CN-122021410-A - Separated flow wall pressure pulsation leading structure modeling method based on spectrum orthogonal decomposition

CN122021410ACN 122021410 ACN122021410 ACN 122021410ACN-122021410-A

Abstract

The application relates to the technical field of hydrodynamic modeling and aerodynamic acoustic analysis, in particular to a method for modeling a separating flow wall pressure pulsation leading structure based on spectrum orthogonal decomposition, wherein the method comprises the steps of acquiring full flow field and/or wall unsteady pressure data of a separating flow under a preset working condition; the method comprises the steps of performing spectrum orthogonal decomposition on unsteady pressure data, performing frequency domain decoupling on multi-scale features, extracting feature values and space dominant feature modes under each feature frequency, constructing a complex wave packet physical parameterization model, fitting the space dominant feature modes, compressing the space dominant feature modes into sparse physical parameter vectors after nonlinear regression and resolving, reconstructing wall pressure pulsation dominant components based on the sparse physical parameter vectors and the feature values, and performing at least one of flow mechanism analysis and pneumatic reduced order modeling according to the dominant components. Therefore, the problems that asymmetric evolution and variable acceleration convection of a large-scale structure in the pressure pulsation of the wall surface of the separation flow cannot be accurately represented in a frequency domain in the related technology, so that the reconstruction accuracy is low and the like are solved.

Inventors

TAN LEI
HAN BINGFU

Assignees

清华大学

Dates

Publication Date: 20260512
Application Date: 20260108

Claims (10)

1. The method for modeling the pressure pulsation dominant structure of the wall surface of the separation flow based on the spectrum orthogonal decomposition is characterized by comprising the following steps of: Acquiring non-steady pressure data of a full flow field and/or a wall surface of a separation flow of the separation flow under a preset working condition; Performing spectrum orthogonal decomposition on the unsteady pressure data, decoupling multi-scale dynamics features in a frequency domain, and extracting feature values and corresponding space dominant feature modes under each feature frequency; Constructing a physical parameterization model based on complex wave packets, fitting the spatial dominant characteristic mode to the physical parameterization model, and compressing the spatial dominant characteristic mode into a sparse physical parameter vector through nonlinear regression calculation; Reconstructing dominant components of the separated flow wall pressure pulse field based on the sparse physical parameter vector and the characteristic values under the characteristic frequencies, and carrying out at least one of flow mechanism analysis and pneumatic reduced modeling according to the dominant components.
2. The method for modeling a separate flow wall pressure pulsation dominant structure based on spectrum orthogonal decomposition according to claim 1, wherein the performing of spectrum orthogonal decomposition on the unsteady pressure data, decoupling multi-scale dynamics features in a frequency domain, extracting feature values and corresponding spatial dominant feature modes at each feature frequency, comprises: Estimating a cross spectral density matrix of the unsteady pressure data; and decoupling the multi-scale dynamic characteristics in the frequency domain according to the cross spectral density matrix, and extracting characteristic values and corresponding space dominant characteristic modes under each characteristic frequency.
3. The method for modeling a separate flow wall pressure pulsation dominant structure based on spectrum orthogonal decomposition according to claim 2, wherein the expression of the cross spectral density matrix is: wherein x is a space coordinate, x' is a space coordinate, f is a frequency; is the Fourier transform of the pressure, E ] The ] is a desired operator, is a conjugate transpose; is a cross spectral density matrix; the solution formula of the cross spectral density matrix is as follows: Wherein lambda (f) is a characteristic value, phi (x, f) is an SPOD characteristic mode, and W is a weighting matrix; and selecting the first-order mode phi (1) with the largest energy ratio as a dominant characteristic mode to be parameterized.
4. The method of modeling a separate flow wall pressure pulsation dominant structure based on spectral orthometric decomposition according to claim 1, wherein the physical parameterization model comprises an amplitude envelope function describing energy distribution and a phase evolution function describing convection characteristics, wherein the amplitude envelope function is configured to characterize asymmetric growth and decay cycles of shear layer instability, and the phase evolution function is configured to characterize variable wavenumber acceleration and constant velocity convection processes of wave packets.
5. The method for modeling a separate flow wall pressure pulsation dominant structure based on spectrum orthogonal decomposition according to claim 4, wherein the expression of the physical parameterization model is: Wherein, the The method is characterized in that the method is a parameterized mode after fitting, x is a flow direction coordinate, and G (x) is an amplitude envelope function; (x) I is an imaginary unit; The expression of the amplitude envelope function is: wherein A is peak amplitude when the wave packet is saturated, mu is flow direction position when the wave packet is saturated, sigma g is space growth scale of instability growth rate, sigma d is space attenuation scale of coherent structure crushing process; the expression of the phase evolution function is as follows: wherein k u =2ax+b is the wave number of linear change, k h is the constant hydrodynamic wave number, S (x) is the Sigmoid smooth switching function, a is the kinematic parameter to be identified, b is the kinematic parameter to be identified, and k h is the kinematic parameter to be identified.
6. The method for modeling a separate flow wall pressure pulsation dominant structure based on spectral orthometric decomposition according to claim 1, wherein said fitting the spatially dominant eigenmodes to the physical parameterized model compresses the spatially dominant eigenmodes into sparse physical parameter vectors by nonlinear regression solution, comprising: Fitting the spatially-dominant feature modality to the physical parameterized model to establish a parameter identification optimization problem, wherein the parameter identification optimization problem comprises defining a sparse physical parameter vector, performing a two-stage regression comprising performing a nonlinear least squares fit on modality magnitudes to determine geometric parameters, and performing a regression on model phases to determine kinematic parameters, the parameter identification objective being to minimize a norm error between the modality and the physical parameterized model; And solving the parameter identification optimization problem through nonlinear regression, and compressing the space dominant characteristic mode into a sparse physical parameter vector according to a solution result of the parameter identification optimization problem.
7. The method for modeling a separate flow wall pressure pulsation dominant structure based on spectral orthometric decomposition of claim 1, wherein the reconstructed separate flow wall pressure pulsation field is spatially and temporally reconstructed using the following formula: Wherein, p' rec (x, t) is a reconstructed pulsating pressure field, N f is the number of selected characteristic frequencies, lambda j is the SPOD characteristic value at the j-th frequency, and M j is the sparse physical parameter vector identified at the j-th frequency.
8. A separate flow wall pressure pulsation dominant structure modeling device based on spectrum orthogonal decomposition, characterized by comprising: the acquisition module is used for acquiring the full flow field of the separation flow under the preset working condition and/or the unsteady pressure data of the wall surface of the separation flow; the decomposition module is used for performing spectrum orthogonal decomposition on the unsteady pressure data, decoupling multi-scale dynamics features in a frequency domain, and extracting feature values under each feature frequency and corresponding space dominant feature modes; The modeling module is used for constructing a physical parameterization model based on a complex wave packet, fitting the spatial dominant characteristic mode to the physical parameterization model, and compressing the spatial dominant characteristic mode into a sparse physical parameter vector through nonlinear regression solution; and the reconstruction module is used for reconstructing a dominant component of the separated flow wall pressure pulse field based on the sparse physical parameter vector and the characteristic value under each characteristic frequency, and carrying out at least one of flow mechanism analysis and pneumatic reduced modeling according to the dominant component.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor executing the program to implement the split stream wall pressure pulsation dominant structure modeling method based on spectral quadrature decomposition of any of claims 1-7.
10. A computer readable storage medium having stored thereon a computer program or instructions, which when executed, is adapted to carry out the method for modeling a split stream wall pressure pulsation dominant structure based on spectral quadrature decomposition according to any of claims 1-7.

Description

Separated flow wall pressure pulsation leading structure modeling method based on spectrum orthogonal decomposition Technical Field The application relates to the technical field of hydrodynamic modeling and aeroacoustic analysis, in particular to a method for modeling a pressure pulsation leading structure of a wall surface of a separation flow based on spectrum orthogonal decomposition. Background High-resolution wind tunnel experiments generate massive high-dimensional unsteady data, and a reduced-order model aims at projecting high-dimensional dynamics to a low-dimensional subspace, but a frequency-band reduced-order method in the related technology cannot accurately represent asymmetric evolution and variable acceleration convection of a large-scale structure in pressure pulsation of a wall surface of a separation flow in a frequency domain, so that reconstruction accuracy is low. Disclosure of Invention The application provides a modeling method of a separated flow wall pressure pulsation leading structure based on spectrum orthogonal decomposition, which aims to solve the problems that in the related art, asymmetric evolution and variable acceleration convection of a large-scale structure in the separated flow wall pressure pulsation cannot be accurately represented in a frequency domain, so that reconstruction accuracy is low and the like. The embodiment of the first aspect of the application provides a method for modeling a pressure pulsation dominant structure of a wall surface of a separation flow based on spectrum orthogonal decomposition, which comprises the following steps of obtaining full flow field of the separation flow under a preset working condition and/or unsteady pressure data of the wall surface of the separation flow, performing spectrum orthogonal decomposition on the unsteady pressure data, decoupling multi-scale dynamics features in a frequency domain, extracting feature values and corresponding space dominant feature modes under each feature frequency, constructing a physical parameterization model based on complex wave packets, fitting the space dominant feature modes to the physical parameterization model, compressing the space dominant feature modes into sparse physical parameter vectors through nonlinear regression calculation, reconstructing dominant components of the pressure pulsation field of the separation flow based on the sparse physical parameter vectors and the feature values under each feature frequency, and performing at least one of flow mechanism analysis and pneumatic order reduction modeling according to the dominant components. Optionally, performing spectral orthogonal decomposition on the unsteady pressure data, decoupling multi-scale dynamics features in a frequency domain, and extracting feature values and corresponding space dominant feature modes under each feature frequency, wherein the method comprises estimating a cross spectral density matrix of the unsteady pressure data, decoupling the multi-scale dynamics features in the frequency domain according to the cross spectral density matrix, and extracting feature values and corresponding space dominant feature modes under each feature frequency. Optionally, the cross spectral density matrix is expressed as: wherein x is a space coordinate, x' is a space coordinate, f is a frequency; is the Fourier transform of the pressure, E ] The ] is a desired operator, is a conjugate transpose; is a cross spectral density matrix. The solution formula of the cross spectral density matrix is: Wherein lambda (f) is a eigenvalue, psi (x, f) is an SPOD eigenvector, and W is a weighting matrix. And selecting the first-order mode phi (1) with the largest energy ratio as a dominant characteristic mode to be parameterized. Optionally, the physical parameterized model comprises an amplitude envelope function describing the energy distribution and a phase evolution function describing the convection characteristics, wherein the amplitude envelope function is configured to characterize asymmetric growth and decay cycles of shear layer instability, and the phase evolution function is configured to characterize variable wavenumber acceleration and constant velocity convection processes of the wave packet. Optionally, the expression of the physical parameterized model is: Wherein, the The method is characterized in that the method is a parameterized mode after fitting, x is a flow direction coordinate, and G (x) is an amplitude envelope function; (x) I is an imaginary unit. The expression of the amplitude envelope function is: Wherein A is peak amplitude when the wave packet is saturated, mu is flow direction position when the wave packet is saturated, sigma g is space growth scale of instability growth rate, and sigma d is space attenuation scale of coherent structure crushing process. The expression of the phase evolution function is: wherein k u =2ax+b is the wave number of linear change, k h is the constant hydrodynamic wav