CN116432556A - Wing surface pressure reconstruction method, electronic equipment and storage medium - Google Patents
Wing surface pressure reconstruction method, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN116432556A CN116432556A CN202310432279.XA CN202310432279A CN116432556A CN 116432556 A CN116432556 A CN 116432556A CN 202310432279 A CN202310432279 A CN 202310432279A CN 116432556 A CN116432556 A CN 116432556A
- Authority
- CN
- China
- Prior art keywords
- wing
- model
- pressure
- surface pressure
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 52
- 238000003860 storage Methods 0.000 title claims abstract description 13
- 238000012360 testing method Methods 0.000 claims abstract description 68
- 238000003062 neural network model Methods 0.000 claims abstract description 50
- 238000009530 blood pressure measurement Methods 0.000 claims abstract description 31
- 238000009826 distribution Methods 0.000 claims abstract description 28
- 238000005457 optimization Methods 0.000 claims abstract description 23
- 238000004088 simulation Methods 0.000 claims abstract description 23
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 21
- 238000012549 training Methods 0.000 claims abstract description 21
- 239000002245 particle Substances 0.000 claims abstract description 17
- 238000013528 artificial neural network Methods 0.000 claims abstract description 16
- 230000006870 function Effects 0.000 claims description 35
- 238000004364 calculation method Methods 0.000 claims description 32
- 238000004590 computer program Methods 0.000 claims description 15
- 210000002569 neuron Anatomy 0.000 claims description 9
- 238000011176 pooling Methods 0.000 claims description 9
- 238000005070 sampling Methods 0.000 claims description 7
- 238000007781 pre-processing Methods 0.000 claims description 5
- 238000013527 convolutional neural network Methods 0.000 claims description 4
- 230000000694 effects Effects 0.000 claims description 4
- 230000004927 fusion Effects 0.000 claims description 4
- 238000012986 modification Methods 0.000 claims description 4
- 230000004048 modification Effects 0.000 claims description 4
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 claims description 3
- 230000003213 activating effect Effects 0.000 claims description 3
- 238000009795 derivation Methods 0.000 claims description 3
- 238000001514 detection method Methods 0.000 claims description 3
- 238000011156 evaluation Methods 0.000 claims description 3
- 238000012795 verification Methods 0.000 claims description 3
- 238000012800 visualization Methods 0.000 claims description 3
- 238000012937 correction Methods 0.000 claims 1
- 238000005259 measurement Methods 0.000 abstract description 5
- 230000008569 process Effects 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 5
- 230000009471 action Effects 0.000 description 3
- 230000014509 gene expression Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 239000012530 fluid Substances 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/28—Design optimisation, verification or simulation using fluid dynamics, e.g. using Navier-Stokes equations or computational fluid dynamics [CFD]
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B64—AIRCRAFT; AVIATION; COSMONAUTICS
- B64F—GROUND OR AIRCRAFT-CARRIER-DECK INSTALLATIONS SPECIALLY ADAPTED FOR USE IN CONNECTION WITH AIRCRAFT; DESIGNING, MANUFACTURING, ASSEMBLING, CLEANING, MAINTAINING OR REPAIRING AIRCRAFT, NOT OTHERWISE PROVIDED FOR; HANDLING, TRANSPORTING, TESTING OR INSPECTING AIRCRAFT COMPONENTS, NOT OTHERWISE PROVIDED FOR
- B64F5/00—Designing, manufacturing, assembling, cleaning, maintaining or repairing aircraft, not otherwise provided for; Handling, transporting, testing or inspecting aircraft components, not otherwise provided for
- B64F5/60—Testing or inspecting aircraft components or systems
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M9/00—Aerodynamic testing; Arrangements in or on wind tunnels
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M9/00—Aerodynamic testing; Arrangements in or on wind tunnels
- G01M9/06—Measuring arrangements specially adapted for aerodynamic testing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/23—Design optimisation, verification or simulation using finite element methods [FEM] or finite difference methods [FDM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/048—Activation functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
- G06T17/20—Finite element generation, e.g. wire-frame surface description, tesselation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/10—Numerical modelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/28—Fuselage, exterior or interior
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Geometry (AREA)
- Data Mining & Analysis (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Computer Hardware Design (AREA)
- Fluid Mechanics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Mathematical Analysis (AREA)
- Algebra (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Computer Graphics (AREA)
- Manufacturing & Machinery (AREA)
- Transportation (AREA)
- Aviation & Aerospace Engineering (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
A wing surface pressure reconstruction method, electronic equipment and storage medium belong to the technical field of wind tunnel tests. The method aims to solve the problems of high efficiency and accuracy of the wind tunnel pressure measurement test. According to the invention, the original pressure data of the wing surface is collected through a three-dimensional wing wind tunnel pressure measurement test and a three-dimensional wing model numerical simulation and preprocessed, and a deep neural network data set for reconstructing the wing surface pressure is generated; the constructed wing surface pressure reconstruction depth neural network model is subjected to a training test by modifying a loss function, fusing a wind tunnel test and a data set with two sources of numerical simulation, and optimizing model super-parameters by adopting a particle swarm optimization algorithm to obtain an optimized depth neural network model; the method is used for a model wind tunnel pressure measurement test, reconstructing holographic pressure distribution on the surface of the wing, predicting pneumatic load distribution data of non-measurement points on the surface of the wing, and evaluating and verifying the predicted holographic pressure distribution data. The invention can be used for conventional pressure measurement test of complex aircraft.
Description
Technical Field
The invention belongs to the technical field of wind tunnel tests, and particularly relates to a wing surface pressure reconstruction method, electronic equipment and a storage medium.
Background
Aerodynamic studies are critical to aircraft aerodynamic feature estimation at various stages of aircraft design. The pressure distribution on the surface of each part of the aircraft provides the original data of aerodynamic load distribution for the calculation of the structural strength of the aircraft and each part, and provides a basis for researching the performance of the aircraft and each part and the flow-around characteristic of a research model, so that the method is a very important ring for aerodynamic research. The main ways to obtain the pressure distribution on the surfaces of various parts of an aircraft are: wind tunnel manometry and computational fluid dynamics numerical simulation calculation methods (CFD, computational Fluid Dynamics), but both methods have drawbacks and disadvantages.
The wind tunnel pressure measurement test technology is one of the conventional test technology capabilities of the production type wind tunnel, is a mark technology for measuring whether the wind tunnel test technology is mature or not, and is a representation of the test capability construction level of the wind tunnel. The reliability of the wind tunnel test is high, and the obtained aerodynamic force/load result is often used as a standard for checking the accuracy of the numerical simulation method. However, because the period of the wind tunnel test is longer, the experimental dependence on the testers is higher, and the rationality of the test scheme directly influences the acquisition efficiency and effect of the pneumatic load. The existing engineering practice generally considers that at least 50-100 pressure measuring holes are required to be arranged on the surface of the airfoil, and the precision of the lift force and pitching moment coefficients obtained through pressure distribution integration is more reliable. In order to obtain the complete flow field information of the airfoil surface, the traditional method generally arranges enough pressure taps on the airfoil surface for wind tunnel test, and more pressure taps are needed for obtaining the pressure distribution of the airfoil full surface through simple interpolation reconstruction. For complex aircrafts, the space position and the test cost are limited, and the acquisition of pressure measurement data is insufficient, so that the precision of the traditional method is insufficient; and the aerodynamic load of the transonic wind tunnel test of the complex aircraft is sensitive to the influence of parameters, and the measurement difficulty of the refined aerodynamic force/load is larger and the period is longer. The numerical simulation calculation has the characteristics of simple implementation, convenience, flexibility and the like, but because the physical model is ambiguous, the simulation of complex flow often has larger access to the real result, and the accuracy consistent with the wind tunnel test cannot be achieved.
Disclosure of Invention
The invention aims to solve the problems that the conventional pressure measurement test is limited by space position and test cost, the pressure measurement holes are difficult to be arranged on the surface of a complex model to obtain complete surface pressure distribution information, and the lift force and moment accuracy obtained by direct integration are not enough; the numerical simulation method is characterized in that the simulation of complex flow often has a certain access to a real result because of the ambiguity of a physical model, and the rule cannot be simply applied to wind tunnel test data.
In order to achieve the above purpose, the present invention is realized by the following technical scheme:
a method of wing surface pressure reconstruction comprising the steps of:
s1, acquiring original pressure data of a wing surface through a three-dimensional wing wind tunnel pressure measurement test and a three-dimensional wing model numerical simulation, preprocessing the obtained original pressure data of the wing surface, and constructing a depth neural network data set for reconstructing the pressure of the wing surface;
s2, constructing a wing surface pressure reconstruction depth neural network model, realizing fusion of two source data samples, namely the test data and the numerical simulation data, in the depth neural network data set for wing surface pressure reconstruction obtained in the step S1 through modification of a loss function, and performing training test on the constructed wing surface pressure reconstruction depth neural network model;
s3, optimizing the super parameters of the wing surface pressure reconstruction depth neural network model constructed in the step S2 by adopting a particle swarm optimization algorithm to obtain an optimized wing surface pressure reconstruction depth neural network model;
s4, utilizing the optimized wing surface pressure reconstruction depth neural network model obtained in the step S3 to be applied to a new aircraft model wind tunnel pressure measurement test, reconstructing holographic pressure distribution of the wing surface of the model, predicting pneumatic load distribution data of non-measuring points of the wing surface of the model, and evaluating and verifying the predicted holographic pressure distribution data.
Further, the specific implementation method of the step S1 includes the following steps:
s1.1, acquiring wing surface original pressure data including Mach number, attack angle, sideslip angle, total pressure, reynolds number and surface pressure coefficient through three-dimensional wing wind tunnel pressure measurement test and three-dimensional wing model numerical simulation calculation;
s1.2, setting the expected precision of the three-dimensional wing model to be 57-81 of the spanwise nodes of the three-dimensional wing, and 161-241 of the chordwise nodes of the three-dimensional wing, and carrying out grid division according to the expected precision of the three-dimensional wing model to obtain grid nodes of the three-dimensional wing model, wherein the grid nodes are used as holographic grid nodes of the three-dimensional wing model;
s1.3, arranging the holographic grid nodes of the three-dimensional wing model obtained in the step S1.2 according to the grid nodes of the three-dimensional wing model in rows and columns to obtain holographic grid nodes of the unfolded three-dimensional wing model;
s1.4, sampling the original pressure data of the wing surface obtained in the step S1.1 according to the holographic grid node coordinate information data of the unfolded three-dimensional wing model obtained in the step S1.3, and obtaining the holographic grid node pressure data of the three-dimensional wing model through interpolation; acquiring two-dimensional unfolding coordinate information data of holographic grid nodes corresponding to the three-dimensional wing model according to the positions of the pressure points of the wind tunnel pressure test;
s1.5, constructing a deep neural network data set for reconstructing the surface pressure of the wing, wherein the deep neural network data set comprises the following components: the data of the depth neural network data set for reconstructing the surface pressure of the wing comprises working condition state parameters, expected precision grid point data, pressure measurement test original surface pressure data, position data of holographic grid nodes of the three-dimensional wing model corresponding to the pressure measurement test original surface pressure data and numerical simulation surface pressure data of holographic grid nodes of the three-dimensional wing model corresponding to the pressure measurement test original surface pressure data.
Further, the specific implementation method of the step S2 includes the following steps:
s2.1, reconstructing a depth neural network model by using wing surface pressure, wherein a two-dimensional U-shaped convolutional neural network structure is adopted, and the model comprises a convolutional layer, a pooling layer and a deconvolution layer, wherein the convolutional layer has a convolutional kernel size of 3 multiplied by 3; the pooling layer is responsible for downsampling the space dimension of the input data, setting 2 multiplied by 2 receptive fields as maximum pooling, and setting the sliding step length as 2; up-sampling is carried out with the convolution kernel size of the deconvolution layer being 2×2;
s2.2, setting the learning rate, the iteration times, the layer number, the number of neurons of each layer and the batch size of the wing surface pressure reconstruction depth neural network model, and activating a function toThe root mean square error is used as a loss function for training
The calculation formula of (2) is as follows:
wherein,,
for loss function->
Is a weight parameter, < >>
For biasing item parameters, ++>
For training sample number, ++>
And->
For the known sample data, +.>
For model predictive value, +.>
For model predictive function +.>
For input value +.>
A forward propagation function formula;
the training adopts a back propagation algorithm, a chain derivative algorithm is utilized to calculate the partial derivative of a loss function between known sample data and a model predicted value on each weight parameter or bias term, then the weight parameter or bias term parameter is reversely updated layer by layer, and the calculation formula of the gradient descent algorithm of the parameters in the wing surface pressure reconstruction depth neural network model is as follows:
wherein,,
is the firstlLayer node->
To the firstl-layer 1 node->
Weight of->
Is->
Layer node->
Bias item of->
For gradient operator->
Is the learning rate;
solving the loss function according to the chained derivation rule
The derivative for each weight or bias term is calculated as:
wherein,,
、/>
is->
Layer node->
Output value of->
Layer node->
Output value of>
Is->
Layer node->
Through->
Calculated result,/->
Equal to->
；
To make the error of each neuron
The calculation formula of (2) is as follows:
substituting the formula (6) into the formula (4) and the formula (5) can obtain:
the calculation formula of the output layer is as follows:
bringing formula (9) into formula (6), the calculation formula for the error of each neuron is:
bringing the formula (10) into the formula (7) and the formula (8) can calculate gradient update values of weights and bias terms in the output layer;
the calculation formula of the hidden layer is obtained according to the complex function derivative formula, and the calculation formula is as follows:
bringing equation (11) into equation (6) yields:
carrying the formula (12) into the formula (7) and the formula (8) to calculate gradient update values of weights and bias terms in the hidden layer; model training is completed through iterative optimization;
s2.3, calibrating the wing surface pressure reconstruction depth neural network model obtained in the step S2.2: correcting the root mean square error loss function: by using penalty coefficients
Correcting to obtain average square error loss function
The calculation formula of (2) is as follows:
wherein,,
simulating the number of sample data for a numerical value; />
The number of the wind tunnel test sample data; />
Penalty coefficients for numerically modeling sample data.
Further, the specific implementation method of the step S3 includes the following steps:
s3.1, firstly, adopting a nonlinear adjustment strategy to optimize inertia factors in a particle swarm optimization algorithm
Improved by adjusting the inertia factor +.>
Disturbance of the global optimum particles, increased stagnation detection, inertial factor->
The calculation formula of (2) is as follows:
wherein,,
representing a maximum inertial factor; />
Representing a minimum inertia factor; />
Representing the current iteration number;
representing the maximum number of iterations.
S3.2, using the particle swarm optimization algorithm improved in the step S3.1 in super-parametric optimization of the wing surface pressure reconstruction depth neural network model constructed in the step S2, taking the learning rate, the iteration times, the layer number, the number of neurons at each layer, the batch size and the weight in the loss function as optimization parameters, and starting from random solutions, searching the optimal super-parameters through iteration;
and S3.3, using the optimal super parameters obtained in the step S3.2 for the wing surface pressure reconstruction depth neural network model constructed in the step S2 to obtain an optimized wing surface pressure reconstruction depth neural network model, and performing training test on the optimized wing surface pressure reconstruction depth neural network model.
Further, the specific implementation method of the step S4 includes the following steps:
s4.1, reconstructing a depth neural network model by utilizing the optimized wing surface pressure obtained in the step S3, retraining and testing, reconstructing holographic pressure distribution on the wing surface of the model, and performing inverse operation on the holographic pressure distribution result on the wing surface of the model to realize the visualization of the holographic pressure distribution on the wing surface of the model;
s4.2, comparing, verifying and evaluating the prediction effect of the wing surface pressure reconstruction depth neural network model before and after optimization;
and S4.3, applying the obtained optimized wing surface pressure reconstruction depth neural network model to a model wind tunnel pressure measurement test, and carrying out holographic surface pressure reconstruction modeling and verification evaluation of limited pressure measuring points on the wing part.
The electronic equipment comprises a memory and a processor, wherein the memory stores a computer program, and the processor realizes the steps of the wing surface pressure reconstruction method when executing the computer program.
A computer readable storage medium having stored thereon a computer program which when executed by a processor implements the method of wing surface pressure reconstruction.
The invention has the beneficial effects that:
according to the wing surface pressure reconstruction method, the characteristics of different data sources are utilized to correlate through fusion of the test data and the CFD simulation data, so that the consistency of pneumatic data is improved, the inherent connection of the pneumatic data is effectively excavated, and the internal deviation of multi-source data is reduced.
The wing surface pressure reconstruction method has commonality, certain universality and popularization, can be used for conventional pressure measurement tests of complex aircrafts, and can be used for expanding a three-dimensional model into a two-dimensional model in a data preprocessing stage, so that the model construction difficulty is reduced, and the model training efficiency is improved; and the problems of complex process and low efficiency of manually setting and adjusting the super parameters are solved through a super parameter optimization algorithm. The numerical simulation is fused with wind tunnel test data, so that the advantages of wind tunnel test actual measurement data are utilized, the advantages of CFD technology can be exerted, the complete pressure distribution is reconstructed by using sparse test pressure measurement data, the spatial resolution of wind tunnel test data is improved, higher-accuracy aerodynamic characteristic data are provided for aerodynamic load design of non-measurement points, and the problem of fine reconstruction of distributed load under the condition of space-limited sparse observation is solved.
Drawings
FIG. 1 is a flow chart of a method of reconstructing airfoil surface pressure according to the present invention;
FIG. 2 is a schematic view of holographic mesh nodes of a three-dimensional airfoil model according to the present invention;
FIG. 3 is a schematic representation of the three-dimensional to two-dimensional interconversion of a three-dimensional airfoil model according to the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be further described in detail below with reference to the accompanying drawings and detailed description. It should be understood that the embodiments described herein are for purposes of illustration only and are not intended to limit the invention, i.e., the embodiments described are merely some, but not all, of the embodiments of the invention. The components of the embodiments of the present invention generally described and illustrated in the figures herein can be arranged and designed in a wide variety of different configurations, and the present invention can have other embodiments as well.
Thus, the following detailed description of the embodiments of the invention, as presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, are intended to fall within the scope of the present invention.
For further understanding of the invention, the following detailed description is to be taken in conjunction with fig. 1-3, in which the following detailed description is given, of the invention:
the first embodiment is as follows:
a method of wing surface pressure reconstruction comprising the steps of:
s1, acquiring original pressure data of a wing surface through a three-dimensional wing wind tunnel pressure measurement test and a three-dimensional wing model numerical simulation, preprocessing the obtained original pressure data of the wing surface, and constructing a depth neural network data set for reconstructing the pressure of the wing surface;
further, the specific implementation method of the step S1 includes the following steps:
s1.1, acquiring wing surface original pressure data including Mach number, attack angle, sideslip angle, total pressure, reynolds number and surface pressure coefficient through three-dimensional wing wind tunnel pressure measurement test and three-dimensional wing model numerical simulation calculation;
s1.2, setting the expected precision of the three-dimensional wing model to be 57-81 of the spanwise nodes of the three-dimensional wing, and 161-241 of the chordwise nodes of the three-dimensional wing, and carrying out grid division according to the expected precision of the three-dimensional wing model to obtain grid nodes of the three-dimensional wing model, wherein the grid nodes are used as holographic grid nodes of the three-dimensional wing model;
s1.3, arranging the holographic grid nodes of the three-dimensional wing model obtained in the step S1.2 according to the grid nodes of the three-dimensional wing model in rows and columns to obtain holographic grid nodes of the unfolded three-dimensional wing model;
s1.4, sampling the original pressure data of the wing surface obtained in the step S1.1 according to the holographic grid node coordinate information data of the unfolded three-dimensional wing model obtained in the step S1.3, and obtaining the holographic grid node pressure data of the three-dimensional wing model through interpolation; acquiring two-dimensional unfolding coordinate information data of holographic grid nodes corresponding to the three-dimensional wing model according to the positions of the pressure points of the wind tunnel pressure test;
further, taking into consideration that the pressure data of adjacent points on the surface of the model have space dependence, sampling the original surface pressure data of numerical simulation according to the sequence of the nodes of the unfolding grid; meanwhile, considering the position information of the pressure measuring holes of the wind tunnel test data, and corresponding to the positions of the holographic grid nodes to obtain the corresponding position information; preprocessing all data according to the requirements;
s1.5, constructing a deep neural network data set for reconstructing the surface pressure of the wing, wherein the deep neural network data set comprises the following components: the data of the depth neural network data set for reconstructing the surface pressure of the wing comprises working condition state parameters, expected precision grid point data, pressure measurement test original surface pressure data, position data of holographic grid nodes of the three-dimensional wing model corresponding to the pressure measurement test original surface pressure data and numerical simulation surface pressure data of holographic grid nodes of the three-dimensional wing model corresponding to the pressure measurement test original surface pressure data;
s2, constructing a wing surface pressure reconstruction depth neural network model, realizing fusion of two source data samples, namely the test data and the numerical simulation data, in the depth neural network data set for wing surface pressure reconstruction obtained in the step S1 through modification of a loss function, and performing training test on the constructed wing surface pressure reconstruction depth neural network model;
further, the specific implementation method of the step S2 includes the following steps:
s2.1, reconstructing a depth neural network model by using wing surface pressure, wherein a two-dimensional U-shaped convolutional neural network structure is adopted, and the model comprises a convolutional layer, a pooling layer and a deconvolution layer, wherein the convolutional layer has a convolutional kernel size of 3 multiplied by 3; the pooling layer is responsible for downsampling the space dimension of the input data, setting 2 multiplied by 2 receptive fields as maximum pooling, and setting the sliding step length as 2; up-sampling is carried out with the convolution kernel size of the deconvolution layer being 2×2;
s2.2, setting learning rate, iteration times, layer number, number of neurons of each layer and batch size of a wing surface pressure reconstruction depth neural network model, activating a function, training by taking root mean square error as a loss function, and obtaining a root mean square error loss function
The calculation formula of (2) is as follows:
wherein,,
for loss function->
Is a weight parameter, < >>
For biasing item parameters, ++>
For training sample number, ++>
And->
For the known sample data, +.>
For model predictive value, +.>
For model predictive function +.>
For input value +.>
A forward propagation function formula;
for the convolutional neural network model, the smaller the error between the predicted value and the actual value is, the better the model is, so that the training process of the network structure is the process of minimizing the loss function, namely, the extreme point of solving the loss number;
the training adopts a back propagation algorithm, a chain derivative algorithm is utilized to calculate the partial derivative of a loss function between known sample data and a model predicted value on each weight parameter or bias term, then the weight parameter or bias term parameter is reversely updated layer by layer, and the calculation formula of the gradient descent algorithm of the parameters in the wing surface pressure reconstruction depth neural network model is as follows:
wherein,,
is the firstlLayer node->
To the firstl-layer 1 node->
Weight of->
Is->
Layer node->
Bias item of->
For gradient operator->
Is the learning rate;
solving the loss function according to the chained derivation rule
The derivative for each weight or bias term is calculated as:
wherein,,
、/>
is->
Layer node->
Output value of->
Layer node->
Output value of>
Is->
Layer node->
Through->
Calculated result,/->
Equal to->
；
To make the error of each neuron
The calculation formula of (2) is as follows:
substituting the formula (6) into the formula (4) and the formula (5) can obtain:
the calculation formula of the output layer is as follows:
bringing formula (9) into formula (6), the calculation formula for the error of each neuron is:
bringing the formula (10) into the formula (7) and the formula (8) can calculate gradient update values of weights and bias terms in the output layer;
the calculation formula of the hidden layer is obtained according to the complex function derivative formula, and the calculation formula is as follows:
bringing equation (11) into equation (6) yields:
carrying the formula (12) into the formula (7) and the formula (8) to calculate gradient update values of weights and bias terms in the hidden layer; model training is completed through iterative optimization;
s2.3, calibrating the wing surface pressure reconstruction depth neural network model obtained in the step S2.2: correcting the root mean square error loss function: by using penalty coefficients
Correcting to obtain average square error loss function
The calculation formula of (2) is as follows:
wherein,,
simulating the number of sample data for a numerical value; />
The number of the wind tunnel test sample data; />
Penalty coefficients for numerically modeling sample data;
s3, optimizing the super parameters of the wing surface pressure reconstruction depth neural network model constructed in the step S2 by adopting a particle swarm optimization algorithm to obtain an optimized wing surface pressure reconstruction depth neural network model;
further, the specific implementation method of the step S3 includes the following steps:
s3.1, firstInertia factors in particle swarm optimization algorithm by adopting nonlinear adjustment strategy
Improved by adjusting the inertia factor +.>
Disturbance is carried out on the global optimal particles, stagnation detection is added, if the change of the global optimal particles is smaller than a certain threshold value after continuous N iterations, the situation that the population is possibly trapped in local optimal at the moment is indicated, stagnation phenomenon occurs, the current particles are disturbed to change the positions of the current particles, and the rest particles are updated by adopting an original method; inertial weight->
Decrementing in case of algorithm stall, when optimizing stability +.>
The value is unchanged; in this improvement, the inertial weight +.>
The calculated expression of (2) is:
wherein,,
representing a maximum inertial factor; />
Representing a minimum inertia factor; />
Representing the current iteration number;
representing the maximum number of iterations.
S3.2, using the particle swarm optimization algorithm improved in the step S3.1 in super-parametric optimization of the wing surface pressure reconstruction depth neural network model constructed in the step S2, taking the learning rate, the iteration times, the layer number, the number of neurons at each layer, the batch size and the weight in the loss function as optimization parameters, and starting from random solutions, searching the optimal super-parameters through iteration;
s3.3, using the optimal super parameters obtained in the step S3.2 for the wing surface pressure reconstruction depth neural network model constructed in the step S2 to obtain an optimized wing surface pressure reconstruction depth neural network model, and performing training test on the optimized wing surface pressure reconstruction depth neural network model;
s4, utilizing the optimized wing surface pressure reconstruction depth neural network model obtained in the step S3 to be applied to a new aircraft model wind tunnel pressure measurement test, reconstructing holographic pressure distribution of the wing surface of the model, predicting pneumatic load distribution data of non-measurement points of the wing surface of the model, and evaluating and verifying the predicted holographic pressure distribution data;
further, the specific implementation method of the step S4 includes the following steps:
s4.1, reconstructing a depth neural network model by utilizing the optimized wing surface pressure obtained in the step S3, retraining and testing, reconstructing holographic pressure distribution on the wing surface of the model, and performing inverse operation on the holographic pressure distribution result on the wing surface of the model to realize the visualization of the holographic pressure distribution on the wing surface of the model;
s4.2, comparing, verifying and evaluating the prediction effect of the wing surface pressure reconstruction depth neural network model before and after optimization;
and S4.3, applying the obtained optimized wing surface pressure reconstruction depth neural network model to an airplane model wind tunnel pressure measurement test, and carrying out holographic surface pressure reconstruction modeling and verification evaluation of limited pressure measuring points on the wing part.
From fig. 3, it can be seen that the three-dimensional to two-dimensional interconversion operation of the three-dimensional wing model mainly depends on different expressions of the surface grid node position information data, the same node position information data three-dimensional can be expressed in a coordinate form, the two-dimensional can be displayed in a number form in different directions, one-to-one correspondence exists, and the three-dimensional to two-dimensional interconversion operation of the three-dimensional wing model is realized through interconversion of different position information data expression modes.
The second embodiment is as follows:
the electronic equipment comprises a memory and a processor, wherein the memory stores a computer program, and the processor realizes the steps of the wing surface pressure reconstruction method when executing the computer program.
The computer device of the present invention may be a device including a processor and a memory, such as a single chip microcomputer including a central processing unit. And the processor is used for executing the computer program stored in the memory to realize the steps of the wing surface pressure reconstruction method.
The processor may be a central processing unit (Central Processing Unit, CPU), other general purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), off-the-shelf programmable gate arrays (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The memory may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program (such as a sound playing function, an image playing function, etc.) required for at least one function, and the like; the storage data area may store data (such as audio data, phonebook, etc.) created according to the use of the handset, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as a hard disk, memory, plug-in hard disk, smart Media Card (SMC), secure Digital (SD) Card, flash Card (Flash Card), at least one disk storage device, flash memory device, or other volatile solid-state storage device.
And a third specific embodiment:
a computer readable storage medium having stored thereon a computer program which when executed by a processor implements the method of wing surface pressure reconstruction.
The computer readable storage medium of the present invention may be any form of storage medium readable by a processor of a computer device, including but not limited to non-volatile memory, ferroelectric memory, etc., having a computer program stored thereon, which when read and executed by the processor of the computer device, implements the steps of a wing surface pressure reconstruction method as described above.
The computer program comprises computer program code which may be in source code form, object code form, executable file or in some intermediate form, etc. The computer readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a U disk, a removable hard disk, a magnetic disk, an optical disk, a computer Memory, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), an electrical carrier signal, a telecommunications signal, a software distribution medium, and so forth. It should be noted that the computer readable medium contains content that can be appropriately scaled according to the requirements of jurisdictions in which such content is subject to legislation and patent practice, such as in certain jurisdictions in which such content is subject to legislation and patent practice, the computer readable medium does not include electrical carrier signals and telecommunication signals.
It is noted that relational terms such as "first" and "second", and the like, are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
Although the present application has been described hereinabove with reference to specific embodiments, various modifications thereof may be made and equivalents may be substituted for elements thereof without departing from the scope of the application. In particular, the features of the embodiments disclosed herein may be combined with each other in any manner so long as there is no structural conflict, and the exhaustive description of these combinations is not given in this specification solely for the sake of brevity and resource saving. Therefore, it is intended that the present application not be limited to the particular embodiments disclosed, but that the present application include all embodiments falling within the scope of the appended claims.
Claims (7)
1. A method of reconstructing the surface pressure of an airfoil, comprising the steps of:
s1, acquiring original pressure data of a wing surface through a three-dimensional wing wind tunnel pressure measurement test and a three-dimensional wing model numerical simulation, preprocessing the obtained original pressure data of the wing surface, and constructing a depth neural network data set for reconstructing the pressure of the wing surface;
s2, constructing a wing surface pressure reconstruction depth neural network model, realizing fusion of two source data samples, namely the test data and the numerical simulation data, in the depth neural network data set for wing surface pressure reconstruction obtained in the step S1 through modification of a loss function, and performing training test on the constructed wing surface pressure reconstruction depth neural network model;
s3, optimizing the super parameters of the wing surface pressure reconstruction depth neural network model constructed in the step S2 by adopting a particle swarm optimization algorithm to obtain an optimized wing surface pressure reconstruction depth neural network model;
s4, utilizing the optimized wing surface pressure reconstruction depth neural network model obtained in the step S3 to be applied to a new aircraft model wind tunnel pressure measurement test, reconstructing holographic pressure distribution of the wing surface of the model, predicting pneumatic load distribution data of non-measuring points of the wing surface of the model, and evaluating and verifying the predicted holographic pressure distribution data.
2. The wing surface pressure reconstruction method according to claim 1, wherein the specific implementation method of step S1 includes the steps of:
s1.1, acquiring wing surface original pressure data including Mach number, attack angle, sideslip angle, total pressure, reynolds number and surface pressure coefficient through three-dimensional wing wind tunnel pressure measurement test and three-dimensional wing model numerical simulation calculation;
s1.2, setting the expected precision of the three-dimensional wing model to be 57-81 of the spanwise nodes of the three-dimensional wing, and 161-241 of the chordwise nodes of the three-dimensional wing, and carrying out grid division according to the expected precision of the three-dimensional wing model to obtain grid nodes of the three-dimensional wing model, wherein the grid nodes are used as holographic grid nodes of the three-dimensional wing model;
s1.3, arranging the holographic grid nodes of the three-dimensional wing model obtained in the step S1.2 according to the grid nodes of the three-dimensional wing model in rows and columns to obtain holographic grid nodes of the unfolded three-dimensional wing model;
s1.4, sampling the original pressure data of the wing surface obtained in the step S1.1 according to the holographic grid node coordinate information data of the unfolded three-dimensional wing model obtained in the step S1.3, and obtaining the holographic grid node pressure data of the three-dimensional wing model through interpolation; acquiring two-dimensional unfolding coordinate information data of holographic grid nodes corresponding to the three-dimensional wing model according to the positions of the pressure points of the wind tunnel pressure test;
s1.5, constructing a deep neural network data set for reconstructing the surface pressure of the wing, wherein the deep neural network data set comprises the following components: the data of the depth neural network data set for reconstructing the surface pressure of the wing comprises working condition state parameters, expected precision grid point data, pressure measurement test original surface pressure data, position data of holographic grid nodes of the three-dimensional wing model corresponding to the pressure measurement test original surface pressure data and numerical simulation surface pressure data of holographic grid nodes of the three-dimensional wing model corresponding to the pressure measurement test original surface pressure data.
3. A method for reconstructing the surface pressure of an airfoil according to claim 1 or 2, wherein the specific implementation method of step S2 comprises the steps of:
s2.1, reconstructing a depth neural network model by using wing surface pressure, wherein a two-dimensional U-shaped convolutional neural network structure is adopted, and the model comprises a convolutional layer, a pooling layer and a deconvolution layer, wherein the convolutional layer has a convolutional kernel size of 3 multiplied by 3; the pooling layer is responsible for downsampling the space dimension of the input data, setting 2 multiplied by 2 receptive fields as maximum pooling, and setting the sliding step length as 2; up-sampling is carried out with the convolution kernel size of the deconvolution layer being 2×2;
s2.2, setting learning rate, iteration times, layer number, number of neurons of each layer and batch size of a wing surface pressure reconstruction depth neural network model, activating a function, training by taking root mean square error as a loss function, and obtaining a root mean square error loss function
The calculation formula of (2) is as follows:
wherein,,
for loss function->
Is a weight parameter, < >>
For biasing item parameters, ++>
For training sample number, ++>
And->
For the known sample data, +.>
For model predictive value, +.>
For model predictive function +.>
For input value +.>
A forward propagation function formula;
the training adopts a back propagation algorithm, a chain derivative algorithm is utilized to calculate the partial derivative of a loss function between known sample data and a model predicted value on each weight parameter or bias term, then the weight parameter or bias term parameter is reversely updated layer by layer, and the calculation formula of the gradient descent algorithm of the parameters in the wing surface pressure reconstruction depth neural network model is as follows:
wherein,,
is the firstlLayer node->
To the firstl-layer 1 node->
Weight of->
Is->
Layer node->
Bias item of->
For gradient operator->
Is the learning rate;
solving the loss function according to the chained derivation rule
The derivative for each weight or bias term is calculated as:
wherein,,
、/>
is->
Layer node->
Output value of->
Layer node->
Output value of>
Is->
Layer node->
Through->
Calculated result,/->
Equal to->
；
To make the error of each neuron
The calculation formula of (2) is as follows:
substituting the formula (6) into the formula (4) and the formula (5) can obtain:
the calculation formula of the output layer is as follows:
bringing formula (9) into formula (6), the calculation formula for the error of each neuron is:
bringing the formula (10) into the formula (7) and the formula (8) can calculate gradient update values of weights and bias terms in the output layer;
the calculation formula of the hidden layer is obtained according to the complex function derivative formula, and the calculation formula is as follows:
bringing equation (11) into equation (6) yields:
carrying the formula (12) into the formula (7) and the formula (8) to calculate gradient update values of weights and bias terms in the hidden layer; model training is completed through iterative optimization;
s2.3, calibrating the wing surface pressure reconstruction depth neural network model obtained in the step S2.2: correcting the root mean square error loss function: by using penalty coefficients
Correction is carried out to obtain an average square error loss function +.>
The calculation formula of (2) is as follows:
wherein,,
simulating the number of sample data for a numerical value; />
The number of the wind tunnel test sample data; />
Is a penalty coefficient.
4. A method for reconstructing the surface pressure of an airfoil according to claim 3, wherein the specific implementation method of step S3 comprises the steps of:
s3.1, firstly, adopting a nonlinear adjustment strategy to optimize inertia factors in a particle swarm optimization algorithm
Improved by adjusting the inertia factor +.>
Disturbance of the global optimum particles, increased stagnation detection, inertial factor->
The calculation formula of (2) is as follows:
wherein,,
representing the maximum inertial factor,/->
Representing the minimum inertial factor,/->
Indicating the number of iterations that are currently performed,
representing a maximum number of iterations;
s3.2, using the particle swarm optimization algorithm improved in the step S3.1 in super-parametric optimization of the wing surface pressure reconstruction depth neural network model constructed in the step S2, taking learning rate, iteration times, layer number, number of neurons of each layer, batch size and penalty coefficient in a loss function as optimization parameters, and starting from random solution, searching the optimal super-parameters through iteration;
and S3.3, using the optimal super parameters obtained in the step S3.2 for the wing surface pressure reconstruction depth neural network model constructed in the step S2 to obtain an optimized wing surface pressure reconstruction depth neural network model, and performing training test on the optimized wing surface pressure reconstruction depth neural network model.
5. The wing surface pressure reconstruction method according to claim 4, wherein the specific implementation method of step S4 includes the steps of:
s4.1, reconstructing a depth neural network model by utilizing the optimized wing surface pressure obtained in the step S3, retraining and testing, reconstructing holographic pressure distribution on the wing surface of the model, and performing inverse operation on the holographic pressure distribution result on the wing surface of the model to realize the visualization of the holographic pressure distribution on the wing surface of the model;
s4.2, comparing, verifying and evaluating the prediction effect of the wing surface pressure reconstruction depth neural network model before and after optimization;
and S4.3, applying the obtained optimized wing surface pressure reconstruction depth neural network model to a model wind tunnel pressure measurement test, and carrying out holographic surface pressure reconstruction modeling and verification evaluation of limited pressure measuring points on the wing part.
6. An electronic device comprising a memory and a processor, the memory storing a computer program, the processor implementing the steps of a wing surface pressure reconstruction method as claimed in any one of claims 1 to 5 when the computer program is executed.
7. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements a wing surface pressure reconstruction method as claimed in any one of claims 1-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310432279.XA CN116432556A (en) | 2023-04-21 | 2023-04-21 | Wing surface pressure reconstruction method, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310432279.XA CN116432556A (en) | 2023-04-21 | 2023-04-21 | Wing surface pressure reconstruction method, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116432556A true CN116432556A (en) | 2023-07-14 |
Family
ID=87090658
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310432279.XA Pending CN116432556A (en) | 2023-04-21 | 2023-04-21 | Wing surface pressure reconstruction method, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116432556A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117993303A (en) * | 2024-04-02 | 2024-05-07 | 上海勘测设计研究院有限公司 | Pneumatic load checking method for deformation of offshore wind power blade |
-
2023
- 2023-04-21 CN CN202310432279.XA patent/CN116432556A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117993303A (en) * | 2024-04-02 | 2024-05-07 | 上海勘测设计研究院有限公司 | Pneumatic load checking method for deformation of offshore wind power blade |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111812732A (en) | Magnetotelluric nonlinear inversion method based on convolutional neural network | |
| CN110119854B (en) | Voltage stabilizer water level prediction method based on cost-sensitive LSTM (least squares) cyclic neural network | |
| CN110222828B (en) | Unsteady flow field prediction method based on hybrid deep neural network | |
| Wang et al. | Data-driven CFD modeling of turbulent flows through complex structures | |
| Duru et al. | CNNFOIL: Convolutional encoder decoder modeling for pressure fields around airfoils | |
| CN114662414B (en) | Oil reservoir production prediction method based on graph wavelet neural network model | |
| CN107862170B (en) | Finite element model correction method based on dynamic polycondensation | |
| CN113568056B (en) | Aviation transient electromagnetic data inversion method based on convolutional neural network | |
| CN116432556A (en) | Wing surface pressure reconstruction method, electronic equipment and storage medium | |
| CN114692501A (en) | Pneumatic data fusion method and device based on multi-precision deep neural network | |
| CN112883522A (en) | Micro-grid dynamic equivalent modeling method based on GRU (generalized regression Unit) recurrent neural network | |
| CN115455793A (en) | High-rise structure complex component stress analysis method based on multi-scale model correction | |
| CN113361194B (en) | Sensor drift calibration method based on deep learning, electronic equipment and storage medium | |
| CN115455838B (en) | High-spatial-resolution flow field reconstruction method for time-course data | |
| CN116227045B (en) | Local stress strain field construction method and system for structural test piece | |
| Khan et al. | Forecasting renewable energy for environmental resilience through computational intelligence | |
| CN114692529B (en) | CFD high-dimensional response uncertainty quantification method and device, and computer equipment | |
| CN114155354B (en) | Method and device for reconstructing capacitance tomography based on graph convolution network | |
| Chen et al. | An airfoil mesh quality criterion using deep neural networks | |
| CN109166128B (en) | Non-equivalent three-dimensional point cloud segmentation method | |
| CN114638048A (en) | Three-dimensional spray pipe flow field rapid prediction and sensitivity parameter analysis method and device | |
| CN110829434B (en) | Method for improving expansibility of deep neural network tidal current model | |
| He et al. | Aerodynamic data fusion with a multi-fidelity surrogate modeling method | |
| CN114139482A (en) | EDA circuit failure analysis method based on depth measurement learning | |
| CN117634365B (en) | Airplane aerodynamic force prediction method, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |