Finite element simulation optimization method and system for crashworthiness of guardrail plate

Abstract

The application belongs to the technical field of computer-aided design in structure or civil engineering, and particularly provides a finite element simulation optimization method and system for crash resistance performance of a guardrail plate, which mainly comprises the following steps: obtaining guardrail plate collision finite element simulation data, dividing the guardrail plate collision finite element simulation data into a stress distribution feature set and a collision condition feature set; performing multi-dimensional feature fusion on the stress distribution feature set to generate a dynamic parameter feature set; generating an anti-collision parameter optimization scheme according to the dynamic parameter feature set; performing anti-collision effect pre-verification on the anti-collision parameter optimization scheme to generate a simulation verification feature set; and generating a final optimization score result according to the collision condition feature set and the simulation verification feature set. The application can realize dynamic adaptation of the parameter optimization direction and the actual collision condition, effectively reduce the risk of deviation of the protection performance, shorten the optimization iteration cycle, and provide more universal optimization basis for guardrail structure design and safety verification.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of computer-aided design in structure or civil engineering, and particularly relates to a finite element simulation optimization method and system for crashworthiness of a guardrail plate. BACKGROUND

[0002] The optimization analysis of the crashworthiness of a guardrail structure is usually realized based on a finite element simulation technology. At present, the mainstream method is to establish a three-dimensional numerical model of the guardrail, simulate the structural deformation and energy absorption characteristics in the vehicle collision process, and then evaluate the protection capability.

[0003] At present, the optimization direction of the crashworthiness of the guardrail structure is usually found by independently adjusting single variables such as material properties and cross-sectional shapes. This processing method ignores the dynamic coupling effect between different mechanical parameters in the collision process, lacks systematic evaluation of the synergistic effect of multiple variables, and may cause deviation of the protection performance of the optimization scheme in the actual complex collision scene, affecting the reliability of the safety evaluation result. SUMMARY

[0004] The application provides a finite element simulation optimization method and system for the crashworthiness of a guardrail plate, which effectively solves the problem that the optimization direction of the crashworthiness of the guardrail structure is found by independently adjusting single variables such as material properties and cross-sectional shapes in the prior art, lacks systematic evaluation of the synergistic effect of multiple variables, and may cause deviation of the protection performance of the optimization scheme in the actual complex collision scene, affecting the reliability of the safety evaluation result, realizes dynamic adaptation of the parameter optimization direction to the actual collision conditions, effectively reduces the risk of deviation of the protection performance, shortens the optimization iteration period, and at the same time ensures the stability and consistency of the safety performance under different collision conditions, provides more universal optimization basis for the design and safety verification of the guardrail structure.

[0005] In order to achieve the above-mentioned purpose, the application adopts the following technical solutions:

[0006] In a first aspect, the application provides a finite element simulation optimization method for the crashworthiness of a guardrail plate, comprising:

[0007] Obtaining guardrail plate collision finite element simulation data, and dividing the guardrail plate collision finite element simulation data into a stress distribution feature set and a collision condition feature set.

[0008] Performing multi-dimensional feature fusion on the stress distribution feature set to generate a dynamic parameter feature set.

[0009] Generating an anti-collision parameter optimization scheme according to the dynamic parameter feature set.

[0010] Performing anti-collision effect pre-verification on the anti-collision parameter optimization scheme to generate a simulation verification feature set.

[0011] According to the collision condition feature set and the simulation verification feature set, a final optimization score result is generated.

[0012] Further, multi-dimensional feature fusion is performed on the stress distribution feature set to generate a dynamic parameter feature set, including:

[0013] A specified number of stress zones are set, all stress zones are divided according to the maximum stress value of each stress zone, and independent features of each stress zone are extracted.

[0014] The stress fluctuation frequency is time-series correlated and analyzed with the independent features of each stress zone to generate a zone-frequency correlation matrix.

[0015] Based on the correlation matrix, an abnormal fluctuation zone is screened, and a dynamic parameter feature set is generated.

[0016] Further, a specified number of stress zones are set, all stress zones are divided according to the maximum stress value of each stress zone, and independent features of each stress zone are extracted, including:

[0017] According to the numerical distribution of the maximum stress value sequence and the preset stress critical value, the guardrail plate is divided into a high stress zone, a medium stress zone, and a low stress zone.

[0018] For the high stress zone: the stress peak interval and the fluctuation frequency are extracted as independent features.

[0019] For the medium stress zone: the stress attenuation rate and the thickness distribution correlation parameter are extracted as independent features.

[0020] For the low stress zone: the stress stability marker is extracted as an independent feature.

[0021] Further, an anti-collision parameter optimization scheme is generated according to the dynamic parameter feature set, including:

[0022] An anti-collision optimization model is used to process the anti-collision parameter optimization scheme, and the anti-collision optimization model increases a multi-modal input layer on the basis of a gradient tree, and generates the anti-collision parameter optimization scheme through a joint splitting rule.

[0023] Further, the anti-collision optimization model increases a multi-modal input layer on the basis of a gradient tree, and generates the anti-collision parameter optimization scheme through a joint splitting rule, including:

[0024] The multi-modal input layer includes a first channel and a second channel; the first channel is used to receive the independent features of the high stress zone and calculate the anti-pressure matching degree of the material replacement type, and the second channel is used to receive the independent features of the medium stress zone and the low stress zone and calculate the deformation suppression rate of the thickness compensation gradient.

[0025] The pressure resistance matching degree is compared with a preset pressure resistance threshold value, and if the pressure resistance matching degree exceeds the pressure resistance threshold value, a matching scheme of a high stress area material replacement type is generated; the deformation inhibition rate is compared with a preset deformation threshold value, and if the deformation inhibition rate is lower than the deformation threshold value, an increment scheme of a thickness compensation gradient is generated based on the medium stress area associated features and the low stress area associated features.

[0026] According to the fluctuation abnormal area, the material replacement scheme and the increment scheme of the thickness compensation are spatially superimposed.

[0027] According to the superimposed result, an anti-collision parameter optimization scheme including a material thickness adjustment value and a stress concentration area correction parameter is output.

[0028] Further, the anti-collision parameter optimization scheme is executed to perform anti-collision effect pre-verification, and a simulation verification feature set is generated, including:

[0029] According to the material thickness adjustment value and the stress concentration area correction parameter, a plurality of candidate parameter combinations are generated.

[0030] According to the plurality of candidate parameter combinations and the collision condition feature set, bidirectional correlation degree analysis screening is performed.

[0031] The screened candidate parameter combinations are verified for effectiveness to obtain verification results.

[0032] The screened candidate parameter combinations are taken as rows, the stress area-frequency association matrix is taken as columns, and the verification results are taken as matrix element values to generate a simulation verification feature set.

[0033] Further, according to the plurality of candidate parameter combinations and the collision condition feature set, bidirectional correlation degree analysis screening is performed, including:

[0034] The candidate parameter combinations are input into the anti-collision optimization model to obtain a pressure resistance matching degree calculation result under the collision speed time series data; the candidate parameter combinations are input into the anti-collision optimization model to obtain a deformation inhibition rate calculation result under the collision contact angle sequence.

[0035] The candidate parameter combinations whose pressure resistance matching degree calculation result exceeds a preset first proportion value and whose deformation inhibition rate calculation result exceeds a preset second proportion value are retained.

[0036] Further, according to the collision condition feature set and the simulation verification feature set, a final optimization score result is generated, including:

[0037] The collision speed time series data and the candidate parameter combinations are time series aligned to generate a speed-parameter time series association matrix; the collision angle time series data and the stress area-frequency association matrix are spatially dimensionally cross-associated to generate an angle-matrix space association matrix.

[0038] The speed-parameter time sequence correlation matrix and the angle-matrix space correlation matrix are processed by using the anti-collision parameter optimization model, the anti-collision parameter optimization model extracts the time correlation features and the space correlation features, and then the features are merged and conflict is eliminated to generate a final optimization score result.

[0039] Further, the feature merging and conflict elimination include:

[0040] The time correlation features and the space correlation features are merged according to the time steps and the space positions to generate comprehensive features.

[0041] The verification results in the feature set include invalid candidate parameter combinations and unstable low-stress area stress stability labels; according to the verification results in the feature set, the conflict of the comprehensive features is eliminated, including:

[0042] If the verification result is marked as an invalid candidate parameter combination, the corresponding time correlation feature is removed.

[0043] If the verification result is marked as an unstable low-stress area stress stability label, the influence range of the space correlation feature is limited.

[0044] In a second aspect, the application provides a finite element simulation optimization system for guardrail plate anti-collision performance, which includes:

[0045] A data division module: acquires guardrail plate collision finite element simulation data, and divides the guardrail plate collision finite element simulation data into a stress distribution feature set and a collision condition feature set;

[0046] A feature fusion module: performs multi-dimensional feature fusion on the stress distribution feature set to generate a dynamic parameter feature set;

[0047] A scheme generation module: generates an anti-collision parameter optimization scheme according to the dynamic parameter feature set;

[0048] An effect verification module: performs anti-collision effect pre-verification on the anti-collision parameter optimization scheme to generate a simulation verification feature set;

[0049] A score output module: generates a final optimization score result according to the collision condition feature set and the simulation verification feature set.

[0050] In a third aspect, the application provides a finite element simulation optimization device for guardrail plate anti-collision performance, which includes a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program to realize the steps of the finite element simulation optimization method for guardrail plate anti-collision performance as described in the first aspect.

[0051] In a fourth aspect, the present application provides a storage medium, wherein the storage medium stores computer program instructions, and the computer program instructions are read and executed by a processor to perform the steps of the guardrail plate crashworthiness finite element simulation optimization method according to the first aspect.

[0052] Advantages of the present application:

[0053] The present application effectively solves the problem in the prior art that the optimization direction of the crashworthiness of the guardrail structure is found by independently adjusting single variables such as material properties and cross-sectional shapes, and the problem that the system evaluation of the multi-variable synergistic effect is lacking, which may lead to deviation of the protection performance in the actual complex collision scene, affecting the reliability of the safety evaluation result, and the problem that the dynamic adaptation of the parameter optimization direction and the actual collision condition is realized, the risk of deviation of the protection performance is effectively reduced, the optimization iteration cycle is shortened, and the stability and consistency of the safety performance under different collision conditions are ensured, thereby providing more universal optimization basis for the design and safety verification of the guardrail structure.

[0054] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art from the description, or can be learned by practice of the present application. The objects and other advantages of the present application can be realized and achieved by the structures indicated in the description and the accompanying drawings. BRIEF DESCRIPTION OF DRAWINGS

[0055] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0056] Figure 1 A flowchart of the guardrail plate crashworthiness finite element simulation optimization method of the present application is shown;

[0057] Figure 2 A module diagram of the guardrail plate crashworthiness finite element simulation optimization system of the present application is shown. DETAILED DESCRIPTION

[0058] In order to solve the problems raised in the background art, the present application integrates the stress distribution of the guardrail plate collision simulation data and the collision condition characteristics, establishes a multi-dimensional dynamic parameter correlation model, and generates an anti-collision parameter optimization scheme based on this, combines a pre-verification mechanism and collision condition matching verification, and finally outputs a comprehensive optimization score, thereby providing more universal optimization basis for the design and safety verification of the guardrail structure.

[0059] In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the following will clearly and completely explain the technical solutions in the embodiments of the present application with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0060] In some embodiments, as shown in FIG. 1, Figure 1 The present application provides a finite element simulation optimization method for the crashworthiness of a guardrail plate, comprising:

[0061] S100. Obtain guardrail plate crash finite element simulation data, and divide the guardrail plate crash finite element simulation data into a stress distribution feature set and a crash condition feature set.

[0062] The dynamic process of a vehicle colliding with a guardrail plate is simulated by a finite element simulation software, and key data in the collision process is output as finite element simulation data.

[0063] The input conditions can include a collision speed V(t) at time t, a collision angle θ(t) at time t, and a vehicle mass m.

[0064] The output data can include stress data and deformation data; wherein the stress data is the Von Mises stress value of each node of the guardrail plate, which can be recorded as a matrix S(x, y, t) according to the time step, wherein x and y represent spatial coordinates, and t represents time; and the deformation data is a displacement matrix D(x, y, t) of each node of the guardrail plate.

[0065] The stress distribution feature set includes a maximum stress value sequence and a stress fluctuation frequency.

[0066] The guardrail plate is divided into multiple zones, such as a 10x10 grid, the maximum stress value of each zone in the time dimension is extracted from the matrix S(x, y, t) to form a maximum stress value sequence S(x, y) = [max(S(x, y, t1)), max(S(x, y, t2)), …, max(S(x, y, tN))], wherein max(S(x, y, tN)) represents the maximum stress value of the zone (x, y) in the time dimension tN. max n The maximum stress value sequence S(x, y) is obtained. n n

[0067] The frequency principal component of the stress change of each stress zone is calculated by Fourier transform to obtain the stress fluctuation frequency of each stress zone.

[0068] ​​​The collision condition feature set includes collision velocity time series data V(t), collision angle time series data θ(t), vehicle mass m, and collision energy E(t). The collision energy E(t) can be calculated according to the functional formula.

[0069] Directly mixing velocity time-series data with stress data can cause the model to mistakenly attribute stress fluctuations caused by velocity changes to material performance issues. By separating velocity time-series data from stress data, the optimization scheme can specifically correct material or structural parameters, thereby avoiding the coupling interference between the two types of data.

[0070] S200. Perform multi-dimensional feature fusion on the stress distribution feature set to generate a dynamic parameter feature set.

[0071] S300. Generate an optimization scheme for collision resistance parameters based on the dynamic parameter feature set.

[0072] S400 performs pre-verification of the collision resistance effect of the collision parameter optimization scheme and generates a simulation verification feature set.

[0073] S500. Generate the final optimized score result based on the collision condition feature set and the simulation verification feature set.

[0074] In some embodiments, S200 performs multi-dimensional feature fusion on the stress distribution feature set to generate a dynamic parameter feature set, including:

[0075] S210. Set a specified number of stress zones, divide all stress zones according to the maximum stress value of each stress zone, and extract the independent features of each stress zone.

[0076] S220. Perform time-series correlation analysis between stress fluctuation frequency and independent characteristics of each stress zone to generate a zone-frequency correlation matrix.

[0077] S230. Screen for fluctuating abnormal areas based on the correlation matrix and generate a dynamic parameter feature set.

[0078] In some embodiments, S210 defines a specified number of stress zones, divides all stress zones according to the maximum stress value of each stress zone, and extracts the independent features of each stress zone, including:

[0079] Based on the numerical distribution of the maximum stress value sequence and the preset stress critical value, the guardrail is divided into high stress zone, medium stress zone, and low stress zone.

[0080] For high-stress areas: extract the stress peak interval and fluctuation frequency as independent features.

[0081] For the medium stress region: extract the stress attenuation rate and thickness distribution correlation parameters as independent features.

[0082] For low-stress regions: extract stress stability markers as independent features.

[0083] The stress zone can be defined by setting a threshold based on the material's yield strength. For example, the critical value for the high stress zone is 250 MPa, and the critical value for the medium stress zone is 150 MPa. The defined stress zones can be high stress zone, medium stress zone, or low stress zone.

[0084] When performing clustering, taking the critical value of 250 MPa for high stress zone and 150 MPa for medium stress zone as an example, the zone with a maximum stress value exceeding 250 MPa in the maximum stress value sequence is classified as high stress zone, the zone with a value between 150 MPa and 250 MPa and the zone with a value below 150 MPa are classified as low stress zone, thus obtaining three stress zones.

[0085] For high-stress areas, the independent features can be the stress peak interval and the fluctuation frequency. The stress peak interval can be obtained by calculating the time difference between adjacent peaks, and the fluctuation frequency represents the number of peaks per unit time.

[0086] For the intermediate stress region, the independent characteristic can be the stress decay rate and the thickness distribution correlation parameter. The formula for the stress decay rate is: Where Ar represents the stress decay rate, σ max σ represents the maximum stress value in the intermediate stress zone over the time range. min Δt represents the minimum stress value in the medium stress zone within a certain time range, and Δt represents the time it takes for the stress to decrease from its maximum value to its minimum value. By calculating the correlation coefficient between stress and thickness, the thickness distribution correlation parameters can be obtained.

[0087] For low-stress areas, an independent feature can be a stress stability marker. Specifically, stability can be calculated using variance, and a stability threshold can be set. If the stability is less than the stability threshold, it is marked as "stable". The stability threshold can be determined based on statistical analysis of historical stress fluctuation data of the guardrail under normal working conditions, industry standards, and practical engineering experience.

[0088] In S220, the correlation between stress fluctuation frequency and independent characteristics of each stress zone, Cov(X,Y), can be calculated using the covariance calculation formula, where X represents the fluctuation frequency sequence and Y represents a certain independent characteristic sequence (such as peak interval).

[0089] In S230, a stress zone-frequency correlation matrix is ​​generated with high stress zone, medium stress zone, and low stress zone as rows and the covariance value of fluctuation frequency and each independent feature as columns.

[0090] Example matrix:

[0091] An anomaly threshold can be set. If the absolute value of the covariance of a certain area exceeds the anomaly threshold, it is determined to be an anomaly zone. The anomaly threshold is determined based on historical simulation data of the guardrail and actual collision test data, or by combining multiple test debugging and industry experience.

[0092] The independent features and covariance values ​​of the outlier region are combined to generate a dynamic parameter feature set that includes the combination of independent features and covariance values ​​of the outlier region.

[0093] In some embodiments, S300 generates a collision resistance parameter optimization scheme based on a dynamic parameter feature set, including: processing the collision resistance parameter optimization scheme using a collision resistance optimization model, wherein the collision resistance optimization model adds a multimodal input layer on the basis of a gradient tree, and generates a collision resistance parameter optimization scheme through joint splitting rules.

[0094] In some embodiments, the collision resistance optimization model adds a multimodal input layer to the gradient tree and generates a collision resistance parameter optimization scheme through joint splitting rules, including:

[0095] S310. The multimodal input layer includes a first channel and a second channel; the first channel is used to receive independent features of the high-stress region and calculate the compressive strength matching degree of the material replacement type, and the second channel is used to receive independent features of the medium-stress region and the low-stress region and calculate the deformation suppression rate of the thickness compensation gradient.

[0096] The input features of the first channel include the stress peak interval and fluctuation frequency in the high-stress zone. The value of comparing the compressive strength of the candidate material with the historical average peak stress in the high-stress zone is marked as the compressive strength matching degree R.

[0097] The second channel's input features include the attenuation rate of the stress zone, the thickness distribution correlation parameters, and the stability markers of the low-stress zone.

[0098] Adjust the thickness parameters of the stress zone of the guardrail panel, input the adjusted thickness parameters into the finite element simulation model, re-simulate the collision process, and output the compensated deformation. Use the reduction ratio of the deformation after thickness compensation as the deformation suppression rate, referring to the following formula: Among them, D r Represents the deformation suppression rate, δ o Represents the original deformation variable, δ c This represents the deformation after compensation.

[0099] Compressive strength matching degree R and deformation inhibition rate D r Discretize it.

[0100] S320. Compare the compressive strength matching degree with the preset compressive strength threshold. If the compressive strength matching degree exceeds the compressive strength threshold, generate a matching scheme for material replacement in the high stress area. Compare the deformation suppression rate with the preset deformation threshold. If the deformation suppression rate is lower than the deformation threshold, generate an incremental scheme for thickness compensation gradient based on the correlation characteristics of the medium stress area and the correlation characteristics of the low stress area.

[0101] The rules for joint splitting include Rule 1 and Rule 2:

[0102] Rule 1: Compare the compressive strength matching degree with a preset compressive strength threshold. If the compressive strength matching degree exceeds the compressive strength threshold, a matching scheme of material replacement type for the high-stress area is generated. For example, if the compressive strength matching degree is less than the preset compressive strength threshold, the material in the high-stress area is triggered to be replaced from Q235 steel to S355 steel.

[0103] Rule 2: Compare the deformation suppression rate with a preset deformation threshold. If the deformation suppression rate is lower than the deformation threshold, an incremental scheme for thickness compensation gradient is generated based on the correlation features of the medium stress zone and the low stress zone. For example, if the deformation suppression rate of the medium stress zone is less than the deformation threshold, the thickness is increased from 8mm to 8.8mm. The low stress zone can determine whether to limit the compensation range based on the stability flag. If it is stable, the thickness of the low stress zone is not adjusted; otherwise, the thickness of the low stress zone is adjusted by a certain proportion.

[0104] S330. Based on the abnormal fluctuation zone, spatially superimpose the material replacement scheme and the incremental scheme of thickness compensation.

[0105] Based on the location of the fluctuation anomaly zone, the material replacement scheme and the incremental thickness compensation scheme are spatially superimposed to avoid conflicts.

[0106] S340. Based on the superimposed results, output an optimized scheme for impact resistance parameters, including material thickness adjustment values ​​and stress concentration zone correction parameters.

[0107] For example, the material thickness adjustment value is such as adjusting the original thickness of S355 steel from 6mm to 8mm, and the stress concentration area correction parameter is such as increasing the chamfer radius of the edge of the high stress area from 3mm to 5mm.

[0108] When training the collision resistance optimization model, a joint loss function can be designed to combine the compression resistance matching degree and deformation suppression rate error. Different weight combinations of compression resistance matching degree and deformation suppression rate error can be tried. The model's compression resistance matching and deformation suppression effects on the test set are compared, the optimal combination is selected, and finally the weight coefficients are adjusted iteratively through backpropagation to minimize the loss value.

[0109] In some embodiments, S400 performs pre-verification of the collision resistance effect of the collision parameter optimization scheme, generating a simulation verification feature set, including:

[0110] S410. Generate multiple sets of candidate parameter combinations based on the material thickness adjustment value and stress concentration zone correction parameters.

[0111] S420. Based on multiple combinations of candidate parameters and collision condition feature sets, perform bidirectional correlation analysis for screening.

[0112] S430. Validate the effectiveness of the selected candidate parameter combinations and obtain the validation results.

[0113] S440. Using the selected candidate parameter combinations as rows, the stress zone-frequency correlation matrix as columns, and the verification results as matrix element values, generate a simulation verification feature set.

[0114] In some embodiments, S420. Based on multiple sets of candidate parameter combinations and collision condition feature sets, a bidirectional correlation analysis is performed for screening, including:

[0115] S421. Input the candidate parameter combination into the collision resistance optimization model to obtain the calculation result of the compressive strength matching degree under the collision velocity time series data; input the candidate parameter combination into the collision resistance optimization model to obtain the calculation result of the deformation suppression rate under the collision contact angle sequence.

[0116] S422. Retain candidate parameter combinations where the calculated compressive strength matching degree exceeds a preset first ratio value and the calculated deformation inhibition rate exceeds a preset second ratio value.

[0117] In S410, multiple sets of candidate parameters can be generated at fixed steps (e.g., ±1mm) for material thickness adjustment values; multiple sets of candidate parameters can also be generated at fixed steps (e.g., ±1mm) for stress concentration zone correction parameters.

[0118] For example, candidate parameter combination 1: the original thickness of 6mm is adjusted to 8mm, and the edge chamfer radius is increased from 3mm to 5mm; candidate parameter combination 2: the original thickness of 6mm is adjusted to 7mm, and the edge chamfer radius is increased from 3mm to 4mm; candidate parameter combination 3: the original thickness of 6mm is adjusted to 10mm, and the edge chamfer radius is increased from 3mm to 6mm.

[0119] In S421, the compressive strength matching degree of the candidate parameter combinations under the collision velocity is calculated to obtain the compressive strength matching degree R. v The formula is: Where csofcm represents the material's compressive strength, the maximum value in the collision velocity time series data, and k v This represents the velocity-stress coefficient, used to normalize the effect of velocity on the compressive strength of a material.

[0120] The deformation suppression ratio D is obtained by calculating the deformation suppression ratio of the candidate parameter combination at the collision angle. θ The formula is: Among them, D θ δ represents the deformation suppression rate at the collision angle. o Represents the original deformation variable, δ c This represents the deformation after compensation.

[0121] The compressive strength of the material needs to be reserved with a safety margin of 1.5 times. The first proportional value can be 1.5; the second proportional value is determined by calculating the safety deformation through the material's elastic limit, and then verified by finite element simulation and calibrated according to industry standards.

[0122] Screening for compressive strength matching degree R v Greater than the first proportional value and deformation suppression rate D θ Candidate parameter combinations that are greater than the second ratio value.

[0123] When verifying the effectiveness, the selected candidate parameter combinations are input into the finite element simulation model, and the collision process is re-simulated to verify whether the following indicators are met simultaneously:

[0124] 1. Whether the maximum stress in the high-stress zone is lower than the compressive strength of the material.

[0125] 2. Whether the deformation in the intermediate stress zone is below the safety threshold.

[0126] The verification results include valid and invalid.

[0127] In some embodiments, S500 generates a final optimization score result based on the collision condition feature set and the simulation verification feature set, including:

[0128] S510. Time-series alignment of collision velocity data with candidate parameter combinations to generate a velocity-parameter time-series correlation matrix; spatial cross-correlation of collision angle data with stress zone-frequency correlation matrix to generate an angle-matrix spatial correlation matrix.

[0129] The velocity-parameter time-series correlation matrix can be generated by combining candidate parameters as columns and velocity time-series data as rows. The generated velocity-parameter time-series correlation matrix is ​​as follows:

[0130]

[0131] The sliding window covariance formula can be used to calculate the correlation value between velocity and candidate parameter combinations at each time step, which is taken as the velocity-parameter correlation value T. ij .

[0132] If the velocity-parameter temporal correlation matrix is: The effective candidate parameter combination 1 has the highest correlation value and the best compressive strength at V(t2).

[0133] Using the angle time series data as rows and the covariance of the stress zone-frequency correlation matrix as columns, the angle-matrix spatial correlation matrix is ​​constructed as follows:

[0134]

[0135] Reference formula for the correlation value between angle and stress zone covariance: S ik =θ(t) i )×Cov k Among them, S ik The spatial correlation value of the k-th stress region at the i-th time step is denoted as the angle-matrix correlation value, Cov. k θ(t) represents the covariance value of the k-th stress region in the stress region-frequency correlation matrix. i ) represents the normalized collision angle.

[0136] If time point t i The collision angles are θ(t1) = 30°, θ(t2) = 45°, and θ(t3) = 60°. The covariances of the high, medium, and low stress zones and the fluctuation frequency are 0.85, 0.60, and 0.10, respectively. Therefore, the angle-matrix spatial correlation matrix is: The correlation value of the high-stress region is highest when θ(t3) = 60°, indicating that the high-stress region should be optimized first for large-angle collisions.

[0137] S520. The collision resistance parameter optimization model is used to process the velocity-parameter temporal correlation matrix and the angle-matrix spatial correlation matrix. After extracting the temporal correlation features and spatial correlation features, the collision resistance parameter optimization model performs feature merging and conflict resolution to generate the final optimization score result.

[0138] The collision resistance parameter optimization model can be based on the gradient tree model with the addition of a multimodal input layer, which includes temporal and spatial correlation branches.

[0139] The time correlation branch is used to receive the velocity-parameter time correlation matrix and extract time correlation features, such as velocity-parameter correlation values; the spatial correlation branch is used to receive the angle-matrix spatial correlation matrix and extract spatial correlation features, such as angle-matrix correlation values.

[0140] In some embodiments, feature merging and conflict resolution are performed in S520, including:

[0141] S521. Merge the temporal correlation features and spatial correlation features according to the time step and spatial location to generate a comprehensive feature.

[0142] Align the temporal and spatial correlation features according to the time step and stress zone to generate a comprehensive feature matrix: F ijk =α×T ij +β×Sik Among them, F ijk T represents the element values ​​in the comprehensive feature matrix. ij Represents the speed-parameter correlation value, S ik Represents the angle-matrix correlation value, where α and β represent T respectively. ij and S ik The weight coefficients can be obtained through training.

[0143] S522. The verification results in the simulation verification feature set include invalid candidate parameter combinations and stress stability markers for unstable low-stress regions; based on the verification results in the simulation verification feature set, conflict resolution is performed on the comprehensive features, including:

[0144] If the verification result is marked as an invalid candidate parameter combination, the corresponding time-related feature is removed.

[0145] For example, if combination 2 is marked as invalid, then the second column in the time correlation matrix is ​​deleted:

[0146]

[0147] Removing invalid combinations can effectively reduce the amount of computation.

[0148] If the verification result is marked as an unstable low-stress zone stress stability marker, then the influence range of the spatial correlation feature is limited.

[0149] For example, if the low-stress region is unstable, only the time-related portion of the low-stress region is retained in the comprehensive characteristic matrix: F ijk =0.6×T ij +0×S ik =0.6×T ij .

[0150] Limiting the influence of low-stress areas can reduce noise interference in unstable regions.

[0151] In S520, the comprehensive feature matrix is ​​summed row by row (candidate combination) and normalized to a percentage:

[0152]

[0153] Among them, Score j The final optimization score represents the j-th candidate parameter combination, used to quantify the overall crashworthiness of this combination. The summation of all time and stress zone eigenvalues ​​for the j-th combination yields the total score for that combination. N represents the total number of time steps, i.e., the number of time discrete points in the collision process. j represents the number of the candidate parameter combination. i represents the time step index and stress zone index, where 1 = high stress zone, 2 = medium stress zone, and 3 = low stress zone. max(∑F) represents the maximum value among all the total scores of the combinations, used to normalize the scoring benchmark.

[0154] For example, if the total score of combination 1 is ∑F = 2.30 and the total score of combination 2 is ∑F = 1.80, then...

[0155] The comprehensive crashworthiness performance of different candidate parameter combinations is directly quantified by a 100-point scoring system; the higher the score, the better the solution.

[0156] For example, combination 1 has significantly better impact resistance than combination 2 and can be prioritized for production or further verification. Combination 2 scored lower and its shortcomings in high-stress areas or high-speed collision scenarios need to be examined and targeted improvements made.

[0157] In some embodiments, such as Figure 2 As shown, this application provides a finite element simulation optimization system for the anti-collision performance of guardrails, which includes:

[0158] Data partitioning module: acquires the finite element simulation data of guardrail collision and partitions the finite element simulation data of guardrail collision into stress distribution feature set and collision condition feature set;

[0159] Feature fusion module: Performs multi-dimensional feature fusion on the stress distribution feature set to generate a dynamic parameter feature set;

[0160] Solution generation module: Generates collision resistance parameter optimization solutions based on dynamic parameter feature sets;

[0161] Effect verification module: Performs pre-verification of the collision resistance effect of the collision parameter optimization scheme and generates a simulation verification feature set;

[0162] Scoring output module: Generates the final optimized scoring result based on the collision condition feature set and the simulation verification feature set.

[0163] In some embodiments, this application provides a finite element simulation optimization apparatus for the anti-collision performance of guardrails, which includes a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program to implement steps such as the finite element simulation optimization method for the anti-collision performance of guardrails.

[0164] In some embodiments, this application provides a storage medium storing computer program instructions, which are read and executed by a processor to perform steps such as a finite element simulation optimization method for the anti-collision performance of guardrails.

[0165] Any references to memory, storage, database, or other media used in the embodiments provided in this invention may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory.

[0166] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0167] Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for optimizing the crash performance of a guardrail panel by finite element simulation, characterized in that, The method comprises the following steps: Obtain guardrail plate collision finite element simulation data, and divide the guardrail plate collision finite element simulation data into a stress distribution feature set and a collision condition feature set; Perform multi-dimensional feature fusion on the stress distribution feature set to generate a dynamic parameter feature set; Generate an anti-collision parameter optimization scheme according to the dynamic parameter feature set; Perform anti-collision effect pre-verification on the anti-collision parameter optimization scheme to generate a simulation verification feature set; Generate a final optimization score result according to the collision condition feature set and the simulation verification feature set; Wherein, performing multi-dimensional feature fusion on the stress distribution feature set to generate a dynamic parameter feature set comprises: Set a specified number of stress zones, divide all stress zones according to the maximum stress value of each stress zone, and extract independent features of each stress zone; Perform time sequence correlation analysis on the stress fluctuation frequency and the independent features of each stress zone to generate a zone-frequency correlation matrix; Screen fluctuation abnormal zones based on the correlation matrix to generate a dynamic parameter feature set; Generate an anti-collision parameter optimization scheme according to the dynamic parameter feature set, which comprises: An anti-collision optimization model is used to process the anti-collision parameter optimization scheme, and the anti-collision optimization model adds a multi-modal input layer based on a gradient tree to generate an anti-collision parameter optimization scheme including material thickness adjustment values and stress concentration area correction parameters through joint splitting rules; Perform anti-collision effect pre-verification on the anti-collision parameter optimization scheme to generate a simulation verification feature set, which comprises: Generate multiple candidate parameter combinations according to the material thickness adjustment values and the stress concentration area correction parameters; Perform bidirectional correlation degree analysis screening according to the multiple candidate parameter combinations and the collision condition feature set; Verify the effectiveness of the screened candidate parameter combinations to obtain verification results; The screened candidate parameter combinations are used as rows, the stress zone-frequency correlation matrix is used as columns, and the verification results are used as matrix element values to generate a simulation verification feature set; Generate a final optimization score result according to the collision condition feature set and the simulation verification feature set, which comprises: Align the collision speed time sequence data with the candidate parameter combinations in time sequence to generate a speed-parameter time sequence correlation matrix; cross-correlate the collision angle time sequence data with the stress zone-frequency correlation matrix in spatial dimensions to generate an angle-matrix spatial correlation matrix; An anti-collision parameter optimization model is used to process the speed-parameter time sequence correlation matrix and the angle-matrix spatial correlation matrix, and the anti-collision parameter optimization model extracts time correlation features and spatial correlation features, then performs feature merging and conflict resolution to generate a final optimization score result.

2. The method of claim 1, wherein, Set a specified number of stress zones, divide all stress zones according to the maximum stress value of each stress zone, and extract independent features of each stress zone, which comprises: According to the numerical distribution of the maximum stress value sequence and the preset stress critical value, the guardrail plate is divided into a high stress zone, a medium stress zone and a low stress zone; For the high stress zone: extract the stress peak interval and the fluctuation frequency as independent features; For the medium stress zone: extract the stress attenuation rate and the thickness distribution correlation parameter as independent features; For the low stress zone: extract the stress stability marker as an independent feature.

3. The method of claim 2, wherein, The anti-collision optimization model increases a multi-modal input layer on the basis of a gradient tree, generates an anti-collision parameter optimization scheme through a joint split rule, and includes the following steps: The multi-modal input layer includes a first channel and a second channel; the first channel is used to receive independent features of a high stress area and calculate a compression resistance matching degree of a material replacement type, and the second channel is used to receive independent features of a medium stress area and independent features of a low stress area and calculate a deformation suppression rate of a thickness compensation gradient; The compression resistance matching degree is compared with a preset compression resistance threshold value, if the compression resistance matching degree exceeds the compression resistance threshold value, a matching scheme of the material replacement type of the high stress area is generated; the deformation suppression rate is compared with a preset deformation threshold value, if the deformation suppression rate is lower than the deformation threshold value, an incremental scheme of the thickness compensation gradient is generated based on the associated features of the medium stress area and the associated features of the low stress area; According to the fluctuation abnormal area, the material replacement scheme and the incremental scheme of the thickness compensation are superimposed in space; According to the superimposed result, an anti-collision parameter optimization scheme including a material thickness adjustment value and a stress concentration area correction parameter is output.

4. The method of claim 3, wherein, According to a plurality of candidate parameter combinations and a collision condition feature set, bidirectional correlation degree analysis and screening are performed, including: The candidate parameter combinations are input into the anti-collision optimization model to obtain a compression resistance matching degree calculation result under collision speed time sequence data; the candidate parameter combinations are input into the anti-collision optimization model to obtain a deformation suppression rate calculation result under a collision contact angle sequence; The candidate parameter combinations whose compression resistance matching degree calculation result exceeds a preset first proportion value and whose deformation suppression rate calculation result exceeds a preset second proportion value are retained.

5. The method of claim 4, wherein, Feature merging and conflict resolution are performed, including: The time-related features and the space-related features are merged according to time steps and spatial positions to generate comprehensive features; The simulation verification feature set includes invalid candidate parameter combinations and unstable low stress area stress stability labels; according to the simulation verification feature set, the comprehensive features are subjected to conflict resolution, including: If the verification result is marked as an invalid candidate parameter combination, the corresponding time-related feature is removed; If the verification result is marked as an unstable low stress area stress stability label, the influence range of the space-related feature is limited.

6. A finite element simulation optimization system for the crash performance of a guardrail panel, characterized by, It includes: A data division module: obtaining guardrail plate collision finite element simulation data, and dividing the guardrail plate collision finite element simulation data into a stress distribution feature set and a collision condition feature set; A feature fusion module: performing multi-dimensional feature fusion on the stress distribution feature set to generate a dynamic parameter feature set; A scheme generation module: generating an anti-collision parameter optimization scheme according to the dynamic parameter feature set; An effect verification module: performing anti-collision effect pre-verification on the anti-collision parameter optimization scheme to generate a simulation verification feature set; A scoring output module: generating a final optimization score result according to the collision condition feature set and the simulation verification feature set; Wherein, performing multi-dimensional feature fusion on the stress distribution feature set to generate a dynamic parameter feature set includes: A specified number of stress areas are set, all stress areas are divided according to the maximum stress value of each stress area, and independent features of each stress area are extracted; Time sequence correlation analysis is performed on stress fluctuation frequency and the independent features of each stress area to generate a zone-frequency correlation matrix; Screening fluctuation abnormal area based on the correlation matrix, generating a dynamic parameter feature set; Generating an anti-collision parameter optimization scheme according to the dynamic parameter feature set, including: Processing the anti-collision parameter optimization scheme by using an anti-collision optimization model, the anti-collision optimization model adding a multi-modal input layer on the basis of a gradient tree, and generating an anti-collision parameter optimization scheme including material thickness adjustment values and stress concentration area correction parameters through a joint split rule; Performing anti-collision effect pre-validation on the anti-collision parameter optimization scheme, generating a simulation verification feature set, including: Generating a plurality of candidate parameter combinations according to the material thickness adjustment values and the stress concentration area correction parameters; Performing bidirectional correlation degree analysis screening according to the plurality of candidate parameter combinations and the collision condition feature set; Validating the screened candidate parameter combinations to obtain a validation result; Generating a simulation verification feature set by taking the screened candidate parameter combinations as rows, taking the stress area-frequency correlation matrix as columns, and taking the validation result as matrix element values; Generating a final optimization score result according to the collision condition feature set and the simulation verification feature set, including: Aligning the collision speed time series data with the candidate parameter combinations in time series to generate a speed-parameter time series correlation matrix; and cross-correlating the collision angle time series data with the stress area-frequency correlation matrix in spatial dimensions to generate an angle-matrix spatial correlation matrix; Processing the speed-parameter time series correlation matrix and the angle-matrix spatial correlation matrix by using the anti-collision parameter optimization model, the anti-collision parameter optimization model extracting time correlation features and spatial correlation features, performing feature merging and conflict resolution, and generating the final optimization score result.

Images