Vehicle body rigidity, weight and cost collaborative optimization method and system
Through structural contribution analysis and hybrid agent model, combined with composite kernel RBF and time-varying RSM model, the multi-objective dynamic balance problem of stiffness, weight and cost in body design is solved, and the practical applicability and economic efficiency of the optimization results are improved.
Patent Information
- Application Number
- CN202510801095.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-16
- Publication Date
- 2025-09-16
AI Technical Summary
Existing technologies make it difficult to achieve a multi-objective dynamic balance between stiffness, weight, and cost in automobile body design. Cost modeling also ignores material price fluctuations and the complexity of manufacturing processes, resulting in insufficient applicability and poor economic efficiency of optimization results in actual production.
Key safety parts are screened through structural contribution analysis, a hybrid agent model is constructed, and a composite kernel RBF model and a time-varying RSM model are used. Combined with the ARIMA model, material price fluctuations are predicted in real time to achieve coordinated optimization of vehicle body stiffness, weight, and cost.
It achieves a multi-objective dynamic balance among stiffness, weight and cost in vehicle body design, reduces the waste of computing resources, and improves the engineering feasibility and industrial adaptability of the optimization results.
Smart Images
Figure CN120654326A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of vehicle engineering and structural optimization, and particularly relates to a method and system for collaborative optimization of vehicle body stiffness, weight and cost. Background Art
[0002] In automotive body design, stiffness, weight, and cost are three key indicators that influence vehicle performance and market competitiveness. These three factors are significantly coupled and constrained. Improving body stiffness typically requires increasing material usage or adopting high-performance materials, leading to higher curb weight and manufacturing costs. While lightweight design improves energy efficiency and handling, it can weaken structural rigidity, impacting safety and comfort. Furthermore, simplifying the structure or process to reduce manufacturing costs can compromise the mechanical properties of the vehicle body. Therefore, achieving a balanced optimization among these three factors has become a core challenge in vehicle body structural design.
[0003] Existing technologies primarily focus on optimizing a single objective, such as stiffness enhancement based on finite element analysis or structural subtractive design based on topology optimization, lacking the coordinated trade-offs and overall considerations between multiple objectives. Furthermore, current optimization methods generally adopt static assumptions in cost modeling, ignoring material price fluctuations, manufacturing process complexity, and the resulting dynamic cost changes. This results in optimization results that are insufficiently applicable and uneconomical in actual production. Therefore, there is an urgent need to propose a method that comprehensively considers multiple key performance indicators such as vehicle body stiffness, weight, and cost, achieving a multi-objective dynamic balance between performance, efficiency, and cost, in order to improve the engineering feasibility and industrial adaptability of vehicle body design. Summary of the Invention
[0004] In order to solve the above problems existing in the prior art, the present invention provides a vehicle body stiffness, weight and cost collaborative optimization method and system. The purpose of the present invention can be achieved through the following technical solutions: A vehicle body stiffness, weight and cost collaborative optimization method, comprising: S1: Obtaining initial safety components and safety component structural parameters, and screening the initial safety components to obtain basic safety components by analyzing the safety component structural parameters through structural contribution analysis; S2: Preset design variables and design responses based on the basic safety components, and output test samples through experimental design; S3: constructing a hybrid surrogate model, and outputting a fitting test sample through the hybrid surrogate model based on the test sample; S4: Preset an optimization target, and optimize the fitting test sample output optimization strategy according to the optimization target.
[0005] Preferably, the analysis and screening process of the structural contribution analysis in step S1 includes: S101: Unifying the safety component structure parameters through regularization to output initial safety component structure normalization parameters; S102: Obtaining an effect coefficient by fitting the normalized parameters of the initial safety component structure through least squares fitting; S103: Convert the effect coefficient into a percentage to obtain a parameter contribution; S104: Preset screening rules, and screen the parameter contribution according to the screening rules to obtain the basic safety components.
[0006] Preferably, the screening rules in step S104 are specifically: S104-1: The parameter contribution includes the vehicle body mass contribution and the overall anti-collision optimization contribution; the vehicle body contribution threshold and the overall anti-collision optimization contribution threshold are preset; S104-2: If the vehicle body mass contribution and the overall crashworthiness optimization contribution are both less than the vehicle body contribution threshold and the overall crashworthiness optimization contribution threshold, the initial safety component is determined to be in a "weak range of both lightweighting potential and crashworthiness optimization potential" and is removed. S104-3: If the vehicle body mass contribution is less than the vehicle body contribution threshold, and the overall crashworthiness optimization contribution is greater than the overall crashworthiness optimization contribution threshold, the initial safety component is determined to be in the "crashworthiness optimization potential advantage range" and is retained; S104-4: If the overall anti-crash optimization contribution is less than the overall anti-crash optimization contribution threshold, directly removing the initial safety component; S104-5: If the vehicle body mass contribution is greater than the vehicle body contribution threshold, and the overall anti-crash optimization contribution is greater than the overall anti-crash optimization contribution threshold, then directly retain the initial safety component.
[0007] Preferably, the output process of the test sample in step S2 is: The plate thickness and material of the basic safety component are obtained and used as design variables, and the vehicle body mass, vehicle body stiffness, and material cost of the basic safety component are used as design responses. Experimental design is used to obtain the output test sample between the design variables and the design responses.
[0008] Preferably, the construction process of the hybrid proxy model in step S3 is: S301: The test sample includes vehicle body mass, vehicle body static torsional stiffness, vehicle body B-pillar lower end peak impact acceleration, and safety component material cost; a nonlinear sensitivity model is constructed between the vehicle body mass and the vehicle body static torsional stiffness using a function strategy; S302: constructing a time-varying cost prediction model based on the peak impact acceleration at the lower end of the B-pillar of the vehicle body and the material cost of the safety component through an intelligent algorithm; S303: Weightedly fuse the nonlinear sensitivity model and the time-varying cost prediction model to output a hybrid agent collaboration model.
[0009] Preferably, the design process of the strategy function in step S301 is: S301-1: constructing an adaptive network using a composite kernel function strategy; fusing the vehicle body mass and the vehicle body static torsional stiffness through a three-layer data fusion mechanism to obtain multi-fidelity data integration; S301-2: Perform a global sensitivity analysis based on the multi-fidelity data integration.
[0010] Preferably, the adaptive network in step S301-1 is represented as: , Where ϕ(r) is the adaptive network, α is the dynamic mixing coefficient, Gaussian and MQ represent radial basis functions, and f NVH represents the static torsional stiffness of the vehicle body, t i is the vehicle body mass.
[0011] Preferably, the data fusion process of the three-layer data fusion mechanism in step S301-1 includes: The vehicle body mass and the vehicle body static torsional stiffness are subjected to FEA simulation to extract high-fidelity data and output high-level parameters; Transferring the high-level parameters to the middle level through transfer learning and automatically adjusting the high-level parameters through covariance analysis, extracting mid-fidelity data of the vehicle body mass and the vehicle body static torsional stiffness through a simplified model, and outputting middle-level parameters; Transferring the middle-level parameters to a lower level through transfer learning and automatically adjusting the middle-level parameters through covariance analysis, and extracting low-fidelity data of the vehicle body mass and the vehicle body static torsional stiffness through an empirical formula; The high-fidelity data, the medium-fidelity data, and the low-fidelity data are integrated and output as the multi-fidelity data integration.
[0012] Preferably, the construction process of the time-varying cost prediction model in step S302 is specifically as follows: S302-1: Preset a dynamic update threshold and an ARIMA model, retrain the ARIMA model based on the dynamic update threshold, and obtain a time fluctuation factor based on the peak impact acceleration at the lower end of the B-pillar of the vehicle body; S302-2: Calculate the predicted cost based on the time fluctuation factor and the safety component material cost.
[0013] A vehicle body stiffness, weight and cost collaborative optimization system includes a parameter screening module, an experiment design module, a model fusion module and a strategy optimization module, including: The parameter screening module is used to obtain the initial safety component and the safety component structural parameters, and screen the initial safety component to obtain the basic safety component by analyzing the safety component structural parameters through structural contribution; The test design module is used to preset design variables and design responses based on the basic safety components and output test samples through test design; The model fusion module is used to construct a hybrid proxy model, and output a fitting test sample based on the test sample through the hybrid proxy model; The strategy optimization module is used to preset an optimization target and optimize the fitting test sample output optimization strategy according to the optimization target.
[0014] The beneficial effects of the present invention are: (1) Through structural contribution analysis, accurately identify safety components that have a significant impact on vehicle body stiffness, weight, and cost, and reduce the waste of computing resources caused by optimizing non-critical safety components.
[0015] (2) A composite kernel RBF model is used for nonlinear indicators such as vehicle body mass and stiffness, and a time-varying RSM model is used for crashworthiness and cost. A weighted fusion mechanism is used to achieve cross-modal collaborative optimization.
[0016] (3) The ARIMA model predicts material price fluctuation factors in real time, and dynamically updates thresholds to trigger model retraining to ensure that cost forecasts are synchronized with market changes. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] To facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.
[0018] Figure 1 The figure is a flow chart of a vehicle body stiffness, weight and cost collaborative optimization method according to the present invention. DETAILED DESCRIPTION
[0019] In order to further illustrate the technical means and effects adopted by the present invention to achieve the predetermined purpose of the invention, the specific implementation methods, structures, features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments.
[0020] See also Figure 1 , a vehicle body stiffness, weight and cost collaborative optimization method, comprising: S1: Obtaining initial safety components and safety component structural parameters, and screening the initial safety components to obtain basic safety components by analyzing the safety component structural parameters through structural contribution analysis; S2: Preset design variables and design responses based on the basic safety components, and output test samples through experimental design; S3: constructing a hybrid surrogate model, and outputting a fitting test sample through the hybrid surrogate model based on the test sample; S4: Preset an optimization target, and optimize the fitting test sample output optimization strategy according to the optimization target.
[0021] Specifically, the initial safety parts in step S1 include head-on collision safety parts and offset collision safety parts; the safety part structural parameters include vehicle body mass, first-order torsional modal frequency, static bending stiffness, static torsional stiffness, maximum intrusion of the front panel, and peak impact acceleration at the lower end of the B-pillar.
[0022] Specifically, the analysis and screening process of the structural contribution analysis in step S1 includes: S101: Unifying the safety component structure parameters through regularization to output initial safety component structure normalization parameters; S102: Obtaining an effect coefficient by fitting the normalized parameters of the initial safety component structure through least squares fitting; S103: Convert the effect coefficient into a percentage to obtain a parameter contribution; S104: Preset screening rules, and screen the parameter contribution according to the screening rules to obtain the basic safety components.
[0023] In this embodiment, the contribution of key optimization design indicators is extracted, including vehicle body mass, first-order torsional modal frequency, static bending stiffness, static torsional stiffness, maximum intrusion of the front panel, and peak impact acceleration at the lower end of the B-pillar.
[0024] Specifically, the screening rules in step S104 are: S104-1: The parameter contribution includes the vehicle body mass contribution and the overall anti-collision optimization contribution; the vehicle body contribution threshold and the overall anti-collision optimization contribution threshold are preset; S104-2: If the vehicle body mass contribution and the overall crashworthiness optimization contribution are both less than the vehicle body contribution threshold and the overall crashworthiness optimization contribution threshold, the initial safety component is determined to be in a "weak range of both lightweighting potential and crashworthiness optimization potential" and is removed. S104-3: If the vehicle body mass contribution is less than the vehicle body contribution threshold, and the overall crashworthiness optimization contribution is greater than the overall crashworthiness optimization contribution threshold, the initial safety component is determined to be in the "crashworthiness optimization potential advantage range" and is retained; S104-4: If the overall anti-crash optimization contribution is less than the overall anti-crash optimization contribution threshold, directly removing the initial safety component; S104-5: If the vehicle body mass contribution is greater than the vehicle body contribution threshold, and the overall anti-crash optimization contribution is greater than the overall anti-crash optimization contribution threshold, then directly retain the initial safety component.
[0025] Specifically, the output process of the test sample in step S2 is: The plate thickness and material of the basic safety component are obtained and used as design variables, and the vehicle body mass, vehicle body stiffness, and material cost of the basic safety component are used as design responses. Experimental design is used to obtain the output test sample between the design variables and the design responses.
[0026] Specifically, the test samples in step S2 include vehicle body mass, vehicle body static torsional stiffness, vehicle body B-pillar lower end peak impact acceleration, and safety component material cost.
[0027] Specifically, the construction process of the hybrid proxy model in step S3 is: S301: constructing a model output nonlinear sensitivity model between the vehicle body mass and the vehicle body static torsional stiffness through a function strategy; S302: constructing a time-varying cost prediction model based on the peak impact acceleration at the lower end of the B-pillar of the vehicle body and the material cost of the safety component through an intelligent algorithm; S303: Weightedly fuse the nonlinear sensitivity model and the time-varying cost prediction model to output a hybrid agent collaboration model.
[0028] Specifically, the design process of the strategy function in step S301 is: S301-1: constructing an adaptive network using a composite kernel function strategy; fusing the vehicle body mass and the vehicle body static torsional stiffness through a three-layer data fusion mechanism to obtain multi-fidelity data integration; S301-2: performing a global sensitivity analysis based on the multi-fidelity data integration; The calculation expression of the global sensitivity analysis is: , Among them, S total i is the global sensitivity index, Var ~xi Indicates the difference between x i Find the variance of all external variables, Ex i (y|x ~i ) means that x is fixed when other variables are fixed. i The conditional expectation of y is the multi-fidelity data integration, x ~i Indicates division by x i The external variable, x irepresents a variable, and Var(y) represents the total variance of the multi-fidelity data integration.
[0029] Specifically, the adaptive network in step S301-1 is represented as: , Where ϕ(r) is the adaptive network, α is the dynamic mixing coefficient, Gaussian and MQ represent radial basis functions, and f NVH represents the static torsional stiffness of the vehicle body, t i is the vehicle body mass.
[0030] Specifically, the data fusion process of the three-layer data fusion mechanism in step S301-1 includes: The vehicle body mass and the vehicle body static torsional stiffness are subjected to FEA simulation to extract high-fidelity data and output high-level parameters; Transferring the high-level parameters to the middle level through transfer learning and automatically adjusting the high-level parameters through covariance analysis, extracting mid-fidelity data of the vehicle body mass and the vehicle body static torsional stiffness through a simplified model, and outputting middle-level parameters; Transferring the middle-level parameters to a lower level through transfer learning and automatically adjusting the middle-level parameters through covariance analysis, and extracting low-fidelity data of the vehicle body mass and the vehicle body static torsional stiffness through an empirical formula; The high-fidelity data, the medium-fidelity data, and the low-fidelity data are integrated and output as the multi-fidelity data integration.
[0031] Specifically, the construction process of the time-varying cost prediction model in step S302 is as follows: S302-1: Preset a dynamic update threshold and an ARIMA model, retrain the ARIMA model based on the dynamic update threshold, and obtain a time fluctuation factor based on the peak impact acceleration at the lower end of the B-pillar of the vehicle body; S302-2: Calculating a predicted cost based on the time fluctuation factor and the material cost of the safety component; The calculation formula for the predicted cost is: , Among them, C material is the predicted cost, β0 is the material cost of the safety component, β i is the cost sensitivity coefficient of the i-th safety component material, x i represents the i-th security component material, and γ(t) is the time fluctuation factor.
[0032] In this embodiment, a vehicle body stiffness, weight, and cost collaborative optimization system includes a parameter screening module, an experiment design module, a model fusion module, and a strategy optimization module, including: The parameter screening module is used to obtain the initial safety component and the safety component structural parameters, and screen the initial safety component to obtain the basic safety component by analyzing the safety component structural parameters through structural contribution; The test design module is used to preset design variables and design responses based on the basic safety components and output test samples through test design; The model fusion module is used to construct a hybrid proxy model, and output a fitting test sample based on the test sample through the hybrid proxy model; The strategy optimization module is used to preset an optimization target and optimize the fitting test sample output optimization strategy according to the optimization target.
[0033] The above description is merely a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Although the present invention has been disclosed as a preferred embodiment as above, it is not intended to limit the present invention. Any person skilled in the art can make some changes or modifications to equivalent embodiments using the technical contents disclosed above without departing from the scope of the technical solution of the present invention. However, any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.
Claims
1. A vehicle body stiffness, weight and cost collaborative optimization method, characterized in that: The following steps are involved: S1: Obtaining initial safety components and safety component structural parameters, and screening the initial safety components to obtain basic safety components by analyzing the safety component structural parameters through structural contribution analysis; S2: Preset design variables and design responses based on the basic safety components, and output test samples through experimental design; S3: constructing a hybrid surrogate model, and outputting a fitting test sample through the hybrid surrogate model based on the test sample; S4: Preset an optimization target, and optimize the fitting test sample output optimization strategy according to the optimization target.
2. The vehicle body stiffness, weight and cost collaborative optimization method according to claim 1, characterized in that: The analysis and screening process of the structural contribution analysis in step S1 includes: S101: Unifying the safety component structure parameters through regularization to output initial safety component structure normalization parameters; S102: Obtaining an effect coefficient by fitting the normalized parameters of the initial safety component structure through least squares fitting; S103: Convert the effect coefficient into a percentage to obtain a parameter contribution; S104: Preset screening rules, and screen the parameter contribution according to the screening rules to obtain the basic safety components.
3. The vehicle body stiffness, weight and cost collaborative optimization method according to claim 2, characterized in that: The screening rules in step S104 are specifically: S104-1: The parameter contribution includes the vehicle body mass contribution and the overall anti-collision optimization contribution; the vehicle body contribution threshold and the overall anti-collision optimization contribution threshold are preset; S104-2: If the vehicle body mass contribution and the overall crashworthiness optimization contribution are both less than the vehicle body contribution threshold and the overall crashworthiness optimization contribution threshold, the initial safety component is determined to be in a "weak range of both lightweighting potential and crashworthiness optimization potential" and is removed. S104-3: If the vehicle body mass contribution is less than the vehicle body contribution threshold, and the overall crashworthiness optimization contribution is greater than the overall crashworthiness optimization contribution threshold, the initial safety component is determined to be in the "crashworthiness optimization potential advantage range" and is retained; S104-4: If the overall anti-crash optimization contribution is less than the overall anti-crash optimization contribution threshold, directly removing the initial safety component; S104-5: If the vehicle body mass contribution is greater than the vehicle body contribution threshold, and the overall anti-crash optimization contribution is greater than the overall anti-crash optimization contribution threshold, then directly retain the initial safety component.
4. The vehicle body stiffness, weight and cost collaborative optimization method according to claim 1, characterized in that: The output process of the test sample in step S2 is as follows: The plate thickness and material of the basic safety component are obtained and used as design variables, and the vehicle body mass, vehicle body stiffness, and material cost of the basic safety component are used as design responses. Experimental design is used to obtain the output test sample between the design variables and the design responses.
5. The vehicle body stiffness, weight and cost collaborative optimization method according to claim 1, characterized in that: The construction process of the hybrid proxy model in step S3 is as follows: S301: The test sample includes vehicle body mass, vehicle body static torsional stiffness, vehicle body B-pillar lower end peak impact acceleration, and safety component material cost; a nonlinear sensitivity model is constructed between the vehicle body mass and the vehicle body static torsional stiffness using a function strategy; S302: constructing a time-varying cost prediction model based on the peak impact acceleration at the lower end of the B-pillar of the vehicle body and the material cost of the safety component through an intelligent algorithm; S303: Weightedly fuse the nonlinear sensitivity model and the time-varying cost prediction model to output a hybrid agent collaboration model.
6. The vehicle body stiffness, weight and cost collaborative optimization method according to claim 5, characterized in that: The design process of the strategy function in step S301 is as follows: S301-1: constructing an adaptive network using a composite kernel function strategy; fusing the vehicle body mass and the vehicle body static torsional stiffness through a three-layer data fusion mechanism to obtain multi-fidelity data integration; S301-2: Perform a global sensitivity analysis based on the multi-fidelity data integration.
7. The vehicle body stiffness, weight and cost collaborative optimization method according to claim 6, characterized in that: The adaptive network in step S301-1 is represented as follows: ， Where ϕ(r) is the adaptive network, α is the dynamic mixing coefficient, Gaussian and MQ represent radial basis functions, and f NVH represents the static torsional stiffness of the vehicle body, t i is the vehicle body mass.
8. The vehicle body stiffness, weight and cost collaborative optimization method according to claim 6, characterized in that: The data fusion process of the three-layer data fusion mechanism in step S301-1 includes: The vehicle body mass and the vehicle body static torsional stiffness are subjected to FEA simulation to extract high-fidelity data and output high-level parameters; Transferring the high-level parameters to the middle level through transfer learning and automatically adjusting the high-level parameters through covariance analysis, extracting mid-fidelity data of the vehicle body mass and the vehicle body static torsional stiffness through a simplified model, and outputting middle-level parameters; Transferring the middle-level parameters to a lower level through transfer learning and automatically adjusting the middle-level parameters through covariance analysis, and extracting low-fidelity data of the vehicle body mass and the vehicle body static torsional stiffness through an empirical formula; The high-fidelity data, the medium-fidelity data, and the low-fidelity data are integrated and output as the multi-fidelity data integration.
9. The vehicle body stiffness, weight and cost collaborative optimization method according to claim 5, characterized in that: The construction process of the time-varying cost prediction model in step S302 is specifically as follows: S302-1: Preset a dynamic update threshold and an ARIMA model, retrain the ARIMA model based on the dynamic update threshold, and obtain a time fluctuation factor based on the peak impact acceleration at the lower end of the B-pillar of the vehicle body; S302-2: Calculate the predicted cost based on the time fluctuation factor and the safety component material cost.
10. A vehicle body stiffness, weight and cost collaborative optimization system, comprising a parameter screening module, an experimental design module, a model fusion module and a strategy optimization module, characterized in that: include: The parameter screening module is used to obtain the initial safety component and the safety component structural parameters, and screen the initial safety component to obtain the basic safety component by analyzing the safety component structural parameters through structural contribution; The test design module is used to preset design variables and design responses based on the basic safety components and output test samples through test design; The model fusion module is used to construct a hybrid proxy model, and output a fitting test sample based on the test sample through the hybrid proxy model; The strategy optimization module is used to preset an optimization target and optimize the fitting test sample output optimization strategy according to the optimization target.