GAN-based High-speed Aerodynamic Heat Flow Field Reconstruction Method and System

Abstract

The present invention discloses a method and system for reconstructing high-speed aerodynamic heat flow fields based on GAN, which relates to the technical field of simulation. The method includes: collecting historical heat flow field data of a high-speed aircraft under different flight conditions to construct a real aerodynamic heat flow field data set; establishing a heat flow field generator and a discriminator through feature extraction and simulation discrimination training; initializing and adaptively adjusting the grid according to the change of the heat flow field distribution, and dividing it into N hierarchical reconstruction regions; adopting a balanced training strategy to alternately train the generator and the discriminator in each region, and building N heat flow field reconstruction networks; monitoring the flight conditions to update the data, and using each reconstruction network for optimization to generate a simulated heat flow field. Furthermore, the technical effects of simplifying parameter configuration, reducing process costs, and improving timeliness are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of simulation, and particularly to a high-speed aerodynamic heat flow field reconstruction method and system based on GAN. Background Art

[0002] Due to the influence of shock waves under complex flight conditions, high-speed aircraft are subjected to strong aerodynamic heating. The construction of the heat flow field distribution of high-speed aircraft is crucial for aircraft structural design, thermal protection system optimization, and flight safety optimization.

[0003] Traditional heat flow field reconstruction methods usually rely on numerical simulations or experimental data (such as wind tunnel or flight experiments). With the continuous complexity of the shapes of high-speed aircraft, the local aerodynamic heating interference of the aircraft gradually intensifies, the phenomenon of turbulent separation and reattachment becomes more significant, and the coupling effect of rarefied gas and real gas effects also becomes more prominent. Under such complex flight conditions, the above traditional methods have technical problems such as complex parameter settings, high computational costs, and poor real-time performance. Summary of the Invention

[0004] The present invention provides a high-speed aerodynamic heat flow field reconstruction method and system based on GAN to solve the technical problems of complex parameter settings, high computational costs, and poor real-time performance in the prior art, and to achieve the technical effects of simplifying parameter configuration, reducing process costs, and improving timeliness.

[0005] In a first aspect, the present invention provides a high-speed aerodynamic heat flow field reconstruction method based on GAN. The high-speed aerodynamic heat flow field reconstruction method based on GAN includes:

[0006] Collect and obtain a real aerodynamic heat flow field data set, where the real aerodynamic heat flow field data set is the historical heat flow field data of a high-speed aircraft under different flight conditions.

[0007] Based on the real aerodynamic heat flow field data set, perform feature extraction training and simulation discrimination training to construct a heat flow field generator and a heat flow field discriminator.

[0008] According to the distribution change data of the target heat flow field, perform grid initialization and adaptive adjustment to obtain a heat flow field distribution grid, and based on the heat flow field distribution grid, divide the reconstruction area of the target heat flow field to obtain N hierarchical reconstruction areas.

[0009] Construct a balanced training strategy, and use the balanced training strategy to alternately balance-train the heat flow field generator and the heat flow field discriminator based on the N hierarchical reconstruction areas to build N heat flow field reconstruction networks.

[0010] Monitor and obtain flight condition update data, and respectively perform heat flow field reconstruction optimization on the flight condition update data based on the N heat flow field reconstruction networks to generate a high-speed aerodynamic simulation heat flow field.

[0011] In a feasible implementation manner, the construction of the heat flow field generator and the heat flow field discriminator includes:

[0012] Perform Gaussian filtering and data alignment processing on the real aerodynamic heat flow field data set to obtain an available real aerodynamic heat flow field data set.

[0013] Extract correlation features from the available real aerodynamic heat flow field data set according to the heat flow field reconstruction correlation factors to obtain a heat flow field correlation feature set.

[0014] Introduce a multi-scale feature fusion module, and use a convolutional neural network and the multi-scale feature fusion module to perform simulation training on the heat flow field correlation feature set to construct a heat flow field generator.

[0015] Output a simulated heat flow field data set through the heat flow field generator, and perform heat flow field discrimination training based on the real aerodynamic heat flow field data set and the simulated heat flow field data set to generate a heat flow field discriminator.

[0016] In a feasible implementation manner, the construction of the heat flow field generator includes:

[0017] Build a generator network architecture according to a convolutional neural network and the multi-scale feature fusion module. The generator network architecture includes a feature input layer, multiple feature convolution extraction layers, a multi-scale fusion layer, and a heat flow field output layer.

[0018] Pass the heat flow field correlation feature set through the feature input layer and import it into the multiple feature convolution extraction layers for multi-scale feature extraction to obtain multi-scale heat flow field feature data.

[0019] Based on the multi-scale fusion layer, splice and fuse the multi-scale heat flow field feature data to obtain heat flow field simulation data, and output it as an output result through the heat flow field output layer.

[0020] Perform heat flow field simulation training based on the heat flow field correlation feature set and the heat flow field simulation data to construct the heat flow field generator.

[0021] In a feasible implementation manner, the generation of the heat flow field discriminator includes:

[0022] Build a generator network architecture. The generator network architecture includes a data input layer, multiple feature convolution extraction layers, a fully connected layer, and a discriminant output layer.

[0023] Transmit the real aerodynamic heat flow field data set and the simulated heat flow field data set through the data input layer to the multiple feature convolution extraction layers for associated feature extraction, so as to obtain heat flow field associated feature data.

[0024] Based on the fully connected layer, conduct identification evaluation and judgment on the heat flow field associated feature data, obtain heat flow field simulation judgment information, and output it as the output result through the discriminant output layer.

[0025] Based on the real aerodynamic heat flow field data set, the simulated heat flow field data set, and the heat flow field simulation judgment information, conduct evaluation judgment training to generate the heat flow field discriminator.

[0026] In a feasible implementation manner, obtaining the heat flow field distribution grid includes:

[0027] Extract features from the initial distribution characteristics of the target heat flow field to determine the heat flow field change characteristics, where the heat flow field change characteristics include temperature gradient, flow field characteristics, and geometric characteristics.

[0028] Initialize the grid according to the heat flow field change characteristics to generate an initial uniform grid.

[0029] Calculate the gradient change of each grid cell in the initial uniform grid based on the distribution change data of the target heat flow field according to the heat flow field change characteristics to determine the grid cell change gradient data.

[0030] Based on the grid cell change gradient data, adaptively adjust the grid density of the initial uniform grid to obtain the heat flow field distribution grid.

[0031] In a feasible implementation manner, obtaining N hierarchical reconstruction regions includes:

[0032] According to the heat flow field distribution grid, determine the grid cell density distribution information.

[0033] Use the grid cell density distribution information to conduct regional clustering analysis on the heat flow field distribution grid to obtain the grid region clustering result.

[0034] Set the reconstruction region segmentation threshold according to the heat flow field reconstruction accuracy requirement.

[0035] Based on the reconstruction region segmentation threshold, divide the grid region clustering result into reconstruction regions to obtain the N hierarchical reconstruction regions.

[0036] In a feasible implementation manner, building N heat flow field reconstruction networks includes:

[0037] Design the mean square loss function of the generator and the cross-entropy loss function of the discriminator according to the heat flow field generator and the heat flow field discriminator.

[0038] Based on the balanced training strategy, analyze the balance coefficients of the N hierarchical reconstruction regions to determine N initial region training balance coefficients.

[0039] Utilize the mean square loss function of the generator and the cross-entropy loss function of the discriminator to perform alternating balance training on the heat flow field generator and the heat flow field discriminator respectively based on the N initial region training balance coefficients, and obtain N heat flow field reconstruction networks.

[0040] In a feasible implementation manner, the obtaining of the N heat flow field reconstruction networks includes:

[0041] Perform simulated discrimination training on the heat flow field generator and the heat flow field discriminator respectively based on the N initial region training balance coefficients to obtain N initial generators and N initial discriminators.

[0042] Utilize the mean square loss function of the generator and the cross-entropy loss function of the discriminator to evaluate the losses of the N initial generators and N initial discriminators, and obtain N generator loss data and N discriminator loss data.

[0043] Based on the N generator loss data and N discriminator loss data, perform balanced alternating training on the N initial generators and N initial discriminators to obtain N heat flow field reconstruction networks.

[0044] In a feasible implementation manner, the obtaining of the N heat flow field reconstruction networks includes:

[0045] Based on the N generator loss data and N discriminator loss data, dynamically adjust the N initial region training balance coefficients to obtain N region training balance dynamic coefficients.

[0046] Iteratively balance and update the N initial generators and N initial discriminators according to the N region training balance dynamic coefficients to obtain N region generators and N region discriminators.

[0047] Connect and combine the N region generators and N region discriminators to obtain the N heat flow field reconstruction networks.

[0048] In a second aspect, the present invention also provides a high-speed aerodynamic heat flow field reconstruction system based on GAN, wherein the high-speed aerodynamic heat flow field reconstruction system based on GAN includes:

[0049] A true data acquisition component for acquiring a real aerodynamic heat flow field data set, where the real aerodynamic heat flow field data set is historical heat flow field data of a high-speed aircraft under different flight conditions.

[0050] A tool training component for performing feature extraction training and simulation discrimination training based on the real aerodynamic heat flow field data set to construct a heat flow field generator and a heat flow field discriminator.

[0051] A mesh generation component for initializing and adaptively adjusting a mesh according to distribution change data of a target heat flow field to obtain a heat flow field distribution mesh, and partitioning a reconstruction area of the target heat flow field based on the heat flow field distribution mesh to obtain N hierarchical reconstruction areas.

[0052] A balance training component for constructing a balance training strategy, and alternately balancing and training the heat flow field generator and the heat flow field discriminator based on the N hierarchical reconstruction areas by using the balance training strategy to build N heat flow field reconstruction networks.

[0053] An update and optimization component for monitoring and acquiring flight condition update data, and respectively performing heat flow field reconstruction and optimization on the flight condition update data based on the N heat flow field reconstruction networks to generate a high-speed aerodynamic simulation heat flow field.

[0054] The present invention discloses a method and system for reconstructing a high-speed aerodynamic heat flow field based on GAN, including: acquiring historical heat flow field data of a high-speed aircraft under different flight conditions to form a real aerodynamic heat flow field data set; subsequently, performing feature extraction training and simulation discrimination training based on the data set to construct a heat flow field generator and a discriminator; on this basis, initializing and adaptively adjusting a mesh according to distribution change data of a target heat flow field to generate a heat flow field distribution mesh, and dividing the target heat flow field into N hierarchical reconstruction areas accordingly; further constructing a balance training strategy, and alternately balancing and training the heat flow field generator and the discriminator in the N hierarchical reconstruction areas by using the strategy to build N heat flow field reconstruction networks; finally, monitoring and acquiring flight condition update data, and respectively performing heat flow field reconstruction and optimization on the update data based on the N heat flow field reconstruction networks to generate a high-speed aerodynamic simulation heat flow field. The method and system for reconstructing a high-speed aerodynamic heat flow field based on GAN disclosed by the present invention solve the technical problems of complex parameter setting, high calculation cost, and poor real-time performance, and achieve the technical effects of simplifying parameter configuration, reducing process cost, and improving timeliness. Description of the Drawings

[0055] Figure 1 It is a schematic flow chart of the method for reconstructing a high-speed aerodynamic heat flow field based on GAN of the present invention.

[0056] Figure 2This is a schematic structural diagram of the GAN-based high-speed aerodynamic heat flow field reconstruction system of the present invention.

[0057] Explanation of reference numerals: Real data acquisition component 11, tool training component 12, grid division component 13, balance training component 14, update and optimization component 15. Detailed implementation manners

[0058] The following will combine the specification drawings and specific implementation manners to elaborate on the above technical solutions in detail to better understand the above technical solutions. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments of the present invention. It should be understood that the present invention is not limited to the example embodiments for explaining the present invention only. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention. In addition, it should be noted that for the sake of description, only parts related to the present invention rather than all are shown in the drawings.

[0059] Embodiment 1, as Figure 1 This is a flowchart of the GAN-based high-speed aerodynamic heat flow field reconstruction method of the present invention. Among them, the GAN-based high-speed aerodynamic heat flow field reconstruction method includes:

[0060] S100: Collect and obtain a real aerodynamic heat flow field data set, and the real aerodynamic heat flow field data set is historical heat flow field data of a high-speed aircraft under different flight conditions.

[0061] Specifically, the real aerodynamic heat flow field data set refers to the actual historical heat flow field data obtained by measuring a high-speed aircraft under different flight conditions (such as Mach number, angle of attack, altitude, etc.). Such as aerodynamic heat flow field data verified to meet the accuracy requirements obtained through wind tunnel experiments, flight experiments or numerical simulations, which include physical parameters such as temperature, pressure, and heat flow distribution.

[0062] By obtaining the real and accurate real aerodynamic heat flow field data set, it provides a reliable data basis for subsequent training of the GAN-based heat flow field generator and heat flow field discriminator, and further helps to ensure that the generator and discriminator can learn the real heat flow field distribution characteristics.

[0063] S200: Based on the real aerodynamic heat flow field data set, perform feature extraction training and simulation discrimination training to construct a heat flow field generator and a heat flow field discriminator.

[0064] Specifically, the feature extraction training is a process of extracting feature information related to the heat flow field distribution from the obtained real aerodynamic heat flow field dataset through a neural network model, so as to obtain a heat flow field generator with the ability to generate heat flow fields; among them, exemplarily, the target features to be extracted include temperature gradient, flow field pattern, geometric features, etc.

[0065] Specifically, the simulation discrimination training is a process of comparing the simulated heat flow field data generated by the above heat flow field generator with the corresponding real data (and the true values of multiple features), so as to improve the ability of the training discriminator to identify the authenticity of the data. In other words, this simulation discrimination training is used to optimize the performance of the generator and the discriminator.

[0066] Through the above steps of feature extraction training, a heat flow field generator based on a neural network is obtained, reducing the manual adjustment of parameters in the heat flow field research. The simulation discrimination training improves the performance of the generator and the discriminator through an adversarial training mechanism, ensuring that the generated simulated heat flow field data has high authenticity and accuracy, and providing high-quality model support for subsequent heat flow field reconstruction.

[0067] In some embodiments, the construction of the heat flow field generator and the heat flow field discriminator includes:

[0068] Performing Gaussian filtering and data alignment processing on the real aerodynamic heat flow field dataset to obtain an available real aerodynamic heat flow field dataset; extracting associated features of the heat flow field from the available real aerodynamic heat flow field dataset according to the heat flow field reconstruction associated factors to obtain a heat flow field associated feature set; introducing a multi-scale feature fusion module, and using a convolutional neural network and the multi-scale feature fusion module to perform simulation training on the heat flow field associated feature set to construct a heat flow field generator; outputting a simulated heat flow field dataset through the heat flow field generator, and performing heat flow field discrimination training based on the real aerodynamic heat flow field dataset and the simulated heat flow field dataset to generate a heat flow field discriminator.

[0069] Specifically, Gaussian filtering is used to smooth the data through a Gaussian function to remove the noise in the data; the data after Gaussian filtering needs to be aligned to adjust data from different sources or in different formats to a unified coordinate system or time series (i.e., generating an available real aerodynamic heat flow field dataset) to ensure the consistency and comparability of the data.

[0070] Specifically, according to the correlation factors required for the reconstruction of the heat flux field (such as boundary layer thickness, Mach number distribution, surface temperature gradient, etc.), the convolutional layer and pooling layer in the convolutional neural network are used to extract representative heat flux field correlation features at different levels in the available real aerodynamic heat flux field dataset, forming a heat flux field correlation feature set. Among them, through the convolutional neural network that is good at processing data with a grid structure (such as the heat flux field distribution map of the aircraft surface or cross-section), the features of the data can be automatically extracted.

[0071] Specifically, the multi-scale feature fusion module is a network structure used to integrate features at different scales, which can capture the change characteristics of the heat flux field at different spatial scales. By introducing the multi-scale feature fusion module, splicing and fusing features at different scales, the unified expression ability of different scale information (such as local eddy current structure and overall flow field trend) can be enhanced. Furthermore, combining the convolutional neural network (CNN) structure with the multi-scale feature fusion module to perform simulation training on the heat flux field correlation feature set, a heat flux field generator is obtained. This heat flux field generator can output a realistic simulated heat flux field dataset according to the input flight state parameters (including aircraft information, environmental information, flight state information, etc.).

[0072] Further, the simulated heat flux field dataset generated by the heat flux field generator is combined with the real aerodynamic heat flux field dataset, and the heat flux field discriminator is trained through the cross-entropy loss function so that it can distinguish real data from simulated data (evaluating the authenticity and structural consistency of the generated data).

[0073] In the above process, data preprocessing and feature extraction can effectively improve the quality of the input data and reduce the influence of noise and inconsistency on model training. The introduction of the multi-scale feature fusion module enables the generator to capture the complex changes of the heat flux field at different spatial scales, improving the authenticity and accuracy of the generated data, thus providing high-quality model support for subsequent heat flux field reconstruction.

[0074] In some implementation manners, the construction of the heat flux field generator includes:

[0075] According to the convolutional neural network and the multi-scale feature fusion module, a generator network architecture is built. The generator network architecture includes a feature input layer, multiple feature convolutional extraction layers, a multi-scale fusion layer, and a heat flux field output layer; the heat flux field correlation feature set is imported through the feature input layer into the multiple feature convolutional extraction layers for multi-scale feature extraction to obtain multi-scale heat flux field feature data; based on the multi-scale fusion layer, the multi-scale heat flux field feature data is spliced and fused to obtain heat flux field simulation data, which is output as the output result through the heat flux field output layer; based on the heat flux field correlation feature set and the heat flux field simulation data, heat flux field simulation training is performed to construct the heat flux field generator.

[0076] Specifically, to construct a heat flux field generator, first, the network architecture of the generator is designed according to a convolutional neural network and a multi-scale feature fusion module. This architecture includes a feature input layer, multiple feature convolution extraction layers, a multi-scale fusion layer, and a heat flux field output layer. Among them, the feature input layer is used to receive the heat flux field correlation feature set as the input end of the network. The multiple feature convolution extraction layers use various convolution layers (such as convolution kernels of sizes 3×3, 5×5, etc.) to perform multi-scale feature extraction on the input correlation features, refining the local and global features of the heat flux field. The multi-scale fusion layer corresponding to the multi-scale feature fusion module is used to splice and fuse the features extracted by each convolution layer, enabling the model to take into account both local details and overall structure information. Exemplarily, this layer adopts a structure similar to Inception and residual connection, making the gradient propagation more stable and enhancing the information expression ability.

[0077] Optionally, batch normalization and ReLU activation functions are introduced after each convolution channel to enhance the non-linear expression ability.

[0078] Specifically, the above-mentioned network architecture of the generator is simulated and trained by using the heat flux field correlation feature set and the generated simulated data to continuously optimize its parameters, ensuring that the heat flux field generator can generate a simulated heat flux field closer to the real data.

[0079] In some implementation manners, the heat flux field discriminator includes:

[0080] Build a generator network architecture, which includes a data input layer, multiple feature convolution extraction layers, a fully connected layer, and a discriminant output layer; transmit the real aerodynamic heat flux field data set and the simulated heat flux field data set through the data input layer to the multiple feature convolution extraction layers for correlation feature extraction to obtain heat flux field correlation feature data; based on the fully connected layer, perform identification evaluation and judgment on the heat flux field correlation feature data to obtain heat flux field simulation judgment information, and output it as the output result through the discriminant output layer; based on the real aerodynamic heat flux field data set, the simulated heat flux field data set, and the heat flux field simulation judgment information, perform evaluation and judgment training to generate the heat flux field discriminator.

[0081] Optionally, a binary classification neural network model is used to construct a heat flow field discriminator. The structure of the discriminator includes a data input layer, multiple feature convolution extraction layers, a fully connected layer, and a discriminant output layer. Among them, the data input layer is the first layer for receiving input data, providing uniformly formatted data input for subsequent feature extraction; the feature convolution extraction layer extracts local features and correlation information from the input data through convolution operations. The above-mentioned multiple convolution layers are used to capture low-level, intermediate-level, and high-level features in the data respectively, so that the spatial variation, texture, and dynamic characteristics of the heat flow field are fully expressed; the fully connected layer is used to stretch the two-dimensional or three-dimensional feature map extracted by convolution into a one-dimensional vector, realizing feature synthesis and information fusion, facilitating subsequent classification or identification judgment; finally, the discriminant output layer combines the identification evaluation result output by the fully connected layer and outputs a discriminant information with an appropriate activation function (such as Softmax or Sigmoid), which is used to characterize whether the input heat flow field data is "real" or "simulated".

[0082] Specifically, first, the real aerodynamic heat flow field dataset and the simulated heat flow field dataset are input through the data input layer and transmitted to multiple feature convolution extraction layers to perform convolution operations, automatically extracting the correlation features in the heat flow field. For example, the input data includes temperature distribution, flow field vortices, and boundary layer characteristics. A series of feature maps obtained after continuous convolution can reflect the local and global heat flow distribution laws in the input data. Then, the heat flow field correlation feature data extracted by the convolution layer is integrated through the fully connected layer, and the fully connected layer is used to evaluate the identification of the integration result (i.e., the heat flow field simulation judgment information), obtaining the heat flow field simulation judgment information, that is, the probability of judging whether the input data is "true" or "false" (for example, [0, 1]). Finally, this result is output through the discriminant output layer as the final classification result of the input heat flow field data.

[0083] Specifically, based on the real aerodynamic heat flow field dataset, the simulated heat flow field dataset, and the obtained heat flow field simulation judgment information, training samples of the discriminator are constructed. Among them, the simulated heat flow field data output by the heat flow field generator is used as a "pseudo sample", and the real aerodynamic heat flow field data is used as a "positive sample"; then, the discriminator is trained by a supervised learning method, and the parameters of the convolution kernel and the fully connected layer are continuously updated through backpropagation, so that the heat flow field discriminator can distinguish real data from simulated data.

[0084] Optionally, the cross-entropy loss function is used to measure the error between the discriminant output and the true class label, and the gradient descent method is combined for parameter optimization until the discriminant accuracy reaches the expected level.

[0085] In addition, a gradient penalty term (such as the gradient penalty in WGAN-GP) can be introduced to improve the smoothness and physical rationality of the heat flow field data output by the generator.

[0086] S300: Initialize and adaptively adjust the grid according to the distribution change data of the target heat flow field to obtain the heat flow field distribution grid, and divide the reconstruction areas of the target heat flow field based on the heat flow field distribution grid to obtain N hierarchical reconstruction areas.

[0087] Specifically, the target heat flow field refers to the heat flow field data obtained in actual or simulation, including the distribution and variation characteristics of parameters such as temperature, flow rate, and heat flux in space, as well as the local or global characteristics of the heat flow field formed thereby; according to the distribution change data of the target heat flow field, a spatial distribution grid can be preliminarily constructed (i.e., grid initialization), which involves determining the positions, spacings, and basic structural forms of grid nodes and provides a basis for subsequent adaptive adjustment.

[0088] Specifically, the grid density is automatically adjusted according to the local change degree of the target heat flow field (i.e., adaptive adjustment). Denser grids are used in areas where the heat flow field changes violently (with a larger gradient) to capture local details; sparser grids are used in areas with gentle changes (with a smaller gradient), so as to achieve the purpose of balanced data expression and optimized calculation efficiency. The heat flow field distribution grid is the spatial distribution structure obtained after grid initialization and adaptive adjustment.

[0089] Furthermore, based on the heat flow field distribution grid, the target heat flow field is divided into several regions (i.e., N hierarchical reconstruction areas). The data within each region has high homogeneity and similar change trends, and different reconstruction strategies can be adopted to gradually achieve a fine restoration of the heat flow field characteristics.

[0090] In some embodiments, obtaining the heat flow field distribution grid includes:

[0091] Extract the characteristics of the initial distribution characteristics of the target heat flow field to determine the heat flow field change characteristics, where the heat flow field change characteristics include temperature gradient, flow field characteristics, and geometric characteristics; perform grid initialization according to the heat flow field change characteristics to generate an initial uniform grid; calculate the gradient change of each grid cell in the initial uniform grid based on the distribution change data of the target heat flow field according to the heat flow field change characteristics to determine the grid cell change gradient data; perform grid density adaptive adjustment on the initial uniform grid based on the grid cell change gradient data to obtain the heat flow field distribution grid.

[0092] Specifically, the heat flow field change characteristics refer to the change characteristics of the heat flow field in spatial distribution, including temperature gradient (the rate of temperature change, reflecting the non-uniformity of local heat flow), flow field characteristics (such as flow velocity, flow direction, and eddy current distribution), and geometric characteristics (the shape, size, and boundary conditions of the region where the heat flow field is located).

[0093] Specifically, the grid cell change gradient data refers to the gradient values calculated from the distribution change data of the target heat flow field within each grid cell or between adjacent cells in the initial uniform grid, which is used to measure the intensity of the change in the heat flow field in a local area; the heat flow field distribution grid is a grid generated after the initial uniform grid undergoes adaptive adjustment of the grid density. This grid has a higher grid density in areas where the heat flow field changes violently, and a lower grid density in areas where the heat flow field changes gently, in order to improve the overall data representation accuracy and calculation efficiency.

[0094] Specifically, after obtaining the change characteristics of the target heat flow field, an initial uniform grid is generated in the entire target area according to the change characteristics of the target heat flow field and in combination with a preset grid rule. For example, grid cells are uniformly generated in the target area with a preset grid granularity; then, by analyzing the distribution change data of the target heat flow field, the gradient change data within each grid cell or between adjacent cells in the initial uniform grid is calculated; for example, the difference in temperature between adjacent grid cells is calculated by the numerical difference method, and the gradient is estimated in combination with the corresponding flow field and geometric information to obtain the change gradient of each grid cell; furthermore, for grid cells with a change gradient greater than a preset threshold (such as greater than 8 °C / m or a violent change in the corresponding flow field / geometric characteristics), the grid is refined and the cell size is reduced (such as from 0.5 m to 0.2 m), and for grid cells with a lower change gradient, the grid spacing is maintained or appropriately enlarged (such as increased to 0.8 m), so as to reduce the calculation density in the low-change area and achieve a balance between local details and global efficiency.

[0095] In summary, through the extraction of the change characteristics of the target heat flow field, the generation of the initial uniform grid, the calculation of the grid cell gradient, and the adaptive adjustment of the grid density based on the gradient data, a heat flow field distribution grid that accurately reflects the change law of the heat flow field distribution can be obtained, providing a solid data foundation and geometric partition basis for subsequent hierarchical region reconstruction, data optimization, and simulation analysis.

[0096] In some embodiments, obtaining the N hierarchical reconstruction regions includes:

[0097] Determining the grid cell density distribution information according to the heat flow field distribution grid; performing regional clustering analysis on the heat flow field distribution grid using the grid cell density distribution information to obtain a grid region clustering result; setting a reconstruction region segmentation threshold according to the heat flow field reconstruction accuracy requirement; and performing reconstruction region division on the grid region clustering result based on the reconstruction region segmentation threshold to obtain the N hierarchical reconstruction regions.

[0098] Specifically, the grid cell density distribution information refers to the distribution density of each grid cell in the grid describing the heat flow field distribution in space, that is, the number and relative spacing of grid cells in different regions. This information can reflect the characteristics of the heat flow field being highly variable or gentle in a certain region; classifying the heat flow field distribution grid according to the grid cell density distribution information can group grid cells with similar density or local variation characteristics into one category, thereby obtaining the grid region clustering result for region reconstruction.

[0099] Specifically, the reconstruction region segmentation threshold is a predefined quantization index for distinguishing the boundaries of different reconstruction regions, reflecting the required data uniformity within the region and the requirement for capturing local details; based on the reconstruction region segmentation threshold, further analysis can be performed on the grid region clustering result. For example, the preliminary clustering result may obtain 5 major regions. According to the refinement requirements of the preset threshold, some of these categories can be further subdivided into multi-level regions, and finally the total number is N hierarchical regions, with high data uniformity within each region, facilitating subsequent reconstruction.

[0100] S400: Construct a balanced training strategy, and use the balanced training strategy to alternately balance-train the heat flow field generator and the heat flow field discriminator based on the N hierarchical reconstruction regions, and build N heat flow field reconstruction networks.

[0101] Specifically, in the generative adversarial network (GAN), the balanced training strategy is a training scheme adopted to balance the training rates of the generator and the discriminator and stabilize the model convergence. This strategy ensures that each reconstruction network achieves the best performance during the heat flow field data reconstruction process through alternating training, loss function balancing, and multi-region adaptation.

[0102] In some embodiments, the building of the N heat flow field reconstruction networks includes:

[0103] According to the heat flow field generator and the heat flow field discriminator, design the mean square loss function of the generator and the cross-entropy loss function of the discriminator; perform a balance coefficient analysis on the N hierarchical reconstruction regions based on the balanced training strategy to determine N initial region training balance coefficients; use the mean square loss function of the generator and the cross-entropy loss function of the discriminator to alternately balance-train the heat flow field generator and the heat flow field discriminator based on the N initial region training balance coefficients to obtain N heat flow field reconstruction networks.

[0104] Specifically, the mean square loss function of the generator is used to measure the difference between the simulated heat flow field data output by the generator and the real heat flow field data, and the cross-entropy loss function of the discriminator is used to measure the degree of classification error of the discriminator for real data and generated data.

[0105] Specifically, the balance coefficient refers to the weight coefficient applied to each hierarchical reconstruction region in the balanced training strategy, which is used to adjust the training speed of the generator and the discriminator in different regions to ensure that the two are in a coordinated and optimized state during the alternating training process. Optionally, the balance coefficient required for the initial training of each region can be determined by statistically analyzing the data change characteristics, error distribution and network convergence in each region. For example, in areas with drastic changes, a larger balance coefficient may be required to speed up the learning of the generator and ensure that the generated results have fine details; while in areas with gentle changes, the balance coefficient is appropriately reduced to prevent overtraining.

[0106] Specifically, the designed generator mean square loss function and discriminator cross entropy loss function are used, combined with the obtained initial balance coefficients of each region, to perform alternating balanced training on the thermal flow field generator and the thermal flow field discriminator; this includes dynamically adjusting the training step size and weight for each hierarchical reconstruction region according to the preset balance coefficient, first using the generator mean square loss function to train the generator for several rounds, and then using the discriminator cross entropy loss function to train the discriminator, repeating the alternating training until the generator and discriminator reach a convergent equilibrium state in each region, and generating the thermal flow field reconstruction network for each region respectively.

[0107] In the above process, the mean square error loss of the generator and the cross entropy loss of the discriminator are combined with the balance coefficient adjustment to help the networks in different regions to achieve smooth convergence in alternating training and avoid training imbalance or mode collapse caused by large differences in local data characteristics; the thermal flow field reconstruction network after training in each region is integrated to achieve the continuity and consistency of the overall thermal flow field while fully restoring the local details.

[0108] In some implementations, obtaining N thermal flow field reconstruction networks includes:

[0109] Based on the N initial region training balance coefficients, the thermal flow field generator and the thermal flow field discriminator are respectively subjected to simulated discriminant training to obtain N initial generators and N initial discriminators; the N initial generators and N initial discriminators are subjected to loss evaluation using the generator mean square loss function and the discriminator cross entropy loss function to obtain N generator loss data and N discriminator loss data; based on the N generator loss data and the N discriminator loss data, the N initial generators and the N initial discriminators are subjected to balanced alternating training to obtain N thermal flow field reconstruction networks.

[0110] Specifically, first, based on the training of the balance coefficient for N initial regions, the heat flux field generators and discriminators in each region are respectively simulated and discriminatively trained. The obtained initial generator can output a heat flux field pattern similar to the real data, while the initial discriminator can preliminarily judge the authenticity of the output data. Then, the obtained N initial generators and N initial discriminators are respectively evaluated for losses using the mean square loss function of the generator and the cross-entropy loss function of the discriminator, and the corresponding generator loss data (mean square error value) and discriminator loss data (cross-entropy error) are obtained. These data reflect the accuracy and stability of the initial training state of the network in each region.

[0111] Furthermore, based on the above-mentioned N generator loss data and N discriminator loss data, an alternating update strategy is adopted to perform balanced alternating training on the initial generators and initial discriminators in each region. That is, first fix the discriminator and continue to train the generator using the mean square loss function of the generator (i.e., the number of training iterations is determined based on the generator loss data) to make its output closer to the real data. Subsequently, fix the generator and then train the discriminator using the cross-entropy loss function of the discriminator (i.e., the number of training iterations is determined based on the discriminator loss data) to improve its discrimination accuracy, so that the generator and the discriminator can achieve the balance and stability of the training process while their respective losses drop to the predetermined target values.

[0112] In some embodiments, obtaining the N heat flux field reconstruction networks includes:

[0113] Based on the N generator loss data and N discriminator loss data, dynamically adjust the training balance coefficients of the N initial regions to obtain N region training balance dynamic coefficients; perform iterative balance updates on the N initial generators and N initial discriminators according to the N region training balance dynamic coefficients to obtain N region generators and N region discriminators; concatenate and merge the N region generators and N region discriminators to obtain the N heat flux field reconstruction networks.

[0114] Optionally, to obtain the foregoing N generator loss data and N discriminator loss data, methods such as weighted average and gradient change rate analysis are used to dynamically adjust the original region training balance coefficients. For example, if the generator loss is significantly higher than the discriminator loss, the training weight of the generator in this region is increased to give it priority for optimization and adjustment, and vice versa.

[0115] Furthermore, according to the adjusted region training balance dynamic coefficients, iterative balance updates are performed on the N initial generators and N initial discriminators. Exemplarily, if the dynamic balance coefficients of a certain region are 0.6 (generator) and 0.4 (discriminator), then more training resources or rounds will be allocated for optimizing the generator parameters in each round of alternating training, so as to effectively compensate for its insufficient accuracy.

[0116] Finally, the updated N region generators and N region discriminators are concatenated and merged to form N independent heat flux field reconstruction networks. Each network corresponds to a hierarchical reconstruction region and can efficiently reconstruct the heat flux field characteristics of that region. Among them, the merging methods can include output-input connection (i.e., the generation result of a certain region is used as the input of the next region), feature fusion (the generated features of multiple regions are spliced in a shared layer), or discriminant cooperation (the results of multiple discriminators are jointly used to inversely guide the generator).

[0117] S500: Monitor and obtain flight condition update data, and respectively perform heat flux field reconstruction optimization on the flight condition update data based on the N heat flux field reconstruction networks to generate a high-speed aerodynamic simulation heat flux field.

[0118] Specifically, the flight condition update data is a set of real-time state parameters of the aircraft during flight (such as airspeed, angle of attack, Mach number, atmospheric density, etc.), which affects the distribution of the heat flux field; through the flight condition update data, a new and high-precision heat flux field distribution map can be re-output to compensate for the deficiencies of traditional numerical simulations in terms of real-time performance or resolution.

[0119] In summary, the GAN-based high-speed aerodynamic heat flux field reconstruction method provided by the present invention has the following technical effects:

[0120] By collecting and obtaining historical heat flux field data of a high-speed aircraft under different flight conditions, a real aerodynamic heat flux field data set is formed; subsequently, based on this data set, feature extraction training and simulation discrimination training are carried out to construct a heat flux field generator and a discriminator; on this basis, according to the distribution change data of the target heat flux field, grid initialization and adaptive adjustment are performed to generate a heat flux field distribution grid, and accordingly the target heat flux field is divided into N hierarchical reconstruction regions; further, a balanced training strategy is constructed, and this strategy is used to alternately balance-train the heat flux field generator and the discriminator in the N hierarchical reconstruction regions to build N heat flux field reconstruction networks; finally, monitor and obtain flight condition update data, and respectively perform heat flux field reconstruction and optimization on the update data based on the N heat flux field reconstruction networks, so as to generate a high-speed aerodynamic simulation heat flux field, thereby achieving the technical effects of simplifying parameter configuration, reducing process costs, and improving timeliness.

[0121] Embodiment 2, as Figure 2 is a schematic structural diagram of the GAN-based high-speed aerodynamic heat flux field reconstruction system of the present invention. For example, Figure 1 The flow schematic diagram of the GAN-based high-speed aerodynamic heat flux field reconstruction method in the present invention can be implemented through a structure such as Figure 2 shown.

[0122] Based on the same concept as the GAN-based high-speed aerodynamic heat flow field reconstruction method in the above embodiments, the GAN-based high-speed aerodynamic heat flow field reconstruction system provided by the present invention includes:

[0123] A real data acquisition component 11, configured to acquire a real aerodynamic heat flow field data set, where the real aerodynamic heat flow field data set is historical heat flow field data of a high-speed aircraft under different flight conditions.

[0124] A tool training component 12, configured to perform feature extraction training and simulation discrimination training based on the real aerodynamic heat flow field data set, and construct a heat flow field generator and a heat flow field discriminator.

[0125] A mesh generation component 13, configured to perform mesh initialization and adaptive adjustment according to the distribution change data of the target heat flow field to obtain a heat flow field distribution mesh, and perform reconstruction area division on the target heat flow field based on the heat flow field distribution mesh to obtain N hierarchical reconstruction areas.

[0126] A balance training component 14, configured to construct a balance training strategy, and perform alternating balance training on the heat flow field generator and the heat flow field discriminator based on the N hierarchical reconstruction areas by using the balance training strategy, and build N heat flow field reconstruction networks.

[0127] An update and optimization component 15, configured to monitor and obtain flight condition update data, and perform heat flow field reconstruction optimization on the flight condition update data respectively based on the N heat flow field reconstruction networks to generate a high-speed aerodynamic simulation heat flow field.

[0128] In some embodiments, the tool training component 12 includes:

[0129] A real aerodynamic heat flow field data preprocessing unit, configured to perform Gaussian filtering and data alignment processing on the real aerodynamic heat flow field data set to obtain an available real aerodynamic heat flow field data set.

[0130] A heat flow field correlation feature extraction unit, configured to perform correlation feature extraction on the available real aerodynamic heat flow field data set according to heat flow field reconstruction correlation factors to obtain a heat flow field correlation feature set.

[0131] A heat flow field generator construction unit, configured to introduce a multi-scale feature fusion module, and perform simulation training on the heat flow field correlation feature set by using a convolutional neural network and the multi-scale feature fusion module to construct a heat flow field generator.

[0132] A heat flow field discriminator generation unit, configured to output a simulated heat flow field data set through the heat flow field generator, and perform heat flow field discrimination training based on the real aerodynamic heat flow field data set and the simulated heat flow field data set to generate a heat flow field discriminator.

[0133] In some implementations, the heat flow field generator construction unit in the tool training component 12 includes:

[0134] A generator network architecture construction unit, configured to construct a generator network architecture according to a convolutional neural network and the multi-scale feature fusion module, where the generator network architecture includes a feature input layer, a plurality of feature convolution extraction layers, a multi-scale fusion layer, and a heat flow field output layer.

[0135] A multi-scale feature extraction unit, configured to import the heat flow field correlation feature set through the feature input layer into the plurality of feature convolution extraction layers for multi-scale feature extraction, to obtain multi-scale heat flow field feature data.

[0136] A heat flow field simulation data generation unit, configured to perform splicing and fusion on the multi-scale heat flow field feature data based on the multi-scale fusion layer, to obtain heat flow field simulation data, and output the heat flow field simulation data as an output result through the heat flow field output layer.

[0137] A heat flow field generator construction unit, configured to perform heat flow field simulation training based on the heat flow field correlation feature set and the heat flow field simulation data, to construct the heat flow field generator.

[0138] In some implementations, the heat flow field discriminator generation unit in the tool training component 12 includes:

[0139] A generator network architecture construction unit, configured to construct a generator network architecture, where the generator network architecture includes a data input layer, a plurality of feature convolution extraction layers, a fully connected layer, and a discriminant output layer.

[0140] An associated feature extraction unit, configured to transmit the real aerodynamic heat flow field data set and the simulated heat flow field data set through the data input layer to the plurality of feature convolution extraction layers for associated feature extraction, to obtain heat flow field associated feature data.

[0141] A heat flow field simulation judgment information generation unit, configured to perform identification evaluation and judgment on the heat flow field associated feature data based on the fully connected layer, to obtain heat flow field simulation judgment information, and output the heat flow field simulation judgment information as an output result through the discriminant output layer.

[0142] A heat flow field discriminator generation unit, configured to perform evaluation and judgment training based on the real aerodynamic heat flow field data set, the simulated heat flow field data set, and the heat flow field simulation judgment information, to generate the heat flow field discriminator.

[0143] In some embodiments, the mesh division component 13 includes:

[0144] A heat flow field change feature extraction unit is used to extract features from the initial distribution characteristics of the target heat flow field and determine the heat flow field change features, where the heat flow field change features include temperature gradient, flow field features, and geometric features.

[0145] An initial uniform grid generation unit is used to initialize the grid according to the heat flow field change features and generate an initial uniform grid.

[0146] A grid cell change gradient calculation unit is used to calculate the gradient change of each grid cell in the initial uniform grid based on the distribution change data of the target heat flow field according to the heat flow field change features and determine the grid cell change gradient data.

[0147] A heat flow field distribution grid generation unit is used to adaptively adjust the grid density of the initial uniform grid based on the grid cell change gradient data to obtain the heat flow field distribution grid.

[0148] In some embodiments, the grid division component 13 further includes:

[0149] A grid cell density distribution information determination unit is used to determine the grid cell density distribution information according to the heat flow field distribution grid.

[0150] A grid region clustering analysis unit is used to perform region clustering analysis on the heat flow field distribution grid using the grid cell density distribution information to obtain a grid region clustering result.

[0151] A reconstruction region segmentation threshold setting unit is used to set a reconstruction region segmentation threshold according to the heat flow field reconstruction accuracy requirement.

[0152] A hierarchical reconstruction region division unit is used to divide the grid region clustering result based on the reconstruction region segmentation threshold to obtain the N hierarchical reconstruction regions.

[0153] In some embodiments, the balance training component 14 includes:

[0154] A loss function design unit is used to design a generator mean square loss function and a discriminator cross-entropy loss function according to the heat flow field generator and the heat flow field discriminator.

[0155] An initial region training balance coefficient determination unit is used to analyze the balance coefficient of the N hierarchical reconstruction regions based on the balance training strategy and determine N initial region training balance coefficients.

[0156] The heat flow field reconstruction network training unit is used to alternately balance-train the heat flow field generator and the heat flow field discriminator based on the training balance coefficients of the N initial regions by using the mean square loss function of the generator and the cross-entropy loss function of the discriminator, so as to obtain N heat flow field reconstruction networks.

[0157] In some implementation manners, the heat flow field reconstruction network training unit in the balance training component 14 includes:

[0158] The initial generator and discriminator training unit is used to perform simulated discrimination training on the heat flow field generator and the heat flow field discriminator respectively based on the training balance coefficients of the N initial regions, so as to obtain N initial generators and N initial discriminators.

[0159] The generator and discriminator loss evaluation unit is used to evaluate the losses of the N initial generators and N initial discriminators by using the mean square loss function of the generator and the cross-entropy loss function of the discriminator, so as to obtain N generator loss data and N discriminator loss data.

[0160] The heat flow field reconstruction network optimization unit is used to perform balanced alternating training on the N initial generators and N initial discriminators based on the N generator loss data and N discriminator loss data, so as to obtain N heat flow field reconstruction networks.

[0161] In some implementation manners, the heat flow field reconstruction network optimization unit in the balance training component 14 further includes:

[0162] The region training balance dynamic coefficient adjustment unit is used to dynamically adjust the N initial region training balance coefficients based on the N generator loss data and N discriminator loss data, so as to obtain N region training balance dynamic coefficients.

[0163] The region generator and discriminator iterative update unit is used to perform iterative balance update on the N initial generators and N initial discriminators according to the N region training balance dynamic coefficients, so as to obtain N region generators and N region discriminators.

[0164] The heat flow field reconstruction network obtaining unit is used to concatenate and merge the N region generators and N region discriminators to obtain the N heat flow field reconstruction networks.

[0165] It should be understood that the embodiments mentioned in this specification focus on their differences from other embodiments. The specific embodiments in the foregoing Embodiment 1 are equally applicable to the GAN-based high-speed aerodynamic heat flow field reconstruction system described in Embodiment 2. For the sake of simplicity of the specification, no further elaboration is made here.

[0166] It should be understood that the disclosed embodiments of the present invention and the above descriptions enable those skilled in the art to implement the present invention. At the same time, the present invention is not limited to the above-mentioned part of the embodiments. It should be understood that ordinary skilled in the art can still modify the technical solutions recorded in the foregoing embodiments or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A high-speed aerodynamic heat flow field reconstruction method based on GAN, characterized in that Including: Collect and obtain a real aerodynamic heat flux field data set, where the real aerodynamic heat flux field data set is historical heat flux field data of a high-speed aircraft under different flight conditions; Based on the real aerodynamic heat flux field data set, perform feature extraction training and simulation discrimination training to construct a heat flux field generator and a heat flux field discriminator; According to the distribution change data of the target heat flux field, perform grid initialization and adaptive adjustment to obtain a heat flux field distribution grid, and based on the heat flux field distribution grid, divide the reconstruction area of the target heat flux field to obtain N hierarchical reconstruction areas; Construct a balanced training strategy, and use the balanced training strategy to alternately balance-train the heat flux field generator and the heat flux field discriminator based on the N hierarchical reconstruction areas to build N heat flux field reconstruction networks; Monitor and obtain flight condition update data, and respectively perform heat flux field reconstruction optimization on the flight condition update data based on the N heat flux field reconstruction networks to generate a high-speed aerodynamic simulation heat flux field; The constructing of the heat flux field generator and the heat flux field discriminator includes: Perform Gaussian filtering and data alignment processing on the real aerodynamic heat flux field data set to obtain an available real aerodynamic heat flux field data set; Extract associated features from the available real aerodynamic heat flux field data set according to the heat flux field reconstruction associated factors to obtain a heat flux field associated feature set; Introduce a multi-scale feature fusion module, and use a convolutional neural network and the multi-scale feature fusion module to perform simulation training on the heat flux field associated feature set to construct a heat flux field generator; Output a simulated heat flux field data set through the heat flux field generator, and perform heat flux field discrimination training based on the real aerodynamic heat flux field data set and the simulated heat flux field data set to generate a heat flux field discriminator; The obtaining of the heat flux field distribution grid includes: Extract features from the initial distribution characteristics of the target heat flux field to determine the heat flux field change characteristics, where the heat flux field change characteristics include temperature gradient, flow field characteristics, and geometric characteristics; Perform grid initialization according to the heat flux field change characteristics to generate an initial uniform grid; Calculate the gradient change of each grid cell in the initial uniform grid according to the heat flux field change characteristics based on the distribution change data of the target heat flux field to determine the grid cell change gradient data; Perform grid density adaptive adjustment on the initial uniform grid based on the grid cell change gradient data to obtain the heat flux field distribution grid; The building of N heat flux field reconstruction networks includes: According to the heat flux field generator and the heat flux field discriminator, design a generator mean square loss function and a discriminator cross-entropy loss function; Perform balance coefficient analysis on the N hierarchical reconstruction areas based on the balanced training strategy to determine N initial area training balance coefficients; Use the generator mean square loss function and the discriminator cross-entropy loss function to alternately balance-train the heat flux field generator and the heat flux field discriminator based on the N initial area training balance coefficients to obtain N heat flux field reconstruction networks.

2. The method for reconstructing a high-speed aerodynamic heat flow field based on GAN according to claim 1, wherein The constructing of the heat flux field generator includes: Build a generator network architecture based on the convolutional neural network and the multi-scale feature fusion module. The generator network architecture includes a feature input layer, multiple feature convolution extraction layers, a multi-scale fusion layer, and a heat flow field output layer; Pass the heat flow field correlation feature set through the feature input layer and import it into the multiple feature convolution extraction layers for multi-scale feature extraction to obtain multi-scale heat flow field feature data; Based on the multi-scale fusion layer, splice and fuse the multi-scale heat flow field feature data to obtain heat flow field simulation data, and output it as the output result through the heat flow field output layer; Based on the heat flow field correlation feature set and the heat flow field simulation data, perform heat flow field simulation training to construct the heat flow field generator.

3. The GAN-based high-speed aerodynamic heat flow field reconstruction method according to claim 1, characterized in that The generated heat flow field discriminator includes: Build a generator network architecture. The generator network architecture includes a data input layer, multiple feature convolution extraction layers, a fully connected layer, and a discriminant output layer; Pass the real aerodynamic heat flow field data set and the simulated heat flow field data set through the data input layer and transmit them to the multiple feature convolution extraction layers for correlation feature extraction to obtain heat flow field correlation feature data; Based on the fully connected layer, perform identification evaluation and judgment on the heat flow field correlation feature data to obtain heat flow field simulation judgment information, and output it as the output result through the discriminant output layer; Based on the real aerodynamic heat flow field data set, the simulated heat flow field data set, and the heat flow field simulation judgment information, perform evaluation and judgment training to generate the heat flow field discriminator.

4. The method for reconstructing a high-speed aerodynamic heat flow field based on GAN according to claim 1, wherein The obtaining of N hierarchical reconstruction regions includes: According to the heat flow field distribution grid, determine the grid cell density distribution information; Use the grid cell density distribution information to perform regional clustering analysis on the heat flow field distribution grid to obtain a grid region clustering result; According to the heat flow field reconstruction accuracy requirement, set a reconstruction region segmentation threshold; Based on the reconstruction region segmentation threshold, divide the grid region clustering result into reconstruction regions to obtain the N hierarchical reconstruction regions.

5. The method for reconstructing high-speed aerodynamic heat flow field based on GAN according to claim 1, wherein The obtaining of N heat flow field reconstruction networks includes: Based on the N initial region training balance coefficients, perform simulation discrimination training on the heat flow field generator and the heat flow field discriminator respectively to obtain N initial generators and N initial discriminators; Use the generator mean square loss function and the discriminator cross-entropy loss function to evaluate the losses of the N initial generators and N initial discriminators to obtain N generator loss data and N discriminator loss data; Based on the N generator loss data and N discriminator loss data, perform balanced alternating training on the N initial generators and N initial discriminators to obtain N heat flow field reconstruction networks.

6. The GAN-based high-speed aerodynamic heat flow field reconstruction method according to claim 5, wherein The obtaining of N heat flow field reconstruction networks includes: Based on the N generator loss data and N discriminator loss data, dynamically adjust the N initial region training balance coefficients to obtain N region training balance dynamic coefficients; According to the N region training balance dynamic coefficients, perform iterative balance updates on the N initial generators and N initial discriminators to obtain N region generators and N region discriminators; The N region generators and N region discriminators are connected in series and merged to obtain the N heat flow field reconstruction networks.

7. The GAN-based high-speed aerodynamic heat flow field reconstruction system is characterized in that For implementing the GAN-based high-speed aerodynamic heat flow field reconstruction method according to any one of claims 1-6, including: A real data acquisition component for acquiring a real aerodynamic heat flow field data set, where the real aerodynamic heat flow field data set is historical heat flow field data of a high-speed aircraft under different flight conditions; A tool training component for performing feature extraction training and simulation discrimination training based on the real aerodynamic heat flow field data set to construct a heat flow field generator and a heat flow field discriminator; A grid division component for initializing and adaptively adjusting a grid according to the distribution change data of a target heat flow field to obtain a heat flow field distribution grid, and dividing the target heat flow field into N hierarchical reconstruction regions based on the heat flow field distribution grid; A balance training component for constructing a balance training strategy and alternately balancing the heat flow field generator and the heat flow field discriminator based on the N hierarchical reconstruction regions by using the balance training strategy to build N heat flow field reconstruction networks; An update and optimization component for monitoring and obtaining flight condition update data, and respectively optimizing the heat flow field reconstruction of the flight condition update data based on the N heat flow field reconstruction networks to generate a high-speed aerodynamic simulation heat flow field.

Images