A vehicle parameter prediction method and device, electronic equipment and storage medium

Abstract

The application provides a prediction method and device of vehicle parameters, an electronic device and a storage medium, wherein the method comprises obtaining a description text of a target vehicle input by a user and extracting vehicle information. Based on the vehicle information, a corresponding vehicle sample data set is matched. Based on the vehicle data set, the torsional stiffness of the target vehicle is predicted. An end-to-end automated process from unstructured text input to standardized parameter output is realized, and the vehicle development efficiency is improved.

Description

Technical Field

[0001] This application relates to the field of vehicle technology, and more specifically, to a method, apparatus, electronic device, and storage medium for predicting vehicle parameters. Background Technology

[0002] With the rapid development of the automotive industry, the requirements for body-in-white performance in the forward development process of modern automobiles are becoming increasingly stringent. As the basic structure of the entire vehicle, the torsional stiffness of the body-in-white directly affects the vehicle's handling stability and collision safety, while the lightweight coefficient relates to the vehicle's energy consumption level and emissions performance. These two core indicators have become important parameters for evaluating the quality of automotive design.

[0003] In the automotive R&D field, traditional analysis methods have long dominated. Engineers need to manually extract key parameters from a large number of technical documents and manually organize historical vehicle data using tools such as Excel. On this basis, they also need to manually perform complex fitting calculations and scheme derivation processes. This manual-driven analysis mode is not only time-consuming and labor-intensive, but also difficult to keep up with the rapid iteration development pace of current vehicle models. With the popularization of new energy vehicles and the shortening of vehicle update cycles, the industry's demand for efficient and accurate stiffness setting and lightweight evaluation tools is becoming increasingly urgent. Summary of the Invention

[0004] The purpose of this application is to provide a method, apparatus, electronic device, and storage medium for predicting vehicle parameters, so as to solve the technical problem of low efficiency in existing vehicle development.

[0005] In a first aspect, the present invention provides a method for predicting vehicle parameters, the method comprising: Obtain the descriptive text of the target vehicle input by the user and extract the vehicle information; Based on vehicle information, the corresponding vehicle sample dataset is matched; Based on the vehicle dataset, the torsional stiffness of the target vehicle is predicted.

[0006] In an optional implementation, vehicle information is extracted in the following way: The intent of the descriptive text is parsed using a large language model to identify vehicle information, which includes at least three of the following: vehicle class, energy type, vehicle type, and size parameters. The dimensions include overall length, overall width, overall height, wheelbase, front track, and rear track.

[0007] In an optional implementation, the corresponding vehicle sample dataset is matched using the following method: Based on the vehicle type and energy type in the vehicle information, perform multi-dimensional condition filtering in the vehicle history database; The vehicle history database stores structured parameters for multiple vehicle models. These structured parameters include at least the following fields: vehicle class, energy type, surface area, body-in-white weight, lightweight coefficient, torsional stiffness test value, and year of market launch. After screening, outlier samples whose torsional stiffness values ​​deviate from the box plot IQR range by more than 1.5 times were removed.

[0008] In an optional implementation, the vehicle sample dataset includes four subsets: the first subset consists of surface area-body-in-white mass data pairs, the second subset consists of lightweight coefficient-torsional stiffness data pairs, the third subset consists of lightweight coefficient-market year time series, and the fourth subset consists of torsional stiffness-market year time series.

[0009] In an optional implementation, the torsional stiffness CT of the target vehicle is predicted in the following manner: CT = (M × 1050) / (A × l); Where M is the predicted mass of the target vehicle, A is the predicted footprint area of ​​the target vehicle, and l is the lightweighting coefficient of the target vehicle.

[0010] In an optional implementation, the predicted footprint area A of the target vehicle is determined by the following method: A = L × ( + ) / 2; Where L is the wheelbase of the target vehicle. , These are the front track and rear track of the target vehicle, respectively.

[0011] In an optional implementation, the lightweighting factor includes L0 / L25 / L50 percentile lightweighting factors, which are used to calculate the three-level torsional stiffness of CT100 / CT75 / CT50, respectively.

[0012] Secondly, the present invention provides a vehicle parameter prediction device, the device comprising: The extraction module is used to obtain the descriptive text of the target vehicle input by the user and extract the vehicle information; The matching module is used to match the corresponding vehicle sample dataset based on vehicle information; The prediction module is used to predict the torsional stiffness of a target vehicle based on a vehicle dataset.

[0013] Thirdly, the present invention provides an electronic device, comprising: a processor, a memory, and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor communicates with the memory via the bus, and the processor executes the machine-readable instructions to perform the steps of the prediction method for any of the vehicle parameters as described in the foregoing embodiments.

[0014] Fourthly, the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the prediction method for any of the vehicle parameters described in the foregoing embodiments.

[0015] This application provides a method, apparatus, electronic device, and storage medium for predicting vehicle parameters. The method includes acquiring a user-input descriptive text of a target vehicle and extracting vehicle information. Based on the vehicle information, a corresponding vehicle sample dataset is matched. Based on the vehicle dataset, the torsional stiffness of the target vehicle is predicted. This achieves an end-to-end automated process from unstructured text input to standardized parameter output, improving vehicle development efficiency. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 A flowchart illustrating a method for predicting vehicle parameters provided in an embodiment of this application; Figure 2 An architecture diagram of an intelligent prediction system provided in this application embodiment; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0018] The technical solutions in the embodiments of this application will now be described with reference to the accompanying drawings.

[0019] Example 1 Figure 1 A flowchart illustrating a method for predicting vehicle parameters provided in an embodiment of this application. Figure 1 As shown, this application provides a method for predicting vehicle parameters, including: S1. Obtain the description text of the target vehicle input by the user and extract the vehicle information.

[0020] In step S1, vehicle information can be extracted in the following way: The intent of the descriptive text is parsed using a large language model to identify vehicle information, which includes at least three of the following: vehicle class, energy type, vehicle type, and size parameters.

[0021] The dimensions include overall length, overall width, overall height, wheelbase, front track, and rear track.

[0022] This interface can receive natural language input from users, such as "calculate the stiffness of a Class A gasoline SUV". It uses a large model to extract vehicle class, energy type, and size parameters, and outputs standard JSON format through regular expressions to achieve the structured transformation of unstructured text.

[0023] The vehicle category here includes vehicle class and vehicle type. Vehicle class includes A, A0, A00, A-SUV, B, C, MPV, etc., while vehicle type includes sedan, SUV, MPV, etc.

[0024] S2. Based on vehicle information, match the corresponding vehicle sample dataset.

[0025] In step S2, the corresponding vehicle sample dataset can be matched in the following way: Based on the vehicle class and energy type in the vehicle information, multi-dimensional conditional filtering is performed in the vehicle history database. The vehicle history database stores structured parameters of multiple vehicle models, which include at least the following fields: vehicle class, energy type, surface area, body-in-white mass, lightweight coefficient, torsional stiffness test value, and year of market launch.

[0026] After screening, outlier samples whose torsional stiffness values ​​deviate from the box plot IQR range by more than 1.5 times were removed.

[0027] The vehicle sample dataset here can include four subsets: the first subset consists of surface area-body-in-white mass data pairs, the second subset consists of lightweight coefficient-torsional stiffness data pairs, the third subset consists of lightweight coefficient-year of market launch time series, and the fourth subset consists of torsional stiffness-year of market launch time series.

[0028] Similarly, the LLM model can be used to retrieve data from user-uploaded database files based on the extracted standard JSON format parameters, automatically filter and match vehicle class, energy type, and brand name to obtain surface area data, mass data, lightweight coefficient data, torsional stiffness test data, etc. of vehicles that meet the filtering conditions.

[0029] Finally, outliers in torsional stiffness were removed, and four datasets were constructed: surface area-mass, lightweight coefficient-stiffness, lightweight coefficient-time, and torsional stiffness-time.

[0030] S3. Based on the vehicle dataset, predict the torsional stiffness of the target vehicle.

[0031] In step S3, the torsional stiffness CT of the target vehicle can be predicted in the following way: CT = (M × 1050) / (A × l); Where M is the predicted mass of the target vehicle, A is the predicted footprint area of ​​the target vehicle, and l is the lightweighting coefficient of the target vehicle.

[0032] The predicted mass of the target vehicle can be obtained by using the least squares algorithm to fit the surface area data and body-in-white mass data selected by the model, thus obtaining the slope, intercept, and formula. Substituting the calculated surface area into the fitting formula yields the fitted mass of the user-input vehicle.

[0033] And, the predicted footprint area A of the target vehicle is determined by the following method: A = L × ( + ) / 2; Where L is the wheelbase of the target vehicle. , These are the front track and rear track of the target vehicle, respectively.

[0034] The lightweighting coefficients include L0 / L25 / L50 percentile lightweighting coefficients, which are used to calculate the torsional stiffness of CT100 / CT75 / CT50 levels, respectively. Specifically, the lightweighting coefficient samples can be sorted in ascending order, and linear interpolation can be used to calculate the three percentile benchmark values ​​of L0 (optimal), L25 (excellent), and L50 (average), forming a quantitative evaluation system.

[0035] In one feasible implementation, the surface area and footprint area can be calculated based on parameters such as total length, total width, total height, wheelbase, front track, and rear track input by the user.

[0036] The surface area of ​​the target vehicle can be calculated using the formula: "total length × total width + total length × total height + total width × total height".

[0037] Furthermore, after calculating the target stiffness, the torsional stiffness obtained by the user may not meet the requirements. The user then queries the large model for methods to improve the stiffness. The large model searches the body-in-white stiffness and lightweight suggestion library and provides specific solutions. During the search, a hard filtering strategy is executed, retaining only core components such as the body-in-white, subframe, and dashboard crossbeams. If the vehicle is identified as an ICE (Internal Combustion Engine) vehicle, optimization suggestions containing "battery" or "BMS" are automatically removed, thereby obtaining a more accurate torsional stiffness.

[0038] It can also generate visualization chart configurations that conform to ECharts specifications based on the predicted data. For example, it can dynamically encapsulate scatter plots and fitted lines of surface area-mass regression analysis, percentile distribution plots of lightweight coefficients, and extrapolation plots of historical stiffness trends, and supports converting chart configurations into standard JSON format output.

[0039] This application provides a method for predicting vehicle parameters that can automatically complete the entire process from parameter extraction and data modeling to report generation, without relying on expert experience for manual Excel calculations. This significantly lowers the threshold for forward design and improves vehicle design efficiency.

[0040] Example 2 In one embodiment of this application, a method for intelligent prediction of automotive lightweighting and torsional stiffness based on Dify and large models is provided, comprising: Based on the Dify workflow, an intent parsing agent is built to extract the structured text of the vehicle model parameters input by the user and output standardized JSON format parameters, including vehicle class, energy type, surface area, and footprint area.

[0041] A dataset intelligence agent is constructed to match samples of the same class and energy type from the vehicle database based on standardized parameters, generating surface area-mass datasets, lightweight coefficient-torsional stiffness datasets, and historical time series datasets.

[0042] The code node executes the calculation of the target vehicle's size information, and calculates the target vehicle's surface area and footprint area based on the vehicle size input by the user.

[0043] By performing linear fitting through code nodes, a regression model of surface area and mass is established, the slope, intercept, and R² are calculated, and the fitted equation and ECharts visualization configuration are output.

[0044] The L0 / L25 / L50 percentile evaluation criteria were calculated based on the lightweight coefficient dataset, and the torsional stiffness of the three-level targets CT100 / CT75 / CT50 was calculated by combining mass and footprint area.

[0045] Linear regression is performed based on historical time series data to predict the changing trends of torsional stiffness and lightweight coefficient in future years.

[0046] Retrieve lightweight solutions from the database and output compliant lightweight recommendations based on energy type and component constraints.

[0047] The entire process uses an automatic verification and feedback mechanism to correct model inference biases and achieve continuous optimization.

[0048] The standardized parameter extraction can employ a large model to extract user intent, vehicle model and brand, vehicle size, test results, test time, and test personnel from user input, and output JSON structured text to achieve the conversion of unstructured text into structured parameters.

[0049] The dataset building agent here supports precise filtering by vehicle class and energy type. Additionally, the dataset building agent supports precise filtering by vehicle type and energy type.

[0050] In one feasible embodiment, the target vehicle size information is calculated by using the total length, total width, total height, wheelbase, front track, and rear track, and then using an internal empirical formula to calculate the target vehicle's surface area and footprint area.

[0051] The linear fitting here can be achieved by using the least squares method to fit the surface area and body-in-white mass curves.

[0052] The final output here can include scatter plots, regression lines, and supports visualization to verify the fitting effect.

[0053] Specifically, the lightweighting coefficient can be evaluated using the percentile method, which sorts the samples from smallest to largest and performs linear interpolation to output the 0th percentile (optimal), 25th percentile, and 50th percentile.

[0054] Specifically, historical trend prediction can perform linear regression on time-stiffness and time-lightweight coefficients to directly output the predicted value for a specified future year.

[0055] The recommended lightweight design only includes the body-in-white, front and rear subframes, dashboard crossbeams, windshield, and sunroof, excluding door panels and interior / exterior trim. The ICE model does not include battery-related optimizations.

[0056] Example 3 Figure 2 This is an architecture diagram of an intelligent prediction system provided in an embodiment of this application. In one embodiment of this application, an intelligent prediction system for automotive lightweighting and torsional stiffness based on Dify and large models (hereinafter referred to as the "intelligent prediction system") is provided, comprising three main components: a Dify orchestration hardware resource layer, an intelligent algorithm software layer, and an engineering database layer.

[0057] Its core lies in using a visual workflow engine to connect the semantic understanding capabilities of a Large Language Model (LLM) with the numerical computation capabilities of a code executor, achieving end-to-end automation from unstructured text input to standardized engineering report output. The specific technical solution is as follows: The Dify orchestration hardware resource layer consists of 5 core functional units, providing computing power support and process scheduling.

[0058] The first unit is the Dify orchestration engine, which serves as the system's central hub. It is responsible for defining the topological connections between agent nodes, data flow paths, and exception handling logic, enabling visual drag-and-drop management of workflows.

[0059] The second unit is the large model inference unit, equipped with a finely tuned large language model (LLM) with built-in semantic parsing capabilities and intent recognition algorithms, responsible for converting unstructured natural language into structured parameters.

[0060] The third unit is the code execution unit, which integrates a Python sandbox environment and comes pre-installed with scientific computing libraries such as NumPy and SciPy. It is responsible for performing high-precision numerical calculations such as linear regression, percentile statistics, and stiffness formula derivation.

[0061] The fourth unit is the document extraction unit, which supports PDF / Word / Excel format parsing. It automatically extracts table data and key paragraph parameters from design documents through OCR technology and layout analysis algorithms.

[0062] The fifth unit is the RAG retrieval unit, which connects to an external vehicle model database and has a built-in dual-path mechanism of vector retrieval and keyword matching. It is responsible for accurately recalling sample data of vehicles of the same class based on contextual semantics.

[0063] The intelligent algorithm software layer is integrated on top of the Dify orchestration engine and includes eight intelligent agent modules that execute specific business logic.

[0064] The first module is an intent-parsing agent that receives natural language input from the user (e.g., "Calculate the stiffness of a Class A gasoline SUV"), extracts vehicle class, energy type, and size parameters using a large model, and outputs the result in standard JSON format via regular expressions, achieving a structured transformation of unstructured text. Vehicle class includes vehicle level and vehicle type; vehicle classes include A, A0, A00, A-SUV, B, C, MPV, etc., and vehicle types include sedans, SUVs, MPVs, etc. The predicted parameter is named the torsional stiffness of the body-in-white.

[0065] The second module constructs an intelligent agent for the dataset. Based on the standard JSON format parameters output from the first module, the LLM model searches the user-uploaded database file, automatically filtering and matching vehicle class, energy type, and brand name to obtain surface area data, mass data, lightweighting coefficient data, and torsional stiffness test data for vehicles that meet the filtering criteria. Outliers in the torsional stiffness are removed, and four datasets are constructed: surface area-mass, lightweighting coefficient-stiffness, lightweighting coefficient-time, and torsional stiffness-time.

[0066] The third module is the dimension calculation module. It performs geometric calculations through code nodes, calculating the surface area and footprint area based on user-input parameters such as total length, total width, total height, wheelbase, front track, and rear track. The formula for surface area is: total length × total width + total length × total height + total width × total height. The formula for footprint area is: wheelbase × (front track + rear track) / 2.

[0067] The fourth module is the linear fitting module. It uses the least squares algorithm to fit the surface area data and body-in-white mass data selected in the second module, obtaining the fitted slope, intercept, and formula. Substituting the surface area obtained from the user-input dimensions calculated in the third module into the fitting formula yields the fitted mass of the user-input vehicle.

[0068] The fifth module is the percentile evaluation module, which sorts the lightweight coefficient samples in ascending order and uses linear interpolation to calculate three percentile benchmark values: L0 (optimal), L25 (good), and L50 (average), forming a quantitative evaluation system.

[0069] The sixth module is the stiffness target calculation module, based on the formula CT=(M×1050) / (A×L), where M represents mass, A represents vehicle footprint area, L represents lightweight coefficient, and CT represents torsional stiffness. Substituting L0, L25, and L50 obtained from the fifth module, the footprint area obtained from the third module, and the fitted mass of the vehicle from the fourth module into the formula, the calculation outputs the corresponding three-level stiffness targets: CT100 (high performance), CT75 (balanced), and CT50 (basic).

[0070] The seventh module is the trend prediction module. It performs linear fitting on the lightweight coefficient-time and torsional stiffness-time datasets obtained from the second module, using the same fitting method as the second module, to obtain the fitted formulas. The time in the dataset is in years (e.g., 2026). The time is extended to the next two years and substituted into the fitting formula to calculate the target values ​​of lightweight coefficient and torsional stiffness for the next two years.

[0071] The eighth module is the solution recommendation module. After calculating the target stiffness, the user may find that the obtained stiffness target result does not meet the requirements. The user queries the large model for methods to improve the stiffness. The large model searches the stiffness and lightweight suggestion library of the white body and provides specific solutions. During the search, a hard filtering strategy is executed, retaining only core components such as the white body, subframe, and instrument panel crossbeam. If the vehicle is identified as a gasoline vehicle (ICE), optimization suggestions containing "battery" or "BMS" are automatically removed.

[0072] The ninth module is the charting module, which is configured to receive output data from the linear fitting module, percentile evaluation module, and trend prediction module, and automatically generate visualization chart configurations that conform to the ECharts specification. This module dynamically encapsulates the scatter plot and fitted line of surface area-mass regression analysis, the percentile distribution plot of lightweight coefficient, and the historical stiffness trend extrapolation plot, and supports converting the chart configurations into standard JSON format output.

[0073] The engineering database layer serves as the data foundation of the system and contains two types of core data assets.

[0074] The engineering database layer includes a vehicle history database, storing detailed parameters for no fewer than 500 vehicle models. Fields cover vehicle class, energy type, surface area, body-in-white mass, lightweight coefficient, torsional stiffness, and year of market launch, and support queries based on multi-dimensional conditions.

[0075] The engineering database layer also includes a lightweight solution knowledge base, which is used to structure and store optimization solutions. Each record contains tags such as "applicable components", "optimization measures", "weight reduction expectations", "cost changes" and "applicable energy type", providing a retrieval source for Module 8.

[0076] This application achieves a fully automated closed loop through the collaboration of the aforementioned hardware units and the streamlined processing of software modules. Ordinary technicians only need to input natural language, and the system can automatically complete the entire process from parameter extraction and data modeling to report generation, without relying on expert experience for manual Excel calculations, which greatly reduces the threshold for forward design.

[0077] Example 4 Please see Figure 3 , Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 3 As shown, the electronic device 300 includes a processor 310, a memory 320, and a bus 330.

[0078] The memory 320 stores machine-readable instructions that can be executed by the processor 310. When the electronic device 300 is running, the processor 310 and the memory 320 communicate via the bus 330. When the machine-readable instructions are executed by the processor 310, the steps of a vehicle parameter prediction calculation method as described in the above method embodiment can be executed. For specific implementation details, please refer to the method embodiment, which will not be repeated here.

[0079] Example 5 This application also provides a computer-readable storage medium storing a computer program. When the computer program is run by a processor, it can execute the steps of a vehicle parameter prediction method as described in the above method embodiments. For specific implementation details, please refer to the method embodiments, which will not be repeated here.

[0080] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0081] Based on the same inventive concept, the embodiments of this application also provide In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some communication interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0082] Furthermore, the units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0083] Furthermore, the functional modules in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0084] It should be noted that if the function is implemented as a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0085] In this document, relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, without necessarily requiring or implying any such actual relationship or order between these entities or operations.

[0086] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for predicting vehicle parameters, characterized in that, The method includes: Obtain the descriptive text of the target vehicle input by the user and extract the vehicle information; Based on the vehicle information, a corresponding vehicle sample dataset is matched; Based on the vehicle dataset, the torsional stiffness of the target vehicle is predicted.

2. The method according to claim 1, characterized in that, Vehicle information can be extracted using the following methods: The intent of the descriptive text is parsed using a large language model to identify vehicle information, which includes at least three of the following: vehicle model class, energy type, vehicle type, and size parameters. The dimensional parameters include overall length, overall width, overall height, wheelbase, front track, and rear track.

3. The method according to claim 1, characterized in that, The corresponding vehicle sample dataset is matched using the following method: Based on the vehicle type and energy type in the vehicle information, multi-dimensional condition filtering is performed in the vehicle history database. The vehicle history database stores structured parameters for multiple vehicle models, including at least the following fields: vehicle class, energy type, surface area, body-in-white weight, lightweight coefficient, torsional stiffness test value, and year of market launch. After screening, outlier samples whose torsional stiffness values ​​deviate from the box plot IQR range by more than 1.5 times were removed.

4. The method according to claim 3, characterized in that, The vehicle sample dataset includes four subsets: the first subset consists of surface area-body-in-white mass data pairs, the second subset consists of lightweight coefficient-torsional stiffness data pairs, the third subset consists of lightweight coefficient-market year time series, and the fourth subset consists of torsional stiffness-market year time series.

5. The method according to claim 4, characterized in that, The torsional stiffness CT of the target vehicle is predicted using the following method: CT = (M × 1050) / (A × l); Where M is the predicted mass of the target vehicle, A is the predicted footprint area of ​​the target vehicle, and l is the lightweighting coefficient of the target vehicle.

6. The method according to claim 5, characterized in that, The predicted footprint area A of the target vehicle is determined using the following method: A = L × ( + ) / 2; Where L is the wheelbase of the target vehicle. , These are the front track and rear track of the target vehicle, respectively.

7. The method according to claim 5, characterized in that, The lightweighting coefficients include L0 / L25 / L50 percentile lightweighting coefficients, which are used to calculate the three levels of torsional stiffness of CT100 / CT75 / CT50, respectively.

8. A vehicle parameter prediction device, characterized in that, The device includes: The extraction module is used to obtain the descriptive text of the target vehicle input by the user and extract the vehicle information; The matching module is used to match the corresponding vehicle sample dataset based on the vehicle information; The prediction module is used to predict the torsional stiffness of the target vehicle based on the vehicle dataset.

9. An electronic device, characterized in that, include: The device includes a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor communicates with the memory via the bus, and the processor executes the machine-readable instructions to perform the steps of the method for predicting vehicle parameters as claimed in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, performs the steps of the method for predicting vehicle parameters as described in any one of claims 1 to 7.

Images