Vehicle collision simulation method based on digital twinning
By constructing a physical model of vehicle collision and using historical collision data to train a digital twin model, the degree of damage of vehicles is evaluated, and the problem of high traditional testing costs is solved, and a fast and economical vehicle collision simulation is achieved.
Patent Information
- Application Number
- CN202510425489.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-07
- Publication Date
- 2025-08-29
AI Technical Summary
Traditional physical collision testing is costly and time-consuming, making it difficult to quickly respond to personalized design needs and meet the ever-elevated safety regulations and requirements. The application of existing digital twin technology in vehicle collision simulation is not yet mature.
Build a physical model of vehicle collision, train a digital twin model of vehicle collision through historical collision data, evaluate the degree of damage to the virtual model of the vehicle after the collision, and provide reference to save actual testing costs.
Simulate vehicle collisions through digital twin technology, quickly evaluate the degree of vehicle damage, reduce actual testing costs, and support rapid design and verification processes.
Smart Images
Figure CN120562092A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of vehicle collision simulation, and in particular to a vehicle collision simulation method based on digital twins. Background Art
[0002] As consumers' attention to vehicle safety continues to grow, automakers need to conduct extensive crash tests to ensure vehicles meet safety standards. Traditional physical crash tests are expensive and time-consuming, and digital twins offer a highly efficient alternative. As an emerging technology framework, digital twins have been successfully applied in various fields, including manufacturing, aviation, and energy, providing a reliable technical foundation for vehicle crash simulation. Consumers' demand for personalized vehicle customization is growing, and manufacturers need to respond quickly and test different design options. Digital twins can accelerate the design and verification process. Global automotive safety regulations are constantly evolving, requiring vehicle manufacturers to provide higher levels of safety. Digital twin technology can help manufacturers better meet these regulatory requirements.
[0003] How to build a vehicle collision digital twin model, train the vehicle collision digital twin model through historical collision data, obtain the trained vehicle collision digital twin model, evaluate the damage degree of the vehicle virtual model after the collision based on the obtained trained vehicle collision digital twin model, and use the damage degree of the vehicle virtual model after the collision as a reference to save the cost of actual vehicle collision is the problem we need to solve. To this end, a vehicle collision simulation method based on digital twin is now provided. Summary of the Invention
[0004] In order to solve the above technical problems, the purpose of the present invention is to provide a vehicle collision simulation method based on digital twins.
[0005] The purpose of the present invention can be achieved through the following technical solution: A vehicle collision simulation method based on digital twins, comprising the following steps:
[0006] Step S1: constructing a vehicle collision physical model;
[0007] Step S2: Obtain historical collision data, and obtain a vehicle collision digital twin model based on the constructed vehicle collision physical model and the acquired historical collision data;
[0008] Step S3: Training the vehicle collision digital twin model using historical collision data to obtain a trained vehicle collision digital twin model;
[0009] Step S4: Evaluate the damage degree of the virtual model of the vehicle after the collision based on the obtained trained vehicle collision digital twin model.
[0010] Furthermore, the process of building a vehicle collision physics model includes:
[0011] The vehicle collision physics model consists of a vehicle layer, a scene layer and a collision layer;
[0012] Draw virtual models of several different types of vehicles and several different types of vehicle collision scenarios through 3D modeling;
[0013] The drawn vehicle virtual model is mapped into the vehicle layer, and the drawn vehicle collision scene virtual model is mapped into the scene layer to obtain a vehicle collision physical model.
[0014] Furthermore, the process of obtaining a vehicle collision digital twin model based on the constructed vehicle collision physical model and the acquired historical collision data includes:
[0015] The historical collision data includes the temperature of the collision parts and the collision force of vehicles of different models when they collide at different speeds in different types of vehicle collision scenarios;
[0016] The acquired historical data is mapped to the collision layer in the vehicle collision physical model to obtain a vehicle collision digital twin model.
[0017] Furthermore, the process of training the vehicle collision digital twin model using historical collision data includes:
[0018] A vehicle model virtual model and a vehicle collision scenario virtual model of a type are randomly selected, and the selected vehicle virtual model and vehicle collision scenario virtual model are mapped to the collision layer. The vehicle virtual model collides at a speed of V km / h under the selected vehicle collision scenario virtual model, where V is a positive number. The vehicle collision digital twin model outputs the temperature and collision force of the collision part of the vehicle virtual model, and compares the temperature and collision force of the collision part of the corresponding vehicle in the historical collision data with the obtained temperature and collision force of the collision part of the vehicle virtual model;
[0019] If the comparison results are the same, the output of the vehicle collision digital twin model is accurate. If the comparison results are different, the output of the vehicle collision digital twin model is inaccurate. The output of the vehicle collision digital twin model is corrected, and the vehicle virtual model is collided again at a speed of V km / h under the selected vehicle collision scenario virtual model. The vehicle collision digital twin model outputs the temperature of the collision part and the collision force of the vehicle virtual model, and so on, until the output of the vehicle collision digital twin model is accurate;
[0020] Similarly, the vehicle collision digital twin model is trained through historical collision data to obtain accurate temperature and collision force of the collision parts of the vehicle virtual models of other models when they collide at other speeds under other types of vehicle collision scene virtual models, and then obtain the trained vehicle collision digital twin model.
[0021] Furthermore, the process of correcting the output of the vehicle collision digital twin model includes:
[0022] The temperature and collision force of the collision part of the vehicle virtual model output by the vehicle collision digital twin model are replaced with the temperature and collision force of the collision part of the corresponding vehicle in the historical collision data.
[0023] Furthermore, the process of evaluating the damage degree of the virtual model of the vehicle after the collision based on the trained vehicle collision digital twin model includes:
[0024] The vehicle virtual models of different models collide at different speeds in the collision layer under different types of vehicle collision scenario virtual models, and the vehicle collision digital twin model outputs the temperature and collision force of the collision part of the vehicle virtual model when the vehicle virtual models of different models collide at different speeds in the collision layer under different types of vehicle collision scenario virtual models;
[0025] Dividing the collision site of the virtual vehicle model into a plurality of sub-regions, labeling the divided sub-regions, obtaining the temperature within each sub-region, and obtaining the maximum temperature, minimum temperature, and temperature curve of the collision site of the virtual vehicle model based on the obtained temperature within each sub-region;
[0026] Integrating the obtained temperature curve of the collision part of the virtual vehicle model to obtain a temperature damage factor;
[0027] Acquire the collision force in each sub-region, and obtain the highest collision force, the lowest collision force, and a collision force curve diagram of the collision part of the vehicle virtual model according to the acquired collision force in each sub-region;
[0028] Integrating the collision force curve of the collision part of the obtained vehicle virtual model to obtain a collision force damage factor;
[0029] Obtaining a damage degree coefficient of the vehicle virtual model based on the obtained maximum temperature, minimum temperature, maximum collision force, minimum collision force, temperature damage factor, and collision force damage factor of the collision part of the vehicle virtual model;
[0030] A damage degree coefficient threshold range is set according to actual conditions, and the damage degree of the vehicle virtual model after the collision is obtained based on the damage degree coefficient of the obtained vehicle virtual model and the set damage degree coefficient threshold range. The damage degree of the vehicle virtual model after the collision is used as a reference to provide a basis for the damage degree of the vehicle after the actual vehicle collision.
[0031] Furthermore, the process of obtaining a temperature curve diagram of the collision part of the vehicle virtual model according to the acquired temperature in each sub-region includes:
[0032] A two-dimensional rectangular coordinate system is constructed with the sub-region number as the horizontal coordinate and the temperature within the sub-region as the vertical coordinate. The obtained temperature in each sub-region is mapped to the constructed two-dimensional rectangular coordinate system to generate corresponding temperature data sample points. Each temperature data sample point is connected in sequence through a curve in the order of the number to obtain a temperature curve diagram of the collision part of the vehicle virtual model.
[0033] Furthermore, the process of obtaining the damage degree of the virtual vehicle model after the collision based on the obtained damage degree coefficient of the virtual vehicle model and the set damage degree coefficient threshold range includes:
[0034] The obtained damage degree coefficient of the vehicle virtual model is compared with the set damage degree coefficient threshold range, and the damage degree of the vehicle virtual model after the collision is obtained according to the comparison result. The damage degree of the vehicle virtual model includes the first damage degree, the second damage degree and the third damage degree.
[0035] Compared with the prior art, the beneficial effects of the present invention are: constructing a vehicle collision physical model, obtaining historical collision data, obtaining a vehicle collision digital twin model based on the constructed vehicle collision physical model and the acquired historical collision data, training the vehicle collision digital twin model through historical collision data, obtaining a trained vehicle collision digital twin model, and evaluating the degree of damage to the virtual model of the vehicle after the collision based on the obtained trained vehicle collision digital twin model. The degree of damage to the virtual model of the vehicle includes first-level damage, second-level damage and third-level damage. The degree of damage to the virtual model of the vehicle after the collision is used as a reference to provide a basis for the degree of damage to the vehicle after the actual vehicle collision, thereby saving the cost of actual vehicle collision and speeding up the process of designing the car design plan and the process of verifying the design plan. BRIEF DESCRIPTION OF THE DRAWINGS
[0036] Figure 1 This is a schematic diagram of the present invention. DETAILED DESCRIPTION
[0037] like Figure 1 As shown, a vehicle collision simulation method based on digital twins includes the following steps:
[0038] Step S1: constructing a vehicle collision physical model;
[0039] Step S2: Obtain historical collision data, and obtain a vehicle collision digital twin model based on the constructed vehicle collision physical model and the acquired historical collision data;
[0040] Step S3: training the vehicle collision digital twin model using historical collision data to obtain a trained vehicle collision digital twin model;
[0041] Step S4: evaluating the damage degree of the virtual model of the vehicle after the collision based on the obtained trained vehicle collision digital twin model;
[0042] It should be further explained that, in the specific implementation process, the process of building a vehicle collision physics model includes:
[0043] The vehicle collision physics model consists of a vehicle layer, a scene layer and a collision layer;
[0044] Draw a number of virtual vehicle models of different models through 3D modeling, and label the drawn virtual vehicle models of different models as i, where i=1, 2, 3, ..., n, where n is a positive integer;
[0045] Draw several different types of virtual models of vehicle collision scenes through 3D modeling, and label the drawn virtual models of different types of vehicle collision scenes as j, where j=1, 2, 3, ..., m, where m is a positive integer;
[0046] The drawn vehicle virtual model is mapped into the vehicle layer, and the drawn vehicle collision scene virtual model is mapped into the scene layer to obtain a vehicle collision physical model.
[0047] It should be further explained that, in the specific implementation process, the process of obtaining the vehicle collision digital twin model based on the constructed vehicle collision physical model and the acquired historical collision data includes:
[0048] The historical collision data includes the temperature of the collision parts and the collision force of vehicles of different models when they collide at different speeds in different types of vehicle collision scenarios;
[0049] The acquired historical data is mapped to the collision layer in the vehicle collision physical model to obtain a vehicle collision digital twin model.
[0050] It should be further explained that, in the specific implementation process, the process of training the vehicle collision digital twin model using historical collision data includes:
[0051] Randomly select a model of vehicle virtual model and a type of vehicle collision scene virtual model, map the selected vehicle virtual model and vehicle collision scene virtual model to the collision layer, the vehicle virtual model collides at a speed of Vkm / h under the selected vehicle collision scene virtual model, V is a positive number, the vehicle collision digital twin model is output, and the temperature and collision force of the collision part of the vehicle virtual model are obtained. The temperature and collision force of the collision part of the corresponding vehicle in the historical collision data are compared with the temperature and collision force of the collision part of the obtained vehicle virtual model. If the temperature and collision force of the collision part of the corresponding vehicle in the historical collision data are the same as the temperature and collision force of the collision part of the obtained vehicle virtual model, then the output of the vehicle collision digital twin model is accurate. If the historical collision data If the temperature and collision force of the collision part of the corresponding vehicle in the vehicle collision scene are different from the temperature and collision force of the collision part of the obtained vehicle virtual model, the output of the vehicle collision digital twin model is inaccurate. The temperature and collision force of the collision part of the vehicle virtual model output by the vehicle collision digital twin model are replaced with the temperature and collision force of the collision part of the corresponding vehicle in the historical collision data. The vehicle virtual model is collided again at a speed of V km / h under the selected vehicle collision scene virtual model. The vehicle collision digital twin model outputs and obtains the temperature and collision force of the collision part of the vehicle virtual model. This process is repeated until the temperature and collision force of the collision part of the vehicle virtual model output by the vehicle collision digital twin model are the same as the temperature and collision force of the collision part of the corresponding vehicle in the historical collision data.
[0052] Similarly, the vehicle collision digital twin model is trained through historical collision data to obtain accurate temperature and collision force of the collision parts of the vehicle virtual models of other models when they collide at other speeds under other types of vehicle collision scene virtual models, and then obtain the trained vehicle collision digital twin model.
[0053] It should be further explained that, in the specific implementation process, the process of evaluating the damage degree of the virtual model of the vehicle after the collision based on the trained vehicle collision digital twin model includes:
[0054] Different types of vehicle virtual models collide at different speeds in the collision layer under different types of vehicle collision scenario virtual models. The speeds used for the collision are 1km / h, 2km / h, 3km / h, ..., bkm / h, where b is a positive number. The speeds used for the collision are labeled and recorded as h, where h=1, 2, 3, ..., v, where v is a positive integer. The vehicle collision digital twin model outputs the temperature and collision force of the collision part of the vehicle virtual model when different types of vehicle virtual models collide at different speeds in the collision layer under different types of vehicle collision scenario virtual models;
[0055] The collision area of the virtual vehicle model is divided into several sub-areas, and the sub-areas are labeled as k, k = 1, 2, 3, ..., p, where p is a positive integer;
[0056] Get the temperature in each sub-region and record the temperature in each sub-region as T ijhk ;
[0057] The maximum temperature of the collision part of the vehicle virtual model is obtained, and the obtained maximum temperature of the collision part of the vehicle virtual model is recorded as T ijhmax ;
[0058] in, ;
[0059] The lowest temperature of the collision part of the vehicle virtual model is obtained, and the lowest temperature of the collision part of the vehicle virtual model is recorded as T ijhmin ;
[0060] in, ;
[0061] A two-dimensional rectangular coordinate system is constructed using the sub-region number as the abscissa and the temperature within the sub-region as the ordinate. The acquired temperature within each sub-region is mapped to the constructed two-dimensional rectangular coordinate system to generate corresponding temperature data sample points. Each temperature data sample point is sequentially connected by a curve in the order of the number to obtain a temperature curve diagram of the collision part of the virtual vehicle model.
[0062] A function of a curve consisting of temperature data sample points is obtained based on the obtained temperature curve of the collision part of the virtual vehicle model. The obtained function of the curve is recorded as w(k). The obtained temperature curve of the collision part of the virtual vehicle model is integrated to obtain a temperature damage factor. The obtained temperature damage factor is recorded as W ijh ;
[0063] in, ;
[0064] Get the collision force in each sub-area, and record the collision force in each sub-area as L ijhk ;
[0065] Obtain the highest collision force of the collision part of the vehicle virtual model, and record the obtained highest collision force of the collision part of the vehicle virtual model as L ijhmax ;
[0066] in, ;
[0067] Obtain the minimum collision force of the collision part of the vehicle virtual model, and record the obtained minimum collision force of the collision part of the vehicle virtual model as L ijhmin ;
[0068] in, ;
[0069] A two-dimensional rectangular coordinate system is constructed using the sub-region numbers as the abscissa and the collision forces within the sub-regions as the ordinate. The collision forces within each sub-region are mapped to the constructed two-dimensional rectangular coordinate system to generate corresponding collision force data sample points. Each collision force data sample point is sequentially connected by a curve in order of the numbers to obtain a collision force curve diagram of the collision part of the vehicle virtual model.
[0070] According to the collision force curve of the collision part of the obtained vehicle virtual model, a function of the curve composed of the collision force data sample points is obtained, and the obtained curve function is recorded as y(k). The collision force curve of the collision part of the obtained vehicle virtual model is integrated to obtain the collision force damage factor, and the obtained collision force damage factor is recorded as Y ijh ;
[0071] in, ;
[0072] The damage coefficient of the vehicle virtual model is obtained based on the maximum temperature, minimum temperature, maximum collision force, minimum collision force, temperature damage factor and collision force damage factor of the collision part of the vehicle virtual model. The damage coefficient of the vehicle virtual model is recorded as WY ijh ;
[0073] in, ;
[0074] Set the damage degree coefficient threshold range according to the actual situation, and record the set damage degree coefficient threshold range as (WY0, WY1);
[0075] Obtaining the degree of damage to the virtual model of the vehicle after the collision based on a comparison result between the obtained damage degree coefficient of the virtual model of the vehicle and a set damage degree coefficient threshold range;
[0076] The damage degree of the vehicle virtual model includes a first-level damage degree, a second-level damage degree, and a third-level damage degree;
[0077] It should be further explained that, in the specific implementation process, the third level of damage is higher than the second level of damage, which is higher than the first level of damage;
[0078] When WY ijh When ≤WY0, the damage degree of the vehicle virtual model after the collision is the first level of damage;
[0079] When WY0<WY ijh When WY1 is less than 1, the damage degree of the virtual model of the vehicle after the collision is the second level of damage;
[0080] When WY ijh When ≥WY1, the damage degree of the vehicle virtual model after the collision is level 3 damage;
[0081] Using the damage degree of the virtual model of the vehicle after the collision as a reference provides a basis for the damage degree of the actual vehicle after the collision, saving costs.
[0082] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.
Claims
1. A vehicle collision simulation method based on digital twins, characterized in that: The following steps are involved: Step S1: constructing a vehicle collision physical model; Step S2: Obtain historical collision data, and obtain a vehicle collision digital twin model based on the constructed vehicle collision physical model and the acquired historical collision data; Step S3: training the vehicle collision digital twin model using historical collision data to obtain a trained vehicle collision digital twin model; Step S4: Evaluate the damage degree of the virtual model of the vehicle after the collision based on the obtained trained vehicle collision digital twin model.
2. The vehicle collision simulation method based on digital twin according to claim 1, characterized in that: The process of building a vehicle collision physics model includes: The vehicle collision physics model consists of a vehicle layer, a scene layer and a collision layer; Draw virtual models of several different types of vehicles and several different types of vehicle collision scenarios through 3D modeling; The drawn vehicle virtual model is mapped into the vehicle layer, and the drawn vehicle collision scene virtual model is mapped into the scene layer to obtain a vehicle collision physical model.
3. The vehicle collision simulation method based on digital twin according to claim 2, characterized in that: The process of obtaining a vehicle collision digital twin model based on the constructed vehicle collision physical model and the acquired historical collision data includes: The historical collision data includes the temperature of the collision parts and the collision force of vehicles of different models when they collide at different speeds in different types of vehicle collision scenarios; The acquired historical data is mapped to the collision layer in the vehicle collision physical model to obtain a vehicle collision digital twin model.
4. The vehicle collision simulation method based on digital twin according to claim 3 is characterized in that: The process of training the vehicle collision digital twin model using historical collision data includes: A vehicle model virtual model and a vehicle collision scenario virtual model of a type are randomly selected, and the selected vehicle virtual model and vehicle collision scenario virtual model are mapped to the collision layer. The vehicle virtual model collides at a speed of V km / h under the selected vehicle collision scenario virtual model, where V is a positive number. The vehicle collision digital twin model outputs the temperature and collision force of the collision part of the vehicle virtual model, and compares the temperature and collision force of the collision part of the corresponding vehicle in the historical collision data with the obtained temperature and collision force of the collision part of the vehicle virtual model; If the comparison results are the same, the output of the vehicle collision digital twin model is accurate. If the comparison results are different, the output of the vehicle collision digital twin model is inaccurate. The output of the vehicle collision digital twin model is corrected, and the vehicle virtual model is collided again at a speed of V km / h under the selected vehicle collision scenario virtual model. The vehicle collision digital twin model outputs the temperature of the collision part and the collision force of the vehicle virtual model, and so on, until the output of the vehicle collision digital twin model is accurate; Similarly, the vehicle collision digital twin model is trained through historical collision data to obtain accurate temperature and collision force of the collision parts of the vehicle virtual models of other models when they collide at other speeds under other types of vehicle collision scene virtual models, and then obtain the trained vehicle collision digital twin model.
5. The vehicle collision simulation method based on digital twin according to claim 4, characterized in that: The process of correcting the output of the vehicle crash digital twin model includes: The temperature and collision force of the collision part of the vehicle virtual model output by the vehicle collision digital twin model are replaced with the temperature and collision force of the collision part of the corresponding vehicle in the historical collision data.
6. The vehicle collision simulation method based on digital twin according to claim 5, characterized in that: The process of evaluating the damage degree of the virtual model of the vehicle after the collision based on the trained vehicle collision digital twin model includes: The vehicle virtual models of different models collide at different speeds in the collision layer under different types of vehicle collision scenario virtual models, and the vehicle collision digital twin model outputs the temperature and collision force of the collision part of the vehicle virtual model when the vehicle virtual models of different models collide at different speeds in the collision layer under different types of vehicle collision scenario virtual models; Dividing the collision site of the virtual vehicle model into a plurality of sub-regions, labeling the divided sub-regions, obtaining the temperature within each sub-region, and obtaining the maximum temperature, minimum temperature, and temperature curve of the collision site of the virtual vehicle model based on the obtained temperature within each sub-region; Integrating the obtained temperature curve of the collision part of the virtual vehicle model to obtain a temperature damage factor; Acquire the collision force in each sub-region, and obtain the highest collision force, the lowest collision force, and a collision force curve diagram of the collision part of the vehicle virtual model according to the acquired collision force in each sub-region; Integrating the collision force curve of the collision part of the obtained vehicle virtual model to obtain a collision force damage factor; Obtaining a damage degree coefficient of the vehicle virtual model based on the obtained maximum temperature, minimum temperature, maximum collision force, minimum collision force, temperature damage factor, and collision force damage factor of the collision part of the vehicle virtual model; A damage degree coefficient threshold range is set according to actual conditions, and the damage degree of the vehicle virtual model after the collision is obtained based on the damage degree coefficient of the obtained vehicle virtual model and the set damage degree coefficient threshold range. The damage degree of the vehicle virtual model after the collision is used as a reference to provide a basis for the damage degree of the vehicle after the actual vehicle collision.
7. The vehicle collision simulation method based on digital twin according to claim 6, characterized in that: The process of obtaining a temperature curve diagram of the collision part of the vehicle virtual model according to the acquired temperature in each sub-area includes: A two-dimensional rectangular coordinate system is constructed with the sub-region number as the horizontal coordinate and the temperature within the sub-region as the vertical coordinate. The obtained temperature in each sub-region is mapped to the constructed two-dimensional rectangular coordinate system to generate corresponding temperature data sample points. Each temperature data sample point is connected in sequence through a curve in the order of the number to obtain a temperature curve diagram of the collision part of the vehicle virtual model.
8. The vehicle collision simulation method based on digital twin according to claim 7, characterized in that: The process of obtaining the damage degree of the vehicle virtual model after the collision according to the obtained damage degree coefficient of the vehicle virtual model and the set damage degree coefficient threshold range includes: The obtained damage degree coefficient of the vehicle virtual model is compared with the set damage degree coefficient threshold range, and the damage degree of the vehicle virtual model after the collision is obtained according to the comparison result. The damage degree of the vehicle virtual model includes the first damage degree, the second damage degree and the third damage degree.