Method and device for determining structural characteristic parameters of energy absorption structure and electronic equipment

Abstract

The invention discloses a method and device for determining structural characteristic parameters of an energy absorption structure and electronic equipment, and the method comprises the steps: inputting candidate structural characteristic parameters of the energy absorption structure and corresponding three-dimensional point cloud data into a first prediction model, and predicting the three-dimensional point cloud data and collision performance characteristics of the energy absorption structure after deformation under a set collision working condition; according to the deformed three-dimensional point cloud data, determining a collision deformation mode of deformation of the energy absorption structure under the set collision working condition, and according to the collision performance characteristics and the collision deformation mode of deformation of the energy absorption structure under the set collision working condition, determining the collision deformation mode of deformation of the energy absorption structure under the set collision working condition; and target structure characteristic parameters of the energy absorption structure are determined from the multiple sets of candidate structure characteristic parameters. According to the method and device for determining the structural characteristic parameters of the energy absorption structure and the electronic equipment, the target structural characteristic parameters enabling the collision safety protection function of the energy absorption structure to be optimal are efficiently determined on the basis of the collision performance characteristics and the collision deformation mode of the energy absorption structure.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle optimization, and in particular to a method, device and electronic equipment for determining structural characteristic parameters of an energy absorbing structure. Background Art

[0002] During the use of vehicles, there is a risk of collision accidents, so the vehicle needs to be equipped with an energy-absorbing structure, such as an energy-absorbing structure made of medium-strength steel or aluminum, to absorb the impact force and reduce the harm of collision accidents. In order to achieve the best collision energy absorption effect, the energy-absorbing structure needs to be optimized during the vehicle design stage to ensure the collision safety protection function of the energy-absorbing structure.

[0003] The prior art constructs a finite element analysis model, modifies the structural characteristic parameters of the energy absorbing structure multiple times, and performs collision simulation to output the corresponding collision performance characteristics, from which the structural characteristic parameters that make the collision performance characteristics of the energy absorbing structure better are selected. However, the collision safety protection function of the energy absorbing structure is related to the collision performance characteristics and collision deformation mode of the energy absorbing structure. If the structural characteristic parameters that make the collision performance characteristics of the energy absorbing structure better cannot make the collision deformation mode of the energy absorbing structure meet the standard, it is necessary to re-perform the collision simulation to obtain more optimal structural characteristic parameters. Therefore, the prior art uses the collision performance characteristics of the energy absorbing structure as a single goal to achieve structural optimization of the energy absorbing structure, and it is difficult to efficiently obtain the optimization result that optimizes the collision safety protection function of the energy absorbing structure, resulting in a long design cycle and high cost for the vehicle energy absorbing structure. Summary of the invention

[0004] The present application provides a method, device and electronic equipment for determining the structural characteristic parameters of an energy absorbing structure, so as to improve the technical problem that the existing technology uses the collision performance characteristics of the energy absorbing structure as a single goal to achieve structural optimization of the energy absorbing structure, and it is difficult to efficiently determine the optimization result that optimizes the collision safety protection function of the energy absorbing structure.

[0005] In view of the above problems, the present application is proposed to provide a method, device and electronic device for determining structural characteristic parameters of an energy absorbing structure that overcomes the above problems or at least partially solves the above problems.

[0006] In a first aspect, a method for determining structural characteristic parameters of an energy absorbing structure is provided, comprising: Acquire multiple groups of candidate structural characteristic parameters of the energy absorbing structure, and three-dimensional point cloud data of the energy absorbing structure corresponding to each group of candidate structural characteristic parameters; For each group of candidate structural characteristic parameters, the candidate structural characteristic parameters and the three-dimensional point cloud data of the energy-absorbing structure corresponding to the candidate structural characteristic parameters are respectively input into a preset prediction model to obtain first prediction data and second prediction data, and the first prediction data is input into a preset recognition model to obtain a collision deformation mode of the energy-absorbing structure corresponding to the candidate structural characteristic parameters under set collision conditions, wherein the first prediction data is the three-dimensional point cloud data of the energy-absorbing structure corresponding to the candidate structural characteristic parameters after deformation under the set collision conditions, and the second prediction data is used to characterize the collision performance characteristics of the energy-absorbing structure corresponding to the candidate structural characteristic parameters when deformed under the set collision conditions; Based on the second prediction data and the collision deformation mode corresponding to each group of candidate structural characteristic parameters, target structural characteristic parameters of the energy absorbing structure are determined from the multiple groups of candidate structural characteristic parameters.

[0007] Optionally, the collision performance characteristics include: energy absorption, specific energy absorption, and collision peak force; based on the second prediction data and the collision deformation mode corresponding to each group of candidate structural characteristic parameters, determining the target structural characteristic parameters of the energy absorbing structure from the multiple groups of candidate structural characteristic parameters includes: Based on the second prediction data of the candidate structural characteristic parameters, the candidate structural characteristic parameters whose energy absorption, specific energy absorption and collision peak force meet the preset screening conditions are screened out from the multiple groups of candidate structural characteristic parameters to obtain a first set of parameters to be optimized; Based on a preset multi-objective optimization model, and the second prediction data and collision deformation mode corresponding to each group of candidate structural characteristic parameters in the first parameter set to be optimized, target structural characteristic parameters of the energy absorbing structure are determined from the first parameter set to be optimized.

[0008] Optionally, based on a preset multi-objective optimization model and second prediction data and collision deformation mode corresponding to each group of candidate structural characteristic parameters in the first parameter set to be optimized, target structural characteristic parameters of the energy absorbing structure are determined from the first parameter set to be optimized, including: Determine the target collision deformation mode; Based on the target collision deformation mode and a preset multi-objective optimization model, candidate structural characteristic parameters whose collision deformation mode is the target collision deformation mode are screened from the first parameter set to be optimized to obtain a second optimization parameter set, and energy absorption, specific energy absorption and collision peak force are taken as optimization targets. Based on the second prediction data of each group of candidate structural characteristic parameters in the second optimization parameter set, the target structural characteristic parameters are determined from the second optimization parameter set.

[0009] Optionally, the structural characteristic parameters include the structural length, effective cross-sectional area and wall thickness of the energy absorbing structure; the multi-objective optimization model includes an optimization function and constraint conditions, wherein the optimization function is: ; The constraints are: ; In the formula, is the structural length of the energy absorbing structure, is the minimum value of the structure length, is the maximum value of the structure length, is the effective cross-sectional area of ​​the energy absorbing structure, The minimum effective cross-sectional area, is the maximum value of the effective cross-sectional area, is the wall thickness of the energy absorbing structure, Minimum wall thickness, is the maximum value of wall thickness, EA is the energy absorption, SEA is the specific energy absorption, PCF is the peak collision force, is the collision deformation mode of the energy-absorbing structure, and c is the target collision deformation mode.

[0010] Optionally, the first prediction model is obtained by the following steps: Acquire first training sample data, the first training sample data including: structural characteristic parameters and three-dimensional point cloud data of at least one energy absorbing structure sample, and deformed three-dimensional point cloud data corresponding to each energy absorbing structure sample, and collision performance characteristic data; A pre-constructed first initial neural network model is trained based on the first training sample data to obtain a first prediction model.

[0011] Optionally, the recognition model is obtained by the following steps: Acquire second training sample data, the second training sample data comprising: deformed three-dimensional point cloud data of at least one energy absorbing structure sample and a collision deformation mode under a set collision working condition; The pre-constructed second initial neural network model is trained based on the second training sample data to obtain a recognition model.

[0012] Optionally, obtaining three-dimensional point cloud data of the energy absorbing structure corresponding to each set of candidate structural characteristic parameters includes: Based on the candidate structural characteristic parameters, a finite element simulation model of the energy absorbing structure is constructed, and based on the finite element simulation model, three-dimensional point cloud data of the energy absorbing structure corresponding to the candidate structural characteristic parameters is obtained.

[0013] In a second aspect, the present application further provides a device for determining structural characteristic parameters of an energy absorbing structure, comprising: an acquisition unit, for acquiring multiple groups of candidate structural characteristic parameters of the energy absorbing structure, and three-dimensional point cloud data of the energy absorbing structure corresponding to each group of candidate structural characteristic parameters; A prediction unit, for inputting the candidate structural characteristic parameters and the three-dimensional point cloud data of the energy-absorbing structure corresponding to the candidate structural characteristic parameters into a preset first prediction model for each group of candidate structural characteristic parameters, to obtain first prediction data and second prediction data, and inputting the first prediction data into a preset recognition model to obtain a collision deformation mode of the energy-absorbing structure corresponding to the candidate structural characteristic parameters under set collision conditions, wherein the first prediction data is the three-dimensional point cloud data of the energy-absorbing structure corresponding to the candidate structural characteristic parameters after deformation under the set collision conditions, and the second prediction data is used to characterize the collision performance characteristics of the energy-absorbing structure corresponding to the candidate structural characteristic parameters after deformation under the set collision conditions; The optimization unit is used to determine the target structural characteristic parameters of the energy absorbing structure from multiple groups of candidate structural characteristic parameters based on the second prediction data and the collision deformation mode corresponding to each group of candidate structural characteristic parameters.

[0014] Optionally, the crash performance characteristics include: energy absorption, specific energy absorption, and crash peak force; the optimization units include: A screening unit is used to screen out candidate structural characteristic parameters whose energy absorption, specific energy absorption and collision peak force meet preset screening conditions from multiple groups of candidate structural characteristic parameters based on the second prediction data of the candidate structural characteristic parameters, so as to obtain a first parameter set to be optimized; The target optimization unit is used to determine the target structural characteristic parameters of the energy absorbing structure from the first parameter set to be optimized based on a preset multi-objective optimization model and the second prediction data and collision deformation mode corresponding to each group of candidate structural characteristic parameters in the first parameter set to be optimized.

[0015] In a third aspect, 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 when the processor executes the computer program, the electronic device executes the method provided in the first aspect.

[0016] In a fourth aspect, the present application further provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the computer executes the method provided in the first aspect.

[0017] In a fifth aspect, the present application also provides a computer program product, including a computer program, which, when executed by a computer, enables the computer to execute the method provided in the first aspect.

[0018] The technical solution provided by this application has at least the following technical effects or advantages: The method, device and electronic device for determining the structural characteristic parameters of the energy-absorbing structure provided in the present application input candidate structural characteristic parameters of the energy-absorbing structure and the corresponding three-dimensional point cloud data into a first prediction model to predict the three-dimensional point cloud data and collision performance characteristics of the energy-absorbing structure after deformation under set collision conditions, determine the collision deformation mode of the energy-absorbing structure under set collision conditions based on the deformed three-dimensional point cloud data, and determine the target structural characteristic parameters of the energy-absorbing structure from multiple groups of candidate structural characteristic parameters based on the collision performance characteristics and collision deformation mode of the energy-absorbing structure under set collision conditions. The method, device and electronic device for determining the structural characteristic parameters of the energy-absorbing structure provided in the present application efficiently determine the target structural characteristic parameters that optimize the collision safety protection function of the energy-absorbing structure based on the collision performance characteristics and collision deformation mode of the energy-absorbing structure.

[0019] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present application. Also, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings: Figure 1 This is a flow chart of a method for determining structural characteristic parameters of an energy absorbing structure in an embodiment of the present application; Figure 2 This is a flowchart of the first prediction model training in the embodiment of the present application; Figure 3 This is a flow chart of the recognition model training in the embodiment of this application; Figure 4 A flow chart of determining target structural characteristic parameters of an energy absorbing structure from multiple groups of candidate structural characteristic parameters in an embodiment of the present application; Figure 5 This is an overall flow chart of energy absorption structure optimization in the embodiment of the present application; Figure 6 A wireframe schematic diagram of a device for determining structural characteristic parameters of an energy-absorbing structure in an embodiment of the present application; Figure 7 Optimize the unit wireframe diagram in the embodiment of this application; Figure 8 It is a wireframe diagram of the target optimization unit in the embodiment of the present application; Fig. 9 It is a wireframe diagram of an electronic device in an embodiment of the present application. DETAILED DESCRIPTION

[0021] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings.

[0022] Various structural schematic diagrams according to embodiments of the present application are shown in the accompanying drawings. These figures are not drawn to scale, and some details are magnified and some details may be omitted for the purpose of clear expression. The shapes of various regions and layers shown in the figures and the relative sizes and positional relationships therebetween are only exemplary, and may deviate in practice due to manufacturing tolerances or technical limitations, and those skilled in the art may further design regions / layers with different shapes, sizes, and relative positions according to actual needs.

[0023] In order to facilitate the clear description of the technical solutions of the embodiments of the present application, in the embodiments of the present application, words such as "first" and "second" are used to distinguish between identical or similar items with substantially the same functions and effects. For example, the first value and the second value are only used to distinguish different values, and their order is not limited. Those skilled in the art can understand that words such as "first" and "second" do not limit the quantity and execution order, and words such as "first" and "second" do not necessarily limit them to be different.

[0024] It should be noted that, in this application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in this application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0025] In order to better understand the above-mentioned technical scheme, the above-mentioned technical scheme will be described in detail below in combination with specific implementation methods. It should be understood that the embodiments of the present disclosure and the specific features in the embodiments are detailed descriptions of the technical scheme of the present application, rather than limitations on the technical scheme of the present application. In the absence of conflict, the embodiments of the present application and the technical features in the embodiments can be combined with each other.

[0026] During the use of vehicles, there is a risk of collision accidents, so the vehicle needs to be equipped with an energy-absorbing structure, such as an energy-absorbing structure made of medium-strength steel or aluminum, to absorb the impact force and reduce the harm of collision accidents. In order to achieve the best collision energy absorption effect, the energy-absorbing structure needs to be optimized during the vehicle design stage to ensure the collision safety protection function of the energy-absorbing structure.

[0027] The prior art constructs a finite element analysis model, modifies the structural characteristic parameters of the energy absorbing structure multiple times, and performs collision simulation to output corresponding collision performance characteristics, from which structural characteristic parameters that make the collision performance characteristics of the energy absorbing structure better are selected.

[0028] However, the collision safety protection function of the energy absorbing structure is related to the collision performance characteristics and collision deformation mode of the energy absorbing structure. If the structural characteristic parameters that make the collision performance characteristics of the energy absorbing structure better cannot make the collision deformation mode of the energy absorbing structure meet the standards, it is necessary to re-perform the collision simulation to obtain better structural characteristic parameters. Therefore, the prior art uses the collision performance characteristics of the energy absorbing structure as a single goal to achieve the structural optimization of the energy absorbing structure, and it is difficult to efficiently obtain the optimization result that optimizes the collision safety protection function of the energy absorbing structure, resulting in a long design cycle and high cost for the vehicle energy absorbing structure.

[0029] Therefore, the optimization of energy-absorbing structures is a hot issue in the field of vehicle collision safety. In addition to the energy-absorbing characteristics of the structure itself, its collision deformation mode is also an important factor affecting the collision safety protection of the whole vehicle. Taking thin-walled energy-absorbing structures as an example, the optimization of thin-walled energy-absorbing structures is basically based on the method based on the agent model. Its main process includes structural design, finite element analysis and model construction and verification, determination of design variables and optimization goals, experimental design, construction of agent models, and multi-objective optimization. The collision deformation mode of thin-walled energy-absorbing structures is not only related to the actual energy absorption during the collision, but also has a certain impact on the posture of the whole vehicle during the collision.

[0030] In view of this, the embodiment of the present application provides a method for determining the structural characteristic parameters of an energy absorbing structure, which efficiently determines the target structural characteristic parameters that optimize the collision safety protection function of the energy absorbing structure based on the collision performance characteristics and collision deformation mode of the energy absorbing structure. Figure 1 , Figure 1 This is a flow chart of a method for determining structural characteristic parameters of an energy absorbing structure in an embodiment of the present application. The method for determining structural characteristic parameters of an energy absorbing structure provided in an embodiment of the present application includes the following operations S101 to S103.

[0031] S101 , obtaining multiple groups of candidate structural characteristic parameters of the energy absorbing structure, and three-dimensional point cloud data of the energy absorbing structure corresponding to each group of candidate structural characteristic parameters.

[0032] The structural characteristic parameters of the energy absorbing structure include: structural length, effective cross-sectional area, and wall thickness. Each structural characteristic parameter has a preset value range, so the energy absorbing structure has a variety of different structural characteristic parameters.

[0033] In some optional implementations, a finite element simulation model of the energy absorbing structure is constructed based on the candidate structural characteristic parameters, and three-dimensional point cloud data of the energy absorbing structure corresponding to the candidate structural characteristic parameters is obtained based on the finite element simulation model.

[0034] In the above operation, initially, an initial finite element simulation model of the energy absorbing structure can be constructed based on the real structural characteristic parameters of the energy absorbing structure; the real collision working condition conditions are input into the initial finite element simulation model to simulate the real collision working condition, and the collision deformation simulation data of the energy absorbing structure is output; the collision deformation simulation data of the energy absorbing structure is compared with the collected real collision deformation data of the energy absorbing structure, and if they match, the construction of the finite element simulation model of the energy absorbing structure is completed; or if they do not match, the modeling parameters of the initial finite element model are modified, such as boundary conditions, constraints, material constitutive (material parameters, failure mode and yield strength). Wherein, the real collision deformation data includes: the collision deformation image sequence, mechanical response curve and collision performance characteristic results of the energy absorbing structure in the historical real collision working condition.

[0035] After determining multiple groups of candidate structural characteristic parameters, the structural length, effective cross-sectional area, and wall thickness of the finite element simulation model of the energy-absorbing structure are set according to each group of candidate structural characteristic parameters, and the three-dimensional point cloud data of the energy-absorbing structure corresponding to each group of candidate structural characteristic parameters can be output based on the finite element simulation model.

[0036] S102. For each group of candidate structural characteristic parameters, the candidate structural characteristic parameters and the three-dimensional point cloud data of the energy-absorbing structure corresponding to the candidate structural characteristic parameters are input into a preset prediction model to obtain first prediction data and second prediction data, and the first prediction data is input into a preset recognition model to obtain a collision deformation mode in which the energy-absorbing structure corresponding to the candidate structural characteristic parameters deforms under set collision conditions.

[0037] The first prediction data is the three-dimensional point cloud data of the energy-absorbing structure corresponding to the candidate structural characteristic parameters after deformation under the set collision condition, and the second prediction data is used to characterize the collision performance characteristics of the energy-absorbing structure corresponding to the candidate structural characteristic parameters after deformation under the set collision condition. The set collision condition can be, for example, a frontal 50% offset collision condition, which is not limited here.

[0038] In some optional implementations, the first prediction model is obtained by the following steps: The first step is to obtain first training sample data, which includes: structural characteristic parameters and three-dimensional point cloud data of at least one energy absorbing structure sample, and deformed three-dimensional point cloud data corresponding to each energy absorbing structure sample, as well as collision performance characteristic data.

[0039] Exemplarily, the structural characteristic parameters of the finite element simulation model of the energy-absorbing structure obtained by the above operation are first set as the structural characteristic parameters of the energy-absorbing structure sample, that is, the finite element simulation model of the energy-absorbing structure sample is obtained, and the finite element simulation model of the energy-absorbing structure sample can be used to output the three-dimensional point cloud data corresponding to the structural characteristic parameters of the energy-absorbing structure sample. This three-dimensional point cloud data refers to the grid node sequence data set of the energy-absorbing structure sample before the collision. Then, the finite element simulation model of the energy-absorbing structure is used to perform collision simulation, and the deformation three-dimensional point cloud data and collision performance characteristic data of the energy-absorbing structure sample under the set collision working conditions can be output. The collision performance characteristic data include energy absorption, specific energy absorption, and collision peak force.

[0040] It is understandable that, usually, the finite element simulation model of the energy absorbing structure obtained by the above operation is obtained by meshing the three-dimensional geometric model of the energy absorbing structure. If the structural characteristic parameters of the energy absorbing structure sample are simple structural parameters such as length, cross-sectional area, and wall thickness, the settings can be directly modified in the finite element simulation model of the energy absorbing structure obtained by the above operation. If it is other more complex parameters, it is necessary to set them in the geometric model of the energy absorbing structure, and then obtain the finite element model through meshing. Therefore, when obtaining the first training sample data, the finite element simulation model constructed based on the structural characteristic parameters of the energy absorbing structure sample outputs the three-dimensional point cloud data corresponding to the structural characteristic parameters of the energy absorbing structure sample, and does not limit the specific construction process of the finite element simulation model of the energy absorbing structure sample.

[0041] The second step is to train the pre-built first initial neural network model based on the first training sample data to obtain a first prediction model.

[0042] like Figure 2 As shown, the structural characteristic parameters and three-dimensional point cloud data of the energy-absorbing structure sample are input into the first initial neural network model, and after feature fusion and feature extraction, the deformed three-dimensional point cloud prediction value and the collision performance characteristic prediction value are output. The deformed three-dimensional point cloud data and the collision performance characteristic data corresponding to the energy-absorbing structure sample in the first sample data are used as the true value, and the regression loss between the predicted value and the true value is calculated. If divergence occurs, the operation of feature fusion is returned, and if convergence occurs, the training is terminated.

[0043] For example, the mean square error (MSE) is used to calculate the regression loss between the predicted value and the true value, and its expression is as follows: In the formula, represents the predicted value of the crash performance characteristic, represents the true value of the crash performance characteristic, is the weighting coefficient, ranging from 0 to 1, and each training data The actual output is , model pair The predicted value is This process is repeated until the regression loss converges, which means the training of the first initial neural network model is completed.

[0044] In some optional implementations, the recognition model in the above operation S102 is obtained by the following steps: The first step is to obtain second training sample data, where the second training sample data includes: deformed three-dimensional point cloud data of at least one energy absorbing structure sample and a corresponding collision deformation mode.

[0045] Exemplarily, a finite element simulation model of the energy absorbing structure sample is constructed through the structural characteristic parameters of the energy absorbing structure sample, and based on the finite element simulation model of the energy absorbing structure sample, three-dimensional point cloud data of the energy absorbing structure sample is output. Then, a collision simulation is performed through the finite element simulation model of the energy absorbing structure sample, and a deformation three-dimensional point cloud sequence of the energy absorbing structure sample and a corresponding collision deformation pattern are output.

[0046] The collision deformation modes of the energy-absorbing structure include layer-by-layer crushing mode, quadrilateral folding crushing mode, octagonal folding crushing mode, anti-folding crushing mode, etc. The finite element simulation model of the energy-absorbing structure sample performs a collision simulation and outputs a set of deformed three-dimensional point cloud sequences and a collision deformation mode. Therefore, Figure 3 As shown, the obtained deformed 3D point cloud sequence is subjected to data enhancement by applying a random transformation function to obtain multiple sets of enhanced deformed 3D point cloud sequences. The specific methods of data enhancement include flipping, rotation, cropping, scaling, translation, jittering, etc., which are not described here. The enhanced deformed 3D point cloud sequence and the corresponding collision deformation mode constitute the second training sample data.

[0047] The second step is to train the pre-built second initial neural network model based on the second training sample data to obtain a recognition model.

[0048] Still Figure 3 As shown, the enhanced deformed three-dimensional point cloud sequence is input into the second initial neural network model, and after the convolutional neural network encoding processing, the mapping vector feature is output, that is, the predicted value of the collision deformation mode, the predicted value of the collision deformation mode is compared with the true value, and the cosine similarity is calculated to measure the similarity between the two vectors. The regression loss is calculated by the comparison loss function until convergence, and the recognition model can be output.

[0049] It is understandable that there are multiple collision deformation modes, and corresponding recognition models can be trained using each collision deformation mode and the corresponding deformed three-dimensional point cloud data. That is to say, multiple recognition models can be trained for multiple collision deformation modes. The recognition model can be used to classify or cluster the structural collision deformation process presented by the structural collision deformation manifestation, map the data to a feature vector dimensional space using a deep neural network, and maximize the distance between the feature vectors of dissimilar data samples while minimizing the distance between the feature vectors of similar data samples, thereby obtaining a specific structural collision deformation mode.

[0050] When executing the operation of S102, if the first prediction data is input into the corresponding recognition model, the collision deformation mode corresponding to the first prediction data is output, and if the first prediction data is input into the mismatched recognition model, the collision deformation mode corresponding to the first prediction data cannot be output. For example, if a group of deformed three-dimensional point clouds of the first prediction data belongs to the layer-by-layer crushing mode, if the group of first prediction data is input into the layer-by-layer crushing mode recognition model, the correct result is output, and if the group of first prediction data is input into the quadrilateral folding crushing mode recognition model, it is difficult to recognize the correct result.

[0051] It can be understood that the recognition model corresponding to each collision deformation mode can be set to directly output whether it belongs to the collision deformation mode corresponding to the current recognition model. In this way, the first prediction data output by the first prediction model is input into the multiple recognition models obtained in the above operation for identification. Under normal circumstances, one of the recognition models will output the collision deformation mode in which the energy-absorbing structure corresponding to the first prediction data is deformed under the set collision working conditions. It is difficult for other recognition models to identify the collision deformation mode based on the first prediction data, and the output results can be eliminated. There is no limitation here.

[0052] The recognition model corresponding to each collision deformation pattern can also output the probability of the collision deformation pattern corresponding to the current recognition model. In this way, the first prediction data output by the first prediction model is input into the multiple recognition models obtained in the above operation for identification respectively. The collision deformation pattern corresponding to the recognition model with the highest probability is taken, and the ones with lower probabilities output by other recognition models can be eliminated. There is no limitation here.

[0053] S103 . Determine target structural characteristic parameters of the energy absorbing structure from multiple groups of candidate structural characteristic parameters based on the second prediction data and the collision deformation mode corresponding to each group of candidate structural characteristic parameters.

[0054] In some optional embodiments, such as Figure 4 As shown, the operation of S103 specifically includes: S401. Based on the second prediction data of the candidate structural characteristic parameters, select the candidate structural characteristic parameters whose energy absorption, specific energy absorption and collision peak force meet the preset screening conditions from multiple groups of candidate structural characteristic parameters to obtain a first set of parameters to be optimized.

[0055] S402 , based on a preset multi-objective optimization model, and second prediction data and collision deformation patterns corresponding to each group of candidate structural characteristic parameters in the first parameter set to be optimized, determine target structural characteristic parameters of the energy absorbing structure from the first parameter set to be optimized.

[0056] In some optional embodiments, based on a preset multi-objective optimization model and second prediction data and collision deformation mode corresponding to each set of candidate structural characteristic parameters in the first set of parameters to be optimized, target structural characteristic parameters of the energy absorbing structure are determined from the first set of parameters to be optimized, including: Determine the target collision deformation mode; Based on the target collision deformation mode and a preset multi-objective optimization model, candidate structural characteristic parameters whose collision deformation mode is the target collision deformation mode are screened from the first parameter set to be optimized to obtain a second optimization parameter set, and energy absorption, specific energy absorption and collision peak force are taken as optimization targets. Based on the second prediction data of each group of candidate structural characteristic parameters in the second optimization parameter set, the target structural characteristic parameters are determined from the second optimization parameter set.

[0057] The structural characteristic parameters include the structural length, effective cross-sectional area and wall thickness of the energy absorbing structure; illustratively, the multi-objective optimization model includes an optimization function and constraint conditions, wherein the optimization function is: ; The constraints are: ; In the formula, is the structural length of the energy absorbing structure, is the minimum value of the structure length, is the maximum value of the structure length, is the effective cross-sectional area of ​​the energy absorbing structure, The minimum effective cross-sectional area, is the maximum value of the effective cross-sectional area, is the wall thickness of the energy absorbing structure, Minimum wall thickness, is the maximum value of wall thickness, EA is the energy absorption, SEA is the specific energy absorption, and PCF is the peak collision force; is the collision deformation mode of the energy absorbing structure, specifically refers to the collision deformation mode of the energy absorbing structure identified by the recognition model based on the deformed three-dimensional point cloud data output by the first prediction model after the deformed three-dimensional point cloud data is input into the recognition model, The target collision deformation mode.

[0058] The collision deformation mode output by the recognition model can be represented by a symbol, and the symbol can be a single number, letter, character, or a combination of numbers, letters, and characters. For example, when a single number is used to represent the collision deformation mode, the recognition model can be set to output the number 1 when it determines that the collision deformation mode of the energy-absorbing structure is a layer-by-layer crushing mode according to the deformed three-dimensional point cloud, output the number 2 when it determines that the collision deformation mode of the energy-absorbing structure is a quadrilateral folding crushing mode according to the deformed three-dimensional point cloud, output the number 3 when it determines that the collision deformation mode of the energy-absorbing structure is an octagonal folding crushing mode according to the deformed three-dimensional point cloud, and output the number 4 when it determines that the collision deformation mode of the energy-absorbing structure is an anti-folding crushing mode according to the deformed three-dimensional point cloud.

[0059] Correspondingly, the target collision deformation mode Can be set to the corresponding symbol. For example, the target collision deformation mode is the layer-by-layer crushing mode. The candidate structural characteristic parameters corresponding to the deformed three-dimensional point cloud that makes the recognition model output 1 are selected from the first set of parameters to be optimized to form the second set of optimized parameters. From the second set of optimized parameters, the candidate structural characteristic parameters that meet the requirements of maximum energy absorption, maximum specific energy absorption, and minimum collision peak force are used as the target candidate structural characteristic parameters. The energy-absorbing structure with the target structural characteristic parameters has the best collision performance characteristics in the layer-by-layer crushing mode.

[0060] The NSGA-II algorithm can also be used to solve the optimization variables (structural length L, effective cross-section S, wall thickness T) in the multi-objective optimization model to obtain the Pareto solution set of the optimization objectives (energy absorption EA, specific energy absorption SEA, collision peak force PCF and collision deformation mode). The minimum distance method is used to make the optimal decision on the optimized Pareto solution set to obtain the optimal solution.

[0061] Among them, the NSGA-II genetic algorithm includes the following steps: Step 1: Set algorithm parameters; Step 2: Generate the initial population (a set of random design variables, i.e., the structural characteristic parameters of the energy-absorbing structure), and obtain the first generation of offspring population through the three basic operations of selection, crossover, and mutation of the genetic algorithm after non-dominated sorting; Step 3: Starting from the second generation, merge the parent population with the child population and perform a fast non-dominated sort; Step 4: Calculate the crowding degree of individuals in each non-dominated layer, and select appropriate individuals to form a new parent population based on the non-dominated relationship and the crowding degree of the individuals; Step 5: Generate a new offspring population through the basic operations of the genetic algorithm; Step 6: Determine whether the convergence conditions are met. If so, generate the final design parameters as candidate structural characteristic parameters; if not, regenerate a new parent population.

[0062] The operations of S101 to S103 in the above-mentioned embodiments, combined with the multi-modal fusion related theories in the field of machine learning, can be applied to the multi-modal machine learning optimization of the energy absorption structure in the whole vehicle collision simulation process. Figure 5 This paper introduces how the method provided in the embodiment of the present application is applied to the optimization of the thin-walled energy absorbing structure of a vehicle. The overall process is as follows: Figure 5 shown.

[0063] 1. Constructing a structural collision analysis dataset The preliminary work of machine learning requires a large amount of data accumulation. It is necessary to construct an analytical data set for energy-absorbing structure collision based on experimental research and with the help of finite element analysis technology.

[0064] Firstly, a corresponding finite element model was constructed according to the initial design of the energy-absorbing structure (real structural performance parameters). The structure was subjected to material stretching and dynamic impact tests, and the rationality of the constructed finite element model was verified by comparing the collision deformation image sequence, mechanical response curves and key collision response indicators.

[0065] Determine the value range of the design variables to be optimized (structural length L, effective cross-sectional area S, wall thickness T), and use the finite element model of the energy-absorbing structure to calculate the structural characteristic parameters, i.e., the design variables It takes the structural deformation data (i.e., deformed finite element mesh node point cloud or deformed 3D point cloud) and energy absorption optimization target (i.e., collision performance characteristics such as energy absorption EA, specific energy absorption SEA, and collision peak force PCF) as the output.

[0066] It is understandable that the collision performance characteristics are used to indicate the collision performance characteristics of the energy absorbing structure that undergoes deformation under set collision conditions. The collision deformation mode of the energy absorbing structure can be analyzed based on the deformed finite element mesh node point cloud, that is, the three-dimensional deformation point cloud.

[0067] The Latin hypercube experimental design method is used to perform scattered point generation experimental design, analyze and establish the functional relationship between the design variables (structural characteristic parameters) and the response variables (i.e., collision performance characteristics) and collision deformation modes, and then obtain the structural collision deformation data. Based on this, a structural collision analysis data set based on finite element simulation data is constructed.

[0068] 2. Building a multimodal machine learning prediction agent model Since the collision performance characteristics and the structural collision deformation pattern representation often do not belong to the same data type, the two cannot be naturally combined for numerical fitting. Therefore, it is necessary to establish a multimodal machine learning prediction agent model, that is, the first prediction model of the above embodiment.

[0069] See also Figure 2 The first prediction model construction process shown in the figure maps the input structural design variables (i.e. structural characteristic parameters) and the initial finite element mesh node point cloud (3D point cloud data) of the two modes into a vector space of uniform dimension for representation through the initial neural network model, and then fuses and extracts the features of this unified feature vector, and finally outputs the collision performance characteristics and the deformed 3D point cloud as the predicted value. Finally, the regression loss between the predicted value and the true value is calculated, and this process is continuously repeated until convergence, which means the training of the model is completed.

[0070] 3. Constructing a collision deformation pattern recognition model The structure deformation output by the multimodal machine learning prediction agent model is a point cloud sequence at different times during the structure deformation process. Therefore, a collision deformation pattern recognition model, i.e., a recognition model, is established to classify or cluster the structure collision deformation process presented by the structure collision deformation expression form, and use a deep neural network to map the data to a feature vector dimensional space, and while minimizing the distance between the feature vectors of similar data samples, maximize the distance between the feature vectors of dissimilar data samples, and then obtain a specific structure collision deformation pattern. The construction operation of the collision deformation pattern recognition model is similar to the above-mentioned embodiments and Figure 3 The construction operations of the provided recognition models are the same and will not be repeated here.

[0071] 4. Multi-objective optimization under the constraints of structural collision deformation mode Based on the constructed multimodal machine learning prediction agent model of the collision deformation process and the structural collision deformation pattern recognition model, the NSGA II genetic algorithm is used to establish a multi-objective optimization method for structural crashworthiness optimization under the constraints of collision deformation patterns. Figure 5As shown in the figure, firstly, a set of random design variables (structural characteristic parameters) are generated as population samples; then, when calculating the response and the corresponding collision deformation representation, the initial state deformation representation of the structure is generated by the design variables, and after preprocessing, it is input into the multimodal machine learning prediction agent model of the structural collision deformation process together with the design variables to obtain the structural deformation prediction (i.e., deformation three-dimensional point cloud) and energy absorption characteristic prediction (i.e., collision performance characteristics) corresponding to the sample point; then, the obtained structural deformation prediction is input into the collision deformation pattern recognition model to obtain the collision deformation pattern; then, multi-objective optimization is performed under the constraints of the collision deformation pattern, and the population information is updated using the NSGA II genetic algorithm until the convergence criterion is met to obtain the optimal design variable value of the energy-absorbing structure, i.e., the target structural characteristic parameters.

[0072] The specific optimization process is as follows: Based on the established finite element model, the structural length L, effective cross-sectional area S, and wall thickness T are used as optimization variables; the collision performance characteristics (energy absorption EA, specific energy absorption SEA, and collision peak force PCF prediction value) output by the multimodal machine learning prediction agent model and the collision deformation mode output by the collision deformation mode recognition model are used as optimization targets to establish a multi-objective optimization theoretical model: ; ; ; Where x is the compression displacement of the energy absorbing structure; is the axial force of the energy absorbing structure, is the mass of the energy absorbing structure, Represents the collision deformation mode of the energy-absorbing structure.

[0073] It is understandable that the collision deformation mode The deformed three-dimensional point cloud output by the first prediction model is input into the recognition model. The collision deformation mode of the energy-absorbing structure output by the recognition model under the set collision conditions is obtained. If the first prediction model and the recognition model are taken as a whole, the collision deformation mode The design variable For the independent variable.

[0074] The NSGA-II algorithm is used to solve the optimization variables (structural length L, effective cross-sectional area S, wall thickness T) in the multi-objective optimization theoretical model, and the Pareto solution set of the optimization objectives (crash performance characteristics (energy absorption EA, specific energy absorption SEA, collision peak force PCF) prediction value, and collision deformation mode prediction value) is obtained. The minimum distance method is used to make the optimal decision for the optimized Pareto solution set to obtain the optimal solution.

[0075] In this way, the point cloud composed of the coordinates of the structural finite element grid nodes at each moment obtained by finite element simulation is used as the representation of the structural collision deformation. Since the drawing of finite element simulation images is usually based on the solution of the grid node displacement data of the structural collision process, it retains almost all the structural collision deformation information, can represent the collision deformation inside and outside the structure, and is not limited by the observation angle. After determining the representation of the structural collision deformation, a multimodal machine learning prediction model and a collision deformation pattern recognition model for the energy-absorbing structure collision process are constructed based on this, and the two are combined with the multi-objective optimization method under the constraints of its structural collision deformation pattern to jointly complete the energy absorption characteristics optimization analysis of the energy-absorbing structure.

[0076] The embodiment of the present application adopts a thin-walled structure collision deformation representation method based on a finite element mesh node coordinate point cloud sequence, which does not require complex parameter analysis, theoretical derivation, etc. of the structural design parameters, and can retain most of the structural collision deformation information.

[0077] In the existing structural crashworthiness optimization analysis process under the constraints of collision deformation mode, it is necessary to solve the structural crashworthiness response and collision deformation mode thousands of times. If the finite element method is used for simulation and solution, the time cost and computing resources consumed are incalculable. This application constructs a proxy model that can simultaneously complete the prediction of structural crashworthiness response and structural collision deformation representation, accelerating the structural crashworthiness optimization process.

[0078] The proxy model constructed in the embodiment of the present application can simultaneously and efficiently predict the structural crashworthiness response and collision deformation process, effectively overcoming the limitation that traditional proxy models cannot process complex unstructured data (such as images, point clouds, etc.), thereby improving optimization efficiency and reducing the consumption of computing resources.

[0079] Based on the same inventive concept, the embodiment of the present application also provides a device for determining structural characteristic parameters of an energy absorbing structure, such as Figure 6 As shown, Figure 6 This is a wireframe diagram of a device for determining structural characteristic parameters of an energy absorbing structure in an embodiment of the present application. The device 600 for determining structural characteristic parameters of an energy absorbing structure includes: An acquisition unit 601 is used to acquire multiple groups of candidate structural characteristic parameters of the energy absorbing structure, and three-dimensional point cloud data of the energy absorbing structure corresponding to each group of candidate structural characteristic parameters; A prediction unit 602 is used to input the candidate structural characteristic parameters and the three-dimensional point cloud data of the energy-absorbing structure corresponding to the candidate structural characteristic parameters into a preset prediction model for each group of candidate structural characteristic parameters, to obtain first prediction data and second prediction data, and to input the first prediction data into a preset recognition model to obtain a collision deformation mode of the energy-absorbing structure corresponding to the candidate structural characteristic parameters under a set collision condition, wherein the first prediction data is the three-dimensional point cloud data of the energy-absorbing structure corresponding to the candidate structural characteristic parameters after deformation under the set collision condition, and the second prediction data is used to characterize the collision performance characteristics of the energy-absorbing structure corresponding to the candidate structural characteristic parameters after deformation under the set collision condition; The optimization unit 603 is used to determine the target structural characteristic parameters of the energy absorbing structure from the multiple groups of candidate structural characteristic parameters based on the second prediction data and the collision deformation mode corresponding to each group of candidate structural characteristic parameters.

[0080] The structural characteristic parameter determination device 600 of the energy absorbing structure provided in the embodiment of the present application can be used to execute the operations of the above-mentioned embodiments S101 to S103, so as to efficiently determine the target structural characteristic parameters that optimize the collision safety protection function of the energy absorbing structure based on the collision performance characteristics and collision deformation mode of the energy absorbing structure.

[0081] In some optional embodiments, the collision performance characteristics include: energy absorption, specific energy absorption, and collision peak force; Figure 7 As shown, the optimization unit 603 includes a screening unit 701 and a target optimization unit 702 .

[0082] The screening unit 701 is used to screen out candidate structural characteristic parameters whose energy absorption, specific energy absorption and collision peak force meet preset screening conditions from multiple groups of candidate structural characteristic parameters based on the second prediction data of the candidate structural characteristic parameters, and obtain a first set of parameters to be optimized.

[0083] The multi-objective optimization unit 702 is used to determine the target structural characteristic parameters of the energy absorbing structure from the first parameter set to be optimized based on a preset multi-objective optimization model and the second prediction data and collision deformation mode corresponding to each group of candidate structural characteristic parameters in the first parameter set to be optimized.

[0084] In some optional embodiments, such as Figure 8 As shown, the multi-objective optimization unit 702 specifically includes: The target collision deformation mode determining unit 801 is used to determine the target collision deformation mode.

[0085] The target structural characteristic parameter determination unit 802 is used to screen candidate structural characteristic parameters whose collision deformation mode is the target collision deformation mode from the first parameter set to be optimized based on the target collision deformation mode and a preset multi-objective optimization model, obtain a second optimization parameter set, and determine from the second optimization parameter set based on the second prediction data of each group of candidate structural characteristic parameters in the second optimization parameter set, taking energy absorption, specific energy absorption and collision peak force as optimization targets.

[0086] An embodiment of the present application further provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a computer, the computer executes the method provided in the above embodiment.

[0087] The present application also provides an electronic device, such as Fig. 9 As shown, the electronic device 900 includes a memory 901, a processor 902, and a computer program 903 stored in the memory 901 and executable on the processor 902. When the processor 902 executes the computer program 903, the electronic device 900 executes the method provided in the above embodiment.

[0088] The electronic device 900 provided in the embodiment of the present application can be used to execute the operations of the above embodiments S101 to S103 to achieve efficient determination of target structural characteristic parameters that optimize the collision safety protection function of the energy absorbing structure based on the collision performance characteristics and collision deformation mode of the energy absorbing structure.

[0089] The embodiments of the present application also provide a computer program product, including a computer program. When the computer program is executed, the computer executes the method provided in the above embodiments.

[0090] Those skilled in the art can understand that all or part of the steps of implementing the above-mentioned method embodiments can be completed by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, the steps of the above-mentioned method embodiments are executed; and the aforementioned storage medium includes: ROM, RAM, disk or optical disk and other media that can store program codes.

[0091] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0092] In the description provided herein, a large number of specific details are described. However, it is understood that the embodiments of the present application can be practiced without these specific details. In some instances, well-known methods, structures and techniques are not shown in detail so as not to obscure the understanding of this description.

Claims

1. A method for determining structural characteristic parameters of an energy absorbing structure, characterized in that: include: Acquire multiple groups of candidate structural characteristic parameters of the energy absorbing structure, and three-dimensional point cloud data of the energy absorbing structure corresponding to each group of the candidate structural characteristic parameters; For each group of candidate structural characteristic parameters, the candidate structural characteristic parameters and the three-dimensional point cloud data of the energy-absorbing structure corresponding to the candidate structural characteristic parameters are respectively input into a preset prediction model to obtain first prediction data and second prediction data, and the first prediction data is input into a preset recognition model to obtain a collision deformation mode of the energy-absorbing structure corresponding to the candidate structural characteristic parameters under set collision conditions, wherein the first prediction data is the three-dimensional point cloud data of the energy-absorbing structure corresponding to the candidate structural characteristic parameters after deformation under the set collision conditions, and the second prediction data is used to characterize the collision performance characteristics of the energy-absorbing structure corresponding to the candidate structural characteristic parameters when deformed under the set collision conditions; Based on the second prediction data and the collision deformation mode corresponding to each group of the candidate structural characteristic parameters, target structural characteristic parameters of the energy absorbing structure are determined from the multiple groups of candidate structural characteristic parameters.

2. The method for determining structural characteristic parameters of an energy absorbing structure according to claim 1, characterized in that: The collision performance characteristics include: energy absorption, specific energy absorption, and collision peak force; the determining the target structural characteristic parameters of the energy absorbing structure from the multiple groups of candidate structural characteristic parameters based on the second prediction data and the collision deformation mode corresponding to each group of candidate structural characteristic parameters includes: Based on the second prediction data of the candidate structural characteristic parameters, the candidate structural characteristic parameters whose energy absorption, specific energy absorption and collision peak force meet the preset screening conditions are screened out from the multiple groups of candidate structural characteristic parameters to obtain a first parameter set to be optimized; Based on a preset multi-objective optimization model, and the second prediction data and collision deformation mode corresponding to each group of candidate structural characteristic parameters in the first parameter set to be optimized, the target structural characteristic parameters of the energy absorbing structure are determined from the first parameter set to be optimized.

3. The method for determining structural characteristic parameters of an energy absorbing structure according to claim 2, characterized in that: Based on a preset multi-objective optimization model, and second prediction data and collision deformation mode corresponding to each group of candidate structural characteristic parameters in the first parameter set to be optimized, target structural characteristic parameters of the energy absorbing structure are determined from the first parameter set to be optimized, including: Determine the target collision deformation mode; Based on the target collision deformation mode and a preset multi-objective optimization model, candidate structural characteristic parameters whose collision deformation mode is the target collision deformation mode are screened from the first parameter set to be optimized to obtain a second optimization parameter set, and taking the energy absorption, specific energy absorption and collision peak force as optimization targets, based on the second prediction data of each group of candidate structural characteristic parameters in the second optimization parameter set, the target structural characteristic parameters are determined from the second optimization parameter set.

4. The method for determining structural characteristic parameters of an energy absorbing structure according to claim 3, characterized in that: The structural characteristic parameters include the structural length, effective cross-sectional area and wall thickness of the energy absorbing structure; the multi-objective optimization model includes an optimization function and constraint conditions, wherein the optimization function is: ; The constraints are: ; In the formula, is the structural length of the energy absorbing structure, is the minimum value of the structure length, is the maximum value of the structure length, is the effective cross-sectional area of ​​the energy absorbing structure, The minimum effective cross-sectional area, is the maximum value of the effective cross-sectional area, is the wall thickness of the energy absorbing structure, Minimum wall thickness, is the maximum value of wall thickness, EA is the energy absorption, SEA is the specific energy absorption, PCF is the peak collision force, is the collision deformation mode of the energy absorbing structure, The collision deformation mode for the target.

5. The method for determining structural characteristic parameters of an energy absorbing structure according to claim 1, characterized in that: The first prediction model is obtained by the following steps: Acquire first training sample data, wherein the first training sample data includes: structural characteristic parameters and three-dimensional point cloud data of at least one energy absorbing structure sample, and deformed three-dimensional point cloud data corresponding to each energy absorbing structure sample, and collision performance characteristic data; A pre-constructed first initial neural network model is trained based on the first training sample data to obtain the first prediction model.

6. The method for determining structural characteristic parameters of an energy absorbing structure according to claim 1, characterized in that: The recognition model is obtained by the following steps: Acquire second training sample data, wherein the second training sample data includes: deformed three-dimensional point cloud data of at least one energy absorbing structure sample and a collision deformation mode under the set collision working condition; The pre-constructed second initial neural network model is trained based on the second training sample data to obtain the recognition model.

7. The method for determining structural characteristic parameters of an energy absorbing structure according to claim 1, characterized in that: Acquiring three-dimensional point cloud data of the energy absorbing structure corresponding to each set of candidate structural characteristic parameters, including: Based on the candidate structural characteristic parameters, a finite element simulation model of the energy absorbing structure is constructed, and based on the finite element simulation model, three-dimensional point cloud data of the energy absorbing structure corresponding to the candidate structural characteristic parameters is obtained.

8. A device for determining structural characteristic parameters of an energy-absorbing structure, characterized in that: include: An acquisition unit, used to acquire multiple groups of candidate structural characteristic parameters of the energy absorbing structure, and three-dimensional point cloud data of the energy absorbing structure corresponding to each group of candidate structural characteristic parameters; A prediction unit, for inputting the candidate structural characteristic parameters and the three-dimensional point cloud data of the energy-absorbing structure corresponding to the candidate structural characteristic parameters into a preset first prediction model for each group of the candidate structural characteristic parameters, to obtain first prediction data and second prediction data, and inputting the first prediction data into a preset recognition model to obtain a collision deformation mode of the energy-absorbing structure corresponding to the candidate structural characteristic parameters under set collision conditions, wherein the first prediction data is the three-dimensional point cloud data of the energy-absorbing structure corresponding to the candidate structural characteristic parameters after deformation under the set collision conditions, and the second prediction data is used to characterize the collision performance characteristics of the energy-absorbing structure corresponding to the candidate structural characteristic parameters when deformed under the set collision conditions; The optimization unit is used to determine the target structural characteristic parameters of the energy absorbing structure from the multiple groups of candidate structural characteristic parameters based on the second prediction data and the collision deformation mode corresponding to each group of candidate structural characteristic parameters.

9. The device for determining structural characteristic parameters of an energy absorbing structure according to claim 8, characterized in that: The collision performance characteristics include: energy absorption, specific energy absorption and collision peak force; the optimization unit includes: A screening unit, configured to screen out candidate structural characteristic parameters whose energy absorption, specific energy absorption and collision peak force meet preset screening conditions from the multiple groups of candidate structural characteristic parameters based on the second prediction data of the candidate structural characteristic parameters, to obtain a first parameter set to be optimized; A target optimization unit is used to determine the target structural characteristic parameters of the energy absorbing structure from the first set of parameters to be optimized based on a preset multi-objective optimization model and the second prediction data and collision deformation mode corresponding to each group of candidate structural characteristic parameters in the first set of parameters to be optimized.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the electronic device executes the method according to any one of claims 1 to 7.

Images