Multi-stage gas compressor pneumatic optimization method based on grid deformation

Abstract

The invention provides a multi-stage gas compressor pneumatic optimization method based on grid deformation, and relates to the field of gas compressors. The method comprises the following steps: performing data processing on a CGNS file of an original geometric flow field grid to obtain coordinates of all grids of the original geometric flow field grid; setting a grid boundary control method for the flow field of the multi-stage gas compressor; the original geometric flow field grid is deformed, and a deformed multi-stage gas compressor flow field grid is obtained; and performing CFD calculation on the basis of the deformed multi-stage compressor flow field grid: if the aerodynamic effect of the compressor reaches a preset target, outputting a geometric figure corresponding to the deformed multi-stage compressor flow field, otherwise, defining the deformed multi-stage compressor flow field grid as an original geometric flow field grid, and repeating deformation until the aerodynamic effect of the compressor reaches the preset target. According to the method, the full-three-dimensional CFD calculation precision is guaranteed, the number of optimized real evaluation samples is reduced, the calculation cost of the optimization iteration process is reduced, and full-three-dimensional rapid pneumatic optimization of the multi-stage gas compressor is achieved.

Description

Technical Field

[0001] The invention relates to the field of compressors, and in particular to an aerodynamic optimization method for a multi-stage compressor based on grid deformation. Background Art

[0002] The full three-dimensional aerodynamic optimization of multi-stage compressors has the characteristics of high optimization variable dimension and expensive evaluation. In the aerodynamic optimization process, each real sample evaluation includes a complete full three-dimensional CFD numerical simulation process, mainly including grid generation, flow field calculation, result processing, etc. Among them, the grid generation and flow field iterative calculation parts are the most time-consuming. In the full three-dimensional aerodynamic optimization of the compressor, the geometric parameterization method is generally used for the deformation of blades and flow channels, and the calculation grid is regenerated after the deformation is completed. For each real calculation, a grid generation calculation is required. As the number of blade rows increases, the area that needs to be controlled by the geometric parameterization method increases, and the calculation time increases sharply, making the geometric parameterization method itself a time-consuming operation. At the same time, the grid of each row of blades needs to be generated in sequence, and the time consumption is also significantly increased. Summary of the invention

[0003] In order to solve the technical problems existing in the prior art, an embodiment of the present invention provides a multi-stage compressor aerodynamic optimization method based on grid deformation, which reduces the number of optimization real evaluation samples, reduces the computational cost of the optimization iteration process, and realizes full three-dimensional rapid aerodynamic optimization of the multi-stage compressor. The technical solution is as follows:

[0004] A multi-stage compressor aerodynamic optimization method based on grid deformation, comprising:

[0005] S1. Acquire the original geometry of the multi-stage compressor, and perform data processing on the CGNS file of the flow field mesh of the original geometry to obtain the coordinates of all meshes of the flow field mesh of the original geometry;

[0006] S2. Based on the RBF grid characteristics, a multi-stage compressor flow field grid boundary control method is set;

[0007] S3, based on the radial basis function parameterization method, the multi-stage compressor flow field grid boundary control method and the coordinates of all grids of the original geometric flow field grid, deforming the original geometric flow field grid to obtain a deformed multi-stage compressor flow field grid;

[0008] S4. Perform CFD calculation based on the deformed multi-stage compressor flow field grid:

[0009] If the aerodynamic effect of the compressor reaches the preset target, the geometric figure corresponding to the deformed multi-stage compressor flow field is output; otherwise, the deformed multi-stage compressor flow field grid is defined as the flow field grid of the original geometry, and S3 is repeatedly executed.

[0010] Optionally, the coordinates of all grid points in the calculation domain of the multi-stage compressor flow field grid in S1 include: X-direction coordinates, Y-direction coordinates and Z-direction coordinates.

[0011] Optionally, the method for setting a multi-stage compressor flow field grid boundary control method based on RBF grid characteristics in S2 includes:

[0012] Set the blade deformation control points and the initial mesh boundary control points to obtain the control point set , the initial mesh vertices are set to the vertices of each mesh block;

[0013] Set the movement range of the parametric deformation control point, set the maximum movement amount, and set the target maximum deformation amount and the maximum number of iterations ;

[0014] Set the basis function and influence radius.

[0015] Optionally, the S3 is based on the radial basis function parameterization method, the multi-stage compressor flow field grid boundary control method and the coordinates of all grids of the original geometric flow field grid, deforming the original geometric flow field grid to obtain the deformed multi-stage compressor flow field grid, including:

[0016] S301, the CGNS file of the flow field grid of the original geometry, determining the flow field grid area to be deformed, and generating the original grid point set ;

[0017] S302, based on the multi-stage compressor flow field mesh boundary control method, performing RBF mesh deformation on the original geometric flow field mesh to obtain a deformed multi-stage compressor flow field mesh;

[0018] S303. According to the deformed multi-stage compressor flow field grid, obtain the parameter of the maximum deformation point of the boundary of the deformed multi-stage compressor flow field grid.

[0019] Optionally, the step S303 of obtaining the parameter of the maximum deformation point of the boundary of the deformed multi-stage compressor flow field grid according to the deformed multi-stage compressor flow field grid includes:

[0020] S3031, separating the deformed multi-stage compressor flow field grid boundary, calculating the absolute deformation of each grid point, and obtaining a set ;

[0021] S3032, define according to the set The maximum value in is the maximum value of the deformation of each point on the mesh boundary , the corresponding grid point is the maximum deformation grid point .

[0022] Optionally, if the aerodynamic efficiency of the compressor reaches a preset target, the geometric figure corresponding to the deformed multi-stage compressor flow field is output in S4; otherwise, the deformed multi-stage compressor flow field mesh is defined as the flow field mesh of the original geometry, and S3 is repeatedly executed, including:

[0023] When the maximum deformation of each point on the mesh boundary Target maximum deformation , or, the current iteration number > Maximum number of iterations , then export the current control point set and the deformed mesh to obtain the deformed geometry;

[0024] Otherwise, the maximum deformed grid point Incorporate control point set , keep the maximum deformation grid point Without deformation, the deformed multi-stage compressor flow field grid is defined as the flow field grid of the original geometry, and S3 is repeatedly executed.

[0025] Optionally, the radial basis function parameterization method includes:

[0026] Radial Basis Function The formula is formula (1):

[0027] ; (1)

[0028] in, For point To point The Euclidean distance between

[0029] point is the coordinate of the ith control point of the RBF, point is the target grid point in space;

[0030] Introducing polynomials , we get the basic mapping formula, which is formula (2), the polynomial Formula (3):

[0031] ; (2)

[0032] ; (3)

[0033] in, is the interpolation function;

[0034] Subscript Indicates the control point number of RBF;

[0035] is the weight of each basis function;

[0036] n represents the total number of control points;

[0037] is a linear polynomial The polynomial coefficients of ;

[0038] Displace the control points Interpolate to the original point and displace the control point The formula for interpolation to the original point is formula (4):

[0039] ; (4)

[0040] Since the polynomial The value at the control point is 0, which is a closed equation. Formula (5) is introduced as a constraint condition. Formula (5):

[0041] ; (5)

[0042] Polynomial The number of times is less than or equal to The number of times;

[0043] Calculate coefficients based on solving linear systems , and the weight of each basis function , interpolate the displacement of each original point and get the parameterized deformation.

[0044] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0045] This method aims to solve the problem of increased optimization time caused by repeated generation of flow field grids in full three-dimensional aerodynamic optimization of multi-stage compressors. A flow field grid deformation method based on radial basis functions and an adaptive point grid boundary constraint method are developed, and a data-driven aerodynamic optimization platform is established based on this method. The multi-stage compressor flow field grid deformation with adaptive boundary constraints replaces the grid generation process in the aerodynamic calculation process, greatly reduces the number of boundary control points, and saves the overall deformation time. While ensuring the accuracy of full three-dimensional CFD calculations, the number of real optimization evaluation samples is reduced, the computational cost of the optimization iteration process is reduced, and full three-dimensional rapid aerodynamic optimization of multi-stage compressors is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0047] Figure 1 is a flow chart of the method provided by an embodiment of the present invention;

[0048] Figure 2 It is a flow chart of the boundary control adaptive point placement algorithm provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] Unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. Similarly, words such as "one", "one" or "the" do not indicate a quantitative limitation, but indicate the existence of at least one. Words such as "include" or "comprise" mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. Words such as "connect" or "connected" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect.

[0051] It should be noted that the terms "up", "down", "left", "right", "front", "back", etc. used in the present invention are only used to indicate relative position relationships. When the absolute position of the object being described changes, the relative position relationship may also change accordingly.

[0052] For the multi-stage compressor optimization problem that requires multiple real evaluations, parameterized deformation and mesh generation operations will significantly increase the overall optimization time. To solve the above problems, a multi-stage compressor aerodynamic optimization method based on mesh deformation is proposed.

[0053] like Figure 1 to Figure 2As shown, a multi-stage compressor aerodynamic optimization method based on grid deformation includes:

[0054] S1. Acquire the original geometry of the multi-stage compressor, and perform data processing on the CGNS file of the flow field mesh of the original geometry to obtain the coordinates of all meshes of the flow field mesh of the original geometry.

[0055] Step S1 mainly processes the data of the multi-stage compressor flow field grid. The Python extension library developed based on the CGNS C language library is used to read and analyze the multi-stage compressor flow field grid CGNS file. The Python interface function in the extension library, the getValueByPath function, is used to extract and reconstruct the data of each data block grid node, including CoordinatesX (X-direction coordinates), CoordinatesY (Y-direction coordinates), and CoordinatesZ (Z-direction coordinates). All grid points in the calculation domain are obtained to facilitate the subsequent flow field grid deformation.

[0056] CGNS (CFD General Notation System) is a standard file format for storing and exchanging CFD data, developed by NASA, Boeing, Lockheed Martin, US Air Force and other institutions. It can be used to store geometric meshes, physical models, numerical results and other information in CFD calculations, and supports a variety of CFD solvers and post-processing software (CFview, Tecplot, etc.). CGNS files use structured binary storage to improve data reading and writing efficiency, and hierarchical storage is conducive to data stratification and management.

[0057] CGNS files consist of two main parts: the file header and the data area. The file header contains the metadata and structural information of the file, such as the CGNS version number, file description, mesh type, physical model, etc. The file header consists of structured text and can be read and written through a text editor or the corresponding functions in the CGNS library. The data area contains actual data, such as mesh coordinates, mesh topology, boundary conditions, flow field data, etc. The data area is a hierarchical structure consisting of multiple nodes and links, and each node can contain one or more data items. The main structure of the data area of ​​the CGNS file is shown in Table 1:

[0058] Table 1 Introduction to the basic structure of CGNS files

[0059]

[0060] Commercial CFD calculation software has the function of reading CGNS files. Some post-processing software such as Tecplot can read and export the grid point position and flow field information of CGNS files, but they cannot directly write the deformed geometry back to CGNS. CGNS officially open-sources the interface for reading and modifying CGNS files. pycgns is a Python extension library for reading and writing CGNS files. It is developed based on the CGNS C language library and provides a series of Python interface functions to easily read and modify the data in CGNS files.

[0061] The CFD calculation tool for aerodynamic optimization of multi-stage compressors uses the commercial software NUMECA, in which the Autogrid5 module is an automatic grid generation tool that generates grid information and stores it in the CGNS file. Use pycgns to read it, and use the getValueByPath function to read the data of each data block grid node, including CoordinatesX (X-direction coordinates), CoordinatesY (Y-direction coordinates), and CoordinatesZ (Z-direction coordinates), to obtain all grid points in the calculation domain, which is convenient for subsequent flow field grid deformation.

[0062] S2. Based on the RBF grid characteristics, a multi-stage compressor flow field grid boundary control method is set;

[0063] Set the blade deformation control points and the initial mesh boundary control points to obtain the control point set , set the initial mesh vertices to the vertices of each mesh block; set the movement range of the parametric deformation control point, and set the maximum movement amount, set the target maximum deformation amount and the maximum number of iterations ; Set the basis function and influence radius.

[0064] For the three-dimensional flow field mesh of the axial compressor, when the parametric method is used for three-dimensional deformation, only the mesh vertices are used as the limit control points, which cannot guarantee the accuracy of the calculation and the same efficiency change trend after deformation. Therefore, in order to improve the accuracy of the calculation after the direct flow field mesh is deformed, it is necessary to further study the boundary control method in the flow field mesh deformation to meet the needs of optimization calculation.

[0065] According to the characteristics of RBF deformation, the overall grid deformation is obtained by the deformation interpolation of the control point, and the influence range and size of the control point are controlled by the influence radius. When the control point is located on the grid node and does not move, the grid at the control point remains unchanged after deformation, and the displacement of the surrounding control points is reduced. Based on this characteristic, control points can be added and fixed on the boundaries of the grid block to reduce the displacement of the grid boundary. However, after the boundary control operation, the number of control points increases sharply, resulting in a significant increase in the calculation time and memory resources required for the boundary control method. This is because, during the RBF deformation process, the Euclidean distance between each grid point and each control point needs to be solved, that is, for each additional control point, all flow field grid points need to be traversed and calculated.

[0066] S3, based on the radial basis function parameterization method, the multi-stage compressor flow field grid boundary control method and the coordinates of all grids of the original geometric flow field grid, the original geometric flow field grid is deformed to obtain a deformed multi-stage compressor flow field grid:

[0067] S301, the CGNS file of the flow field grid of the original geometry, determining the flow field grid area to be deformed, and generating the original grid point set ;

[0068] S302, based on the multi-stage compressor flow field mesh boundary control method, performing RBF mesh deformation on the original geometric flow field mesh to obtain a deformed multi-stage compressor flow field mesh;

[0069] S303, according to the deformed multi-stage compressor flow field grid, obtain the parameter of the maximum deformation point of the boundary of the deformed multi-stage compressor flow field grid:

[0070] S3031, separating the deformed multi-stage compressor flow field grid boundary, calculating the absolute deformation of each grid point, and obtaining a set ;

[0071] S3032, define according to the set The maximum value in is the maximum value of the deformation of each point on the mesh boundary , the corresponding grid point is the maximum deformation grid point .

[0072] S4. Perform CFD calculation based on the deformed multi-stage compressor flow field grid:

[0073] If the aerodynamic effect of the compressor reaches the preset target, the geometric figure corresponding to the deformed multi-stage compressor flow field is output; otherwise, the deformed multi-stage compressor flow field grid is defined as the flow field grid of the original geometry, and S3 is repeatedly executed, including:

[0074] When the maximum deformation of each point on the mesh boundary Target maximum deformation , or, the current iteration number > Maximum number of iterations , then export the current control point set and the deformed mesh to obtain the deformed geometry;

[0075] Otherwise, the maximum deformed grid point Incorporate control point set , keep the maximum deformation grid point Without deformation, the deformed multi-stage compressor flow field grid is defined as the flow field grid of the original geometry, and S3 is repeatedly executed.

[0076] With respect to the mesh deformation of the multi-stage compressor flow field based on radial basis function, the present invention mainly includes the following implementation methods:

[0077] The radial basis function (RBF) deformation method can handle meshes of arbitrary shapes, with relatively smooth deformation and without changing the topological structure of the mesh. It is mostly used for unsteady calculations. The basic idea of ​​this method is to achieve mesh deformation by representing each point in the mesh as a linear combination of radial basis functions, and then recalculating the radial basis function coefficients of each point according to the position of the deformed control points. RBF direct operation flow field mesh deformation technology is mostly used in refined small deformations of outflow or single-row blades. There is a research gap in the mesh deformation direction of multi-row blades and multi-stage compressors. Therefore, this implementation makes the following improvements to RBF applied to compressors:

[0078] The principle of the RBF method is based on a mapping : . Radial Basis Function The formula is formula (1):

[0079] ; (1)

[0080] in, For point To point The Euclidean distance between

[0081] point is the coordinate of the ith control point of the RBF, point is the target grid point in space;

[0082] Formula (1) shows that the value of the RBF basis function depends only on the Euclidean distance from the target point to the source node. In order to achieve translation and rotation deformation, the polynomial is introduced , we get the basic mapping formula, which is formula (2), the polynomial Formula (3):

[0083] ; (2)

[0084] ; (3)

[0085] in, is the interpolation function;

[0086] Subscript Indicates the control point number of RBF;

[0087] is the weight of each basis function;

[0088] n represents the total number of control points;

[0089] and is a linear polynomial The polynomial coefficients of ;

[0090] To perform the deformation operation, the control points are displaced Interpolate to the original point and solve the displacement equation of the source node , shift the control point The formula for interpolation to the original point is formula (4):

[0091] ; (4)

[0092] Since the polynomial The value at the control point is 0, which is a closed equation. Formula (5) is introduced as a constraint condition. Formula (5):

[0093] ; (5)

[0094] Polynomial The number of times is less than or equal to The number of times;

[0095] Calculate coefficients based on solving linear systems , and the weight of each basis function , interpolate the displacement of each original point and obtain the parameterized deformation;

[0096] After reading the CGNS file to obtain the mesh nodes, since RBF deformation will not change the mesh topology and the mesh block information preset by NUMECA, in order to ensure the quality of the mesh and the continuity between the topology and mesh blocks, all mesh points need to be introduced into the RBF deformation control.

[0097] Among them, the RBF direct flow field mesh deformation process includes:

[0098] (1) Read the CGNS file and use the pycgns library function to obtain the position of each grid point, the grid block array structure, the CGNS file tree, and the position of the grid node through the grid block name index. After obtaining the coordinates of all grid points, flatten and combine each grid block. Since RBF solves each grid point in the deformation area independently, the grid points can be imported in any order to form the overall flow field grid.

[0099] (2) Determine the grid blocks that need to be deformed. To ensure the quality of the grid, select all grid blocks in the blade discharge flow domain and add them to the overall RBF deformation. Introduce the moving blade grid block as a whole, move the blade leading edge control point, and move the blade inlet grid block and blade grid block to ensure the quality of the grid.

[0100] (3) Set the grid and deformation control points. In order to make the blade geometric deformation intuitive and effective and ensure the accuracy of the CFD calculation of the flow field grid after deformation, it is necessary to select control points in the innermost layer of the blade grid (blade wall grid) and the boundary layer of each grid block. The control points on the blade are used to directly control the grid deformation, and the control points on the boundary grid layer are used to ensure the overall flow field grid structure. When performing RBF parametric deformation, keep the control points at the vertices fixed to limit the frame deformation of the flow field grid, and move the control points on the blade to perform parametric deformation of the blade;

[0101] (4) Perform RBF deformation. After determining the grid and control points to be deformed, first calculate the Euclidean distance from each grid point to each control point, determine the radial basis function and influence radius, then represent each point as a linear combination of radial basis functions, recalculate the radial basis function coefficients of each point based on the position of the deformed control points, and calculate the new coordinates of each point;

[0102] (5) After obtaining the deformed overall flow field grid, the overall flow field grid is divided into blocks according to the array structure of each grid block in the CGNS file, and the corresponding grid blocks in the file tree are directly modified based on the new grid node coordinate data. The modified file tree is exported to the CGNS file to complete the deformation of the overall flow field grid of the multi-stage compressor.

[0103] The steps of the multi-stage compressor flow field grid boundary control method combined with RBF deformation mainly include:

[0104] Initialization module, reads CGNS file, determines the flow field grid area to be deformed, and generates the original grid point set ; Set the blade deformation control points and initial mesh boundary control points to generate a control point set , set the initial mesh vertices to the vertices of each mesh block; set the movement range of the parametric deformation control point, and set the movement amount to the maximum value, set the target maximum deformation , maximum number of iterations ; Determine the basis function and influence radius, and complete the initialization work before RBF deformation.

[0105] Perform RBF mesh deformation.

[0106] Get the maximum deformation point of the boundary. After mesh deformation, separate the mesh boundary and calculate the absolute deformation of each mesh point to get the set ,gather The maximum value is the maximum value of the deformation of each point on the grid boundary , get the maximum deformation grid point .

[0107] Determine the maximum value of boundary mesh deformation Is it less than the target maximum deformation? , whether the current generation has reached the maximum number of iterations :

[0108] When the maximum deformation of each point on the mesh boundary Target maximum deformation Or, the current generation number > Maximum number of iterations , then export the current control point set and the deformed mesh to obtain the deformed geometry; otherwise, the maximum deformed mesh point Incorporate control point set , keep the maximum deformation grid point Without deformation, the deformed multi-stage compressor flow field grid is defined as the flow field grid of the original geometry, and the maximum deformation point of the boundary is obtained again.

[0109] The principle of the present invention is to process the data of the multi-stage compressor flow field grid. The Python extension library developed based on the CGNS C language library is used to read and analyze the multi-stage compressor flow field grid CGNS file, and the data of each data block grid node, including CoordinatesX (X-direction coordinates), CoordinatesY (Y-direction coordinates), and CoordinatesZ (Z-direction coordinates) are extracted and reconstructed through the Python interface function in the extension library-getValueByPath function, so as to obtain all grid points in the calculation domain, which is convenient for the subsequent execution of flow field grid deformation. Based on the radial basis function (RBF) parameterization method and the coordinates of all grid points in the computational domain, the grid of the multi-stage compressor flow field is deformed. After obtaining the grid node data of the CGNS file of the multi-stage compressor flow field grid, the grid blocks that need to be deformed are determined, and all grid points are introduced into the RBF deformation control to ensure the quality of the grid and the continuity between the topology and grid blocks; the grid and deformation control points are set, the RBF deformation operation is performed, and the new coordinates of each grid node are calculated; according to the array structure of each grid block in the CGNS file, the file tree formed by the new coordinates of the grid nodes is exported to the CGNS file, and the construction of the new CGNS file is completed, thereby realizing the overall flow field grid deformation of the multi-stage compressor. Implement the boundary control of the multi-stage compressor flow field grid deformation, establish the grid boundary control adaptive point distribution algorithm, obtain the maximum deformation point of the grid boundary in each RBF deformation iteration process, merge the maximum deformation grid point into the control point set, and fix it; when the boundary control reaches the set accuracy or reaches the maximum number of iterations, export the current control point set for subsequent optimization operations, export the deformed grid, and generate a new CGNS file. This limits the total number of control points during mesh deformation of the multi-stage compressor flow field, and reduces the calculation time and memory resources of the mesh deformation of the multi-stage compressor while meeting the calculation accuracy of aerodynamic optimization.

[0110] The main steps are as follows: according to the original geometry, a full three-dimensional CFD calculation flow field grid is generated by the calculation software, and the initial calculation is performed to obtain the initial calculation results. Among them, the aerodynamic calculation tool uses the commercial software NUMECA, and the Autogrid5 module is used to generate the initial grid. The flow field calculation is performed using the Fine module. The blade parameterized deformation control points are set. All flow field grids except the overall inlet and outlet sections are introduced. The deformation of the moving and stationary blade flow field grids is controlled separately, and the grid boundary control uses the adaptive point distribution boundary control method. The flow field grid is directly deformed and parameterized. Through the adaptive point distribution boundary control method, the moving blade and stationary blade grid blocks are iterated separately, focusing on ensuring the maximum deformation at the rotor / stationary interface. The displacement of the blade control point is set to the maximum movement range. After the adaptive boundary control iteration, the adaptive boundary control point set is stored. In the subsequent optimization calculation, the adaptive control point set is directly imported, the overall grid deformation is calculated, and the optimization variables and parameters are set. The data-driven algorithm based on the pre-screening strategy is used for optimization. The proxy model selects the Kriging model, and the absolute value of efficiency is used as the pre-screening criterion.

[0111] This method aims to solve the problem of increased optimization time caused by repeated generation of flow field grids in full three-dimensional aerodynamic optimization of multi-stage compressors. A flow field grid deformation method based on radial basis functions and an adaptive point grid boundary constraint method are developed, and a data-driven aerodynamic optimization platform is established based on this method. The multi-stage compressor flow field grid deformation with adaptive boundary constraints replaces the grid generation process in the aerodynamic calculation process, greatly reduces the number of boundary control points, and saves the overall deformation time. While ensuring the accuracy of full three-dimensional CFD calculations, the number of real optimization evaluation samples is reduced, the computational cost of the optimization iteration process is reduced, and full three-dimensional rapid aerodynamic optimization of multi-stage compressors is achieved.

[0112] The following points need to be explained:

[0113] (1) The drawings of the embodiments of the present invention only involve structures related to the embodiments of the present invention. Other structures may refer to conventional designs.

[0114] (2) For the sake of clarity, in the drawings used to describe the embodiments of the present invention, the thickness of layers or regions is exaggerated or reduced, that is, these drawings are not drawn according to the actual scale. It is understood that when an element such as a layer, film, region or substrate is referred to as being "on" or "under" another element, the element may be "directly" "on" or "under" the other element or there may be intermediate elements.

[0115] (3) In the absence of conflict, the embodiments of the present invention and the features therein may be combined with each other to obtain new embodiments.

[0116] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. The protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A multi-stage compressor aerodynamic optimization method based on grid deformation, It is characterized in that include: S1. Acquire the original geometry of the multi-stage compressor, and perform data processing on the CGNS file of the flow field mesh of the original geometry to obtain the coordinates of all meshes of the flow field mesh of the original geometry; S2. Based on the RBF grid characteristics, a multi-stage compressor flow field grid boundary control method is set; S3, based on the radial basis function parameterization method, the multi-stage compressor flow field grid boundary control method and the coordinates of all grids of the original geometric flow field grid, deforming the original geometric flow field grid to obtain a deformed multi-stage compressor flow field grid; S4. Perform CFD calculation based on the deformed multi-stage compressor flow field grid: If the aerodynamic effect of the compressor reaches the preset target, the geometric figure corresponding to the deformed multi-stage compressor flow field is output; otherwise, the deformed multi-stage compressor flow field grid is defined as the flow field grid of the original geometry, and S3 is repeatedly executed.

2. The multi-stage compressor aerodynamic optimization method based on grid deformation according to claim 1, It is characterized in that The coordinates of all grid points in the calculation domain of the multi-stage compressor flow field grid in S1 include: X-direction coordinates, Y-direction coordinates and Z-direction coordinates.

3. The multi-stage compressor aerodynamic optimization method based on grid deformation according to claim 2, It is characterized in that The method for setting the grid boundary control of the multi-stage compressor flow field based on the RBF grid characteristics of S2 includes: Set the blade deformation control points and the initial mesh boundary control points to obtain the control point set , the initial mesh vertices are set to the vertices of each mesh block; Set the movement range of the parametric deformation control point, set the maximum movement amount, and set the target maximum deformation amount and the maximum number of iterations ; Set the basis function and influence radius.

4. The multi-stage compressor aerodynamic optimization method based on grid deformation according to claim 3, It is characterized in that The S3 is based on the radial basis function parameterization method, the multi-stage compressor flow field grid boundary control method and the coordinates of all grids of the original geometric flow field grid, deforming the original geometric flow field grid to obtain the deformed multi-stage compressor flow field grid, including: S301, the CGNS file of the flow field grid of the original geometry, determining the flow field grid area to be deformed, and generating the original grid point set ; S302, based on the multi-stage compressor flow field mesh boundary control method, performing RBF mesh deformation on the original geometric flow field mesh to obtain a deformed multi-stage compressor flow field mesh; S303. According to the deformed multi-stage compressor flow field grid, obtain the parameter of the maximum deformation point of the boundary of the deformed multi-stage compressor flow field grid.

5. The multi-stage compressor aerodynamic optimization method based on grid deformation according to claim 4, It is characterized in that The step S303 of obtaining the parameters of the maximum deformation point of the boundary of the deformed multi-stage compressor flow field grid according to the deformed multi-stage compressor flow field grid includes: S3031, separating the deformed multi-stage compressor flow field grid boundary, calculating the absolute deformation of each grid point, and obtaining a set ; S3032, define according to the set The maximum value in is the maximum value of the deformation of each point on the mesh boundary , the corresponding grid point is the maximum deformation grid point .

6. The multi-stage compressor aerodynamic optimization method based on grid deformation according to claim 5, It is characterized in that If the aerodynamic efficiency of the compressor reaches a preset target, the geometric figure corresponding to the deformed multi-stage compressor flow field is output in S4; otherwise, the deformed multi-stage compressor flow field grid is defined as the flow field grid of the original geometry, and S3 is repeatedly executed, including: When the maximum deformation of each point on the mesh boundary Target maximum deformation , or, the current iteration number > Maximum number of iterations , then export the current control point set and the deformed mesh to obtain the deformed geometry; Otherwise, the maximum deformed grid point Incorporate control point set , keep the maximum deformation grid point Without deformation, the deformed multi-stage compressor flow field grid is defined as the flow field grid of the original geometry, and S3 is repeatedly executed.

7. The multi-stage compressor aerodynamic optimization method based on grid deformation according to claim 6, It is characterized in that The radial basis function parameterization method comprises: Radial Basis Function The formula is formula (1): ；（1） in, For point To point The Euclidean distance between point is the coordinate of the ith control point of the RBF. is the target grid point in space; Introducing polynomials , we get the basic mapping formula, which is formula (2), the polynomial Formula (3): ；（2） ；（3） in, is the interpolation function; Subscript Indicates the control point number of RBF; is the weight of each basis function; n represents the total number of control points; is a linear polynomial The polynomial coefficients of ; Displace the control points Interpolate to the original point and displace the control point The formula for interpolation to the original point is formula (4): ；（4） Since the polynomial The value at the control point is 0, which is a closed equation. Formula (5) is introduced as a constraint condition: ;（5） Polynomial The number of times is less than or equal to The number of times; Calculate coefficients based on solving linear systems , and the weight of each basis function , interpolate the displacement of each original point and get the parameterized deformation.

Images