Vehicle weight target setting method and device

Abstract

This application discloses a method and apparatus for setting a vehicle weight target. The method includes: acquiring weight characteristic data of competing vehicles corresponding to the development model; constructing a machine learning model based on the weight characteristic data, wherein the machine learning model is used to predict the overall vehicle weight of the development model or competing vehicles; determining the lightweighting rate of the competing vehicles based on the prediction results of the machine learning model, and determining the lightweighting rate boundary values ​​of the competing vehicles at different industry levels based on the lightweighting rate, wherein the lightweighting rate reflects the lightweighting level of the competing vehicles; and determining the overall vehicle weight target of the development model based on the lightweighting rate boundary values. This application solves the technical problem that existing methods for setting vehicle weight targets can only output a single target value that is the industry average or optimal, and cannot be dynamically set according to vehicle positioning and classification. Furthermore, they rely on static fitting models, making it difficult to efficiently respond to iterative competitor data, resulting in insufficient accuracy in setting weight target values.

Description

Technical Field

[0001] This application relates to the field of automotive R&D technology, and more specifically, to a method and apparatus for setting a target vehicle weight. Background Technology

[0002] Vehicle weight, as a parameter describing a car's static properties, directly affects its power performance, energy economy, safety, and handling stability. Studies show that for every 10% reduction in vehicle weight, fuel consumption of gasoline vehicles can decrease by 6-8%, and the range of new energy vehicles can increase by 6-8%. Simultaneously, weight reduction can increase tire life by 7%, shorten braking distance by 3-4%, reduce acceleration time by 6-8%, and optimize vehicle layout.

[0003] Due to the complexity of the vehicle system and the diversity of influencing factors, predicting the target vehicle weight presents certain challenges. In related technologies, setting the target vehicle weight mainly relies on two types of methods: First, using a polynomial fitting method based on a few geometric parameters such as the vehicle's length, width, and height. However, this method lacks sufficient feature dimensions and fails to cover key variables such as battery energy and energy density, making it poorly applicable to new energy vehicles. Second, calculating the battery weight independently and then superimposing the remaining target value. While applicable to new energy vehicles, this method outputs an industry-optimal value, lacking adaptability to non-top-tier models, potentially leading to excessively high development costs or wasted resources. Furthermore, both methods are based on static regression models. When competing models iterate rapidly, data needs to be re-collected and repeatedly fitted, resulting in slow response and low efficiency.

[0004] There is currently no effective solution to the above problems. Summary of the Invention

[0005] This application provides a method and apparatus for setting a vehicle weight target, which at least solves the technical problem that the vehicle weight target setting methods in the related art can only output a single target value that is the industry average or the best, and cannot be dynamically set according to the vehicle model positioning and classification. Furthermore, they rely on static fitting models and are difficult to respond efficiently to the iteration of competitor data, resulting in insufficient accuracy in setting the weight target value.

[0006] According to one aspect of the embodiments of this application, a method for setting a vehicle weight target is provided, comprising: acquiring weight characteristic data of competitor vehicles corresponding to the development model; constructing a machine learning model based on the weight characteristic data, wherein the machine learning model is used to predict the vehicle weight of the development model or competitor vehicles; determining the lightweighting rate of competitor vehicles based on the prediction results of the machine learning model, and determining the lightweighting rate boundary values ​​of competitor vehicles at different industry levels based on the lightweighting rate, wherein the prediction results are used to represent the predicted vehicle weight of competitor vehicles, and the lightweighting rate is used to reflect the lightweighting level of competitor vehicles; and determining the vehicle weight target of the development model based on the lightweighting rate boundary values.

[0007] Optionally, the weight characteristic data of competing vehicles corresponding to the developed model can be obtained, including: obtaining the vehicle weight data, original features and derived features of competing vehicles according to preset screening principles, wherein the original features are used to represent the basic structural parameters of competing vehicles, and the derived features are used to represent the spatial features and energy configuration features of competing vehicles; and determining the weight characteristic data based on the vehicle weight data, original features and derived features.

[0008] Optionally, the method further includes: determining interactive feature terms based on the original features, wherein the interactive feature terms are the product or ratio between the original features; adding the interactive feature terms to the weight feature data, and filtering the weight feature data according to preset requirements to obtain target weight feature data; and standardizing the target weight feature data.

[0009] Optionally, a machine learning model is constructed based on weight feature data, including: dividing the weight feature data into a training set and a test set according to a preset ratio; training multiple initial models using the training set, and determining the evaluation metrics of the initial models on the test set after training, wherein the evaluation metrics include at least one of the following: mean absolute error, root mean square error, and coefficient of determination, the coefficient of determination being used to reflect the explanatory power of the initial models for changes in the overall vehicle weight; and determining a machine learning model from the initial models based on the evaluation metrics for predicting the overall vehicle weight of the development model or competing models.

[0010] Optionally, the lightweighting rate of the competitor vehicle is determined based on the prediction results of the machine learning model, including: predicting the weight feature data through the machine learning model to obtain a first predicted value; and determining the lightweighting rate of the competitor vehicle based on the first predicted value and the actual overall weight of the competitor vehicle.

[0011] Optionally, determining the lightweighting rate boundary values ​​of competitor vehicles in different industry levels based on the lightweighting rate includes: classifying competitor vehicles into a first industry level, a second industry level, a third industry level, and a fourth industry level based on the lightweighting rate and a preset classification ratio, wherein the product positioning capability corresponding to the first industry level is greater than that corresponding to the second industry level, the product positioning capability corresponding to the second industry level is greater than that corresponding to the third industry level, and the product positioning capability corresponding to the third industry level is greater than that corresponding to the fourth industry level; determining the minimum lightweighting rate of competitor vehicles in the first industry level, and determining the maximum and minimum lightweighting rates of competitor vehicles in the second, third, and fourth industry levels respectively, to obtain the lightweighting rate boundary values.

[0012] Optionally, determining the vehicle weight target of the development model based on the lightweighting rate boundary value includes: acquiring the original weight characteristic data of the development model; predicting the original weight characteristic data through a machine learning model to obtain a second predicted value, wherein the second predicted value is used to represent the predicted vehicle weight of the development model; determining the target industry level to which the development model belongs, and determining the vehicle weight target of the development model based on the lightweighting rate boundary value and the second predicted value under the target industry level.

[0013] According to another aspect of the embodiments of this application, a method for setting automotive component weight targets is also provided, comprising: acquiring weight characteristic data of automotive components of competing vehicles corresponding to the development model, wherein the automotive components include at least one of the following: automotive systems, subsystems, and parts; constructing a machine learning model based on the weight characteristic data, wherein the machine learning model is used to predict the weight of automotive components of the development model or competing vehicles; determining the lightweighting rate of automotive components of competing vehicles based on the prediction results of the machine learning model, and determining the lightweighting rate boundary values ​​of automotive components of competing vehicles at different industry levels based on the lightweighting rate, wherein the prediction results are used to represent the weight prediction results of automotive components of competing vehicles, and the lightweighting rate is used to reflect the lightweighting level of automotive components of competing vehicles; and determining the automotive component weight target of the development model based on the lightweighting rate boundary values.

[0014] According to another aspect of the embodiments of this application, a vehicle weight target setting device is also provided, comprising: a first acquisition module, configured to acquire weight characteristic data of competing vehicles corresponding to the development model; a first construction module, configured to construct a machine learning model based on the weight characteristic data, wherein the machine learning model is used to predict the overall vehicle weight of the development model or competing vehicles; a first determination module, configured to determine the lightweighting rate of competing vehicles based on the prediction results of the machine learning model, and determine the lightweighting rate boundary values ​​of competing vehicles at different industry levels based on the lightweighting rate, wherein the prediction results are used to represent the overall vehicle weight prediction results of competing vehicles, and the lightweighting rate is used to reflect the lightweighting level of competing vehicles; and a second determination module, configured to determine the overall vehicle weight target of the development model based on the lightweighting rate boundary values.

[0015] According to another aspect of the embodiments of this application, an electronic device is also provided, including: a memory and a processor, wherein the memory is used to store program instructions; and the processor is connected to the memory and used to execute the above-described method for setting the vehicle weight target or the method for setting the weight target of automotive components.

[0016] According to another aspect of the embodiments of this application, a non-volatile storage medium is also provided, the non-volatile storage medium including a stored computer program, wherein the device where the non-volatile storage medium is located executes the above-mentioned vehicle weight target setting method or automobile component weight target setting method by running the computer program.

[0017] According to another aspect of the embodiments of this application, a computer program product is also provided, including computer instructions, which, when executed by a processor, implement the above-described method for setting the overall vehicle weight target or the method for setting the weight target of automotive components.

[0018] In this embodiment, the weight characteristic data of competing vehicles corresponding to the development model are obtained; a machine learning model is constructed based on the weight characteristic data, wherein the machine learning model is used to predict the overall vehicle weight of the development model or competing vehicles; the lightweighting rate of the competing vehicles is determined based on the prediction results of the machine learning model, and the lightweighting rate boundary values ​​of the competing vehicles under different industry levels are determined based on the lightweighting rate, wherein the prediction results are used to represent the overall vehicle weight prediction results of the competing vehicles, and the lightweighting rate is used to reflect the lightweighting level of the competing vehicles; the overall vehicle weight target of the development model is determined based on the lightweighting rate boundary values, thereby achieving the purpose of setting differentiated overall vehicle weight targets according to product positioning levels, thus realizing the technical effect of improving the accuracy, flexibility and iteration efficiency of weight target setting, and solving the technical problem that the overall vehicle weight target setting methods in related technologies can only output a single target value of industry average or optimal, cannot be dynamically set according to vehicle positioning levels, and rely on static fitting models, making it difficult to efficiently respond to competitor data iteration, resulting in insufficient accuracy of weight target value setting. Attached Figure Description

[0019] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0020] Figure 1 This is a hardware structure diagram of a computer terminal for implementing a method for setting a vehicle weight target or a method for setting a vehicle component weight target, according to an embodiment of this application.

[0021] Figure 2 This is a flowchart of a method for setting a vehicle weight target according to an embodiment of this application;

[0022] Figure 3 This is a flowchart illustrating a model dynamic update mechanism according to an embodiment of this application;

[0023] Figure 4 This is a flowchart of a method for setting a target weight for automotive components according to an embodiment of this application;

[0024] Figure 5 This is a structural diagram of a vehicle weight target setting device according to an embodiment of this application;

[0025] Figure 6This is a structural diagram of an automotive component weight target setting device according to an embodiment of this application. Detailed Implementation

[0026] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0027] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0028] First, some nouns or terms that appear in the explanation of the embodiments of this application shall be interpreted as follows:

[0029] Vehicle weight: refers to the total mass of a vehicle in its prepared state, including the mass of all necessary components such as the body, chassis, power system, fuel / battery, coolant, lubricant, spare tire, and tools. It is a core parameter for measuring the static attributes of a vehicle and directly affects its power, economy, safety, and handling stability.

[0030] Competitor vehicles: These refer to mass-produced vehicles that are comparable to the development vehicle in key dimensions such as wheelbase, drive type, power type, market positioning, and year on the market (preferably within the last three years). They are used to build training datasets or as a benchmark for lightweighting levels.

[0031] Lightweight ratio: refers to the relative deviation between the actual weight of a vehicle and its baseline weight (predicted value). The larger the value, the higher the lightweight level of the vehicle, which means that it has achieved better weight reduction effect with the same configuration.

[0032] Machine learning models refer to intelligent algorithm models trained based on historical competitor vehicle data, used to predict the weight of the whole vehicle or components. These include, but are not limited to, linear regression, lasso regression, random forest, gradient boosting tree, etc. Their input is weight feature data, and their output is the predicted weight value. The best model is selected through evaluation metrics, such as MAE (Mean Absolute Error), RMSE (Root Mean Square Error), and R² (Coefficient of Determination).

[0033] Automotive components: refer to any level of functional or structural units that constitute a vehicle, including vehicle systems (such as chassis systems), subsystems (such as suspension systems and braking systems), and parts (such as front subframes and battery packs). The method proposed in this application is applicable to the setting of weight targets for components at any level.

[0034] To address the issue of poor accuracy in detection algorithms in related technologies, this application provides a method for setting a vehicle weight target and a method for setting a vehicle component weight target. This method can be implemented in… Figure 1 The computer terminal shown is described below.

[0035] The vehicle weight target setting method and the automotive component weight target setting method provided in this application can be executed on a mobile terminal, computer terminal or similar computing device. Figure 1 A hardware block diagram of a computer terminal for implementing methods for setting vehicle weight targets and methods for setting automotive component weight targets is shown. Figure 1 As shown, the computer terminal 10 may include one or more processors (shown as 102a, 102b, ..., 102n in the figure) (the processor may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.), a memory 104 for storing data, and a transmission module 106 for communication functions connected via wired and / or wireless networks. In addition, it may also include: a display, a keyboard, a cursor control device, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the I / O interface), a network interface, and a BUS bus. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0036] It should be noted that the aforementioned one or more processors and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element within the computer terminal 10. As involved in the embodiments of this application, the data processing circuits serve as a processor control mechanism (e.g., selection of a variable resistor termination path connected to an interface).

[0037] The memory 104 can be used to store software programs and modules for application software, such as the program instructions / data storage device corresponding to the vehicle weight target setting method and the automotive component weight target setting method in the embodiments of this application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the aforementioned vehicle weight target setting method and automotive component weight target setting method. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0038] The transmission module 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission module 106 includes a network interface controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission module 106 may be a radio frequency (RF) module, used for wireless communication with the Internet.

[0039] The display can be, for example, a touchscreen liquid crystal display (LCD) that allows the user to interact with the user interface of the computer terminal 10.

[0040] It should be noted here that, in some optional embodiments, the above... Figure 1 The computer terminal shown may include hardware elements (including circuitry), software elements (including computer code stored on a computer-readable medium), or a combination of both hardware and software elements. It should be noted that... Figure 1This is only one instance of a specific particular instance, and is intended to illustrate the types of components that may exist in the aforementioned computer terminal.

[0041] Under the above operating environment, the embodiments of this application provide a method for setting a target weight for a whole vehicle and an embodiment for setting a target weight for automotive components. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Also, although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than that shown here.

[0042] Figure 2 This is a flowchart of a method for setting a vehicle weight target according to an embodiment of this application, such as... Figure 2 As shown, the method includes the following steps:

[0043] Step S202: Obtain the weight characteristic data of competing vehicles corresponding to the developed model.

[0044] Step S204: Construct a machine learning model based on weight feature data, wherein the machine learning model is used to predict the overall weight of the developed model or competing models.

[0045] Step S206: Determine the lightweighting rate of the competitor's vehicle based on the prediction results of the machine learning model, and determine the lightweighting rate boundary value of the competitor's vehicle under different industry levels based on the lightweighting rate. The prediction results are used to represent the predicted weight of the competitor's vehicle, and the lightweighting rate is used to reflect the lightweighting level of the competitor's vehicle.

[0046] Step S208: Determine the target vehicle weight for the developed model based on the lightweighting rate boundary value.

[0047] Through steps S202 to S208 above, the goal of setting differentiated vehicle weight targets based on product positioning and grading is achieved. This improves the accuracy, flexibility, and iteration efficiency of weight target setting, thereby solving the technical problem that related technologies' vehicle weight target setting methods can only output a single target value that is the industry average or optimal, cannot be dynamically set according to vehicle positioning and grading, and rely on static fitting models, making it difficult to efficiently respond to competitor data iterations, resulting in insufficient accuracy in weight target value setting. A detailed explanation follows.

[0048] In step S202 above, the weight characteristic data of the competing vehicles corresponding to the development model (the model to be developed) are obtained, including: obtaining the vehicle weight data, original features and derived features of the competing vehicles according to the preset screening principles, wherein the original features are used to represent the basic structural parameters of the competing vehicles, and the derived features are used to represent the spatial features and energy configuration features of the competing vehicles; and determining the weight characteristic data based on the vehicle weight data, original features and derived features.

[0049] In this embodiment, the weight characteristic data of the competitor vehicle includes the overall vehicle weight and weight-related characteristic data. The weight-related characteristic data includes original features and derived features. The original features include basic structural parameters such as length, width, height, wheelbase, track width, and battery capacity, while the derived features include spatial configuration features such as the overall vehicle footprint, volume, and battery energy density.

[0050] Furthermore, the calculation method for the vehicle footprint in the derived features is: vehicle footprint = length × width; the vehicle volume in the derived features is difficult to calculate directly due to the influence of the outer shape, and its estimation method is: vehicle volume = length × width × height; the calculation method for the battery energy density in the derived features is: battery energy density = battery energy / battery weight, where the battery energy and battery weight are the total energy and total weight of the vehicle's power battery pack.

[0051] It should be noted that each feature in the weight-related feature data is an independent variable, and the total vehicle weight is a dependent variable, used to construct the predictive relationship of the machine learning model.

[0052] Specifically, the preset filtering principles for competitor vehicle data are as follows:

[0053] 1) The competing vehicles were launched within the last five years, preferably within the last three years;

[0054] 2) The difference between the wheelbase of the competing vehicle and the wheelbase of the developed vehicle is within ±100mm, preferably within ±50mm;

[0055] 3) Preferably, the driving mode, overall performance and configuration of the competing vehicles should be as similar as possible to the level of the developed model; in particular, the overall performance includes power performance, economic performance, safety performance and environmental performance, etc.; in particular, economic performance includes the vehicle's energy consumption, range and energy recovery efficiency, etc.; power performance includes power responsiveness, climbing performance and acceleration performance, etc.; safety performance includes passive safety, active safety and braking safety, etc.; environmental performance includes pollutant emissions and energy efficiency, etc.

[0056] 4) If the vehicle being developed is a gasoline-powered vehicle or a hybrid vehicle, preferably, the engine displacement of the competing vehicle should be the same as that of the vehicle being developed;

[0057] 5) If the vehicle being developed is a new energy vehicle, preferably, the battery type of the competing vehicle should be consistent with that of the vehicle being developed.

[0058] In particular, the selection of original and derived features is configurable. The subset of features to be used in modeling can be dynamically selected and the number of features used can be determined based on the type of vehicle being developed (such as fuel vehicle or new energy vehicle), the target level (whole vehicle, system or component) and data availability, so as to achieve a balance between model accuracy and computational efficiency.

[0059] Optionally, after obtaining the weight feature data of competing vehicles corresponding to the development model, the process also includes filtering and handling missing data. The specific implementation method is as follows: determine interactive feature items based on the original features, where the interactive feature items are the product or ratio between the original features; add the interactive feature items to the weight feature data, and filter the weight feature data according to preset requirements to obtain the target weight feature data; and standardize the target weight feature data.

[0060] In this embodiment, missing data is handled by using the mean method, interpolation method, or trend method to complete the missing values ​​in the weight data. Then, a clustering algorithm is used to identify and remove noise from the completed vehicle weight data to improve data quality, ensure data integrity, and enhance the reliability of subsequent model analysis. The specific analysis is as follows:

[0061] 1. Mean Method: For features with missing values ​​in weight data, the mean of all non-missing values ​​of this feature is used to fill in the missing values. This method is simple, intuitive, and computationally inexpensive, and is suitable for situations where the proportion of missing data is small and the data approximately follows a normal distribution.

[0062] Specifically, first, the mean of the non-missing values ​​in the weight feature is calculated, and then all missing values ​​are replaced with this mean. The expression involved in this mean-based method is as follows:

[0063]

[0064] In the formula, This represents the average value of the weight feature sequence (excluding missing values). Indicates the length of the feature sequence. This represents the feature value in the feature sequence.

[0065] 2. Interpolation method: Construct relationships based on non-missing feature data, and use functions to infer missing values. Common interpolation methods include linear interpolation, polynomial interpolation, and spline interpolation.

[0066] To simplify calculations, this application uses a linear interpolation method to fill in missing data, the specific expression of which is as follows:

[0067]

[0068] In the formula, This indicates missing values ​​in the weight feature sequence. and This represents the numerical values ​​of the feature sequences adjacent to the missing value.

[0069] 3. Trend Method: This method utilizes the statistical relationships or trends between other relevant features in the weighted feature data and the missing features to be filled in the missing data to build a predictive model. The model is then used to estimate and fill in the missing values. This method is suitable for situations where there is a strong correlation between features and a large number of missing values.

[0070] 4. Clustering algorithms: such as K-means clustering algorithm, as an effective unsupervised method to identify and remove noise in data. The working principle is to detect anomalies based on distance, that is, normal sample data will be closely clustered around the cluster center, while noisy sample data will be far away from the center of its cluster.

[0071] By calculating the distance from each sample (such as a complete set of competitor vehicle weight feature data) to the center of a cluster (such as a set of competitor vehicle samples with similar overall vehicle structure and configuration features in the feature space), and setting a threshold based on the statistical distribution, excessively deviating noise in the weight feature data can be effectively identified and removed. Specific steps include:

[0072] 1) Predetermine the number of clusters. The SSE (Sum of Squared Errors within Clusters) formula is as follows:

[0073]

[0074] In the formula, x represents a standardized weight feature vector. Let K represent the center of the k-th cluster. Run the K-means algorithm, using the selected K to cluster the standardized data, and obtain the center of each cluster. and the cluster label for each sample .

[0075] 2) Calculate the distance from the sample to the cluster center. The specific expression is as follows:

[0076]

[0077] In the formula, This represents the i-th weight feature vector. Indicates sample The cluster center to which it belongs Indicates sample To the cluster center European distance, Represents the dimension of the feature vector. Indicates sample The value that can be taken on feature dimension j, Indicates cluster center The value taken in feature dimension j.

[0078] 3) Set a noise threshold and remove noise: Normal samples are closer to the cluster center. For each cluster, if the sample distance is greater than the threshold, it is labeled as noise. Remove all samples labeled as noise from the dataset to obtain the processed weight feature data.

[0079] In particular, to improve the model's fitting ability and prediction performance, interactive feature terms can be added to the weight feature data. These interactive feature terms are the product or ratio between the original features, such as length × height, axis distance × width, length / axis distance, etc.

[0080] In particular, in order to comprehensively consider the amount of data in the weight feature database and the model's fitting ability, the acquired weight feature data can be filtered according to actual needs to determine an appropriate amount of target feature weight data.

[0081] Finally, the selected target feature weight data is standardized to eliminate the influence of units and feature range.

[0082] In step S204 above, constructing a machine learning model based on weight feature data includes: dividing the weight feature data into a training set and a test set according to a preset ratio; training multiple initial models using the training set, and determining the evaluation metrics of the initial models on the test set after training, wherein the evaluation metrics include at least one of the following: mean absolute error, root mean square error, and coefficient of determination, the coefficient of determination being used to reflect the explanatory power of the initial model for changes in the overall vehicle weight; and determining a machine learning model from the initial models based on the evaluation metrics for predicting the overall vehicle weight of the development model or competing models.

[0083] In the embodiments of this application, the initial model includes, but is not limited to, linear regression, ridge regression, lasso regression, random forest, gradient boosting tree, and other models; the preset ratio of the training set to the test set can be 8:2 or 7:3.

[0084] Specifically, during model training, 5-fold cross-validation can be performed on the training set of each model to more objectively and stably measure the model's generalization ability. For example, the weight feature data of competitor vehicles used for training can be randomly divided into 5 mutually exclusive subsets of roughly equal size. In each round of validation, one subset is selected in turn as the validation set, and the other 4 subsets are used as the training set. The model is then built and trained, and its evaluation metrics (such as MAE, RMSE, R²) on the validation set are recorded. A total of 5 rounds of training and evaluation are performed, and the arithmetic mean of the 5 evaluation results is taken as the overall performance index of the model.

[0085] Furthermore, multiple initial models are compared by integrating various evaluation metrics, and the model that performs best on the test set is selected as the final machine learning model.

[0086] Specifically, evaluation metrics are primarily used to assess the prediction accuracy of machine learning models. They measure the difference between model predictions and actual values ​​from different perspectives, including Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Coefficient of Determination (R²). The specific calculation formulas are as follows:

[0087]

[0088] In the formula, This indicates the actual weight of the competitor's vehicle. This represents the predicted total vehicle weight of competing vehicles. This represents the average value of the actual vehicle weight of competing vehicles, and n represents the number of samples in the test set.

[0089] Specifically, the specific judgment methods for the above evaluation indicators are as follows:

[0090] 1) MAE: represents the average absolute error between the predicted value and the actual value, with a value range of [0, +∞). 0 represents a perfect prediction, and the closer the value is to 0, the better the prediction effect.

[0091] 2) RMSE: Represents the square root of the mean of the squared errors between the predicted and actual values. The value ranges from [0, +∞), where 0 represents a perfect prediction and the closer the value is to 0, the better the prediction effect.

[0092] 3) R²: Represents the proportion of variance explained by the model, that is, the goodness of fit between the model's predicted values ​​and the actual values. The theoretical range of values ​​is (-∞, 1], usually between 0 and 1. 1 indicates a perfect fit, and the closer the value is to 1, the better the prediction effect.

[0093] In step S206 above, determining the lightweighting rate of the competitor vehicle based on the prediction results of the machine learning model includes: predicting the weight feature data through the machine learning model to obtain a first predicted value; and determining the lightweighting rate of the competitor vehicle based on the first predicted value and the actual overall weight of the competitor vehicle.

[0094] In this embodiment, the machine learning model, after training and optimization, has the ability to accurately predict the overall weight of a vehicle. Specifically, to quantify the lightweighting level of each competing vehicle relative to the industry average, the lightweighting rate of the competing vehicles can be calculated based on the prediction results of the machine learning model.

[0095] First, the standardized weight characteristic data of all competing vehicles (including length, width, wheelbase, track width, battery weight, battery energy density, vehicle footprint, vehicle volume, etc.) are input into the trained machine learning model, and the first predicted value is output to reflect the predicted weight of the competing vehicles.

[0096] Subsequently, the lightweighting rate of competitor vehicles was calculated based on the following formula:

[0097]

[0098] In the formula, the baseline weight is the predicted weight output by the machine learning model, i.e., the first predicted value.

[0099] In particular, a higher lightweighting rate indicates a higher level of lightweighting of the competing vehicle.

[0100] In the above process, the lightweighting rate obtained by combining data-driven dynamic benchmarks (machine learning predictions) reflects the lightweighting level of competing vehicles at the industry average level. It breaks through the limitations of traditional methods that only compare absolute weight or fixed benchmarks, and transforms static evaluation into a relative optimization evaluation based on multi-dimensional feature correlation and intelligence. This provides a scientific basis for setting differentiated and graded weight targets for subsequent vehicle development.

[0101] Furthermore, the lightweighting rates of competing vehicles are sorted in descending order, and their lightweighting levels are graded according to a preset division ratio defined by actual needs. This includes: classifying competing vehicles into four industry levels—first, second, third, and fourth—based on their lightweighting rates and preset division ratios. The product positioning capability corresponding to the first industry level is greater than that corresponding to the second industry level, the second industry level is greater than that corresponding to the third industry level, and the third industry level is greater than that corresponding to the fourth industry level. The minimum lightweighting rate of competing vehicles in the first industry level is determined, as are the maximum and minimum lightweighting rates in the second, third, and fourth industry levels, respectively, thus obtaining the lightweighting rate boundary values.

[0102] In this embodiment, the vehicle lightweighting level can be classified according to the product competition strategy LACU method. All competing vehicles are sorted from largest to smallest lightweighting rate, and are divided into the following four levels:

[0103] 1) L (Leader) Leadership Position: This refers to the highest industry level, where the vehicle's lightweighting level is at the forefront of the industry. It is defined by dividing the competition into 10% of vehicles, specifically selecting the lightweighting rate values ​​corresponding to the top 10% of vehicles, and choosing the lowest lightweighting rate value as the leader. ;

[0104] 2) A (Among Leader) Leading Position: This refers to the second industry level, where the vehicle's lightweighting level is at the forefront of the industry. This is defined by dividing the competition into 20% of vehicles, specifically taking the lightweighting rate values ​​corresponding to the top 10%~30% of vehicles, and selecting the largest and smallest lightweighting rate values ​​as [values ​​to be filled in]. , ;

[0105] 3) C (Competitive) Competitive Position: This refers to the third industry level, where the vehicle's lightweighting level is competitive within the industry. It is defined by 30% of competing vehicles, specifically taking the lightweighting rate values ​​corresponding to the top 30%~60% of vehicles by number, and selecting the largest and smallest lightweighting rate values ​​as... , ;

[0106] 4) U (Uncompetitive): This is the fourth industry level, where the vehicle's lightweighting level is not competitive within the industry. It is defined as 40% of competing vehicles, specifically taking several lightweighting rate values ​​corresponding to the top 60%~100% of vehicles by number, and selecting the largest and smallest lightweighting rate values ​​as [values ​​to be filled in]. , .

[0107] In particular, the number of lightweight levels can be flexibly adjusted and the proportion of each level can be adjusted according to product positioning strategy, market competition situation or R&D resource constraints. As long as the sum of all the proportions is 100%, personalized and customized weight targets can be set for different development models.

[0108] In particular, when the number of graded samples calculated according to the preset division ratio is not an integer (e.g., a total of 97 vehicles, 10% corresponds to 9.7 vehicles), the boundary lightweighting rate can be converted by linear interpolation: the lightweighting rate values ​​of two adjacent samples are weighted and averaged according to the decimal part weight to determine the boundary threshold of that grade.

[0109] In the above process, by calculating the lightweight rate of competing vehicles based on machine learning predictions and classifying them according to continuous interpolation, an objective, fair, and high-precision quantitative evaluation of the lightweight level of vehicles was achieved. This provides a data-driven and reproducible scientific basis for setting differentiated and feasible graded weight targets for vehicle development.

[0110] In step S208 above, determining the vehicle weight target of the development model based on the lightweighting rate boundary value includes: acquiring the original weight feature data of the development model; predicting the original weight feature data through a machine learning model to obtain a second predicted value, wherein the second predicted value is used to represent the predicted vehicle weight of the development model; determining the target industry level to which the development model belongs, and determining the vehicle weight target of the development model based on the lightweighting rate boundary value and the second predicted value under the target industry level.

[0111] In this embodiment of the application, based on the lightweighting rate grading system constructed in step S206, and combined with the weight characteristic data of the development vehicle, a precise overall vehicle weight target can be set for the development vehicle. Specifically:

[0112] First, collect the original weight feature data of the development vehicle, including its original features and derived features. In particular, the weight feature data of the development vehicle should have the same dimensions as the weight feature data of competing vehicles, and the input order of the feature data should be the same as the input order of the model training to ensure the uniformity of the input feature space.

[0113] Secondly, based on the selected machine learning model, the standardized weight characteristic data of the development vehicle are predicted, and a second predicted value is output. This value represents the expected curb weight of the vehicle without an active lightweighting strategy under the current industry technical benchmark, that is, the benchmark weight of the whole vehicle at the industry average level.

[0114] Subsequently, based on the aforementioned lightweighting rate grading system, the lightweighting rate boundary values ​​corresponding to the L (leading position), A (advanced position), C (competitive position), and U (non-competitive position) industry levels of competing vehicles were extracted, namely... (L-level lightweighting rate lower limit threshold) , (Upper and lower limits of Class A lightweighting rate) , (Upper and lower limits of Class C lightweight ratio) and , (Upper and lower limits of U-class lightweighting rate).

[0115] Finally, based on the target industry level of the developed vehicle model, and combining the corresponding lightweighting rate boundary value and baseline weight (second predicted value), the target vehicle weight value for the corresponding level is calculated. The L and A classes are used as examples for illustration:

[0116]

[0117]

[0118]

[0119] In the formula, This represents the baseline weight of the vehicle model under development, i.e., the second predicted value; This indicates the target vehicle weight when the lightweighting level of the developed model is at the industry-leading level (Level L); , This indicates the target range for vehicle weight when the lightweighting level of the developed model is in a leading position (Class A), including the minimum and maximum vehicle weight.

[0120] In the above process, by combining the industry benchmark weight predicted by the machine learning model with the configurable lightweighting rate grading boundary, the intelligent, graded and flexible vehicle weight target setting based on the enterprise's product positioning was realized for the first time, which significantly improved the scientific nature, flexibility and engineering feasibility of lightweight design.

[0121] In particular, to address the challenges of data timeliness brought about by the rapid iteration of automotive technologies, the dense launch of competing models, and the continuous evolution of lightweighting technologies, this application also constructs a dynamic update mechanism for machine learning models to ensure that the accuracy of weight prediction results and their engineering guidance value remain in sync with the latest industry standards. Figure 3 As shown, the system periodically (e.g., every quarter or after each new model release) automatically or semi-automatically executes the following processes:

[0122] 1. Data collection and filtering: Obtain data on newly launched competitor vehicles from public databases, industry reports, and internal R&D platforms within the past 1-3 years, and perform intelligent filtering and cleaning based on preset filtering principles;

[0123] 2. Feature Alignment and Standardization: Align the new data with the historical training set in terms of feature dimensions, standardize the processing flow, and ensure consistency of the input space;

[0124] 3. Model retraining and validation: Merge the updated dataset with the original training set, and retrain and select the optimal model using the same cross-validation and evaluation mechanism (such as 5-fold cross-validation, MAE / RMSE / R² metrics);

[0125] 4. Version Management and Deployment: Mark the new model with a version number (e.g., V2.1.0) and compare the performance improvement of the old model on the test set (e.g., RMSE reduction ≥3% or R² improvement ≥1%). Only trigger the model to go online and update when the performance is significantly improved, and notify the R&D system to switch the prediction interface synchronously.

[0126] This mechanism effectively solves the problems of model obsolescence and increased prediction bias caused by the rapid iteration of competitors, enabling the weight target setting system constructed in this application to have continuous self-evolution capabilities. It is not only applicable to the development of current vehicle models, but also has long-term applicability and technological foresight for product planning over the next few years, significantly enhancing the practical value of the system and the extensibility of intellectual property protection.

[0127] In this embodiment, by fusing multidimensional original and derived features (including battery energy, volume, energy density, etc.), and combining a machine learning model to output a dynamic industry benchmark weight, a four-level configurable hierarchical boundary (L / A / C / U) is constructed based on competitor lightweighting rate ranking and linear interpolation, achieving accurate mapping of different weight target ranges corresponding to different product positioning. Simultaneously, a dynamic model update mechanism is introduced, enabling data-driven self-evolution and ensuring the prediction system maintains high accuracy as industry technology rapidly iterates. This method is not only uniformly applicable to both fuel-powered and new energy vehicles but can also be seamlessly extended to subsystem levels such as chassis and suspension, significantly improving the flexibility, feasibility, and long-term effectiveness of vehicle lightweighting design, demonstrating significant technological advancement and industrial application value.

[0128] Figure 4 This is a flowchart of a method for setting a target weight for automotive components according to an embodiment of this application, such as... Figure 4 As shown, the method includes the following steps:

[0129] Step S402: Obtain the weight characteristic data of the automotive components of the competing vehicles corresponding to the developed vehicle model, wherein the automotive components include at least one of the following: automotive system, subsystem and parts.

[0130] Step S404: Construct a machine learning model based on weight feature data, wherein the machine learning model is used to predict the weight of automotive components of the developing model or competing models.

[0131] Step S406: Determine the lightweighting rate of the automotive components of the competitor's vehicle based on the prediction results of the machine learning model, and determine the lightweighting rate boundary values ​​of the automotive components of the competitor's vehicle under different industry levels based on the lightweighting rate. The prediction results are used to represent the weight prediction results of the automotive components of the competitor's vehicle, and the lightweighting rate is used to reflect the lightweighting level of the automotive components of the competitor's vehicle.

[0132] Step S408: Determine the target weight of automotive components for the developed vehicle model based on the lightweighting rate boundary value.

[0133] Through steps S402 to S408, intelligent, hierarchical, and configurable weight targets are set at the vehicle system, subsystem, and component levels, enabling lightweight design to extend from the whole vehicle to key components and achieving a unified technical effect of full-chain collaborative optimization and engineering feasibility.

[0134] In this embodiment, since the weight of automotive components is highly correlated with performance, the weight characteristic data includes not only original and derived characteristics directly related to the weight of automotive components, but also performance characteristics. Taking the chassis system as an example, original characteristics include vehicle curb weight, wheelbase, track width, unsprung mass, battery weight, etc.; derived characteristics include vehicle footprint area (wheelbase × track width), floor area, etc.; performance characteristics include steering torque, load-bearing capacity, durability indicators, etc.

[0135] It should be noted that the method for setting the weight target of automotive components is similar to the aforementioned method for setting the weight target of the entire vehicle in terms of core technical means. Both are based on a unified technical framework of predicting the baseline weight using machine learning, ranking and classifying the lightweighting rate, calculating boundary values ​​through interpolation, and generating the target range through positioning mapping. The only difference lies in the targeted adjustments to the input feature dimensions (such as adding performance features and subsystem parameters) and the component-level adaptation. The model construction logic, classification mechanism, dynamic update process, and target calculation formula remain consistent. Therefore, to avoid redundancy, the modeling methods, lightweighting rate definitions, boundary value determination, and target calculation methods involved in the embodiments of this application will not be elaborated upon. The technical principles and beneficial effects can be directly referred to in the aforementioned section on setting the weight target of the entire vehicle.

[0136] According to an embodiment of this application, a vehicle weight target setting device is provided. It should be noted that the vehicle weight target setting device of this application embodiment can be used to execute the vehicle weight target setting method provided in the embodiment of this application. The vehicle weight target setting device provided in the embodiment of this application will be described below.

[0137] Figure 5 This is a structural diagram of a vehicle weight target setting device provided according to an embodiment of this application. Figure 5 As shown, the device includes:

[0138] The first acquisition module 50 is used to acquire the weight characteristic data of competing vehicles corresponding to the developed vehicle model;

[0139] The first building module 52 is used to build a machine learning model based on weight feature data, wherein the machine learning model is used to predict the overall weight of the developed vehicle model or competing vehicles.

[0140] The first determining module 54 is used to determine the lightweighting rate of the competitor vehicle based on the prediction results of the machine learning model, and to determine the lightweighting rate boundary value of the competitor vehicle under different industry levels based on the lightweighting rate. The prediction results are used to represent the predicted weight of the competitor vehicle, and the lightweighting rate is used to reflect the lightweighting level of the competitor vehicle.

[0141] The second determining module 56 is used to determine the overall vehicle weight target of the development model based on the lightweighting rate boundary value.

[0142] Through the first acquisition module, first construction module, first determination module, and second determination module in the above-mentioned vehicle weight target setting device, the purpose of setting differentiated vehicle weight targets according to product positioning levels is achieved. This improves the accuracy, flexibility, and iteration efficiency of weight target setting, thereby solving the technical problem that the vehicle weight target setting methods in related technologies can only output a single target value that is the industry average or the best, and cannot be dynamically set according to vehicle positioning levels. Furthermore, they rely on static fitting models and are difficult to respond efficiently to competitor data iterations, resulting in insufficient accuracy in setting weight target values.

[0143] In the vehicle weight target setting device provided in this application embodiment, the first acquisition module is further used to acquire the vehicle weight data, original features and derived features of the competitor vehicle according to a preset screening principle. The original features are used to represent the basic structural parameters of the competitor vehicle, and the derived features are used to represent the spatial features and energy configuration features of the competitor vehicle. The weight feature data is determined based on the vehicle weight data, original features and derived features.

[0144] In the vehicle weight target setting device provided in this application embodiment, the first acquisition module is further used to determine interactive feature items based on the original features, wherein the interactive feature items are the product or ratio between the original features; add the interactive feature items to the weight feature data, and filter the weight feature data according to preset requirements to obtain target weight feature data; and perform standardization processing on the target weight feature data.

[0145] In the vehicle weight target setting device provided in this application embodiment, the first construction module is further configured to divide the weight feature data into a training set and a test set according to a preset ratio; train multiple initial models through the training set, and determine the evaluation index of the initial model on the test set after training is completed, wherein the evaluation index includes at least one of the following: mean absolute error, root mean square error and coefficient of determination, the coefficient of determination is used to reflect the explanatory power of the initial model for changes in vehicle weight; and determine a machine learning model from the initial model based on the evaluation index to predict the vehicle weight of the development model or competitor vehicles.

[0146] In the vehicle weight target setting device provided in this application embodiment, the first determining module is further configured to divide the weight feature data into a training set and a test set according to a preset ratio; train multiple initial models through the training set, and determine the evaluation index of the initial model on the test set after training is completed, wherein the evaluation index includes at least one of the following: mean absolute error, root mean square error and coefficient of determination, the coefficient of determination is used to reflect the explanatory power of the initial model for changes in vehicle weight; and determine a machine learning model from the initial model based on the evaluation index to predict the vehicle weight of the development model or competitor vehicles.

[0147] In the vehicle weight target setting device provided in this application embodiment, the first determining module is further configured to classify competing vehicles into a first industry level, a second industry level, a third industry level, and a fourth industry level based on the lightweighting rate and a preset division ratio, wherein the product positioning capability corresponding to the first industry level is greater than the product positioning capability corresponding to the second industry level, the product positioning capability corresponding to the second industry level is greater than the product positioning capability corresponding to the third industry level, and the product positioning capability corresponding to the third industry level is greater than the product positioning capability corresponding to the fourth industry level; determine the minimum lightweighting rate of the competing vehicles in the first industry level, and determine the maximum and minimum lightweighting rates of the competing vehicles in the second, third, and fourth industry levels respectively, to obtain lightweighting rate boundary values.

[0148] In the vehicle weight target setting device provided in this application embodiment, the second determining module is further used to obtain the original weight feature data of the development model; predict the original weight feature data through a machine learning model to obtain a second predicted value, wherein the second predicted value is used to represent the vehicle weight prediction result of the development model; determine the target industry level to which the development model belongs, and determine the vehicle weight target of the development model based on the lightweighting rate boundary value under the target industry level and the second predicted value.

[0149] According to an embodiment of this application, a vehicle component weight target setting device is also provided. It should be noted that the vehicle component weight target setting device of this application embodiment can be used to execute the vehicle component weight target setting method provided in the embodiment of this application. The vehicle component weight target setting device provided in the embodiment of this application will be described below.

[0150] Figure 6 This is a structural diagram of an automotive component weight target setting device provided according to an embodiment of this application. Figure 6 As shown, the device includes:

[0151] The second acquisition module 60 is used to acquire the weight characteristic data of automotive components of competing vehicles corresponding to the development model, wherein the automotive components include at least one of the following: automotive system, subsystem and parts.

[0152] The second building module 62 is used to build a machine learning model based on weight feature data, wherein the machine learning model is used to predict the weight of automotive components of the developing model or competing models.

[0153] The third determining module 64 is used to determine the lightweighting rate of the automotive components of the competitor's vehicle based on the prediction results of the machine learning model, and to determine the lightweighting rate boundary value of the automotive components of the competitor's vehicle under different industry levels based on the lightweighting rate. The prediction results are used to represent the weight prediction results of the automotive components of the competitor's vehicle, and the lightweighting rate is used to reflect the lightweighting level of the automotive components of the competitor's vehicle.

[0154] The fourth determination module 66 is used to determine the target weight of automotive components for the developed vehicle model based on the lightweighting rate boundary value.

[0155] Through the second acquisition module, second construction module, third determination module and fourth determination module in the above-mentioned automotive component weight target setting device, intelligent, hierarchical and configurable weight target setting is realized at the automotive system, subsystem and component levels, enabling lightweight design to extend from the whole vehicle to key components, and achieving a unified technical effect of full-link collaborative optimization and engineering feasibility.

[0156] This application also provides an electronic device, including: a memory and a processor, wherein the memory is used to store program instructions; the processor is connected to the memory and is used to execute the above-described method for setting the vehicle weight target and the method for setting the weight target of automotive components.

[0157] It should be noted that the aforementioned electronic equipment is used to perform Figure 2 The method for setting the target vehicle weight shown or Figure 4 The method for setting the target weight of automotive components shown above also applies to this electronic device, and will not be repeated here.

[0158] This application embodiment also provides a non-volatile storage medium, which includes a stored computer program, wherein the device containing the non-volatile storage medium executes the above-described vehicle weight target setting method and automotive component weight target setting method by running the computer program.

[0159] It should be noted that the aforementioned non-volatile storage media is used for execution. Figure 2 The method for setting the target vehicle weight shown or Figure 4 The method for setting the target weight of automotive components shown above also applies to this non-volatile storage medium, and will not be repeated here.

[0160] This application also provides a computer program product, including computer instructions, which, when executed by a processor, implement the above-described method for setting the overall vehicle weight target and the method for setting the weight target of automotive components.

[0161] It should be noted that the above-mentioned computer program product is used to execute Figure 2 The method for setting the target vehicle weight shown or Figure 4 The method for setting the target weight of automotive components shown above also applies to this computer program product, and will not be repeated here.

[0162] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0163] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0164] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0165] 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 units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0166] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0167] If the integrated unit is implemented as a software functional unit 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 all 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 a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0168] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for setting a target vehicle weight, characterized in that, include: Obtain weight characteristic data of competing vehicles corresponding to the developed model; A machine learning model is constructed based on the weight feature data, wherein the machine learning model is used to predict the overall vehicle weight of the developed vehicle model or the competing vehicle. The lightweighting rate of the competitor vehicle is determined based on the prediction results of the machine learning model, and the lightweighting rate boundary value of the competitor vehicle under different industry levels is determined based on the lightweighting rate. The prediction results are used to represent the predicted weight of the competitor vehicle, and the lightweighting rate is used to reflect the lightweighting level of the competitor vehicle. The target vehicle weight for the developed model is determined based on the aforementioned lightweighting rate boundary value.

2. The method according to claim 1, characterized in that, Obtain weight characteristic data of competing vehicles corresponding to the developed model, including: The vehicle weight data, original features, and derived features of the competing vehicles are obtained according to a preset screening principle. The original features are used to represent the basic structural parameters of the competing vehicles, and the derived features are used to represent the spatial features and energy configuration features of the competing vehicles. The weight feature data is determined based on the vehicle weight data, the original features, and the derived features.

3. The method according to claim 2, characterized in that, The method further includes: Interactive feature terms are determined based on the original features, wherein the interactive feature terms are the product or ratio of the original features; The interactive feature items are added to the weight feature data, and the weight feature data is filtered according to preset requirements to obtain the target weight feature data. The target weight feature data is standardized.

4. The method according to claim 1, characterized in that, Constructing a machine learning model based on the weight feature data includes: The weight feature data is divided into a training set and a test set according to a preset ratio; Multiple initial models are trained using the training set, and an evaluation metric for the initial model on the test set is determined after training. The evaluation metric includes at least one of the following: mean absolute error, root mean square error, and coefficient of determination. The coefficient of determination is used to reflect the explanatory power of the initial model for changes in vehicle weight. Based on the evaluation metrics, a machine learning model is determined from the initial model to predict the overall vehicle weight of the development vehicle or the competitor vehicle.

5. The method according to claim 1, characterized in that, Determining the lightweight ratio of the competitor vehicle based on the prediction results of the machine learning model includes: The weight feature data is predicted using the machine learning model to obtain a first predicted value; The lightweighting rate of the competitor vehicle is determined based on the first predicted value and the actual overall weight of the competitor vehicle.

6. The method according to claim 5, characterized in that, Determining the lightweighting rate boundary values ​​of the competing vehicles at different industry levels based on the aforementioned lightweighting rate includes: Based on the lightweighting rate and the preset division ratio, the competing vehicles are divided into a first industry level, a second industry level, a third industry level, and a fourth industry level. The product positioning capability corresponding to the first industry level is greater than that corresponding to the second industry level, the product positioning capability corresponding to the second industry level is greater than that corresponding to the third industry level, and the product positioning capability corresponding to the third industry level is greater than that corresponding to the fourth industry level. The minimum lightweighting rate of the competing vehicle in the first industry level is determined, and the maximum and minimum lightweighting rates of the competing vehicle in the second, third, and fourth industry levels are determined respectively, to obtain the lightweighting rate boundary value.

7. The method according to claim 5, characterized in that, Determining the overall vehicle weight target of the development model based on the aforementioned lightweighting rate boundary value includes: Obtain the original weight characteristic data of the developed vehicle model; The machine learning model is used to predict the original weight feature data to obtain a second predicted value, wherein the second predicted value is used to represent the predicted weight of the developed vehicle model. The target industry level of the developed vehicle model is determined, and the overall vehicle weight target of the developed vehicle model is determined based on the lightweighting rate boundary value under the target industry level and the second predicted value.

8. A method for setting a target weight for automotive components, characterized in that, include: Obtain weight characteristic data of automotive components of competing vehicles corresponding to the developed vehicle model, wherein the automotive components include at least one of the following: automotive systems, subsystems, and parts; A machine learning model is constructed based on the weight feature data, wherein the machine learning model is used to predict the weight of automotive components of the developed vehicle model or the competing vehicle. The lightweighting rate of the automotive components of the competitor vehicle is determined based on the prediction results of the machine learning model, and the lightweighting rate boundary value of the automotive components of the competitor vehicle is determined based on the lightweighting rate under different industry levels. The prediction results are used to represent the weight prediction results of the automotive components of the competitor vehicle, and the lightweighting rate is used to reflect the lightweighting level of the automotive components of the competitor vehicle. The target weight of automotive components for the developed vehicle model is determined based on the aforementioned lightweighting rate boundary value.

9. A vehicle weight target setting device, characterized in that, include: The first acquisition module is used to acquire the weight characteristic data of competing vehicles corresponding to the developed vehicle model; The first construction module is used to construct a machine learning model based on the weight feature data, wherein the machine learning model is used to predict the overall vehicle weight of the development model or the competitor's vehicle. The first determining module is used to determine the lightweighting rate of the competitor vehicle based on the prediction results of the machine learning model, and to determine the lightweighting rate boundary value of the competitor vehicle under different industry levels based on the lightweighting rate, wherein the prediction results are used to represent the predicted weight of the competitor vehicle, and the lightweighting rate is used to reflect the lightweighting level of the competitor vehicle. The second determining module is used to determine the target vehicle weight of the development model based on the lightweighting rate boundary value.

10. An electronic device, characterized in that, include: A memory and a processor, wherein the memory is used to store program instructions; The processor, connected to the memory, is used to execute the vehicle weight target setting method according to any one of claims 1 to 7 or the automotive component weight target setting method according to claim 8.

11. A non-volatile storage medium, characterized in that, The non-volatile storage medium includes a stored computer program, wherein the device containing the non-volatile storage medium executes the vehicle weight target setting method according to any one of claims 1 to 7 or the automotive component weight target setting method according to claim 8 by running the computer program.

12. A computer program product comprising computer instructions, characterized in that, When the computer instructions are executed by the processor, they implement the vehicle weight target setting method according to any one of claims 1 to 7 or the automotive component weight target setting method according to claim 8.

Images