characterization. There are relatively few methods for characterizing collision pulses. Current technologies generally use deep learning methods to extract collision pulse features through neural networks. Some methods use deep learning (such as CNN and LSTM) to automatically extract pulse features, but there are still two major flaws:

[0006] Poor interpretability: Neural networks are "black box" models, and feature extraction can only be diagnosed and adjusted from the input and output ends. There's no guarantee that the extracted features will accurately respond to the collision pulse. Neural network hidden layer features lack a clear correlation with physical quantities (such as peak acceleration and velocity change), making it difficult for engineers to verify the rationality of the features.

[0007] Pulse feature extraction takes a long time: Neural networks often require complex network structures and deep neural network layers to extract features from long sequence data. Complex and deep neural networks take too long to extract long sequence data, making it difficult to meet the rapid requirements of iterative training, parameter tuning, and engineering applications of occupant injury prediction models. Summary of the Invention

[0008] In view of this, the present application aims to propose a method, device, electronic device and storage medium for predicting occupant injury to solve at least one of the above problems.

[0009] To achieve the above objectives, the technical solution of this application is implemented as follows:

[0010] In a first aspect, the present application provides a method for predicting occupant injury, comprising:

[0011] Determining the discontinuity positions of two phases in a pulse curve based on the acquired original operating condition crash pulse to generate reconstructed crash pulse curves corresponding to multiple crash conditions, wherein the original operating condition crash pulse is used to represent the occupant position under the actual crash mechanism;

[0012] Performing a second-order half-sine wave fitting process on the reconstructed collision pulse curve to obtain a second-order half-sine simplified pulse curve, and extracting interpretable waveform features;

[0013] The interpretable waveform features and the obtained restraint system parameters are input into a pre-built occupant injury prediction model to output an occupant injury result, wherein the occupant injury prediction model is trained by the interpretable waveform features, restraint system parameters and occupant injury results.

[0014] In a second aspect, based on the same inventive concept, the present application further provides an occupant injury prediction device, comprising:

[0015] a pulse curve generation module configured to determine the discontinuity position of two phases in the pulse curve based on the acquired original working condition collision pulse to generate reconstructed collision pulse curves corresponding to multiple collision working conditions, wherein the original working condition collision pulse is used to represent the occupant position under the actual collision mechanism;

[0016] a fitting processing module configured to perform a second-order half-sine wave fitting process on the reconstructed collision pulse curve to obtain a second-order half-sine simplified pulse curve, and extract interpretable waveform features;

[0017] The injury prediction module is configured to input the interpretable waveform features and the obtained restraint system parameters into a pre-built occupant injury prediction model to output an occupant injury result, wherein the occupant injury prediction model is trained by the interpretable waveform features, restraint system parameters and occupant injury results.

[0018] In a third aspect, based on the same inventive concept, the present application also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method described in the first aspect when executing the program.

[0019] In a fourth aspect, based on the same inventive concept, the present application also provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the method described in the first aspect.

[0020] Compared with the prior art, the occupant injury prediction method, device, electronic device, and storage medium described in this application have the following beneficial effects:

[0021] The occupant injury prediction method described in this application achieves high-precision and strong generalization of occupant injury prediction through physically constrained pulse reconstruction technology and interpretable pulse simplification methods, thereby solving the bottleneck of traditional engineering applications that rely on small sample data and black box models. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The accompanying drawings, which constitute part of this application, are intended to provide a further understanding of this application. The exemplary embodiments and descriptions of this application are intended to explain this application and do not constitute an improper limitation on this application. In the accompanying drawings:

[0023] Figure 1 This is a flow chart of a method for predicting occupant injury according to an embodiment of the present application;

[0024] Figure 2 This is a schematic diagram of the original working condition collision pulse and the reconstructed collision pulse curve according to the embodiment of the present application;

[0025] Figure 3 This is a schematic diagram of the reconstructed collision pulse and the second-order half-sine simplified pulse curve according to the embodiment of the present application;

[0026] Figure 4 This is a schematic structural diagram of an occupant injury prediction device according to an embodiment of the present application;

[0027] Figure 5 This is a schematic diagram of the hardware structure of the electronic device described in an embodiment of the present application. DETAILED DESCRIPTION

[0028] In order to make the objectives, technical solutions and advantages of this application more clear, this application is further described in detail below in combination with specific embodiments and with reference to the accompanying drawings.

[0029] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the embodiments of the present application should have the usual meanings understood by people with ordinary skills in the field to which this application belongs. The "first", "second" and similar words used in the embodiments of the present application do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0030] The embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0031] See also Figure 1 As shown, this embodiment provides a method for predicting occupant injuries, which specifically includes the following steps:

[0032] Step S101: Determine the discontinuity positions of two stages in a pulse curve based on the acquired original working condition collision pulse to generate reconstructed collision pulse curves corresponding to multiple collision working conditions, wherein the original working condition collision pulse is used to represent the occupant position under the actual collision mechanism.

[0033] Specifically, step S101 of this embodiment includes the following steps:

[0034] Step S110: Obtain the original working condition collision pulse curve. A real-world collision simulation is performed based on the vehicle simulation model to obtain the original working condition collision pulse for the occupant position consistent with the actual collision mechanism. The two-stage discontinuity locations of the collision pulse curve are determined (the valley points corresponding to the maximum valley-to-peak value in the original collision pulse curve are used as the two-stage discontinuity locations).

[0035] Step S120 , confirming the two-stage discontinuity position of the collision pulse curve: taking the valley point position corresponding to the maximum valley-peak value in the original collision pulse curve as the two-stage discontinuity position.

[0036] Specifically, based on the two-stage characteristics of the original crash pulse and the corresponding characteristics of occupant dynamics, a two-stage discontinuity is identified on the original crash pulse curve. The discontinuity corresponds to the point where the front energy-absorbing structure (such as the longitudinal beam) is completely crushed. At this point, kinetic energy absorption shifts from local deformation to global body deformation. After the discontinuity, the head / thorax acceleration curves of the rear dummy (or occupant) begin to rise significantly, reflecting the secondary application of inertial loads.

[0037] Specifically, the entire collision process goes through two main stages: After the collision, the front structure of the vehicle (such as the anti-collision beam and longitudinal beam) begins to collapse, and the collision load is transmitted to the rear row through the body frame. This stage is the first stage (initial pulse), which is characterized by a rapid increase in pulse amplitude, reflecting the rigidity and initial energy absorption of the front structure. The entire process lasts for a short time. After the front energy-absorbing structure is completely crushed, the collision load continues to be transmitted through the passenger compartment floor, B-pillar and other paths, which may be accompanied by overall vehicle deceleration or a secondary collision. This stage is the second stage (subsequent pulse), which is characterized by a possible increase in pulse amplitude, reflecting the rigid resistance and energy redistribution of the passenger compartment. The discontinuity between the two stages corresponds to the point where the front energy-absorbing structure (such as the longitudinal beam) is crushed. At this time, kinetic energy absorption shifts from local deformation to global vehicle deformation. After the discontinuity point, the pulse value at the passenger position begins to rise significantly, reflecting the secondary loading of the inertial load.

[0038] In this embodiment, the method for confirming the two-stage discontinuity position is as follows:

[0039] ① Traverse the original collision pulse curve and identify valid peak points and valley points. The specific method for identifying valid peak points and valley points includes searching all peak and valley points, eliminating local peak and valley points, and clearing adjacent consecutive peak points or valley points. After all these steps, valid peak points and valley points that occur at intervals are obtained, namely peak-valley pairs.

[0040] ② Determine the maximum valley-peak point. Define the valley point and the absolute value of the difference between the adjacent valid peak points after the valley point as the valley-peak value. Traverse all valley points and obtain the valley-peak value corresponding to each valley point. The valley point corresponding to the maximum valley-peak value is used as the discontinuity position between the two stages of the collision pulse.

[0041] Step S130: generating reconstructed collision pulse curves for various collision conditions.

[0042] ① Split the original collision pulse into two phases. Using the two-phase discontinuity position obtained in the above steps, the original collision pulse is divided into two phases: starting time 0s to the discontinuity (the first phase, i.e., the energy-absorbing structure collapse phase), and the discontinuity to the pulse cutoff (the second phase, i.e., the passenger compartment load transfer phase). The pulse cutoff is defined as the moment on the second half of the pulse curve where the pulse value first drops to 0.05g.

[0043] ② Simulate various collision conditions and generate reconstructed collision pulse curves. Adjust the first and second phase pulse values ​​according to the simulation conditions. Generally, this is done by strengthening (or weakening) the impact of the energy-absorbing structure's crushing phase (or the passenger compartment load transfer phase) to simulate various collision conditions.

[0044] First, the pulse curve reconstruction matrix is ​​determined. To reconstruct the energy distribution and waveform characteristics of the pulse in the two phases, the pulse curves in the two phases are strengthened (or weakened) respectively. Specifically, the pulse value in the specified phase is weighted by 0.5-1 times, 0.7-1 times, 1-2 times, and 1-3 times. The energy-absorbing structure crushing phase (or the passenger compartment load transfer phase) is processed separately. From the original collision pulse curve, collision pulse curves representing eight new collision conditions can be reconstructed.

[0045] Next, a two-stage weighted ratio sequence is calculated. To ensure the continuity of the entire reconstructed collision pulse curve, the initial pulse values ​​at the start, end, and discontinuity points remain unchanged. Therefore, a weighted ratio sequence for each of the two stages' reconstruction methods needs to be calculated. Specifically, the median index of a given stage is taken, and the weighted ratio of the pulse value at that index is the maximum (or minimum) value. The corresponding weighted ratios for the remaining pulse values ​​in that stage are linearly generated according to the index position. The calculation formula is as follows:

[0046]

[0047] Where, Indicates the weighted ratio of the corresponding index position, represents the median magnification ( ), Indicates the total index length included in the calculated stage.

[0048] Finally, the collision pulse curve is reconstructed using a weighted ratio sequence. The weighted ratio of the corresponding index is multiplied by the pulse sequence data to obtain the reconstructed collision pulse curve. However, during the simulation of various collisions, the overall energy (curve integral) of the reconstructed collision pulse curve changes. To ensure the authenticity and effectiveness of the simulated collision conditions, the overall pulse energy must be kept consistent. By calculating the curve integral of the reconstructed collision pulse curve and the original collision pulse curve over the pulse duration (before the pulse cutoff), the reconstructed collision pulse curve is adjusted according to the ratio of the two integral values ​​to ensure that the pulse energies of the two are consistent. Through the above steps, reconstructed collision pulse curves for various working conditions that conform to the actual collision mechanism are obtained, which are used to enrich the correlation model between the collision mechanism and various damage indicators.

[0049] Step S101 of this embodiment is to identify the two-stage discontinuity position on the original collision pulse curve based on the two-stage characteristics of the original collision pulse and the corresponding characteristics of the occupant dynamics, and generate reconstructed collision pulse curves for multiple collision conditions (the original collision pulse and the reconstructed collision pulse curves are as follows: Figure 2 As shown in the figure, the effective training samples are increased by more than 8 times, which significantly improves the generalization ability of the model, especially in the application effect during the development of new models.

[0050] Step S102 : performing second-order half-sine wave fitting processing on the reconstructed collision pulse curve to obtain a second-order half-sine simplified pulse curve, and extracting interpretable waveform features.

[0051] Specifically, step S102 of this embodiment includes the following steps:

[0052] Step S201: Fit the collision pulse curve to a second-order half-sine wave. Based on the method for determining the two-stage discontinuity position of the collision pulse curve in step S101, the collision pulse curve is divided into two stages: the first stage (energy-absorbing structure crushing stage) and the second stage (passenger compartment load transfer stage).

[0053] The collision pulse curves of the two stages are fitted with a second-order half-sine wave, including: confirmation of the second-order simplified waveform design scheme (first-stage half-sine wave, second-stage half-sine wave), boundary condition setting (first-stage pulse starting point, second-stage pulse end point, discontinuity position pulse difference), and minimum penalty term fitting (boundary conditions, pulse fitting difference). The energy of the collision pulse before and after simplification must remain consistent.

[0054] Confirmation of the second-order simplified waveform design scheme: Since the two stages do not completely conform to the waveform distribution characteristics of the first half-sine wave, it is necessary to test and confirm the simplified waveforms of the two stages. The collision pulse curves of the two stages are fitted into the first half-sine wave and the second half-sine wave respectively. The fitting formula is as follows:

[0055]

[0056]

[0057] Where, 、 They represent the fitting results of the first half sine wave and the second half sine wave respectively. Indicates amplitude, Indicates the time when the pulse occurs, Indicates phase, Indicates a period;

[0058] Boundary condition settings for the first and second stages:

[0059] The starting point constraint, the pulse value of the first stage starting point is: ;

[0060] End point constraint, the second stage end point pulse value is: ;

[0061] Discontinuity point constraint, the difference between the two-stage discontinuity position pulses is =0;

[0062] Where, 、 Represent the first and second stage pulse curve functions respectively, Indicates the corresponding time of the two-stage discontinuity position, Indicates the second stage pulse cut-off time;

[0063] Construct a comprehensive penalty function. The specific formula is as follows:

[0064]

[0065] Where, 、 Represents the weight coefficient, which is used to adjust the importance of boundary conditions and fitting errors; represents the boundary condition penalty term; Represents the fitting error penalty term, and the final fitting goal is to minimize the penalty term .

[0066] Maintain energy consistency: By calculating the curve integral values ​​of the fitted second-order half-sine simplified pulse curve and the collision pulse curve within the pulse duration (before the pulse cutoff time), the fitted second-order half-sine simplified pulse curve is adjusted according to the ratio of the two integral values ​​to make the pulse energies of the two consistent.

[0067] (2) Verification of the effectiveness of simplified pulses. The collision pulse curves before and after simplification are input into the finite element model of occupant damage simulation as the collision load response. Based on the output damage curves of each part (head, neck, chest, legs, etc.), the matching degree of the damage curves before and after the simplified pulse is calculated. In this embodiment, the cross-correlation coefficient CCF is used to evaluate the similarity of the two pulse signals. Therefore, the CCF value of each damage curve is used to evaluate the effectiveness of the collision response of the second-order half-sine simplified pulse, and verify the effectiveness of the simplified pulse of the two-stage simplified waveform setting scheme in the collision mechanism. The specific calculation formula is as follows:

[0068]

[0069] in, ;

[0070] Where, represents the standard normalized value of the cross coefficient; represents the cross-correlation coefficient; Indicates the total number of collected signals; represents the time shift parameter, represents the signal Relative to the signal The number of displacement points is due to the fact that there is no time shift between the second-order half-sine simplified pulse and the collision pulse curve. Take 0; Respectively represent signals 、 In this embodiment, when CCF>0.9, it is determined that the second-order half-sine simplified pulse can effectively reflect the collision response of the collision pulse.

[0071] Different stages are fitted with different half-sine wave forms, respectively, and input into the occupant injury simulation finite element model, and the CCF value of each damage index is calculated. It is finally determined that the first-stage collision pulse is fitted with a quadratic half-sine wave, and the second-stage collision pulse is fitted with a primary half-sine wave. This simplified waveform setting scheme has the best fitting effect on the damage curves of various parts before and after the simplified pulse.

[0072] (3) Extraction of interpretable waveform features of the second-order half-sine simplified pulse. The second-order half-sine simplified pulse, which has been verified to be effective, is determined by six parameters: position parameters (discontinuity point position time, second-order wave cutoff time) and shape parameters (first-order wave amplitude, first-order wave phase, second-order wave amplitude, second-order wave phase). These six waveform parameters are used as interpretable waveform features of the original collision pulse curve, which can fully and effectively characterize the collision process response of the complex collision pulse curve.

[0073] Step S201 of this embodiment is based on the second-order half-sine wave pulse simplification to reconstruct the collision pulse curve under various working conditions, and the following is obtained: Figure 3 The second-order half-sine simplified pulse curve shown in the figure further extracts the interpretable waveform features of the second-order half-sine simplified pulse curve. Compared with the traditional black box model, the waveform feature interpretability is significantly improved, enabling the prediction model to effectively extract the pulse interpretable waveform features using the same standard on collision pulse curves of different scales.

[0074] Step S103: Input the interpretable waveform features and the obtained restraint system parameters into a pre-built occupant injury prediction model to output the occupant injury results, wherein the occupant injury prediction model is trained by the interpretable waveform features, the restraint system parameters, and the occupant injury results.

[0075] Specifically, step S103 of this embodiment includes the following steps:

[0076] Step S301: Obtain an occupant injury dataset. Based on the reconstructed collision pulse curves and restraint system parameters (seatbelt ignition timing, retractor force limit, seatbelt D-ring position) for various collision scenarios obtained in the previous step, an occupant injury expansion matrix is ​​generated. Finite element simulation calculations are then performed based on multiple vehicle occupant finite element simulation models to obtain occupant injury results for each restraint system instance (including head, neck, and chest injury values. Head injuries are evaluated using the head injury criterion HIC15 and the cumulative 3ms composite acceleration value; neck extension is used as the evaluation indicator for neck injuries; and chest compression is used as a quantitative indicator for chest injuries, which directly reflects the degree of chest deformation of the dummy occupant during a frontal collision).

[0077] The interpretable waveform features of the reconstructed collision pulse curve, the restraint system parameters, and the occupant injury results together constitute the occupant injury dataset, which is used to train the occupant injury prediction model.

[0078] This embodiment uses a total of about 1,500 constraint solution data from 8 car models. The data of one car model is randomly selected as the test set, and 70% of the remaining data is used as the training set, and 30% is used as the validation set and randomly shuffled.

[0079] Step S302: Train the occupant injury prediction model. Interpretable waveform features (discontinuity point location, second-order wave cutoff, first-order wave amplitude, first-order wave phase, second-order wave amplitude, and second-order wave phase) and restraint system parameters (seatbelt ignition time, retractor force limit, and seatbelt D-ring position) are used as feature inputs, and the damage values ​​of various parts are used as target quantities. These constitute the input and output of the occupant injury model.

[0080] Machine learning algorithms (Lasso regression, random forest regression RF, support vector machine regression SVR) are used to train the occupant injury prediction model to obtain the occupant injury prediction model corresponding to each indicator. The model training process mainly adjusts the hyperparameters for each machine learning algorithm until the required prediction accuracy is obtained on the validation set and test set. It is generally believed that a prediction error of less than 10% meets the prediction requirements.

[0081] The main hyperparameters that need to be adjusted in machine learning algorithms are as follows: Lasso regression (regularization coefficient α, optimization tolerance tol), random forest regression RF (number of trees n, maximum tree depth d, minimum number of samples for node splitting s, maximum number of features considered for each tree f), support vector machine regression SVR (penalty coefficient C, kernel function kernel, kernel coefficient γ, insensitive band ε).

[0082] Prediction accuracy mainly refers to the mean absolute percentage error (MAPE): MAPE measures the ratio of prediction error to actual value in percentage form, which is convenient for comparing model performance across data sets or different dimensions. However, it may be distorted when the true value is close to zero. The smaller the value, the higher the prediction accuracy of the model and the smaller the deviation between the prediction result and the true value.

[0083]

[0084] Where, represents the predicted value, represents the simulation calculation value, Indicates the number of calculated data items.

[0085] Step S303: Predicting occupant damage values ​​for new vehicle models. The collision pulse curve of the new vehicle model is simplified to a second-order sinusoidal pulse. The resulting interpretable waveform features are directly combined with the restraint system parameter matrix and input into a trained occupant damage prediction model. This rapidly obtains occupant damage outputs for various parts of multiple restraint system solutions. The outputs of multiple damage indicators are evaluated to determine the optimal restraint system design for the vehicle model, providing a reference for the occupant protection R&D phase.

[0086] The occupant injury prediction method described in this embodiment achieves high-precision and highly generalized occupant injury prediction through physically constrained pulse reconstruction technology and an interpretable pulse simplification method, thereby solving the bottleneck of traditional engineering applications that rely on small sample data and black box models.

[0087] 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.

[0088] Based on the same inventive concept, corresponding to any of the above-mentioned embodiments and methods, an embodiment of the present application further provides an occupant injury prediction device.

[0089] like Figure 4 As shown, the occupant injury prediction device includes:

[0090] a pulse curve generating module 11 configured to determine the discontinuity positions of two phases in the pulse curve based on the acquired original working condition collision pulse, so as to generate reconstructed collision pulse curves corresponding to various collision working conditions, wherein the original working condition collision pulse is used to represent the occupant position under the actual collision mechanism;

[0091] The fitting processing module 12 is configured to perform a second-order half-sine wave fitting process on the reconstructed collision pulse curve to obtain a second-order half-sine simplified pulse curve and extract interpretable waveform features;

[0092] The damage prediction module 13 is configured to input the interpretable waveform features and the obtained restraint system parameters into a pre-built occupant damage prediction model to output the occupant damage results, wherein the occupant damage prediction model is trained by the interpretable waveform features, the restraint system parameters and the occupant damage results.

[0093] For the convenience of description, the above devices are described as being divided into various modules according to their functions. Of course, when implementing the embodiments of the present application, the functions of each module can be implemented in the same or multiple software and / or hardware.

[0094] The apparatus of the above embodiment is used to implement the corresponding method in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be described in detail here.

[0095] Based on the same inventive concept, corresponding to any of the above-mentioned embodiments and methods, an embodiment of the present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and runnable on the processor, wherein when the processor executes the program, the method described in any of the above embodiments is implemented.

[0096] Figure 5 10 is a schematic diagram showing a more specific hardware structure of an electronic device provided in this embodiment. The device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040 are communicatively connected to each other within the device via the bus 1050.

[0097] The processor 1010 can be implemented using a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0098] The memory 1020 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage devices, dynamic storage devices, etc. The memory 1020 can store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 1020 and is called and executed by the processor 1010.

[0099] The input / output interface 1030 is used to connect to an input / output module to enable information input and output. The input / output module can be configured as a component within the device (not shown) or can be externally connected to the device to provide corresponding functions. Input devices may include a keyboard, mouse, touch screen, microphone, various sensors, etc. Output devices may include a display, speaker, vibrator, indicator light, etc.

[0100] The communication interface 1040 is used to connect to a communication module (not shown) to enable communication between the device and other devices. The communication module can communicate via wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, Wi-Fi, Bluetooth, etc.).

[0101] The bus 1050 comprises a pathway for transmitting information between various components of the device, such as the processor 1010 , the memory 1020 , the input / output interface 1030 , and the communication interface 1040 .

[0102] It should be noted that although the above device only shows the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040, and the bus 1050, in a specific implementation, the device may also include other components necessary for normal operation. In addition, it will be understood by those skilled in the art that the above device may only include the components necessary to implement the embodiments of this specification, and does not necessarily include all the components shown in the figure.

[0103] The electronic device of the above embodiment is used to implement the corresponding method in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be described in detail here.

[0104] Based on the same inventive concept, corresponding to any of the above-mentioned embodiments and methods, the present application also provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the method described in any of the above embodiments.

[0105] The computer-readable media of this embodiment includes permanent and non-permanent, removable and non-removable media that can be used to store information by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, tape disk storage or other magnetic storage devices, or any other non-transmission media that can be used to store information that can be accessed by a computing device.

[0106] The computer instructions stored in the storage medium of the above embodiment are used to enable the computer to execute the method described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0107] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present application (including the claims) is limited to these examples. Within the scope of the present application, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the embodiments of the present application as described above, which are not provided in detail for the sake of simplicity.

[0108] In addition, to simplify the description and discussion, and to avoid obscuring the understanding of the embodiments of the present application, well-known power / ground connections to integrated circuit (IC) chips and other components may or may not be shown in the provided figures. Furthermore, devices may be shown in block diagram form to avoid obscuring the understanding of the embodiments of the present application, and this also takes into account the fact that the implementation details of these block diagram devices are highly dependent on the platform on which the embodiments of the present application will be implemented (i.e., these details should be fully understood by those skilled in the art). Where specific details (e.g., circuits) are set forth to describe the exemplary embodiments of the present application, it will be apparent to those skilled in the art that the embodiments of the present application can be implemented without these specific details or with variations therefrom. Therefore, these descriptions should be considered illustrative rather than restrictive.

[0109] Although the present invention has been described in conjunction with specific embodiments thereof, many alternatives, modifications, and variations of these embodiments will be apparent to those skilled in the art based on the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may utilize the discussed embodiments.

[0110] The embodiments of the present application are intended to cover all such substitutions, modifications, and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present application should be included in the scope of protection of this application.

Claims

1. A method for predicting occupant injury, characterized in that: include: Determining the discontinuity positions of two phases in a pulse curve based on the acquired original operating condition crash pulse to generate reconstructed crash pulse curves corresponding to multiple crash conditions, wherein the original operating condition crash pulse is used to represent the occupant position under the actual crash mechanism; Performing a second-order half-sine wave fitting process on the reconstructed collision pulse curve to obtain a second-order half-sine simplified pulse curve, and extracting interpretable waveform features; The interpretable waveform features and the obtained restraint system parameters are input into a pre-built occupant injury prediction model to output an occupant injury result, wherein the occupant injury prediction model is trained by the interpretable waveform features, restraint system parameters and occupant injury results.

2. The method according to claim 1, characterized in that The two-stage discontinuous position determination method includes: By traversing the original working condition collision pulse curve, identifying effective peak points and valley points that appear at intervals in the curve; Define the valley point and the absolute value of the difference between the adjacent valid peak points after the valley point as the valley-peak value, traverse all valley points to obtain the valley-peak value corresponding to each valley point, and take the valley point position corresponding to the maximum valley-peak value as the discontinuity position of the two stages of the collision pulse.

3. The method according to claim 2, characterized in that The method for generating the reconstructed collision pulse curve includes: The original working condition collision pulse is divided into two stages according to the determined discontinuity position, wherein the first stage is the energy absorbing structure crushing stage, and the second stage is the passenger compartment load transfer stage; The pulse curves of the two stages are processed respectively, and weighted ratio sequences of different reconstruction modes of the two stages are calculated, and the collision pulse curve is reconstructed according to the weighted ratio sequences.

4. The method according to claim 1, wherein The reconstructed collision pulse curve is subjected to a second-order half-sine wave fitting process to obtain a second-order half-sine simplified pulse curve, comprising: The reconstructed collision pulse curve is divided into the first stage and the second stage, and the following boundary conditions are set, including: The starting point constraint, the pulse value of the first stage starting point is: ; End point constraint, the second stage end point pulse value is: ; Discontinuity point constraint, the difference between the two-stage discontinuity position pulses is ; Construct a comprehensive penalty function. The specific formula is as follows: ； The reconstructed collision pulse curves of the two stages are fitted into a first half-sine wave and a second half-sine wave respectively. The fitting formulas are as follows: ； ； Where, 、 Represent the first and second stage pulse curve functions respectively, Indicates the corresponding time of the two-stage discontinuity position, Indicates the second stage pulse cut-off time, 、 represents the weight coefficient, represents the boundary condition penalty term; represents the fitting error penalty term, represents the minimization penalty term, 、 They represent the fitting results of the first half sine wave and the second half sine wave respectively. Indicates amplitude, Indicates the time when the pulse occurs, Indicates phase, Indicates a period; The final fitting determined that the first-stage collision pulse was fitted with a quadratic half-sine wave, and the second-stage collision pulse was fitted with a primary half-sine wave, resulting in a second-order half-sine simplified pulse curve.

5. The method according to claim 4, characterized in that Also includes: The collision pulse curves before and after simplification are input into the occupant damage simulation finite element model as the collision load response. Based on the output damage curves of various occupant parts, the matching degree of the damage curves before and after the simplified pulse is calculated to evaluate the effectiveness of the collision response of the second-order half-sine simplified pulse. The specific formula is as follows: ； in, ; Where, represents the standard normalized value of the cross coefficient, represents the cross-correlation coefficient, Indicates the total number of collected signals. represents the time shift parameter, Respectively represent signals 、 The autocorrelation of .

6. The method according to claim 1, wherein: The interpretable waveform features include position parameters and shape parameters, wherein the position parameters include the discontinuity point position time and the second-order wave cutoff time, and the shape parameters include the first-order wave amplitude, the first-order wave phase, the second-order wave amplitude and the second-order wave phase.

7. The method according to claim 1, wherein: Obtain an occupant injury dataset, use interpretable waveform features and restraint system parameters as feature inputs, and the damage values ​​of each part as target quantities to form the input and output of the occupant injury model to train the occupant injury prediction model. Multiple occupant injury indicators are evaluated based on the output results of the trained occupant injury prediction model.

8. A passenger injury prediction device, characterized in that: include: a pulse curve generation module configured to determine the discontinuity position of two phases in the pulse curve based on the acquired original working condition collision pulse to generate reconstructed collision pulse curves corresponding to multiple collision working conditions, wherein the original working condition collision pulse is used to represent the occupant position under the actual collision mechanism; a fitting processing module configured to perform a second-order half-sine wave fitting process on the reconstructed collision pulse curve to obtain a second-order half-sine simplified pulse curve, and extract interpretable waveform features; The injury prediction module is configured to input the interpretable waveform features and the obtained restraint system parameters into a pre-built occupant injury prediction model to output an occupant injury result, wherein the occupant injury prediction model is trained by the interpretable waveform features, restraint system parameters and occupant injury results.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method according to any one of claims 1 to 7 when executing the program.

10. A non-transitory computer-readable storage medium, characterized in that in, The non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to cause a computer to execute the method according to any one of claims 1 to 7.

Images