Vehicle front-end cooling module design method and system based on air intake amount prediction

Abstract

The application discloses a kind of based on air intake volume prediction vehicle front end cooling module design method and system, belong to automobile cooling system air volume matching and application technical field, including: obtain the key factor data of current vehicle front end cooling module design scheme under the influence radiator air intake volume;For key factor data, respectively using convolutional neural network and graph convolutional neural network extract the non-space feature and spatial feature of the cooling module currently designed;The extracted features are input into the air intake volume prediction model, and the predicted air intake volume under the current design scheme is output;Constantly adjust key factor data to generate different design schemes, and predict the air intake volume under different design schemes, according to the predicted air intake volume adjustment and screening, obtain the optimal cooling module design scheme.The application adopts the radiator air intake volume prediction mode based on neural network, simultaneously with feedback mechanism, improve the cooling module selection process, speed up the design iteration of front grille, improve research and development efficiency.

Description

Technical Field

[0001] This invention belongs to the field of automotive cooling system airflow matching and application technology, and particularly relates to a design method and system for a vehicle front-end cooling module based on intake airflow prediction. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] In the early stages of automotive vehicle research and development, to ensure the proper functioning of the vehicle's thermal management system, it is necessary to match radiators of appropriate size and to pair them with suitable fans. Furthermore, to ensure sufficient airflow to the front radiators, the design of the front grille must also be considered simultaneously. Therefore, in the systematic design of the vehicle's front cooling module (also known as the heat dissipation module, cooling module, etc.), it is necessary to perform extensive calculations of radiator airflow for different radiator and fan arrangements, combined with the continuously iterating front grille design, to ensure that thermal management performance requirements are met.

[0004] In the early stages of cooling module design, simulation is currently the primary method used for development. This involves creating a three-dimensional geometric model of the entire vehicle using computational fluid dynamics (CFD) software, including key components such as the cooling module (e.g., radiator), fan, and front grille. Computer simulations are then performed on the heat dissipation of each component in the model to calculate the radiator's airflow. This process involves multiple simulation-adjustment-resimulation cycles to complete the cooling module design. However, this method requires a complete geometric model, is complex, time-consuming, and labor-intensive. It cannot handle the large-scale airflow calculations required for various radiator and fan combinations. Furthermore, in the early stages of styling design, frequent changes to the front grille design cannot quickly reflect the impact on radiator airflow, hindering the efficient development of effective design solutions. Summary of the Invention

[0005] To address the shortcomings of the existing technologies, this invention provides a design method and system for a vehicle front-end cooling module based on air intake volume prediction. It adopts a radiator air intake volume prediction method based on neural networks, and is combined with a feedback mechanism to solve the problems of long cycle, large workload, and inability to quickly evaluate and provide feedback in traditional radiator air intake volume simulation optimization design methods. This improves the cooling module selection process, accelerates the design iteration of the front grille, and increases R&D efficiency.

[0006] In a first aspect, the present invention provides a design method for a vehicle front-end cooling module based on air intake volume prediction.

[0007] A design method for a vehicle front-end cooling module based on intake air volume prediction, comprising: Obtain data on key factors affecting radiator airflow under the current vehicle front cooling module design scheme; these key factors include vehicle speed, front grille styling data, radiator PQ performance data, fan PQ performance data, and overall vehicle layout status data. For key factor data, convolutional neural networks and graph convolutional neural networks are used to extract the non-spatial and spatial features of the currently designed cooling module, respectively. The extracted features are input into the air intake volume prediction model, and the predicted air intake volume under the current design scheme is output. The key factor data is continuously adjusted to generate different design schemes, and the air intake volume under different design schemes is predicted. Based on the predicted air intake volume, the optimal cooling module design scheme is obtained through adjustment and screening.

[0008] A further technical solution utilizes convolutional neural networks to extract non-spatial features of the cooling module, including: The vehicle speed, radiator PQ performance data, and fan PQ performance data are all one-dimensional parameter data, which are directly used as the physical characteristics of the cooling module; among them, PQ performance data refers to the slope of the PQ characteristic curve, where P represents static pressure and Q represents airflow. The front grille styling data and the overall vehicle layout data are both three-dimensional point cloud data. Convolutional neural networks are used to perform dimensionality reduction processing on the three-dimensional point cloud data to extract the opening ratio features of the front grille styling and the geometric features of various components in the overall vehicle layout. Among them, the components are components in the front cooling module of the vehicle, including the radiator, fan and front grille.

[0009] A further technical solution utilizes graph neural networks to extract spatial features of the cooling module, including: The components are treated as nodes, and adjacent nodes are connected by edges. Node attributes and edge attributes are set to model the spatial relationship diagram of the cooling module. The node attributes include geometric features and physical features, and the edge attributes represent the spatial attributes between nodes, including distance, distance weight, relative angle, and occlusion state. Graph neural networks are used to extract features from spatial relationship graphs. Each node, after fusing its own geometric and physical attributes, passes messages through edge connections to obtain spatial features that fuse the spatial attributes between nodes.

[0010] A further technical solution involves inputting the extracted features into the air intake volume prediction model. After feature fusion, the predicted air intake volume under the current design scheme is output, including: By integrating several extracted non-spatial features using a multilayer perceptron, the global features of the entire cooling module are obtained. By using global pooling, the spatial features of all nodes are pooled to obtain the spatial features of the entire cooling module. The global features and spatial features are concatenated, and the predicted air intake is output through a fully connected layer.

[0011] A further technical solution is that the distance is calculated as follows: ; in, , They represent the first and The three-dimensional centroid coordinates of a node or component; The distance weight is calculated as follows: ; in, This is an adjustable parameter, representing the distance attenuation factor. This represents the correlation coefficient between types.

[0012] A further technical solution involves constructing a multi-objective optimization function by taking the target air volume as the objective and various key factor data as variables when calculating the optimal cooling module design. A genetic algorithm is used to solve the multi-objective optimization function. During the solution process, the air intake volume prediction model function is used as the fitness function to continuously iterate and optimize the data of various key factors until the target air output volume is reached, thereby obtaining the optimal key factor data and generating the optimal cooling module design scheme.

[0013] Secondly, the present invention provides a vehicle front-end cooling module design system based on air intake volume prediction.

[0014] A vehicle front-end cooling module design system based on intake air volume prediction, comprising: The data acquisition module is used to acquire data on key factors affecting the radiator air intake under the current vehicle front cooling module design scheme; among which, key factor data includes vehicle speed, front grille styling data, radiator PQ performance data, fan PQ performance data, and overall vehicle layout status data. The feature extraction module is used to extract the non-spatial and spatial features of the currently designed cooling module based on key factor data, using convolutional neural networks and graph convolutional neural networks respectively. The air intake volume prediction module is used to input the extracted features into the air intake volume prediction model and output the predicted air intake volume under the current design scheme. The design optimization module is used to continuously adjust key factor data to generate different design schemes and predict the air intake volume under different design schemes. Based on the predicted air intake volume, the module is adjusted and screened to obtain the optimal cooling module design scheme.

[0015] Thirdly, the present invention also provides an electronic device, comprising: a memory for storing executable instructions; and a processor for implementing the above-described design method for a vehicle front-end cooling module based on air intake volume prediction when executing the executable instructions stored in the memory.

[0016] Fourthly, the present invention also provides a computer-readable storage medium storing executable instructions for causing a processor to execute the executable instructions to implement the above-described vehicle front-end cooling module design method based on air intake volume prediction.

[0017] Fifthly, the present invention also provides a computer program product comprising executable instructions stored in a computer-readable storage medium; wherein, when the processor of an electronic device reads the executable instructions from the computer-readable storage medium and executes the executable instructions, it implements the above-described design method for a vehicle front-end cooling module based on air intake volume prediction.

[0018] The above one or more technical solutions have the following beneficial effects: 1. This invention proposes a design method and system for vehicle front-end cooling modules based on air intake volume prediction. By collecting and analyzing various influencing factors, it uses neural network technology to predict the air intake volume of the radiator, and simultaneously uses a feedback mechanism to optimize the design and performance of the cooling system. This solves the problems of long cycle, large workload, and inability to quickly evaluate and provide feedback in traditional radiator air intake volume simulation optimization design methods, improves the cooling module selection process, accelerates the design iteration of the front grille, and improves the accuracy and efficiency of automotive cooling system design.

[0019] 2. The method proposed in this invention, by fusing non-spatial features (such as vehicle speed, PQ performance, etc.) and spatial features (topological relationships between components), especially the modeling of spatial relationships by graph neural networks, can more accurately capture complex airflow interaction effects. The predictive model uses extracted features to predict air intake volume. This model incorporates a physical constraint loss function during training to ensure that the prediction results conform to physical laws, reducing non-physical prediction errors and further improving the accuracy of the prediction model. Furthermore, when adjusting the design based on the predicted air intake volume, the combination of genetic algorithms and multi-objective optimization functions can simultaneously optimize airflow, cost, wind resistance, and other indicators, outputting a Pareto optimal solution, effectively improving the accuracy and efficiency of automotive cooling system design.

[0020] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0021] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0022] Figure 1 This is an overall flowchart of the vehicle front-end cooling module design method based on air intake volume prediction as described in an embodiment of the present invention. Detailed Implementation

[0023] It should be noted that the following detailed descriptions are exemplary and are intended only to describe specific embodiments and to provide further explanation of the invention, and are not intended to limit the scope of exemplary embodiments of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0024] Example 1 To address the problems of long cycles, heavy workloads, and inability to quickly evaluate and provide feedback in traditional radiator airflow simulation optimization design methods, which consequently affect the design and development efficiency of cooling modules, this embodiment proposes a vehicle front-end cooling module design method based on airflow prediction. First, data on key factors affecting radiator airflow are collected, such as vehicle speed, front grille styling data, radiator data, fan data, and overall vehicle layout data. Considering that vehicle speed information, radiator drag (PQ) pressure drop data, and fan PQ performance data are all one-dimensional parameter data, they can be directly used as input to train the radiator airflow prediction model. However, the front grille styling data and overall vehicle layout data are three-dimensional parameter data, requiring further dimensionality reduction processing to obtain one-dimensional feature data. For example, by using neural networks to process the corresponding 3D point cloud data, feature parameters such as opening ratio, shape, size, engine compartment layout, and the positional relationship between various components and surrounding components in the module can be identified and extracted. The original 3D data is reduced to 1D parameter data. Then, through data cleaning, normalization, and standardization preprocessing operations, it is made suitable for the input requirements of the neural network. After constructing a neural network-based air intake volume prediction model, the model is trained using an existing air volume database. The trained radiator air intake volume model is used to predict the actual situation. A feedback mechanism is established based on the prediction results to provide guidance for the selection of the cooling system and the design of the front grille. This improves the cooling module selection process, accelerates the design iteration of the front grille, and improves R&D and design efficiency.

[0025] The vehicle front-end cooling module design method based on intake air volume prediction proposed in this embodiment is as follows: Figure 1 As shown, the specific steps include: Step S1: Obtain data on key factors affecting radiator airflow under the current vehicle front-end cooling module design scheme.

[0026] In this embodiment, a series of key information is first collected, namely, the key factor data affecting the radiator's air intake volume. This key factor data includes vehicle speed, front grille design data, radiator PQ performance data, fan PQ performance data, and overall vehicle layout data. This data is the basis for predicting the radiator's air intake volume. Among them, PQ performance data refers to the slope of the PQ characteristic curve, where P represents static pressure and Q represents air volume.

[0027] Step S2: For the key factor data, use convolutional neural networks and graph convolutional neural networks to extract the non-spatial and spatial features of the currently designed cooling module.

[0028] Specifically, after acquiring the aforementioned key factor data, the data undergoes preprocessing and feature extraction. One-dimensional parameter data, such as vehicle speed, radiator drag (PQ) pressure drop data, and fan PQ performance data, can be directly used as physical features of the cooling module and input directly into the model. However, three-dimensional data, such as front grille data and overall vehicle layout, requires point cloudification and feature extraction for dimensionality reduction to adapt to the input format of the neural network model. In this embodiment, a convolutional neural network (CNN) is used to identify feature parameters, extracting the opening ratio features of the front grille shape and the geometric features of various components in the vehicle layout (components in the front-end cooling module, including the radiator, fan, and front grille), reducing the three-dimensional data to one-dimensional parameter data, which is then used as input for training the airflow prediction model.

[0029] Preferably, the data is preprocessed through data cleaning, normalization, and standardization to make it suitable for the input requirements of the neural network.

[0030] As one implementation method, after extracting the above-mentioned non-spatial features of the cooling module using a convolutional neural network, considering the limitations of traditional CNN processing of point clouds: it is impossible to explicitly model the spatial topological relationship between the radiator, grille, and fan, and this spatial relationship is also one of the factors affecting the final air intake, this embodiment also introduces a graph neural network (GNN) to extract the spatial features of the cooling module.

[0031] Specifically, firstly, components (such as radiators, fans, front grilles, etc.) are treated as nodes, and adjacent nodes are connected by edges. Node and edge attributes are then defined to model the spatial relationship graph of the cooling module. Node attributes include the geometric and physical features extracted in the previous step, while edge attributes characterize the spatial attributes between nodes, including distance, distance weight, relative angle, and occlusion state. Furthermore, the distance, i.e., the Euclidean distance between two nodes / two components, can be expressed as: ; In the above formula, , They represent the first and The three-dimensional centroid coordinates of a node or component.

[0032] Considering physical characteristics: the greater the distance between two components, the weaker their interaction (such as the effect of airflow), and this weakening follows an exponential decay law. Therefore, this embodiment designs a distance weight, which can be expressed as: ; In the above formula, This is an adjustable parameter, representing the distance attenuation factor. Indicates the type correlation coefficient, such as fan to heatsink. From grille to radiator Other components to the heat sink .

[0033] Additionally, the relative angle is the angle between the fan airflow direction and the heatsink normal; the obstruction mark is {0,1}, where 1 indicates that there is obstruction between the two components, and 0 indicates that there is no obstruction between the two components.

[0034] Secondly, a graph neural network is used to extract features from the spatial relationship graph. After each node integrates its own geometric and physical attributes, it passes messages through edge connections to obtain spatial features that integrate the spatial attributes between nodes.

[0035] Step S3: Input the extracted features into the air intake volume prediction model and output the predicted air intake volume under the current design scheme.

[0036] In this embodiment, a radiator air intake prediction model based on neural networks is first built. A suitable neural network architecture, such as a multilayer perceptron (MLP), convolutional neural network (CNN), or recurrent neural network (RNN), is selected, and the model architecture is defined, including an input layer, a hidden layer, and an output layer. The input layer receives preprocessed data, the hidden layer extracts features through the neural network, and the output layer generates prediction results.

[0037] Secondly, the airflow prediction model constructed above is pre-trained. Specifically, based on the existing vehicle model database and airflow data, or by obtaining data from existing airflow simulation results, feature data is obtained after preprocessing, and a dataset is constructed accordingly. This dataset is divided into a training set and a test set. The prediction model is trained using the training set. During the training process, the model architecture is defined, the loss function and optimizer are selected, and the model parameters are adjusted through an iterative process to minimize the loss function and improve the model's prediction accuracy.

[0038] As one implementation method, a composite loss function is used for iterative training during the training process of the air intake volume prediction model. This composite loss function is a weighted sum of regression loss, physical constraint loss, and uncertainty quantification loss. The regression loss can be expressed as: ; In the above formula, This indicates the actual air intake volume. Indicates the predicted air intake volume. Indicates deviation; Physical constraint losses include energy conservation constraint losses and boundary condition constraint losses, which are expressed as follows: ; ; in, air density, Indicates vehicle speed. This represents the radiator pressure drop, which is inversely proportional to the airflow. The relationship between the two can be described by the fan power and efficiency. This indicates the theoretical maximum air volume; Uncertainty quantification loss can be expressed as: ; in, This represents the predicted average intake air volume. Indicates the prediction confidence level. This is the variance scaling factor.

[0039] Finally, after model training is complete, the model's performance is evaluated using a test set to ensure good generalization ability and avoid overfitting. Preferably, the model can be optimized based on the evaluation results by adjusting model parameters or structure to further improve model performance.

[0040] After the above training and optimization, the trained prediction model is used to predict the air intake volume of the car radiator based on input information such as vehicle speed, front grille data, radiator data, fan data, and overall vehicle layout. Specifically, the extracted features are input into the air intake volume prediction model, and after feature fusion, the predicted air intake volume under the current design scheme is output. This involves: first, using a multilayer perceptron (MLP) to integrate several extracted non-spatial features to obtain the global features of the entire cooling module; second, using global pooling to pool the spatial features of all nodes to obtain the spatial features of the entire cooling module; and finally, concatenating the global features with the spatial features and outputting the predicted air intake volume through a fully connected layer.

[0041] Step S4: Continuously adjust the key factor data to generate different design schemes, predict the air intake volume under different design schemes, adjust and screen according to the predicted air intake volume, and obtain the optimal cooling module design scheme.

[0042] Specifically, based on the predicted air intake volume, the influence of factors such as the corresponding grille opening ratio, styling features, and different front-end radiator combinations on the air volume is analyzed, and this is used as feedback to guide the selection of the cooling system and the design of the front grille.

[0043] As one implementation method, when calculating the optimal cooling module design scheme, a multi-objective optimization function is constructed with the target air volume as the objective (cost, wind resistance, and other indicators can also be introduced according to the actual situation) and multiple key factor data as variables. A genetic algorithm is used to solve the multi-objective optimization function. During the solution process, the air intake prediction model function is used as the fitness function to continuously iterate and optimize multiple key factor data until the target air volume is reached, obtain the optimal key factor data, and generate the optimal cooling module design scheme.

[0044] The above methods not only reduce a lot of simulation calculations in the early stage of selecting radiators and fans for the cooling system, but also provide a basis for the design and optimization of the front grille to meet performance requirements, which helps to improve the rapid update and iteration of the styling design.

[0045] Example 2 This embodiment provides a vehicle front-end cooling module design system based on intake air volume prediction, including: The data acquisition module is used to acquire data on key factors affecting the radiator air intake under the current vehicle front cooling module design scheme; among which, key factor data include vehicle speed, front grille styling data, radiator PQ performance data, fan PQ performance data, and overall vehicle layout status data. The feature extraction module is used to extract the non-spatial and spatial features of the currently designed cooling module based on key factor data, using convolutional neural networks and graph convolutional neural networks respectively. The air intake volume prediction module is used to input the extracted features into the air intake volume prediction model and output the predicted air intake volume under the current design scheme. The design optimization module is used to continuously adjust key factor data to generate different design schemes and predict the air intake volume under different design schemes. Based on the predicted air intake volume, the module is adjusted and screened to obtain the optimal cooling module design scheme.

[0046] Furthermore, non-spatial features of the cooling module are extracted using a convolutional neural network, including: The acquired vehicle speed, radiator PQ performance data, and fan PQ performance data are all one-dimensional parameter data, directly used as physical characteristics of the cooling module; among them, PQ performance data refers to the slope of the PQ characteristic curve, where P represents static pressure and Q represents airflow; the acquired front grille styling data and overall vehicle layout state data are all three-dimensional point cloud data, which are dimensionality-reduced using convolutional neural networks to extract the opening ratio features of the front grille styling and the geometric features of various components in the overall vehicle layout; among them, the components are those in the vehicle's front-end cooling module, including the radiator, fan, and front grille.

[0047] Simultaneously, spatial features of the cooling module are extracted using graph neural networks, including: The components are treated as nodes, and adjacent nodes are connected by edges. Node attributes and edge attributes are set to model the spatial relationship diagram of the cooling module. The node attributes include geometric features and physical features, and the edge attributes represent the spatial attributes between nodes, including distance, distance weight, relative angle, and occlusion state. Graph neural networks are used to extract features from spatial relationship graphs. Each node, after fusing its own geometric and physical attributes, passes messages through edge connections to obtain spatial features that fuse the spatial attributes between nodes.

[0048] Based on the above, the extracted features are input into the air intake prediction model. After feature fusion, the predicted air intake under the current design scheme is output. That is, the multilayer perceptron is used to integrate several extracted non-spatial features to obtain the global features of the entire cooling module; global pooling is used to pool the spatial features of all nodes to obtain the spatial features of the entire cooling module; the global features and spatial features are concatenated and the predicted air intake is output through a fully connected layer.

[0049] The proposed solution in this embodiment utilizes deep learning technology to automatically learn and extract key features affecting radiator airflow, reducing preliminary modeling and simulation work and further improving the efficiency and accuracy of radiator selection and front grille design. Furthermore, the model exhibits excellent generalization ability, adapting to different operating conditions and vehicle models, and has broad application prospects. Through continuous model optimization and iteration, prediction accuracy can be further improved, providing strong support for the design and improvement of automotive cooling systems.

[0050] Example 3 This embodiment provides an electronic device, including: a memory for storing executable instructions; and a processor for executing the executable instructions stored in the memory to implement the method provided in this embodiment.

[0051] Example 4 This embodiment also provides a computer-readable storage medium storing executable instructions, which, when executed by a processor, will cause the processor to execute the method described above in this embodiment.

[0052] Example 5 This embodiment provides a computer program product including executable instructions, which are computer instructions; the executable instructions are stored in a computer-readable storage medium. When the processor of an electronic device reads the executable instructions from the computer-readable storage medium and executes the executable instructions, the electronic device performs the method described in this embodiment.

[0053] The steps and methods involved in Embodiments 2 to 5 above correspond to those in Embodiment 1. For specific implementation details, please refer to the relevant description section of Embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood as including any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and enabling the processor to perform any of the methods in this invention.

[0054] Those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computer devices. Optionally, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by a computer device, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. The present invention is not limited to any particular combination of hardware and software.

[0055] The above description is only a preferred embodiment of the present invention. Although the specific implementation of the present invention has been described in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that, based on the technical solution of the present invention, various modifications or variations that can be made by those skilled in the art without creative effort are still within the scope of protection of the present invention.

Claims

1. A design method for a vehicle front-end cooling module based on intake air volume prediction, characterized in that, include: Obtain data on key factors affecting radiator airflow under the current vehicle front cooling module design scheme; these key factors include vehicle speed, front grille styling data, radiator PQ performance data, fan PQ performance data, and overall vehicle layout status data. For key factor data, convolutional neural networks and graph convolutional neural networks are used to extract the non-spatial and spatial features of the currently designed cooling module, respectively. Non-spatial features of the cooling module are extracted using a convolutional neural network, including: The vehicle speed, radiator PQ performance data, and fan PQ performance data are all one-dimensional parameter data, which are directly used as the physical characteristics of the cooling module; among them, PQ performance data refers to the slope of the PQ characteristic curve, where P represents static pressure and Q represents airflow. The front grille styling data and the overall vehicle layout data are both three-dimensional point cloud data. Convolutional neural networks are used to perform dimensionality reduction on the three-dimensional point cloud data to extract the opening ratio features of the front grille styling and the geometric features of various components in the overall vehicle layout. Among them, the components are components in the front cooling module of the vehicle, including the radiator, fan and front grille. Spatial features of the cooling module are extracted using a graph neural network, including: The components are treated as nodes, and adjacent nodes are connected by edges. Node attributes and edge attributes are set to model the spatial relationship diagram of the cooling module. The node attributes include geometric features and physical features, and the edge attributes represent the spatial attributes between nodes, including distance, distance weight, relative angle, and occlusion state. Graph neural networks are used to extract features from spatial relationship graphs. Each node, after fusing its own geometric and physical attributes, passes messages through edge connections to obtain spatial features that fuse the spatial attributes between nodes. The extracted features are input into the air intake volume prediction model, and the predicted air intake volume under the current design scheme is output. The key factor data is continuously adjusted to generate different design schemes, and the air intake volume under different design schemes is predicted. Based on the predicted air intake volume, the optimal cooling module design scheme is obtained through adjustment and screening.

2. The vehicle front-end cooling module design method based on air intake volume prediction as described in claim 1, characterized in that, The distance is calculated as follows: ； in, , They represent the first and The three-dimensional centroid coordinates of a node or component; The distance weight is calculated as follows: ； in, This is an adjustable parameter, representing the distance attenuation factor. This represents the correlation coefficient between types.

3. The vehicle front-end cooling module design method based on air intake volume prediction as described in claim 1, characterized in that, The extracted features are input into the air intake volume prediction model. After feature fusion, the predicted air intake volume under the current design scheme is output, including: By integrating several extracted non-spatial features using a multilayer perceptron, the global features of the entire cooling module are obtained. By using global pooling, the spatial features of all nodes are pooled to obtain the spatial features of the entire cooling module. The global features and spatial features are concatenated, and the predicted air intake is output through a fully connected layer.

4. The vehicle front-end cooling module design method based on air intake volume prediction as described in claim 1, characterized in that, When calculating the optimal cooling module design, a multi-objective optimization function is constructed with the target air volume as the objective and various key factor data as variables. A genetic algorithm is used to solve the multi-objective optimization function. During the solution process, the air intake volume prediction model function is used as the fitness function to continuously iterate and optimize the data of various key factors until the target air output volume is reached, thereby obtaining the optimal key factor data and generating the optimal cooling module design scheme.

5. A vehicle front-end cooling module design system based on intake air volume prediction, characterized in that, The vehicle front-end cooling module design method based on airflow prediction as described in any one of claims 1-4 includes: The data acquisition module is used to acquire data on key factors affecting the radiator air intake under the current vehicle front cooling module design scheme; among which, key factor data includes vehicle speed, front grille styling data, radiator PQ performance data, fan PQ performance data, and overall vehicle layout status data. The feature extraction module is used to extract the non-spatial and spatial features of the currently designed cooling module based on key factor data, using convolutional neural networks and graph convolutional neural networks respectively. The air intake volume prediction module is used to input the extracted features into the air intake volume prediction model and output the predicted air intake volume under the current design scheme. The design optimization module is used to continuously adjust key factor data to generate different design schemes and predict the air intake volume under different design schemes. Based on the predicted air intake volume, the module is adjusted and screened to obtain the optimal cooling module design scheme.

6. An electronic device, characterized in that, include: Memory, used to store executable instructions; The processor, when executing executable instructions stored in the memory, implements the vehicle front-end cooling module design method based on air intake volume prediction as described in any one of claims 1-4.

7. A computer-readable storage medium, characterized in that, The device stores executable instructions that, when executed by a processor, implement the vehicle front-end cooling module design method based on air intake volume prediction as described in any one of claims 1-4.

8. A computer program product, characterized in that, The computer program product includes executable instructions stored in a computer-readable storage medium; When the processor of the electronic device reads the executable instructions from the computer-readable storage medium and executes the executable instructions, it implements the vehicle front-end cooling module design method based on air intake volume prediction as described in any one of claims 1-4.

Images