Method and device for synchronously predicting aerodynamic performance through multiple parameters, equipment and storage medium

Abstract

The invention discloses a method and device for synchronously predicting aerodynamic performance through multiple parameters, equipment and a storage medium. The method comprises the steps of extracting spatial feature information of a model by reading a to-be-predicted model file; spatial feature information of the model is transmitted into the neural network model for prediction, and multi-parameter synchronous prediction of aerodynamic performance of the to-be-predicted model, including a surface pressure field, a velocity field and an appearance wind resistance value, is realized. According to the method, the problems of large calculation amount and high time cost of traditional simulation are effectively solved; a traditional machine learning method is poor in multi-parameter synchronous prediction generalization ability and low in prediction precision; a traditional machine learning method has the problems of high calculation resource loss and high time cost due to multi-parameter separate training and repeated feature extraction. According to the method, multi-parameter synchronous training is adopted, key features are extracted in a unified mode, then task-divided synchronous processing is conducted, a large amount of computing resources and time cost are saved in the aspect of predicting aerodynamic multi-parameter performance, and the method has high practical application value.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of fluid dynamics simulation, and in particular to a method and apparatus, device, and storage medium for synchronously predicting aerodynamic performance with multiple parameters. Background Art

[0002] Aerodynamic drag research plays a key role in shaping the performance and efficiency of vehicles in the automotive and aerospace industries. Aerodynamic performance is essential to improving vehicle range, reducing wind noise, improving fuel economy, reducing emissions and driving towards carbon neutrality goals.

[0003] Traditional aerodynamic prediction methods use computational fluid dynamics (CFD) simulation technology, which relies on intensive and time-consuming calculations. In the specific design process, during the initial styling surface (CAS, Concept A Surface) stage of the car, engineers need to calculate the drag coefficient of the car, find the noise source by looking at the velocity cloud map and pressure cloud map of CFD fluid simulation, and make styling improvements for the airflow separation point, so as to propose targeted optimization solutions.

[0004] As a new prediction tool, artificial intelligence methods have also been initially applied in the field of aerodynamic performance prediction. Traditional artificial intelligence methods usually use U-Net to extract the spatial features of the model and output a parameter in the drag coefficient, velocity field, or pressure field. Traditional models can only perform single-task training and prediction for a single parameter, and show poor generalization ability and low prediction accuracy when dealing with simultaneous prediction of multiple parameters. In order to ensure prediction accuracy, traditional artificial intelligence methods need to train multiple parameters separately and extract features repeatedly when predicting multiple parameters, resulting in large loss of computing resources and high time costs.

[0005] At present, new energy vehicles are developing rapidly. Automobile manufacturers need to quickly iterate automobile products and optimize product performance to improve their own product competitiveness in order to better occupy market share and increase profits. Using artificial intelligence methods to replace traditional CFD methods is of great significance for saving product iteration time and cost, improving product competitiveness, occupying market share and increasing profits. Summary of the invention

[0006] In view of this, the embodiments of the present application provide a method and apparatus, device, and storage medium for synchronously predicting aerodynamic performance with multiple parameters, which can at least solve the aforementioned technical problems.

[0007] According to a first aspect of an embodiment of the present application, a method for synchronously predicting aerodynamic performance with multiple parameters is provided, the method comprising: Read the model file and extract the spatial feature information of the model; Storing the spatial feature information of the model; The spatial feature information of the model is passed into the neural network for training to determine the best neural network model; The characteristic data of the model to be predicted is input into the optimal neural network model, and the aerodynamic performance of the model to be predicted is predicted with multiple parameters simultaneously.

[0008] In some exemplary embodiments, reading the model file and extracting the spatial feature information of the model includes: Read the model prediction point position, speed, pressure and air resistance coefficient in the model file; Extract the model signed distance function (SDF, Signed Distance Funciton) signed space distance field, model vertex adjacency, and the normal vector of the predicted point relative to the model surface.

[0009] In some exemplary embodiments, the reading of the model prediction point position, speed, pressure and air resistance coefficient in the model file includes: Read the .vtk file including the vertex position and velocity of each vertex of the velocity field around the car to obtain velocity data; Read the .ply file including the car shape mesh and vertices to obtain the mesh information, read the .npy file corresponding to the .ply file to obtain the pressure data; Read the .csv file containing the air drag coefficient corresponding to the car model file name to obtain the air drag coefficient.

[0010] In some exemplary embodiments, extracting the model SDF signed space distance field, model vertex adjacency, and normal vector of the predicted point relative to the model surface includes: Calculate the SDF signed space distance field through the car model watertight mesh; Obtain the vertex adjacency matrix through the adjacency relationship between mesh vertices; Calculate the normal vector of the vertex relative to the car surface through the mesh and the vertex.

[0011] In some exemplary embodiments, obtaining the adjacency matrix of the vertices through the adjacency relationship between the mesh vertices includes: taking each vertex as the center of a circle and considering the vertices within a preset radius to construct an adjacency matrix of the vertices; Calculating the normal vector of a vertex relative to the surface of the car through the mesh and the vertex includes: calculating the distance component of each vertex relative to the nearest point on the surface of the car in the .ply file, and normalizing the distance component into the vertex normal vector.

[0012] In some exemplary embodiments, storing the model feature space feature information includes: Store the predicted point position, predicted point velocity, pressure, model air resistance coefficient, model SDF signed space distance field, model vertex adjacency, and normal vector of the predicted point relative to the model surface in a .npz file; Storage methods include images, videos, numbers, arrays, matrices, lists, and dictionaries.

[0013] In some exemplary embodiments, the neural network comprises: Graph neural network model or its variant, Fourier neural operator or its variant, convolutional neural network or its variant.

[0014] In some exemplary embodiments, inputting the model feature data to be predicted into the optimal neural network model and performing multi-parameter synchronous prediction on the aerodynamic performance of the model to be predicted includes: The characteristic data of the model to be predicted is input into the optimal neural network model, and the velocity component, velocity synthesis value, pressure component, pressure synthesis value, global air resistance coefficient and local air resistance coefficient of the model to be predicted are predicted.

[0015] According to a second aspect of an embodiment of the present application, a device for synchronously predicting aerodynamic performance by multiple parameters is provided, the device comprising: A data preprocessing unit, used to read the model file and extract the spatial feature information of the model; A storage unit, used for storing the spatial feature information of the model; A model training unit is used to input the spatial feature information of the model into the neural network for training and determine the best neural network model; The prediction unit is used to input the characteristic data of the model to be predicted into the optimal neural network model, and perform multi-parameter synchronous prediction on the aerodynamic performance of the model to be predicted.

[0016] According to a third aspect of an embodiment of the present application, an electronic device is provided, comprising a processor and a memory for storing processor executable instructions, wherein the processor is configured to execute the steps of the above-mentioned method for synchronously predicting aerodynamic performance with multiple parameters when calling the executable instructions in the memory.

[0017] According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, wherein a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, the steps of the above-mentioned method for synchronously predicting aerodynamic performance with multiple parameters are implemented.

[0018] The method for predicting aerodynamic performance based on artificial intelligence in the embodiment of the present application obtains the shape data of the object and the velocity field and pressure field and air resistance coefficient around the object, extracts and stores the shape features of the object, reads the feature information and transmits it to the SCA (Spatial and Channel-wise Attention) + GNN (Graph Neural Network) + FNO (Fourier Neural Operator) network model for training, and uses the best training model to predict the model to be predicted. This method is based on the shape structure and has a wide range of applications. It can be applied to the fields of ships, automobiles, aerospace, etc., and can simultaneously process multiple tasks and simultaneously predict the velocity field, pressure field and air resistance coefficient. It has a high accuracy rate and extremely short prediction time, effectively solving the problems of large amount of calculation and high time cost of traditional simulation, and has high practical application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following is a brief introduction to the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0020] Figure 1 A schematic diagram of the process flow of a method for synchronously predicting aerodynamic performance with multiple parameters according to the first embodiment of the present application is shown; Figure 2 A schematic diagram of the normal vector and pressure of the predicted point in an embodiment of the present application is shown; Figure 3 A schematic diagram of the predicted point speed synthesis quantity of an embodiment of the present application is shown; Figure 4 A schematic diagram of a training loss curve of an embodiment of the present application is shown; Figure 5 A schematic diagram of a method for predicting aerodynamic performance of a model to be predicted according to an embodiment of the present application is shown; Figure 6 A schematic diagram showing the absolute error between the predicted result of the surface pressure of the automobile model and the actual pressure in an embodiment of the present application is shown; Figure 7 A schematic diagram showing the absolute error between the predicted result of the vehicle model velocity field and the actual velocity field and the predicted points where the absolute error is greater than 10% in an embodiment of the present application; Figure 8 A schematic diagram showing a flow chart of a method for synchronously predicting aerodynamic performance with multiple parameters according to a second embodiment of the present application; Fig. 9 A schematic diagram showing the composition structure of a device for synchronously predicting aerodynamic performance with multiple parameters according to an embodiment of the present application is shown; Fig.10 A schematic diagram of the structure of an electronic device according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0021] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution in the embodiment of the present invention will be described clearly and completely below. The exemplary embodiments described below and shown in the drawings are intended to teach the principles of the present invention so that those skilled in the art can implement and use the present invention in several different environments and for several different applications. Therefore, the scope of protection of the present invention is defined by the appended claims, and the exemplary embodiments are not intended to be, and should not be considered to be, a restrictive description of the scope of protection of the present invention. Unless otherwise specifically stated, the order of components and assembly steps and the numerical values ​​described in the embodiments do not limit the scope of the present invention. Moreover, any numerical range stated herein is intended to include all sub-ranges contained therein, and the numerical range represented by "value A to value B" refers to a range including the endpoint values ​​A and B. It can be understood by those skilled in the art that the terms "first", "second", "step" and the like in the present invention are only used to distinguish different steps, devices or modules, etc., and do not represent any specific technical meanings, nor do they represent the necessary logical order between them. For example, steps two and three can be swapped or run in parallel.

[0022] The essence of the technical solution of the embodiment of the present application is explained in detail below with reference to the accompanying drawings.

[0023] Figure 1 The schematic diagram of the method flow of multi-parameter synchronous prediction of aerodynamic performance in the first embodiment of the present application is shown as follows: Figure 1 As shown, the method for synchronously predicting aerodynamic performance with multiple parameters in an embodiment of the present application includes the following steps: Step 101: Read the model file and extract the spatial feature information of the model; In the embodiment of the present application, the reading of the model file and the extraction of the spatial feature information of the model include: Read the model prediction point position, speed, pressure and air resistance coefficient in the model file; Extract the model signed distance function SDF, signed space distance field, model vertex adjacency, and normal vector of the predicted point relative to the model surface.

[0024] Figure 2 This is a schematic diagram of the normal vector and pressure of the predicted point in the embodiment of the present application, such as Figure 2 As shown, the gradient from red to blue represents a decreasing pressure value, and the green line represents the normal vector of the predicted point.

[0025] Figure 3 Schematic diagram of the predicted point speed synthesis quantity of the embodiment of the present application; Figure 3 As shown, the predicted point position distribution and the composite speed of the velocity components in the x, y, and z directions of the predicted point. The gradual change from red to blue represents a decreasing speed value.

[0026] SDF represents the minimum distance from a point to a surface, and uses positive and negative to distinguish whether a point is inside or outside the surface. If the point is inside the surface, the distance is negative, if the point is outside the surface, the distance is positive, and if the point is on the surface, the distance is 0.

[0027] As an implementation method, the reading of the model prediction point position, speed, pressure and air resistance coefficient in the model file includes: Read the .vtk file including the vertex positions and velocities of each vertex in the velocity field around the car to obtain velocity data; read the .ply file including the car shape mesh and vertices to obtain mesh information, read the .npy file corresponding to the .ply file to obtain pressure data; read the .csv file including the air resistance coefficient corresponding to the car model file name to obtain the air resistance coefficient.

[0028] As an implementation method, the extraction of the model SDF signed space distance field, the model vertex adjacency relationship, and the normal vector of the prediction point relative to the model surface includes: The SDF signed space distance field is calculated by the car model watertight grid, where the resolution of the car model watertight grid is ; Obtain the adjacency matrix of the vertex through the adjacency relationship between the mesh vertices; Calculate the normal vector of the vertex relative to the car surface through the mesh and the vertex.

[0029] Specifically, obtaining the adjacency matrix of vertices through the adjacency relationship between mesh vertices includes: taking each vertex as the center of a circle, considering vertices within a preset radius, and constructing an adjacency matrix of vertices; in the embodiment of the present application, the preset radius is 0.028m.

[0030] Calculating the normal vector of a vertex relative to the surface of the car through the mesh and the vertex includes: calculating the distance component of each vertex relative to the nearest point on the surface of the car in the .ply file, and normalizing the distance component into the vertex normal vector.

[0031] Step 102: storing the spatial feature information of the model; In the embodiment of the present application, storing the model feature space feature information includes: The predicted point position, predicted point speed, pressure, model air resistance coefficient, model SDF signed space distance field, model vertex adjacency, and normal vector of the predicted point relative to the model surface are stored in a .npz file; the storage method includes image, video, value, array, matrix, list, and dictionary.

[0032] Step 103: The spatial feature information of the model is passed into the neural network for training to determine the best neural network model; In an embodiment of the present application, the spatial feature information includes the SDF signed spatial distance field of the model, the adjacency relationship of the model vertices, the normal vector of the predicted point relative to the model surface, and the distance of the predicted point relative to the first fluid disturbance point.

[0033] The neural network includes: a graph neural network model or its variants, a Fourier neural operator or its variants, a convolutional neural network or its variants.

[0034] In the embodiment of the present application, spatial and channel-wise attention (SCA) is a hybrid attention mechanism, which includes two attention mechanisms: spatial attention and channel wise attention. The attention mechanism is a data processing method in machine learning, which reflects the degree of attention to different information by weighting the information. This enables the network model to learn features better.

[0035] Graph Neural Networks (GNN) can capture the complex relationships between nodes and the overall topological structure of the graph. The features of neighboring nodes are weighted and aggregated through graph convolution to update the feature representation of the current node, allowing GNN to capture the local structural information of the graph.

[0036] An operator is a mapping from function space to function space. The Fourier Neural Operator (FNO) maps feature data to Fourier space and returns it to the physical space as output after solving it. It can better capture global information and has good noise resistance.

[0037] Convolutional Neural Networks (CNN) include convolution calculations, which have the characteristics of local perception, parameter sharing, and multi-core. The local information obtained by convolution can be expanded to the global through multi-layer full connection. The parameter sharing of convolution greatly reduces the amount of calculation, and multiple features can be extracted at one time through multi-core. This enables convolutional neural networks to better and faster capture local and global features.

[0038] Specifically, 90% of the data is set as the training set, 10% of the data is set as the test set, the number of iterations is set to 120, the batch size is set to 5, and the loss function is the L2 loss function. Get the SDF information stored in .npz, predict the point position, normal vector, adjacency, pressure value, and velocity value, and pass them into the SCA+GNN+FNO neural network for training. Figure 4 Schematic diagram of the training loss curve of the embodiment of the present application, such as Figure 4 As shown, as the number of iterations increases, the loss decreases; the model with the highest accuracy is saved during the training process. In this embodiment, the loss of the best model on the test set is 0.04553.

[0039] Step 104: inputting the characteristic data of the model to be predicted into the optimal neural network model, and performing multi-parameter synchronous prediction on the aerodynamic performance of the model to be predicted.

[0040] In the embodiment of the present application, the step of inputting the model feature data to be predicted into the optimal neural network model and performing multi-parameter synchronous prediction on the aerodynamic performance of the model to be predicted includes: The characteristic data of the model to be predicted is input into the optimal neural network model, and the velocity component, velocity synthesis value, pressure component, pressure synthesis value, global air resistance coefficient and local air resistance coefficient of the model to be predicted are predicted.

[0041] Specifically, read the mesh vertices in the .ply file of the car model to be predicted as the pressure value to be predicted points, and obtain the velocity field to be predicted points from the .vtk file of the model to be predicted. Extract the shape features, load the best model for prediction, and output the pressure value, velocity component and air resistance coefficient of the point to be predicted. Figure 5 Schematic diagram of a method for predicting aerodynamic performance of a model to be predicted according to an embodiment of the present application. Figure 5 As shown, the method for performing multi-parameter synchronous prediction of the aerodynamic performance of the model to be predicted in the embodiment of the present application includes the following steps: Step 501: Obtain model feature data to be predicted; The model feature data includes the normal vector and adjacency relationship of the predicted point and the SDF signed space distance field.

[0042] Step 502: Process the normal vector and adjacency relationship of the predicted point through the SCA hybrid attention mechanism; Step 503: Processing the SDF signed space distance field data through the SCA hybrid attention mechanism; Step 504: Input the data processed in step 502 into the graph neural network model GNN; Step 505: input the data processed in step 503 into the Fourier neural operator FNO; Step 506: Obtain corresponding features of the pressure prediction points and splice the features; Step 507: Obtain the corresponding features of the speed prediction points and splice the features; Step 508: extracting into a value through the fully connected layer, corresponding to the drag coefficient; Step 509: Mapping to a one-dimensional output corresponding to a pressure value through a fully connected layer; Step 510: Mapping to a three-dimensional output through a fully connected layer, corresponding to a three-dimensional velocity component; Step 511: Output the prediction result.

[0043] Figure 6 This is a schematic diagram of the absolute error between the predicted surface pressure of the car model according to an embodiment of the present application and the actual pressure, where the gradient from green to black represents a decreasing error. Figure 7 This is a schematic diagram of the absolute error between the predicted result of the vehicle model velocity field and the actual velocity field and the predicted points where the absolute error is greater than 10%. The gradual change from green to black represents a decreasing error. Figure 6 and Figure 7 The final pressure values ​​and velocity components are stored as .npy files and plotted as shown in Figure 6 , Figure 7 The absolute error image of the prediction results is shown, and the air resistance coefficient is stored in a .csv file with the corresponding file name.

[0044] Figure 8 Schematic diagram of the process of the method for synchronously predicting aerodynamic performance with multiple parameters according to the second embodiment of the present application. Figure 8 As shown, the method for multi-parameter synchronous prediction of aerodynamic performance of the second embodiment of the present application takes the automobile model as an example, including data preprocessing, data storage, data reading, training neural network model, saving the best neural network model, predicting the model to be predicted, saving the prediction results and other processes. Among them, the data preprocessing includes reading the model file and extracting its spatial feature information, obtaining the model shape features and storing them. Specifically including: reading the model file and the model prediction point position and the prediction point speed, pressure, and model air resistance coefficient; extracting the model SDF signed space distance field and the model vertex adjacency relationship, as well as the normal vector of the prediction point relative to the model surface; saving the processed model features; reading the model features; passing the model features into the SCA+GNN+FNO neural network for training; saving the best neural network model; extracting the features of the model to be predicted; loading the saved neural network model and predicting the model to be predicted, and outputting the prediction results.

[0045] The specific steps include: Step 1001: Read car model data; Step 1002: extracting the signed space distance field of the SDF of the automobile shape, the adjacency relationship of the model vertices, and the normal vector of the predicted point relative to the model surface; Step 1003: storing feature information; Step 1004: Obtain characteristic information and corresponding prediction point speed, pressure, and air resistance coefficient; Step 1005: training the SCA+GNN+FNO neural network model; Step 1006: Save the optimal neural network model parameter weights; Step 1007: Read the car model to be predicted; Step 1008: extracting the signed space distance field of the SDF of the automobile shape, the adjacency relationship of the model vertices, and the normal vector of the predicted point relative to the model surface, and storing them; Step 1009: Acquire feature information and corresponding prediction point location information; Step 1010: Load the saved neural network model and predict the model to be predicted; Step 1011: Output prediction results.

[0046] In this embodiment, it takes about 0.12 seconds to read the processed features and predict the speed, pressure and air resistance coefficient of a single car model. Adding the time for feature extraction, storage and re-reading, a single car model takes about 14 seconds. If multiple car models are pre-processed in parallel and then predicted, the average prediction time for each car model is about 4 seconds, which is much lower than the time required for traditional CFD solutions.

[0047] The method for predicting aerodynamic performance based on artificial intelligence in the embodiment of the present application is applied to the field of fluid aerodynamic simulation, including data preprocessing, data storage, data reading, training neural network model, saving the best neural network model, predicting the model to be predicted, saving the prediction results and other steps. By learning the connection between the shape features of the object and the results of fluid mechanics simulation, the velocity field, pressure field and air resistance coefficient around the object to be predicted are quickly predicted, and two neural network models are used to process global and local information respectively, so as to realize multi-parameter combined prediction. Specifically, by obtaining the shape data of the object and the velocity field, pressure field and air resistance coefficient around the object, the shape features of the object are extracted and stored, the feature information is read and transmitted to the SCA+GNN+FNO network model for training, and the model to be predicted is predicted using the best training model. In the embodiment of the present application, SCA can be replaced by other attention mechanisms. The method is based on the shape structure and has a wide range of applications, and can be applied to ships, automobiles, aerospace and other fields. It can also process multiple tasks simultaneously, and simultaneously predict the velocity field, pressure field and air resistance coefficient, with a high accuracy and extremely short prediction time, which effectively solves the problems of large amount of calculation and high time cost of traditional simulation, and has high practical application value.

[0048] Fig. 9 Schematic diagram of the structure of the device for synchronously predicting aerodynamic performance with multiple parameters according to an embodiment of the present application. Fig. 9 As shown, the device of the embodiment of the present application includes: The data preprocessing unit 91 is used to read the model file and extract the spatial feature information of the model; A storage unit 92, used to store the spatial feature information of the model; A model training unit 93, used to transfer the spatial feature information of the model to the neural network for training and determine the best neural network model; The prediction unit 94 is used to input the characteristic data of the model to be predicted into the optimal neural network model, and perform multi-parameter synchronous prediction on the aerodynamic performance of the model to be predicted.

[0049] In an exemplary embodiment, the data preprocessing unit 91, the storage unit 92, the model training unit 93, the prediction unit 94, etc. can be implemented by one or more central processing units (CPU), graphics processing units (GPU), application specific integrated circuits (ASIC), DSP, programmable logic devices (PLD), complex programmable logic devices (CPLD), field programmable gate arrays (FPGA), general-purpose processors, controllers, microcontrollers (MCU), microprocessors, or other electronic components.

[0050] In the embodiments of the present application, Fig. 9 The specific manner in which each unit in the illustrated device for synchronously predicting aerodynamic performance with multiple parameters performs operations has been described in detail in the embodiment of the method, and will not be elaborated on in detail here.

[0051] In an exemplary embodiment, an electronic device is also provided, including a processor and a memory for storing processor executable instructions, wherein the processor is configured to execute the steps of the above-mentioned method for synchronously predicting aerodynamic performance with multiple parameters when calling the executable instructions in the memory.

[0052] Fig.10Schematic diagram of the structure of the electronic device according to the embodiment of the present application. Fig.10 As shown, the electronic device 800 supports multi-screen output, and the electronic device 800 may include one or more of the following components: a processing component 802, a memory 804, a power component 806, a multimedia component 808, an audio component 810, an input / output (I / O) interface 812, a sensor component 814, and a communication component 816.

[0053] The processing component 802 generally controls the overall operation of the electronic device 800, such as operations associated with display, phone calls, data communications, camera operations, and recording operations. The processing component 802 may include one or more processors 820 to execute instructions to complete all or part of the steps of the above-mentioned method. In addition, the processing component 802 may include one or more modules to facilitate the interaction between the processing component 802 and other components. For example, the processing component 802 may include a multimedia module to facilitate the interaction between the multimedia component 808 and the processing component 802.

[0054] The memory 804 is configured to store various types of data to support operations on the device 800. Examples of such data include instructions for any application or method operating on the electronic device 800, contact data, phone book data, messages, pictures, videos, etc. The memory 804 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0055] The power component 806 provides power to the various components of the electronic device 800. The power component 806 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for the electronic device 800.

[0056] The multimedia component 808 includes a screen that provides an output interface between the electronic device 800 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundaries of the touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 808 includes a front camera and / or a rear camera. When the device 800 is in an operating mode, such as a shooting mode or a video mode, the front camera and / or the rear camera may receive external multimedia data. Each front camera and rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.

[0057] The audio component 810 is configured to output and / or input audio signals. For example, the audio component 810 includes a microphone (MIC), and when the electronic device 800 is in an operation mode, such as a call mode, a recording mode, and a voice recognition mode, the microphone is configured to receive an external audio signal. The received audio signal can be further stored in the memory 804 or sent via the communication component 816. In some embodiments, the audio component 810 also includes a speaker for outputting audio signals.

[0058] I / O interface 812 provides an interface between processing component 802 and peripheral interface modules, such as keyboards, click wheels, buttons, etc. These buttons may include but are not limited to: a home button, a volume button, a start button, and a lock button.

[0059] The sensor assembly 814 includes one or more sensors for providing various aspects of status assessment for the electronic device 800. For example, the sensor assembly 814 can detect the open / closed state of the device 800, the relative positioning of components, such as the display and keypad of the electronic device 800, and the sensor assembly 814 can also detect the position change of the electronic device 800 or a component of the electronic device 800, the presence or absence of contact between the user and the electronic device 800, the orientation or acceleration / deceleration of the electronic device 800, and the temperature change of the electronic device 800. The sensor assembly 814 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. The sensor assembly 814 may also include an optical sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 814 may also include an accelerometer, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

[0060] The communication component 816 is configured to facilitate wired or wireless communication between the electronic device 800 and other devices. The electronic device 800 can access a wireless network based on a communication standard, such as Wi-Fi, 2G or 3G, or a combination thereof. In an exemplary embodiment, the communication component 816 receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 816 also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.

[0061] In an exemplary embodiment, the electronic device 800 may be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the steps of the method for simultaneously predicting aerodynamic performance with multiple parameters of the above-described embodiment.

[0062] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, wherein a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, the steps of the method for synchronously predicting aerodynamic performance by multiple parameters are implemented. For example, the non-transitory computer-readable storage medium may be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like.

[0063] It should be understood that "one embodiment" or "an embodiment" mentioned throughout the specification means that specific features, structures or characteristics related to the embodiment are included in at least one embodiment of the present invention. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. It should be understood that in various embodiments of the present invention, the size of the sequence number of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present application. The above-mentioned sequence numbers of the embodiments of the present application are for description only and do not represent the advantages and disadvantages of the embodiments.

[0064] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0065] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or do not exist.

[0066] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0067] The above is only an embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for synchronously predicting aerodynamic performance using multiple parameters, characterized in that: The method comprises: Read the model file and extract the spatial feature information of the model; Storing the spatial feature information of the model; The spatial feature information of the model is passed into the neural network for training to determine the best neural network model; The characteristic data of the model to be predicted is input into the optimal neural network model, and the aerodynamic performance of the model to be predicted is predicted with multiple parameters simultaneously.

2. The method for synchronously predicting aerodynamic performance using multiple parameters according to claim 1, characterized in that: The step of reading the model file and extracting the spatial feature information of the model includes: Read the model prediction point position, speed, pressure and air resistance coefficient in the model file; Extract the model signed distance function SDF, signed space distance field, model vertex adjacency, and normal vector of the predicted point relative to the model surface.

3. The method for synchronously predicting aerodynamic performance by multiple parameters according to claim 2, characterized in that: The reading of the model prediction point position, speed, pressure and air resistance coefficient in the model file includes: Read the .vtk file including the vertex position and velocity of each vertex of the velocity field around the car to obtain velocity data; Read the .ply file including the car shape mesh and vertices to obtain the mesh information, read the .npy file corresponding to the .ply file to obtain the pressure data; Read the .csv file containing the air drag coefficient corresponding to the car model file name to obtain the air drag coefficient.

4. The method for synchronously predicting aerodynamic performance by multiple parameters according to claim 2, characterized in that: The extraction model SDF has a signed space distance field, a model vertex adjacency relationship, and a normal vector of a prediction point relative to a model surface, including: Calculate the SDF signed space distance field through the car model watertight mesh; Obtain the vertex adjacency matrix through the adjacency relationship between mesh vertices; Calculate the normal vector of the vertex relative to the car surface through the mesh and the vertex.

5. The method for synchronously predicting aerodynamic performance using multiple parameters according to claim 4, characterized in that: The step of obtaining the adjacency matrix of the vertices through the adjacency relationship between the mesh vertices includes: taking each vertex as the center of a circle and considering the vertices within a preset radius to construct an adjacency matrix of the vertices; The method of calculating the normal vector of a vertex relative to the surface of the car through the mesh and the vertex includes: calculating the distance component of each vertex relative to the nearest point on the surface of the car in the .ply file, and unitizing the distance component into the vertex normal vector.

6. The method for synchronously predicting aerodynamic performance using multiple parameters according to claim 1, characterized in that: The storing of the model feature space feature information includes: Store the predicted point position, predicted point velocity, pressure, model air resistance coefficient, model SDF signed space distance field, model vertex adjacency, and normal vector of the predicted point relative to the model surface in a .npz file; Storage methods include images, videos, numbers, arrays, matrices, lists, and dictionaries.

7. The method for synchronously predicting aerodynamic performance using multiple parameters according to claim 1, characterized in that: The neural network comprises: Graph neural network model or its variant, Fourier neural operator or its variant, convolutional neural network or its variant.

8. The method for synchronously predicting aerodynamic performance using multiple parameters according to claim 1, characterized in that: The step of inputting the model characteristic data to be predicted into the optimal neural network model and performing multi-parameter synchronous prediction on the aerodynamic performance of the model to be predicted includes: The characteristic data of the model to be predicted is input into the optimal neural network model, and the velocity component, velocity synthesis value, pressure component, pressure synthesis value, global air resistance coefficient and local air resistance coefficient of the model to be predicted are predicted.

9. A device for synchronously predicting aerodynamic performance using multiple parameters, characterized in that: The device comprises: A data preprocessing unit, used to read the model file and extract the spatial feature information of the model; A storage unit, used for storing spatial feature information of the model; A model training unit is used to input the spatial feature information of the model into the neural network for training and determine the best neural network model; The prediction unit is used to input the characteristic data of the model to be predicted into the optimal neural network model, and perform multi-parameter synchronous prediction on the aerodynamic performance of the model to be predicted.

10. An electronic device comprising a processor and a memory for storing instructions executable by the processor, characterized in that: The processor is configured to execute the steps of the method for synchronously predicting aerodynamic performance with multiple parameters as claimed in any one of claims 1 to 8 when calling the executable instructions in the memory.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method for synchronously predicting aerodynamic performance with multiple parameters according to any one of claims 1 to 8 are implemented.

Images