Multi-level optimization design method for multi-stage axial flow compressor based on genetic algorithm

Abstract

The present application relates to the technical field of aero-engine compressor design, and particularly relates to a multistage axial compressor multilevel optimization design method based on a genetic algorithm, which comprises the following steps: determining one-dimensional optimization variables of the compressor according to design indexes, performing performance optimization through a genetic algorithm combined with a one-dimensional characteristic calculation model of the axial compressor, and obtaining a one-dimensional design scheme of the compressor; further, determining two-dimensional optimization variables of the compressor according to an isoperimetric design principle, performing performance optimization again through the genetic algorithm combined with a two-dimensional characteristic calculation model of the axial compressor, and finally determining a two-dimensional design scheme of the compressor that meets the design requirements; inputting the two-dimensional design scheme of the compressor into a three-dimensional modeling calculation model of the axial compressor, obtaining a three-dimensional geometric model of the completed compressor, and performing three-dimensional numerical simulation verification on the scheme. The present application shortens the compressor design process, reduces the input of the designers, and improves the efficiency of the optimization design.

Description

Technical Field

[0001] The present invention relates to the technical field of aero-engine compressor design. Background Art

[0002] Axial-flow compressors are core components of important equipment such as aircraft engines and large industrial gas turbines. For example, in aircraft engines, axial-flow compressors transfer mechanical energy to the flowing gas, completing the gas compression process in the engine's thermodynamic cycle. They also increase the air pressure entering the combustion chamber, thereby improving the engine's thermodynamic cycle efficiency. The quality of axial-flow compressor design directly impacts engine performance, making axial-flow compressor design a critical core technology that must be mastered.

[0003] In existing technology, axial flow compressors consist of multiple stages, each consisting of a row of moving blades and a row of stationary blades, with inlet guide vanes (IGVs) and outlet guide vanes (OGVs) installed at the inlet and outlet, respectively. The working fluid in the compressor is first accelerated by the moving blades, decelerated in the stationary blade passages, and the dynamic pressure head of the airflow is converted into static pressure head. This process is repeated in multiple stages of blades until the required total pressure ratio is achieved. Due to the inherent adverse pressure gradient, high three-dimensionality, and high degree of unsteadiness of the airflow in axial flow compressors, the development of axial flow compressors is extremely difficult. Multi-stage, high-pressure ratio axial flow compressors are particularly difficult to match between stages and optimize performance under all operating conditions. Although the design system is constantly being improved, due to the simultaneous improvement of design specifications, multi-stage axial flow compressor design remains a bottleneck technology for engines, generally requiring multiple rounds of modifications and repeated iterations to meet design requirements.

[0004] The traditional design approach for axial compressors involves a continuous, iterative process of narrowing the design scope from one-dimensional to three-dimensional, based on design requirements. The design process primarily consists of one-dimensional flow path design, one-dimensional characteristic analysis, S2 flow surface design, three-dimensional blade shaping, two-dimensional flow performance analysis, and three-dimensional numerical simulation performance verification. During this process, the one-dimensional flow path design, S2 flow surface design, and three-dimensional blade shaping require a large number of input parameters to the corresponding programs. In particular, the two-dimensional meridional flow surface design phase requires the designer to specify the spanwise parameter distributions for each row, including but not limited to the total pressure ratio and isentropic efficiency spanwise distribution for the moving blade row, and the total pressure recovery coefficient and outlet airflow angle spanwise distribution for the stationary blade row. The three-dimensional shaping phase then requires the specification of the spanwise cross-sectional profile geometry for each blade row, including but not limited to the profile centerline type, maximum relative thickness, maximum relative thickness location, leading and trailing edge thickness, angle of attack, and trailing angle. The two-dimensional meridional flow surface design results are then combined with the three-dimensional blade shaping results to perform a two-dimensional flow analysis and calculation to verify that the design objectives are met. With the increase in the number of compressor stages and the continuous improvement of design indicators, the work of given parameter distribution and iterative debugging is a huge test for designers.

[0005] Generally speaking, the more compressor stages there are, the higher the input variable dimension (i.e., the number of variables), and the greater the design difficulty. For an eight-stage axial-flow compressor, for example, designers must rely on empirically determined parameter dimensions reaching into the thousands. The design process for an axial-flow compressor is equivalent to finding the permutations and combinations that meet design requirements from a thousand-dimensional set of parameters. Obviously, finding the optimal solution from such a high-dimensional set of parameters in a short period of time is difficult. As design requirements continue to rise, the challenges facing designers' experience and skills also intensify, leading to an increasing number of iterations and a longer design cycle. Summary of the Invention

[0006] The purpose of the present invention is to avoid the shortcomings of the existing technology and provide a multi-level optimization design method for a multi-stage axial flow compressor based on a genetic algorithm, which greatly reduces the design dimensions of the axial flow compressor and shortens the design cycle while ensuring design reliability and combining the genetic algorithm optimization algorithm.

[0007] To achieve the above object, the technical solution adopted by the present invention is: a multi-level optimization design method for a multi-stage axial flow compressor based on a genetic algorithm, comprising the following steps:

[0008] Step 1: Pressure ratio based on the axial flow compressor design point ,efficiency , the overall surge margin of the whole machine The total pressure at the outlet of the axial flow compressor is determined based on the overall design indicators of the compressor and the initial size constraints of the compressor. , outlet total temperature And the whole machine rim work ;

[0009] Then, according to the whole machine rim work The size of the axial flow compressor is used to determine the number of stages, and the load coefficient and reaction force of each stage of the axial flow compressor are allocated. The inlet and outlet axial velocity of each stage of the moving blade, the radius of the middle section, and the aspect ratio and consistency of each row are selected. Then, the flow path coordinates of the axial flow compressor, the velocity triangle at the middle section of each stage of the moving blade and the stator blade, the inlet and outlet airflow angle, and the chord length are obtained by using the one-dimensional design calculation model of the axial flow compressor.

[0010] The initial values ​​of the angle of attack and the angle of lagging at the middle section of each level of moving blades and stationary blades are set to 0. The inlet and outlet structural angles at the middle section of each level of moving blades and stationary blades are determined according to the inlet and outlet airflow angles. At the same time, the maximum relative thickness of the blades and the position of the maximum relative thickness are given.

[0011] Thus, based on the one-dimensional characteristic calculation model of the axial flow compressor, the one-dimensional total pressure ratio, the overall surge margin and efficiency of the axial flow compressor are obtained;

[0012] Taking efficiency as the optimization objective, total pressure ratio and overall surge margin as optimization constraints, and based on the initial size constraints of the compressor, the flow path coordinates of the axial compressor, the inlet and outlet structural angles at the middle section of each stage of the moving and stationary blades, the density, chord length, the maximum relative thickness of the blades, and the maximum relative thickness position are used as optimization variables. Parameter optimization is performed based on a genetic algorithm until the optimization objective meets the overall design indicators. The updated optimization variables obtained are the one-dimensional design scheme of the compressor.

[0013] Step 2: Based on the principle of equal annular volume design and the one-dimensional design scheme of the compressor, determine the inlet and outlet structural angles of at least 11-21 radial sections of the axial flow compressor blades at each stage, and specify the maximum relative thickness and the position of the maximum relative thickness of the blades;

[0014] Thus, based on the two-dimensional characteristic calculation model of the axial flow compressor, the two-dimensional integrated surge margin, total pressure ratio and efficiency of the axial flow compressor are obtained;

[0015] Taking efficiency as the optimization objective, total pressure ratio and overall engine margin as optimization constraints, and the inlet and outlet structural angles of at least 11-21 sections, the maximum relative thickness of the blades, and the maximum relative thickness position as optimization variables, a genetic algorithm is used to optimize the parameters until the overall design indicators are met. The updated optimization variables obtained are the two-dimensional design scheme of the compressor.

[0016] Step 3: Input the two-dimensional design scheme of the compressor into the three-dimensional modeling calculation model of the axial flow compressor to obtain the designed three-dimensional geometric model of the compressor.

[0017] Furthermore, the outlet total pressure of the axial flow compressor in step 1 is , outlet total temperature and the whole machine rim work The calculation process is:

[0018] First, calculate the total outlet pressure for:

[0019] ,

[0020] Where, is the total inlet pressure of the axial compressor;

[0021] Then, according to the isentropic efficiency relationship of the axial flow compressor, the total outlet temperature is determined as:

[0022] ,

[0023] Where, is the total inlet temperature of the axial compressor;

[0024] Finally, the wheel rim power of the whole machine is determined for:

[0025] ,

[0026] Where, is the constant-pressure specific heat capacity of the axial compressor.

[0027] Furthermore, the flow channel coordinates described in step 1 are obtained by fitting the fourth-order Bezier curve equation to ensure the smoothness and continuity of the flow channel coordinates. The specific formula for fitting the flow channel coordinates is:

[0028] ,

[0029] Where, is the flow channel coordinate, is the control parameter, and , There are five control points of the flow channel coordinates, and the optimization boundaries of the five control points are within ±20% of the initial value of each control point.

[0030] Furthermore, the optimization boundaries of the blade inlet and outlet structural angles, density and chord length at the middle sections of the moving blades and stationary blades at each stage in the optimization variables described in step one are between ±20% of the initial values ​​of the blade inlet and outlet structural angles, density and chord length at the middle sections of the moving blades and stationary blades at each stage.

[0031] Furthermore, in the genetic algorithm of steps one and two, if the optimization goal is not achieved or the optimization history is not fully converged, the best design is selected from the current optimization population, and the boundaries of the optimization variables are updated based on the selected design scheme. The boundaries of the optimization variables are selected as 50% of the size of the boundaries of the variables optimized last time. The optimization is continued until the optimization goal is achieved, the one-dimensional or two-dimensional design of the compressor is terminated, and a one-dimensional or two-dimensional design scheme of the compressor that meets the optimization goal is output.

[0032] Furthermore, the genetic algorithm in step 1 and the one-dimensional characteristic calculation model of the axial flow compressor are transferred via a Python script to achieve parameter transfer;

[0033] Similarly, the genetic algorithm in step 2 and the two-dimensional characteristic calculation model of the axial flow compressor are also transferred through Python scripts to realize parameter transfer, thus realizing the construction and connection of the global optimization design platform of the axial flow compressor.

[0034] Furthermore, in step 2, the specific steps of obtaining the two-dimensional integrated surge margin, total pressure ratio, and efficiency of the axial flow compressor based on the two-dimensional characteristic calculation model of the axial flow compressor are as follows:

[0035] First, according to the principle of equal annular design, the absolute circumferential velocity of the blade root and blade tip of each stage of the axial flow compressor is determined. and , respectively expressed as:

[0036] ,

[0037] ,

[0038] Then, assuming that the radial distribution of the axial velocity on the blade is the same, the velocity triangles at the blade root and blade tip are determined. Based on the Pythagorean theorem, the inlet and outlet airflow angles at the blade root, the middle section of each blade, and the blade tip are further determined.

[0039] Then, according to the Lieblien reference lagging angle model, the inlet and outlet structural angles of at least 11-21 sections in the radial direction of each stage of the axial flow compressor blade are determined. The specific calculation derivation formulas include:

[0040] ,

[0041] Where, is the reference trailing angle of the blade, is the shape factor of the blade, is the relative thickness factor of the blade, The reference trailing angle at 10% blade relative thickness under 0 camber, is the curvature coefficient of the blade, is the bending angle of the blade;

[0042] in,

[0043] ,

[0044] Where, For consistency;

[0045] ,

[0046] Where, is the relative thickness of the leaf;

[0047] ,

[0048] Where, is the camber coefficient of the NACA 65 series blades, is an index, and and The expressions are:

[0049] ,

[0050] ,

[0051] ,

[0052] ,

[0053] Where, is the inlet airflow angle at the middle section of each stage of moving and stationary blades; is the outlet airflow angle at the middle section of each stage of moving and stationary blades;

[0054] Next, the inlet and outlet structural angles of at least 11-21 sections of the axial compressor blades at each stage in the radial direction are fitted to obtain:

[0055] ,

[0056] Where x is the relative radius of the blade root, the middle section and the tip of each blade at each level in the radial direction; y is the inlet structural angle of the blade root, the middle section and the tip of each blade at each level; a, b and c are fitting coefficients;

[0057] At this time, the inlet and outlet structural angles of all radial sections of blades at each stage are determined through nonlinear uniform interpolation, and then the maximum relative thickness, maximum relative thickness position, leading and trailing edge relative thickness and maximum relative deflection position in the two-dimensional characteristic calculation model of the axial flow compressor are given. Based on the two-dimensional characteristic calculation model of the axial flow compressor, the two-dimensional comprehensive surge margin, total pressure ratio and efficiency of the axial flow compressor are calculated.

[0058] Furthermore, the initial values ​​of the maximum relative thickness position and the maximum relative deflection position of the blade are given as 50% of the chord length, and the initial value of the relative thickness of the leading and trailing edges is given as 10% of the maximum relative thickness; the optimization boundaries of the maximum relative thickness position, the maximum relative deflection position, and the relative thickness of the leading and trailing edges of the blade are between ±10% of the initial values ​​of the maximum relative thickness position, the maximum relative deflection position, and the relative thickness of the leading and trailing edges.

[0059] Furthermore, the optimization constraints in the genetic algorithm of step 2 also include: blocking flow, diffusion factor, and load coefficient.

[0060] The beneficial effects of the present invention are as follows: the present invention provides a multi-level optimization design method for a multi-stage axial compressor based on a genetic algorithm, provides an implementation step and design process for optimizing the design of an axial compressor by combining a one-dimensional and two-dimensional analysis program with a genetic algorithm, and can build a corresponding optimization design platform; through the method described in the present invention, the use of a two-dimensional flow design program and a three-dimensional modeling program that give a large number of parameters according to the design experience of the designer in the traditional axial compressor design process, and the repeated calling of the one-dimensional design program are avoided; and the compressor design scheme is determined by a global optimization genetic algorithm combined with constraints extracted from design experience; through parallel optimization, combined with the automation of the optimization design platform, the compressor design process is shortened, the input of designers is reduced, and the efficiency of optimization design is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 It is a schematic flow diagram of the present invention;

[0062] Figure 2 This is a schematic diagram of the meridional flow surface corresponding to the one-dimensional design scheme for initialization of a three-stage axial flow compressor when implementing the present invention;

[0063] Figure 3 1. It is a schematic diagram comparing the meridional flow surface of a three-stage axial flow compressor before and after optimization in the one-dimensional design stage when implementing the present invention;

[0064] Figure 4 This is a comparison diagram of the compressor converted flow-pressure ratio characteristics before and after optimization of the one-dimensional design stage of a three-stage axial flow compressor in the implementation of the present invention;

[0065] Figure 5 This is a comparison diagram of the compressor pressure ratio-efficiency characteristics before and after optimization of the one-dimensional design stage of a three-stage axial flow compressor in the implementation of the present invention;

[0066] Figure 6 is the convergence history of the optimization process of the one-dimensional design phase of a three-stage axial flow compressor when implementing the present invention;

[0067] Figure 7 This is a comparison diagram of the compressor converted flow-pressure ratio characteristics before and after optimization in the two-dimensional design stage of a three-stage axial flow compressor when implementing the present invention;

[0068] Figure 8 This is a comparison diagram of the compressor pressure ratio-efficiency characteristics before and after optimization of the two-dimensional design stage of a three-stage axial flow compressor in the implementation of the present invention;

[0069] Figure 9 It is the optimization convergence history of a three-stage axial flow compressor in the two-dimensional design stage when the present invention is implemented. DETAILED DESCRIPTION

[0070] The principles and features of the present invention are described below with reference to the accompanying drawings. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention.

[0071] To achieve the above object, the present invention provides the following specific implementation: a multi-level optimization design method for a multi-stage axial flow compressor based on a genetic algorithm, comprising the following steps:

[0072] S01, pressure ratio based on axial flow compressor design point ,efficiency , the overall surge margin of the whole machine The total pressure at the outlet of the axial flow compressor is determined based on the overall design indicators of the compressor and the initial size constraints of the compressor. , outlet total temperature And the whole machine rim work , specifically:

[0073] Calculated total outlet pressure for:

[0074] ,

[0075] Where, is the total inlet pressure of the axial compressor;

[0076] Then, according to the isentropic efficiency relationship of the axial flow compressor, the total outlet temperature is determined as:

[0077] ,

[0078] Where, is the total temperature at the inlet of the axial compressor;

[0079] Finally, the wheel rim power of the whole machine is determined for:

[0080] ,

[0081] Where, is the constant-pressure specific heat capacity of the axial compressor.

[0082] S02, according to the whole machine rim power The size of the axial flow compressor is used to determine the number of stages, and the load coefficient and reaction force of each stage of the axial flow compressor are allocated. The inlet and outlet axial velocity of each stage of the moving blade, the radius of the middle section, and the aspect ratio and consistency of each row are selected. Then, the flow path coordinates of the axial flow compressor, the velocity triangle at the middle section of each stage of the moving blade and the stator blade, the inlet and outlet airflow angle, and the chord length are obtained by using the one-dimensional design calculation model of the axial flow compressor.

[0083] The flow channel coordinates are obtained by fitting the fourth-order Bezier curve equation to ensure the smoothness and continuity of the flow channel coordinates. The specific formula for fitting the flow channel coordinates is:

[0084] ,

[0085] Where, is the flow channel coordinate, is the control parameter, and , There are five control points of the flow channel coordinates, and the optimization boundaries of the five control points are between ±20% of the initial value of each control point.

[0086] S03. Initial values ​​of the angle of attack and the angle of lagging at the middle section of each level of the moving blades and the stationary blades are set to 0. Based on the inlet and outlet airflow angles, the inlet and outlet structural angles at the middle section of each level of the moving blades and the stationary blades are determined. At the same time, the maximum relative thickness of the blades and the position of the maximum relative thickness are determined.

[0087] Therefore, based on the one-dimensional characteristic calculation model of the axial flow compressor, the one-dimensional total pressure ratio, the overall surge margin and efficiency of the axial flow compressor are obtained.

[0088] S04. Using efficiency as the optimization objective, total pressure ratio and overall surge margin as optimization constraints, and subject to the initial size constraints of the compressor, using the flow path coordinates of the axial compressor, the inlet and outlet structural angles at the middle sections of the rotor and stator blades at each stage, the density, the chord length, the maximum relative thickness of the blades, and the maximum relative thickness position as optimization variables, parameter optimization is performed using a genetic algorithm until the optimization objective meets the overall design specifications. The updated optimization variables obtained are the one-dimensional design scheme of the compressor.

[0089] The optimization boundaries of the blade inlet and outlet structural angles, density, and chord length at the middle sections of the moving and stationary blades at each level among the optimization variables are within ±20% of the initial values ​​of the blade inlet and outlet structural angles, density, and chord length at the middle sections of the moving and stationary blades at each level;

[0090] The initial values ​​of the maximum relative thickness position and the maximum relative deflection position of the blade are given as 50% of the chord length, and the initial values ​​of the relative thickness of the leading and trailing edges are given as 10% of the maximum relative thickness; the optimization boundaries of the maximum relative thickness position, the maximum relative deflection position, and the relative thickness of the leading and trailing edges of the blade are within ±10% of the initial values ​​of the maximum relative thickness position, the maximum relative deflection position, and the relative thickness of the leading and trailing edges;

[0091] S05. If the optimization goal is not achieved or the optimization history is not fully converged, the best design solution is selected from the current optimization population, and the optimization variable boundary is updated based on the best design solution. The optimization variable boundary is selected to be 50% of the size of the previous optimization variable boundary. The optimization is continued until the optimization goal is achieved, the one-dimensional compressor design is terminated, and a one-dimensional compressor design solution that meets the optimization goal is output.

[0092] S06. Parameter transfer between the genetic algorithm and the one-dimensional characteristic calculation model of the axial flow compressor is achieved through Python script transfer to achieve the construction and connection of the global optimization design platform of the axial flow compressor;

[0093] S07. Determine the absolute circumferential velocity at the root and tip of each blade of an axial flow compressor based on the principle of constant annular mass design and the one-dimensional design of the compressor. and , respectively expressed as:

[0094] ,

[0095] ,

[0096] Then, assuming that the radial distribution of the axial velocity on the blade is the same, the velocity triangle at the blade root and the blade tip is determined, and then according to the Pythagorean theorem, the inlet and outlet airflow angles at the root of each blade, the middle section of each blade, and the tip of each blade are further determined.

[0097] S08. Determine the inlet and outlet structural angles of at least 11-21 sections of the axial compressor blades in the radial direction according to the Lieblien reference lagging angle model. The specific calculation derivation formulas include:

[0098] ,

[0099] Where, is the reference trailing angle of the blade, is the shape factor of the blade, is the relative thickness factor of the blade, The reference trailing angle at 10% blade relative thickness under 0 camber, is the curvature coefficient of the blade, is the bending angle of the blade;

[0100] in,

[0101] ,

[0102] Where, For consistency;

[0103] ,

[0104] Where, is the relative thickness of the leaf;

[0105] ,

[0106] Where, is the camber coefficient of the NACA 65 series blades, is an index, and and The expressions are:

[0107] ,

[0108] ,

[0109] ,

[0110] ,

[0111] Where, is the inlet airflow angle at the middle section of each stage of moving and stationary blades; It is the outlet airflow angle at the middle section of each stage of moving blades and stationary blades.

[0112] S09. Fit the inlet and outlet structural angles of at least 11-21 sections of the axial compressor blades at each stage in the radial direction to obtain:

[0113] ,

[0114] Where x is the relative radius of the blade root, the middle section and the tip of each blade at each level in the radial direction; y is the inlet structural angle of the blade root, the middle section and the tip of each blade at each level; a, b and c are fitting coefficients;

[0115] At this time, through nonlinear uniform interpolation, the inlet and outlet structural angles of all radial sections of blades at all levels are calculated, and then the maximum relative thickness, maximum relative thickness position, leading and trailing edge relative thickness and maximum relative deflection position in the two-dimensional characteristic calculation model of the axial flow compressor are given. Based on the two-dimensional characteristic calculation model of the axial flow compressor, the two-dimensional comprehensive surge margin, total pressure ratio and efficiency of the axial flow compressor are calculated.

[0116] S10. Taking efficiency as the optimization target, total pressure ratio and overall machine comprehensive margin as optimization constraints, and inlet and outlet structural angles of at least 11-21 sections, maximum relative thickness of blades and maximum relative thickness position as optimization variables, perform parameter optimization based on a genetic algorithm until the overall design indicators are met. The updated optimization variables obtained are the two-dimensional design scheme of the compressor.

[0117] S11. If the optimization goal is not achieved or the optimization history is not fully converged, the best design is selected from the current optimization population, and the boundaries of the optimization variables are updated based on the best design scheme. The boundaries of the optimization variables are selected as 50% of the boundary size of the last optimization variable. The optimization is continued until the optimization goal is achieved, the two-dimensional design of the compressor is terminated, and a two-dimensional design scheme of the compressor that meets the optimization goal is output.

[0118] S12, the genetic algorithm and the axial flow compressor two-dimensional characteristic calculation model are used to realize parameter transfer through Python script transfer, thus realizing the construction and linking of the global optimization design platform of the axial flow compressor.

[0119] S13. Input the two-dimensional design scheme of the compressor into the three-dimensional modeling calculation model of the axial flow compressor to obtain a designed three-dimensional geometric model of the compressor.

[0120] Example 2: The same as Example 1, except that the optimization constraints in the genetic algorithm in step 2 also include: blocking flow, diffusion factor, and load coefficient.

[0121] like Figure 1-9 As shown, in order to further illustrate the technical solution and technical effect of the present invention, a three-stage axial flow compressor is designed as an example. Figure 2 As shown in the figure, the meridional flow surface diagram corresponding to the one-dimensional design scheme of the three-stage axial compressor initialization is shown, where IGV is the inlet guide vane, R1, R2, and R3 represent the first-stage, second-stage, and third-stage moving blades respectively, and S1, S2, and S3 represent the first-stage, second-stage, and third-stage stationary blades respectively;

[0122] Therefore, the following specific examples are provided:

[0123] The multi-level optimization design method of a multi-stage axial flow compressor based on a genetic algorithm of the present invention is as follows: Figure 1 As shown, the following steps are included:

[0124] Step 1: Pressure ratio based on axial flow compressor design ,efficiency , the overall margin of the whole machine The total pressure at the outlet of the axial flow compressor is determined based on the overall design index of 29.8 kg / s and the initial size constraint of the compressor. , outlet total temperature And the whole machine rim work ;

[0125] Then, according to the whole machine rim work The size of the axial flow compressor is determined to be 3 stages, and the load coefficient and reaction force of each stage of the axial flow compressor are allocated. The inlet and outlet axial velocity of each stage of the moving blade, the radius of the middle section, and the aspect ratio and consistency of each row are selected. Then, the one-dimensional design calculation model of the axial flow compressor is used to obtain the flow channel coordinates of the axial flow compressor, the velocity triangle at the middle section of each stage of the moving blade and the stationary blade, the inlet and outlet airflow angle, and the chord length;

[0126] The initial values ​​of the angle of attack and the angle of lagging at the middle section of each level of moving blades and stationary blades are set to 0. The inlet and outlet structural angles at the middle section of each level of moving blades and stationary blades are determined according to the inlet and outlet airflow angles. At the same time, the maximum relative thickness of the blades and the position of the maximum relative thickness are given.

[0127] Thus, based on the one-dimensional characteristic calculation model of the axial flow compressor, the one-dimensional total pressure ratio, the overall surge margin and efficiency of the axial flow compressor are obtained, such as Figure 3 shown by the dotted line;

[0128] Taking efficiency as the optimization objective, total pressure ratio and overall surge margin as optimization constraints, and based on the initial size constraints of the compressor, the flow channel coordinates of the axial compressor, the inlet and outlet structural angles at the middle section of each stage of the moving and stationary blades, the consistency, chord length, the maximum relative thickness of the blades, and the maximum relative thickness position are used as optimization variables. The optimization variables are updated based on a genetic algorithm and transferred through a Python script until the optimization objective meets the overall design indicators. The updated optimization variables are the one-dimensional design scheme of the compressor.

[0129] like Figure 3 As shown, a schematic diagram comparing the compressor meridian flow surface before and after optimization is shown, wherein the dotted line is a schematic diagram of the flow path of the three-stage axial flow compressor before optimization, and the solid line is a schematic diagram of the flow path of the three-stage axial flow compressor after optimization. Figure 3 In the figure, the flow path of the three-stage axial compressor has undergone significant changes before and after optimization;

[0130] like Figure 4 Figure 2 shows a comparison of the converted flow-pressure ratio characteristics of the three-stage axial compressor at the design speed before and after optimization. The axial coordinate represents the converted flow, and the vertical coordinate represents the pressure ratio. The dashed line represents the pressure ratio characteristic calculated using the axial compressor one-dimensional characteristic calculation model before optimization, while the solid line represents the pressure ratio characteristic calculated using the axial compressor one-dimensional characteristic calculation model after optimization. Figure 4 The optimized pressure ratio characteristics are significantly improved compared with those before optimization, and meet the design indicators of pressure ratio and comprehensive margin;

[0131] like Figure 5 Figure 2 shows a comparison of the efficiency-compression ratio characteristics of the three-stage axial compressor's one-dimensional design scheme at the design speed before and after optimization. The axial coordinate represents efficiency, and the vertical coordinate represents pressure ratio. The dashed line represents the efficiency-compression ratio characteristic calculated using the one-dimensional axial compressor characteristic calculation model before optimization, while the solid line represents the efficiency-compression ratio characteristic calculated using the one-dimensional axial compressor characteristic calculation model after optimization. Figure 5The optimized efficiency-pressure ratio characteristics have been greatly improved compared with those before optimization, so that the optimized one-dimensional design scheme of the axial flow compressor exceeds the design target by 1.014% when the pressure ratio and surge margin of its design point meet the standards.

[0132] like Figure 6 As shown in Figure 3, the convergence history of the optimization process of the one-dimensional design stage of the three-stage axial flow compressor is shown in Figure 3, where the axial coordinate is the number of iterations and the vertical coordinate is the inverse of the efficiency. Figure 6 In the optimization process, as the number of iterations increases, the design point efficiency of the one-dimensional design scheme continues to converge to the maximum value, and the optimization process converges smoothly;

[0133] Step 2: Based on the principle of equal annular volume design and the one-dimensional design scheme of the compressor, determine the inlet and outlet structural angles of at least 11-21 radial sections of the axial flow compressor blades at each stage, and specify the maximum relative thickness and the position of the maximum relative thickness of the blades;

[0134] Therefore, based on the two-dimensional characteristic calculation model of the axial flow compressor, the two-dimensional comprehensive surge margin, total pressure ratio and efficiency of the axial flow compressor are obtained, such as Figure 7 shown by the dotted line;

[0135] Taking efficiency as the optimization objective, total pressure ratio and overall engine margin as optimization constraints, and the inlet and outlet structural angles of at least 11-21 sections, the maximum relative thickness of the blades, and the maximum relative thickness position as optimization variables, the optimization variables are updated using a genetic algorithm and transferred through a Python script until the overall design indicators are met. The updated optimization variables are the two-dimensional compressor design solution.

[0136] like Figure 7 Figure 2 shows a comparison of the converted flow-pressure ratio characteristics of the three-stage axial compressor 2D design scheme at the design speed before and after optimization. The axial coordinate represents the converted flow rate, and the vertical coordinate represents the pressure ratio. The dashed line represents the pressure ratio characteristic calculated by the 2D axial compressor characteristic calculation model before optimization, while the solid line represents the pressure ratio characteristic calculated by the 2D axial compressor characteristic calculation model after optimization. Figure 7 The pressure ratio characteristics of the optimized two-dimensional design scheme have been greatly improved compared with those before optimization, and the design indicators of pressure ratio and comprehensive margin are met;

[0137] like Figure 8Figure 2 shows a comparison of the efficiency-compression ratio characteristics of the three-stage axial compressor at the design speed before and after optimization. The axial coordinate represents efficiency, and the vertical coordinate represents pressure ratio. The dashed line represents the efficiency-compression ratio characteristic calculated using the 2D axial compressor characteristic calculation model before optimization of the 2D axial compressor design, while the solid line represents the efficiency-compression ratio characteristic calculated using the 2D axial compressor characteristic calculation model after optimization. Figure 8 The optimized efficiency-pressure ratio characteristics are significantly improved compared with those before optimization, and meet the design index of efficiency;

[0138] like Figure 9 As shown in Figure 3, the convergence history of the optimization process of the two-dimensional design stage of the three-stage axial flow compressor is shown in Figure 3, where the axial coordinate is the number of iterations and the vertical coordinate is the inverse of the efficiency. Figure 9 In the optimization process, as the number of iterations increases, the design point efficiency continues to converge to the maximum value, and the optimization process converges smoothly;

[0139] Step 3: Input the two-dimensional design scheme of the compressor into the three-dimensional modeling calculation model of the axial flow compressor to obtain the designed three-dimensional geometric model of the compressor.

[0140] Based on the above examples, the method described in this invention can avoid the two-dimensional flow and three-dimensional modeling stages of axial-flow compressor design, which rely heavily on the designer's experience and extensive parameter input. This allows for a simpler and faster approach to achieving a compressor design that meets design objectives. This method also establishes a global (genetic algorithm) compressor optimization platform, reducing the design complexity and shortening the design cycle. Furthermore, it further explores design potential through optimization methods.

[0141] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A multi-level optimization design method for a multi-stage axial flow compressor based on a genetic algorithm, characterized in that: The following steps are involved: Step 1: Pressure ratio based on the axial flow compressor design point ,efficiency , the overall surge margin of the whole machine The total pressure at the outlet of the axial flow compressor is determined based on the overall design indicators of the compressor and the initial size constraints of the compressor. , outlet total temperature And the whole machine rim work ; Then, according to the whole machine rim work The size of the axial flow compressor is used to determine the number of stages, and the load coefficient and reaction force of each stage of the axial flow compressor are allocated. The inlet and outlet axial velocity of each stage of the moving blade, the radius of the middle section, and the aspect ratio and consistency of each row are selected. Then, the flow path coordinates of the axial flow compressor, the velocity triangle at the middle section of each stage of the moving blade and the stator blade, the inlet and outlet airflow angle, and the chord length are obtained by using the one-dimensional design calculation model of the axial flow compressor. The initial values ​​of the angle of attack and the angle of lagging at the middle section of each level of moving blades and stationary blades are set to 0. The inlet and outlet structural angles at the middle section of each level of moving blades and stationary blades are determined according to the inlet and outlet airflow angles. At the same time, the maximum relative thickness of the blades and the position of the maximum relative thickness are given. Thus, based on the one-dimensional characteristic calculation model of the axial flow compressor, the one-dimensional total pressure ratio, the overall surge margin and efficiency of the axial flow compressor are obtained; Taking efficiency as the optimization objective, total pressure ratio and overall surge margin as optimization constraints, and based on the initial size constraints of the compressor, the flow path coordinates of the axial compressor, the inlet and outlet structural angles at the middle section of each stage of the moving and stationary blades, the density, chord length, the maximum relative thickness of the blades, and the maximum relative thickness position are used as optimization variables. Parameter optimization is performed based on a genetic algorithm until the optimization objective meets the overall design indicators. The updated optimization variables obtained are the one-dimensional design scheme of the compressor. The flow channel coordinates are obtained by fitting the fourth-order Bezier curve equation to ensure the smoothness and continuity of the flow channel coordinates. The specific formula for fitting the flow channel coordinates is: ， Where, is the flow channel coordinate, is the control parameter, and , are the five control points of the flow channel coordinates, and the optimization boundaries of the five control points are within ±20% of the initial value of each control point; Step 2: Based on the principle of equal annular volume design and the one-dimensional design scheme of the compressor, determine the inlet and outlet structural angles of 11-21 radial sections of the axial flow compressor blades at each stage, and specify the maximum relative thickness and the position of the maximum relative thickness of the blades; Thus, based on the two-dimensional characteristic calculation model of the axial flow compressor, the two-dimensional integrated surge margin, total pressure ratio and efficiency of the axial flow compressor are obtained; Using efficiency as the optimization objective, total pressure ratio and overall engine margin as optimization constraints, and the inlet and outlet structural angles, maximum relative blade thickness, and maximum relative blade thickness position of sections 11-21 as optimization variables, a genetic algorithm was used to optimize the parameters until the overall design indicators were met. The updated optimization variables obtained were the two-dimensional compressor design scheme. Step 3: Input the two-dimensional design scheme of the compressor into the three-dimensional modeling calculation model of the axial flow compressor to obtain the designed three-dimensional geometric model of the compressor.

2. The multi-level optimization design method for a multi-stage axial flow compressor based on a genetic algorithm according to claim 1, characterized in that: The total outlet pressure of the axial flow compressor in step 1 , outlet total temperature and the whole machine rim work The calculation process is: First, calculate the total outlet pressure for: ， Where, is the total inlet pressure of the axial compressor; Then, according to the isentropic efficiency relationship of the axial flow compressor, the total outlet temperature is determined as: ， Where, is the total temperature at the inlet of the axial compressor; Finally, the wheel rim power of the whole machine is determined for: ， Where, is the constant-pressure specific heat capacity of the axial compressor.

3. The multi-level optimization design method for a multi-stage axial flow compressor based on a genetic algorithm according to claim 1, characterized in that: The optimization boundaries of the blade inlet and outlet structural angles, consistency and chord lengths at the middle sections of the moving blades and stationary blades at each stage in the optimization variables described in step 1 are between ±20% of the initial values ​​of the blade inlet and outlet structural angles, consistency and chord lengths at the middle sections of the moving blades and stationary blades at each stage.

4. The multi-level optimization design method for a multi-stage axial flow compressor based on a genetic algorithm according to claim 1, wherein: In the genetic algorithm of steps 1 and 2, if the optimization goal is not achieved or the optimization history is not fully converged, the best design is selected from the current optimization population, and the boundaries of the optimization variables are updated based on the best design scheme. The boundaries of the optimization variables are selected to be 50% of the size of the boundaries of the variables optimized last time. The optimization is continued until the optimization goal is achieved, the one-dimensional or two-dimensional design of the compressor is terminated, and a one-dimensional or two-dimensional design scheme of the compressor that meets the optimization goal is output.

5. The multi-level optimization design method for a multi-stage axial flow compressor based on a genetic algorithm according to claim 1, characterized in that: The genetic algorithm in step 1 and the one-dimensional characteristic calculation model of the axial flow compressor are connected by transferring parameters through a python script. Similarly, the genetic algorithm in step 2 and the two-dimensional characteristic calculation model of the axial flow compressor are also transferred through Python scripts to realize parameter transfer, thus realizing the construction and connection of the global optimization design platform of the axial flow compressor.

6. The multi-level optimization design method for a multi-stage axial flow compressor based on a genetic algorithm according to claim 1, characterized in that: In step 2, the specific steps of obtaining the two-dimensional integrated surge margin, total pressure ratio, and efficiency of the axial flow compressor based on the two-dimensional characteristic calculation model of the axial flow compressor are as follows: First, according to the principle of equal annular design, the absolute circumferential velocity of the blade root and blade tip of each stage of the axial flow compressor is determined. and , respectively expressed as: ， ， Then, assuming that the radial distribution of the axial velocity on the blade is the same, the velocity triangles at the blade root and blade tip are determined. Based on the Pythagorean theorem, the inlet and outlet airflow angles at the blade root, the middle section of each blade, and the blade tip are further determined. Then, according to the Lieblien reference lagging angle model, the inlet and outlet structural angles of the 11-21 sections in the radial direction of the axial flow compressor blades at each stage are determined. The specific calculation derivation formulas include: ， Where, is the reference trailing angle of the blade, is the shape factor of the blade, is the relative thickness factor of the blade, The reference trailing angle at 10% blade relative thickness under 0 camber, is the curvature coefficient of the blade, is the bending angle of the blade; in, ， Where, For consistency; ， Where, is the relative thickness of the leaf; ， Where, is the camber coefficient of the NACA 65 series blades, is an index, and and The expressions are: ， ， ， ， Where, is the inlet airflow angle at the middle section of each stage of moving and stationary blades; is the outlet airflow angle at the middle section of each stage of moving and stationary blades; Next, the inlet and outlet structural angles of 11-21 sections of the axial compressor blades at each stage in the radial direction are fitted to obtain: ， Where x is the relative radius of the blade root, the middle section and the tip of each blade at each level in the radial direction; y is the inlet structural angle of the blade root, the middle section and the tip of each blade at each level; a, b and c are fitting coefficients; At this time, the inlet and outlet structural angles of all radial sections of blades at each stage are determined through nonlinear uniform interpolation, and then the maximum relative thickness, maximum relative thickness position, leading and trailing edge relative thickness and maximum relative deflection position in the two-dimensional characteristic calculation model of the axial flow compressor are given. Based on the two-dimensional characteristic calculation model of the axial flow compressor, the two-dimensional comprehensive surge margin, total pressure ratio and efficiency of the axial flow compressor are calculated.

7. The multi-level optimization design method for a multi-stage axial flow compressor based on a genetic algorithm according to claim 6, characterized in that: The initial values ​​of the maximum relative thickness position and the maximum relative deflection position of the blade are given as 50% of the chord length, and the initial value of the relative thickness of the leading and trailing edges is given as 10% of the maximum relative thickness; the optimization boundaries of the maximum relative thickness position, the maximum relative deflection position, and the relative thickness of the leading and trailing edges of the blade are within ±10% of the initial values ​​of the maximum relative thickness position, the maximum relative deflection position, and the relative thickness of the leading and trailing edges.

8. The multi-level optimization design method for a multi-stage axial flow compressor based on a genetic algorithm according to any one of claims 1 to 7, characterized in that: The optimization constraints in the genetic algorithm of step 2 also include: blocking flow, diffusion factor, and load coefficient.

Images