CN115828797A - Submarine hydrodynamic load rapid forecasting method based on reduced order model - Google Patents
Submarine hydrodynamic load rapid forecasting method based on reduced order model Download PDFInfo
- Publication number
- CN115828797A CN115828797A CN202310113949.1A CN202310113949A CN115828797A CN 115828797 A CN115828797 A CN 115828797A CN 202310113949 A CN202310113949 A CN 202310113949A CN 115828797 A CN115828797 A CN 115828797A
- Authority
- CN
- China
- Prior art keywords
- submarine
- hydrodynamic
- parameter
- model
- snapshot 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
- 238000013277 forecasting method Methods 0.000 title claims abstract description 11
- 238000000034 method Methods 0.000 claims abstract description 32
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 22
- 238000004364 calculation method Methods 0.000 claims abstract description 20
- 238000004088 simulation Methods 0.000 claims abstract description 14
- 238000005457 optimization Methods 0.000 claims abstract description 13
- 230000008569 process Effects 0.000 claims abstract description 6
- 238000013473 artificial intelligence Methods 0.000 claims abstract description 4
- 238000013528 artificial neural network Methods 0.000 claims abstract description 4
- 239000003795 chemical substances by application Substances 0.000 claims abstract description 4
- 238000012549 training Methods 0.000 claims abstract description 4
- 239000011159 matrix material Substances 0.000 claims description 26
- 239000013598 vector Substances 0.000 claims description 23
- 239000012530 fluid Substances 0.000 claims description 5
- 230000008859 change Effects 0.000 claims description 4
- 238000003491 array Methods 0.000 claims description 3
- 230000010354 integration Effects 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 230000003068 static effect Effects 0.000 claims description 3
- 230000007613 environmental effect Effects 0.000 abstract description 5
- 238000013461 design Methods 0.000 abstract description 4
- 238000011156 evaluation Methods 0.000 abstract description 3
- 238000004458 analytical method Methods 0.000 abstract description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 abstract description 2
- 238000010276 construction Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 238000007794 visualization technique Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000004904 shortening Methods 0.000 description 1
- 230000005477 standard model Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a submarine hydrodynamic load rapid forecasting method based on a reduced order model, which comprises the steps of obtaining a parameter subspace and obtaining a snapshot data resource pool of a whole sample; performing singular value decomposition on a snapshot data resource pool of a sample, introducing an intrinsic orthogonal decomposition method, decomposing the physical quantity distribution of the whole submarine hydrodynamic space into the sum of products of a time-average flow field and a plurality of modes and coefficients thereof, and selecting the first two modes as main modes of the submarine hydrodynamic flow field to construct a reduced-order model; training an agent model taking a neural network as a core by adopting an artificial intelligence hyper-parameter optimization method based on Bayesian optimization to obtain a forecasting model; and inputting a parameter combination in the forecasting model, and quickly obtaining the physical quantity information of all grid nodes and the hydrodynamic load of the submarine. The method can solve the problem that high-precision simulation calculation is required under each environmental working condition in conventional simulation analysis, and improves the water power load evaluation working efficiency in the design and optimization process.
Description
Technical Field
The invention belongs to the technical field of submarine hydrodynamic load forecasting, and particularly relates to a submarine hydrodynamic load rapid forecasting method based on a reduced order model.
Background
The hydrodynamic load of the submarine is under hydrodynamic action when the submarine runs in water, is a decisive influence factor of key performance indexes such as rapidity and maneuverability of the submarine, and the rapid and accurate acquisition of the hydrodynamic load of the submarine under different marine environmental conditions has important significance on submarine design and performance evaluation.
At present, a Computational Fluid Dynamics (CFD) method is usually adopted for calculating and evaluating hydrodynamic load of a submarine in design safety, the CFD method is a method for establishing a simulation model on a computer and solving a partial differential equation set for describing Fluid motion, mass transfer and heat transfer, hydrodynamic load data of a certain specific working condition (navigational speed and attitude) can only be obtained by one-time calculation and solution, recalculation is needed once working condition parameters change, a large amount of solution time is consumed for single high-precision CFD calculation, and rapid prediction of hydrodynamic load cannot be realized.
A Reduced Order Model (ROM) is a simplified model approximate original model constructed on the basis of a model with reasonable accuracy, and the core idea is that a large-scale complex system model is converted into a smaller-scale approximate system model under a certain condition, so that the error between the approximate system model and the original system model is small enough, and the calculation speed is high enough.
Disclosure of Invention
In view of the above, the invention provides a submarine hydrodynamic load rapid forecasting method based on a reduced order model, and aims to solve the problems of large calculation amount, long time consumption and multiple working conditions requiring multiple calculations in the submarine hydrodynamic load forecasting process by using the conventional CFD method.
The method comprises the following steps:
s1, analyzing a CFD simulation engineering file, acquiring definition parameters of an entrance boundary condition, and acquiring a parameter subspace by adopting an improved Latin hypercube sampling method through specifying upper and lower limit values of parameter change;
s2, completing CFD simulation solution based on working condition parameters in the parameter subspace, writing snapshot data of the sample under each working condition after the working condition is calculated, and obtaining a snapshot data resource pool of the whole sample;
s3, carrying out singular value decomposition on the snapshot data resource pool of the sample, introducing an intrinsic orthogonal decomposition method, decomposing the physical quantity distribution of the whole submarine hydrodynamic space into the sum of time-average flow field and products of a plurality of modes and coefficients thereof, and selecting the first two-order mode as a main mode of the submarine hydrodynamic flow field to construct a reduced-order model;
s4, training an agent model taking a neural network as a core by adopting an artificial intelligence hyper-parameter optimization method based on Bayesian optimization to obtain a forecasting model;
and S5, inputting any entrance boundary parameter combination into the reduced model, obtaining the physical quantity information of all grid nodes through the forecasting model, and performing static pressure value integration on the submarine surface calculation boundary to obtain the complete submarine hydrodynamic load.
Further, step S2 specifically includes:
the contents of the snapshot data comprise coordinate information, speed and pressure on grid nodes, a full-working-condition sample snapshot data resource pool containing entry boundary parameters is formed after all working conditions are calculated, and the form of the whole sample snapshot data resource pool is as follows:
in the above formularFor all the number of operating conditions in the parameter subspace,nin order to calculate the total number of domain mesh nodes,
is as followsrUnder one calculation conditionnThe velocity and pressure values at the nodes of each grid,v r is as followsrThe inlet boundary incoming flow speed under each calculated condition,a r is as followsrAnd calculating the inflow angle of the inlet boundary under the working condition.
Further, step S3 specifically includes:
because the sample snapshot data resource pool is not in a square matrix form and cannot obtain the eigenvalue and the eigenvector thereof directly through matrix decomposition, the form of the sample snapshot data resource pool is firstly rewritten into the following form:
P=ΦΣΨ T
whereinΦAndΨare square matrices, respectively matrixPThe left and right singular vectors of (a),Σis a matrixPThe elements on the main diagonal line of the singular array are singular values, and the elements except the main diagonal line are 0;
will matrixPCarrying out characteristic value decomposition after carrying out square matrix:
(P T P)Ψ i =λ i Ψ i
whereinλ i AndΨ i are respectively a square matrixP T PTo (1) aiA characteristic value andia feature vector;
the feature value based decomposition synchronization further comprises:
(PP T )Φ i =β i Φ i
whereinβ i AndΦ i are respectively a square matrixPP T To (1)iA characteristic value andithe feature vectors are obtained, and then the matrix is obtainedPLeft singular vector ofΦAnd right singular vectorsΨ；
ByPP T =ΦΣΨ T ΣΦ T =ΨΣ 2 Φ T Comprises the following steps:
whereinδ i As a singular arrayΣThe elements on the middle main diagonal line are arranged in the order from big to small to obtainPSingular arrays ofΣ；
After the characteristic value and the characteristic vector of the sample snapshot data resource pool P are obtained, unknown quantities in a fluid control equation are decomposed by an intrinsic orthogonal decomposition method, including velocity and pressure, in a fundamental mode decomposition mode, velocity and pressure values on all grid nodes in a submarine hydrodynamic calculation domain are decomposed into a mode of the sum of a time mean value and a disturbance value, wherein the disturbance value is expressed by the product sum of a plurality of modes and mode coefficients thereof:
wherein
A predicted value of the parameter is calculated for the hydrodynamic force,p 0 to be the average of the snapshots of the simulated data samples,α r is a mode shape coefficient of an orthogonal basis,
in the form of a basis vector, the vector,Rand selecting the first two-order modes of the speed field and the pressure field as the main modes of the submarine hydrodynamic flow field to construct a reduced-order model.
Further, step S4 specifically includes:
modal coefficient to orthogonal basisα r And introducing a Bayesian hyper-parameter optimization method into the speed and angle parameters on the entrance boundary to construct a Gaussian process proxy model, and forming an incidence relation between the modal coefficient and the entrance boundary parameter.
The technical scheme provided by the invention has the beneficial effects that:
the invention provides a method based on a reduced order model for rapidly acquiring hydrodynamic loads of submarines under different marine environmental conditions, the hydrodynamic loads and the spatial flow field distribution conditions of the submarines under other marine environmental conditions can be rapidly acquired by constructing the reduced order model through a small amount of simulation engineering result data and forming a high-precision proxy model through machine learning, and the method can solve the problem that high-precision simulation calculation needs to be carried out once under each environmental working condition in the conventional simulation analysis method, thereby remarkably shortening the computer processing time and improving the water-dynamic load evaluation working efficiency in the design and optimization process.
Drawings
FIG. 1 is a flow chart of a submarine hydrodynamic load rapid forecasting method based on a reduced order model according to the invention;
FIG. 2 is a submarine hydrodynamic load forecasting model construction method according to an embodiment of the present invention;
fig. 3 is a submarine hydrodynamic load extraction and visualization method according to an embodiment of the invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be further described with reference to the accompanying drawings.
Referring to fig. 1, fig. 2 and fig. 3, fig. 1 is a block flow diagram of a submarine hydrodynamic load rapid forecasting method based on a reduced order model according to the present invention; FIG. 2 is a submarine hydrodynamic load forecasting model construction method according to an embodiment of the invention; fig. 3 is a submarine hydrodynamic load extraction and visualization method according to an embodiment of the invention.
S1, analyzing a CFD simulation engineering file of the submarine hydrodynamic load, acquiring a definition mode of an incoming flow speed and an incoming flow angle in an entrance boundary condition, specifying upper and lower limit values of parameter change, acquiring a parameter subspace by adopting an improved Latin hypercube sampling method, and writing out and storing the parameter subspace in an XML general file format.
And S2, driving the CFD simulation engineering file to complete the simulation solution of the required working conditions based on the working condition parameters in the parameter subspace, writing the snapshot data of the sample under each working condition after the calculation of each working condition is finished, wherein the written content comprises coordinate information, speed, pressure and other physical quantity information on all grid nodes in the calculation engineering file, and outputting the physical quantity information in a specific arrangement form to form a full-working-condition sample snapshot data resource pool containing the entrance boundary parameters.
The form of the entire sample snapshot data resource pool is as follows:
in the above formularFor all the number of operating conditions in the parameter subspace,nin order to calculate the total number of domain grid nodes,
is as followsrUnder one calculation conditionnThe velocity and pressure values at the nodes of each grid,v r is as followsrThe inlet boundary incoming flow speed under each calculated condition,a r is as followsrAnd calculating the inflow angle of the inlet boundary under the working condition.
S3, singular value decomposition is carried out on the snapshot data resource pool of the sample, and as the sample snapshot data resource pool is not in a square matrix form and cannot directly obtain the eigenvalue and the eigenvector thereof through matrix decomposition, the form of the sample snapshot data resource pool is firstly rewritten into the following form:
P=ΦΣΨ T
whereinΦAndΨare square matrices, respectively matrixPThe left and right singular vectors of (a),Σis a matrixPThe elements on the main diagonal of the singular matrix of (1) are singular values, and the elements except the main diagonal are all 0.
Will matrixPCarrying out characteristic value decomposition after carrying out square matrix:
(P T P)Ψ i =λ i Ψ i
whereinλ i AndΨ i are respectively a square matrixP T PTo (1) aiA characteristic value andia feature vector.
The feature value based decomposition synchronization further comprises:
(PP T )Φ i =β i Φ i
whereinβ i AndΦ i are respectively a square matrixPP T To (1) aiA characteristic value andithe feature vectors are obtained, and then the matrix is obtainedPLeft singular vector ofΦAnd right singular vectorΨ。
ByPP T =ΦΣΨ T ΣΦ T =ΨΣ 2 Φ T Comprises the following steps:
whereinδ i As a singular arrayΣThe elements on the middle main diagonal line are arranged in the order from large to small to obtainPSingular arrays ofΣ。
Sample snapshot data resource pool is obtainedPThe eigenvalue and eigenvector of the method are introduced into an intrinsic orthogonal decomposition method, a small part of space mode is used for representation, projection reconstruction is carried out on each basis vector, and the minimum residual error is constructedA reduced order model; carrying out fundamental mode decomposition on unknown quantities including speed and pressure in a fluid control equation by an intrinsic orthogonal decomposition method, and decomposing speed and pressure values on all grid nodes in a submarine hydrodynamic calculation domain into a form of sum of a time mean value and a disturbance value, wherein the disturbance value is represented by the sum of products of a plurality of modes and mode coefficients thereof:
wherein
A predicted value of the parameter is calculated for the hydrodynamic force,p 0 to be the average of the snapshots of the simulated data samples,α r is a mode shape coefficient of an orthogonal basis,
in the form of a basis vector, the vector,Rand selecting the first two-order modes of the speed field and the pressure field as the main modes of the submarine hydrodynamic flow field to construct a reduced-order model.
And S4, training the agent model taking the neural network as the core by adopting an artificial intelligence hyper-parameter optimization method based on Bayesian optimization to obtain a forecasting model, and forecasting the maximum profit combination according to the completed hyper-parameter combination.
Modal coefficient to orthogonal basisα r And introducing a Bayesian super-parameter optimization method into the speed and angle parameters on the entrance boundary to construct a Gaussian process proxy model, forming an incidence relation between a modal coefficient and the entrance boundary parameters, and obtaining hydrodynamic load response values on all spatial grid nodes by inputting the entrance boundary parameters.
And S5, inputting any entry boundary parameter combination into the forecasting model, namely quickly obtaining physical quantity information of all grid nodes, performing static pressure value integration on the submarine surface calculation boundary to obtain a complete submarine hydrodynamic load, combining a cloud picture drawing algorithm to obtain speed and pressure distribution images on the submarine surface and a typical section, and through suboff submarine standard model hydrodynamic resistance calculation test, the time consumption for obtaining the submarine body hydrodynamic resistance by using the forecasting model under the same computer configuration can be shortened to one thousandth of the time consumption for CFD simulation calculation.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (4)
1. A submarine hydrodynamic load rapid forecasting method based on a reduced order model is characterized by comprising the following steps:
s1, analyzing a CFD simulation engineering file, acquiring definition parameters of an entrance boundary condition, and acquiring a parameter subspace by adopting an improved Latin hypercube sampling method through specifying upper and lower limit values of parameter change;
s2, completing CFD simulation solution based on working condition parameters in the parameter subspace, writing snapshot data of the sample under each working condition after the working condition is calculated, and obtaining a snapshot data resource pool of the whole sample;
s3, carrying out singular value decomposition on the snapshot data resource pool of the sample, introducing an intrinsic orthogonal decomposition method, decomposing the physical quantity distribution of the whole submarine hydrodynamic space into the sum of time-average flow field and products of a plurality of modes and coefficients thereof, and selecting the first two-order mode as a main mode of the submarine hydrodynamic flow field to construct a reduced-order model;
s4, training the agent model with the neural network as the core by adopting an artificial intelligence hyper-parameter optimization method based on Bayesian optimization to obtain a forecasting model;
and S5, inputting any entrance boundary parameter combination into the reduced model, obtaining the physical quantity information of all grid nodes through the forecasting model, and performing static pressure value integration on the submarine surface calculation boundary to obtain the complete submarine hydrodynamic load.
2. The submarine hydrodynamic load rapid forecasting method based on the reduced order model according to claim 1, wherein the step S2 specifically comprises:
the contents of the snapshot data comprise coordinate information, speed and pressure on grid nodes, a full-working-condition sample snapshot data resource pool containing entry boundary parameters is formed after all working conditions are calculated, and the form of the whole sample snapshot data resource pool is as follows:
in the above formularFor all the number of operating conditions in the parameter subspace,nin order to calculate the total number of domain mesh nodes,
is as followsrUnder one calculation conditionnThe velocity and pressure values at the nodes of each grid,v r is as followsrThe inlet boundary incoming flow speed under each calculated condition,a r is as followsrAnd calculating the inflow angle of the inlet boundary under the working condition.
3. The submarine hydrodynamic load rapid forecasting method based on the reduced order model according to claim 2, wherein the step S3 specifically comprises:
firstly, the form of the sample snapshot data resource pool is rewritten into the following form:
P=ΦΣΨ T
whereinΦAndΨare square matrices, respectively matrixPThe left and right singular vectors of (a),Σis a matrixPThe elements on the main diagonal line of the singular array are singular values, and the elements except the main diagonal line are 0;
will matrixPCarrying out characteristic value decomposition after carrying out square matrix:
(P T P)Ψ i =λ i Ψ i
whereinλ i AndΨ i are respectively a square matrixP T PTo (1) aiA characteristic value andia feature vector;
the eigenvalue-based decomposition synchronization also includes:
(PP T )Φ i =β i Φ i
whereinβ i AndΦ i are respectively a square matrixPP T To (1) aiA characteristic value andithe feature vectors are obtained, and then the matrix is obtainedPLeft singular vector ofΦAnd right singular vectorΨ；
ByPP T =ΦΣΨ T ΣΦ T =ΨΣ 2 Φ T Comprises the following steps:
whereinδ i As a singular arrayΣThe elements on the middle main diagonal line are arranged in the order from big to small to obtainPSingular arrays ofΣ；
Sample snapshot data resource pool is obtainedPAfter the characteristic values and the characteristic vectors are obtained, performing fundamental mode decomposition on unknown quantities in a fluid control equation by an intrinsic orthogonal decomposition method, wherein the unknown quantities comprise speed and pressure, and the speed and pressure values on all grid nodes in a submarine hydrodynamic calculation domain are decomposed into a form of the sum of a time mean value and a disturbance value, wherein the disturbance value is expressed by the sum of products of a plurality of modes and mode coefficients thereof:
wherein
A predicted value of the parameter is calculated for the hydrodynamic force,p 0 to be the average of the snapshots of the simulated data samples,α r is a mode shape coefficient of an orthogonal basis,
is a function of the basis vector and is,Rand selecting the first two-stage modes of the velocity field and the pressure field as the main modes of the submarine hydrodynamic flow field to construct a reduced-order model.
4. The submarine hydrodynamic load rapid forecasting method based on the reduced order model according to claim 3, wherein the step S4 specifically comprises:
modal coefficient to orthogonal basisα r And introducing a Bayesian hyper-parameter optimization method into the speed and angle parameters on the entrance boundary to construct a Gaussian process proxy model, and forming an incidence relation between the modal coefficient and the entrance boundary parameter.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310113949.1A CN115828797A (en) | 2023-02-15 | 2023-02-15 | Submarine hydrodynamic load rapid forecasting method based on reduced order model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310113949.1A CN115828797A (en) | 2023-02-15 | 2023-02-15 | Submarine hydrodynamic load rapid forecasting method based on reduced order model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115828797A true CN115828797A (en) | 2023-03-21 |
Family
ID=85521410
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310113949.1A Pending CN115828797A (en) | 2023-02-15 | 2023-02-15 | Submarine hydrodynamic load rapid forecasting method based on reduced order model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115828797A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117252130A (en) * | 2023-11-17 | 2023-12-19 | 中国空气动力研究与发展中心计算空气动力研究所 | Near-field pressure distribution prediction method, system and related equipment for aircraft |
| CN117744540A (en) * | 2024-02-19 | 2024-03-22 | 青岛哈尔滨工程大学创新发展中心 | Underwater operation hydrodynamic characteristic trend prediction method of underwater unmanned aircraft |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060004551A1 (en) * | 2004-06-30 | 2006-01-05 | Freund Roland W | Method and apparatus for structure-preserving reduced-order modeling |
| US20100250205A1 (en) * | 2009-03-31 | 2010-09-30 | Airbus Espana, S.L. | Method and system for a quick calculation of aerodynamic forces on an aircraft in transonic conditions |
| CN103902782A (en) * | 2014-04-11 | 2014-07-02 | 北京理工大学 | POD (proper orthogonal decomposition) and surrogate model based order reduction method for hypersonic aerodynamic thermal models |
| CN113919243A (en) * | 2021-09-11 | 2022-01-11 | 西北工业大学 | Supercritical fluid heat transfer associated agent model construction method based on machine learning |
| CN114065662A (en) * | 2021-11-12 | 2022-02-18 | 西北工业大学 | Method suitable for rapidly predicting airfoil flow field with variable grid topology |
| CN114943192A (en) * | 2022-04-14 | 2022-08-26 | 西北工业大学 | Method for calculating unsteady steady solution of airfoil flow based on reduced order model and gradient optimization |
| CN115293050A (en) * | 2022-08-22 | 2022-11-04 | 上海大学 | Method, device and system for establishing fluid flow reduced-order model and storage medium |
| CN115577655A (en) * | 2022-10-26 | 2023-01-06 | 中国科学院沈阳自动化研究所 | Method for predicting rotor wing and airframe aerodynamic force of tilt rotor wing aircraft |
-
2023
- 2023-02-15 CN CN202310113949.1A patent/CN115828797A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060004551A1 (en) * | 2004-06-30 | 2006-01-05 | Freund Roland W | Method and apparatus for structure-preserving reduced-order modeling |
| US20100250205A1 (en) * | 2009-03-31 | 2010-09-30 | Airbus Espana, S.L. | Method and system for a quick calculation of aerodynamic forces on an aircraft in transonic conditions |
| CN103902782A (en) * | 2014-04-11 | 2014-07-02 | 北京理工大学 | POD (proper orthogonal decomposition) and surrogate model based order reduction method for hypersonic aerodynamic thermal models |
| CN113919243A (en) * | 2021-09-11 | 2022-01-11 | 西北工业大学 | Supercritical fluid heat transfer associated agent model construction method based on machine learning |
| CN114065662A (en) * | 2021-11-12 | 2022-02-18 | 西北工业大学 | Method suitable for rapidly predicting airfoil flow field with variable grid topology |
| CN114943192A (en) * | 2022-04-14 | 2022-08-26 | 西北工业大学 | Method for calculating unsteady steady solution of airfoil flow based on reduced order model and gradient optimization |
| CN115293050A (en) * | 2022-08-22 | 2022-11-04 | 上海大学 | Method, device and system for establishing fluid flow reduced-order model and storage medium |
| CN115577655A (en) * | 2022-10-26 | 2023-01-06 | 中国科学院沈阳自动化研究所 | Method for predicting rotor wing and airframe aerodynamic force of tilt rotor wing aircraft |
Non-Patent Citations (1)
| Title |
|---|
| 曾广会;朱军;刘洁;: "潜艇操纵水动力小波神经网络预报方法研究" * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117252130A (en) * | 2023-11-17 | 2023-12-19 | 中国空气动力研究与发展中心计算空气动力研究所 | Near-field pressure distribution prediction method, system and related equipment for aircraft |
| CN117252130B (en) * | 2023-11-17 | 2024-02-06 | 中国空气动力研究与发展中心计算空气动力研究所 | Near-field pressure distribution prediction method, system and related equipment for aircraft |
| CN117744540A (en) * | 2024-02-19 | 2024-03-22 | 青岛哈尔滨工程大学创新发展中心 | Underwater operation hydrodynamic characteristic trend prediction method of underwater unmanned aircraft |
| CN117744540B (en) * | 2024-02-19 | 2024-04-30 | 青岛哈尔滨工程大学创新发展中心 | Underwater operation hydrodynamic characteristic trend prediction method of underwater unmanned aircraft |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115828797A (en) | Submarine hydrodynamic load rapid forecasting method based on reduced order model | |
| Ni et al. | Discussions on non-probabilistic convex modelling for uncertain problems | |
| Lan et al. | Emulation of higher-order tensors in manifold Monte Carlo methods for Bayesian inverse problems | |
| Mdrafi et al. | Joint learning of measurement matrix and signal reconstruction via deep learning | |
| CN112560966A (en) | Polarimetric SAR image classification method, medium and equipment based on scattergram convolution network | |
| CN116227359A (en) | Flow field prediction method based on attention and convolutional neural network codec | |
| CN113221475A (en) | Grid self-adaption method for high-precision flow field analysis | |
| CN116451584A (en) | Thermal stress prediction method and system based on neural network | |
| Marmin et al. | Deep gaussian processes for calibration of computer models (with discussion) | |
| CN115983148A (en) | CFD simulation cloud picture prediction method, system, electronic device and medium | |
| Xing et al. | Shared-gaussian process: Learning interpretable shared hidden structure across data spaces for design space analysis and exploration | |
| Zhang et al. | MIMO non-parametric modeling of ship maneuvering motion for marine simulator using adaptive moment estimation locally weighted learning | |
| Pawar et al. | Towards multi-fidelity deep learning of wind turbine wakes | |
| CN116502539B (en) | VOCs gas concentration prediction method and system | |
| Kang et al. | Investigation on the nonintrusive multi-fidelity reduced-order modeling for PWR rod bundles | |
| CN116227650B (en) | Lithium battery temperature distribution prediction model construction method based on orthogonal enhancement type local maintenance projection algorithm | |
| CN114692529B (en) | CFD high-dimensional response uncertainty quantification method and device, and computer equipment | |
| Wei et al. | Study of non-intrusive model order reduction of neutron transport problems | |
| CN117373709B (en) | Reactor core measurement method based on model order reduction and mixed data assimilation | |
| Zakharova et al. | Methods of Constructing a Visual Map of Generalized Computational Experiment | |
| Wang et al. | E-LMC: Extended Linear Model of Coregionalization for Spatial Field Prediction | |
| House et al. | A POD-based input-output reduced-order modeling framework for acceleration of coupled-solver workflows | |
| CN115048846B (en) | Model order reduction and stability judgment method and system for tube bundle fluid bullet system | |
| BISTRIAN | Mathematical considerations on Randomized Orthogonal Decomposition method for developing twin data models | |
| de Sousa Almeida et al. | Nonintrusive Operator Inference for Chaotic Systems |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20230321 |