A design method and device for a multi-stage axial flow compressor of a gas turbine

Abstract

The present application provides a design method and device for a multi-stage axial flow compressor of a gas turbine, including: performing dimensionality reduction processing on multiple performance parameters in an initial three-dimensional hot flow design scheme, a one-dimensional design scheme, and corresponding performance parameters in a two-dimensional S2 stream surface design scheme to obtain a dimensionality reduction data transfer model based on a neural network; based on the obtained dimensionality reduction data transfer model through training, performing dimensionality increase design again to obtain a target three-dimensional hot flow design scheme, and detecting whether the operating conditions performance of the target three-dimensional hot flow design scheme reaches the design target; if it meets the standard, determining the cold state geometric model and structural design of the compressor according to the three-dimensional hot flow design scheme to obtain a compressor that meets the design target, thereby ensuring the consistency of multi-dimensional design results and improving the accuracy of design results.

Description

Technical Field

[0001] The present application relates to the technical field of compressor design, and in particular, to a design method and device for a multi-stage axial compressor of a gas turbine. Background Art

[0002] A compressor is one of the three core components of a gas turbine, and its successful development is the first milestone on the R & D route of a gas turbine. With the continuous development of the world's heavy industry level, multi-stage axial compressors of gas turbines are continuously developing in the direction of high efficiency, high load, and high surge margin. How to quickly and accurately obtain the aerodynamic design scheme of a compressor, and even establish a multi-dimensional design system for compressors in a serialized and modular manner, is a technical problem that must be solved in the face of future gas turbine technological revolutions and industrial transformations.

[0003] At the present stage, the aerodynamic design method based on the traditional compressor design process has been relatively mature. However, low-dimensional simulation tools that rely on empirical regression seriously limit the result consistency in the dimensionality-up design process of the compressor and the accuracy of low-dimensional design, resulting in over-reliance on three-dimensional simulation in the design and prolonging the design cycle. Therefore, how to quickly and accurately design a compressor according to the design goal has become an urgent problem to be solved. Summary of the Invention

[0004] In view of this, the purpose of the present application is to provide a design method and device for a multi-stage axial compressor of a gas turbine. During the initial dimensionality-up design process, low-dimensional simulation tools need to avoid using radial models with relatively weak robustness, such as radial mixing and end-region blockage. After completing one-dimensional inverse problem design and forward problem simulation, two-dimensional S2 stream surface inverse problem design and forward problem simulation, three-dimensional blade modeling design and three-dimensional single-stage forward problem simulation, performance parameters in the three-dimensional simulation results are extracted, and a dimensionality reduction data transfer model is obtained through training a neural network. During the next dimensionality-up design, the low-dimensional simulation tool will combine the obtained dimensionality reduction data, correct the correction coefficients in the one-dimensional simulation and the correction coefficients in the two-dimensional S2 stream surface simulation, give a more accurate performance evaluation, and provide relatively accurate boundary conditions for the three-dimensional single-stage simulation, and iterate multiple times to obtain a compressor that meets the design goal. The present invention improves the result consistency between multi-dimensional simulations through dimensionality reduction data transfer, and thus, relying on the rapidity of low-dimensional simulation, shortens the design iteration cycle while ensuring the accuracy of the simulation.

[0005] The embodiment of the present application provides a design method for a multi-stage axial compressor of a gas turbine, and the design method includes:

[0006] According to the design goal of the compressor to be designed, during the initial dimensionality-up design process, determine in sequence a one-dimensional design scheme, a two-dimensional S2 stream surface design scheme, and an initial three-dimensional hot condition through-flow design scheme for the compressor to be designed;

[0007] Input multiple initial three-dimensional performance parameters in the initial three-dimensional hot flow design scheme, multiple corresponding one-dimensional performance parameters in the one-dimensional design scheme, and multiple corresponding two-dimensional S2 flow surface performance parameters in the two-dimensional S2 flow surface design scheme into a pre-established dimensionality reduction data transfer model, and output multiple one-dimensional correction coefficients in the one-dimensional direct problem simulation process and multiple two-dimensional S2 flow surface correction coefficients in the two-dimensional S2 flow surface direct problem simulation process; wherein, the one-dimensional correction coefficients and the two-dimensional S2 flow surface correction coefficients are obtained by comparing parameters after dimensionality reduction processing of multiple initial three-dimensional performance parameters in the initial three-dimensional hot flow design scheme.

[0008] Based on multiple one-dimensional correction coefficients and multiple two-dimensional S2 flow surface correction coefficients, determine a target three-dimensional hot flow design scheme, and detect whether the operating conditions performance of the target three-dimensional hot flow design scheme reaches the design target.

[0009] If the operating conditions performance of the target three-dimensional hot flow design scheme reaches the design target, then determine the cold-state geometric model and structural design of the compressor according to the target three-dimensional hot flow design scheme, and obtain a compressor that meets the design target.

[0010] In a possible implementation manner, the determining the cold-state geometric model and structural design of the compressor according to the target three-dimensional hot flow design scheme, and obtaining a compressor that meets the design target includes:

[0011] Based on the target three-dimensional hot flow design scheme, determine a three-dimensional cold-state blade model through cold-hot state conversion and structural design.

[0012] Detect whether the combined stress of the blade bending stress and tensile stress of the three-dimensional blades in the three-dimensional cold-state blade model is less than the ratio of the yield limit to the safety factor and whether the tensile stress in the three-dimensional blades is less than the ratio of the yield limit to the safety factor.

[0013] If the combined stress of the blade bending stress and tensile stress of the three-dimensional cold-state blade model in the target three-dimensional hot flow design scheme is greater than the ratio of the yield limit to the safety factor or the tensile stress is greater than the ratio of the yield limit to the safety factor, then adjust the geometric model in the three-dimensional hot flow design scheme until the combined stress of the blade bending stress and tensile stress of the three-dimensional cold-state blade model is less than the ratio of the yield limit to the safety factor and the tensile stress is less than the ratio of the yield limit to the safety factor, and obtain a compressor that meets the design target.

[0014] In a possible implementation manner, the one-dimensional design scheme is determined through the following steps, including:

[0015] Based on the design objective, determine a plurality of one-dimensional design parameters for the compressor to be designed;

[0016] Based on the plurality of one-dimensional design parameters, perform one-dimensional design at the mean radius of the compressor to determine a one-dimensional design scheme;

[0017] According to the basic parameters of the elementary stage and the meridional flow path parameters in the one-dimensional design scheme, perform a forward problem simulation to determine the performance parameters of the one-dimensional design scheme and the aerodynamic and thermodynamic parameters at the mean radius of the compressor.

[0018] In a possible implementation manner, determine a two-dimensional S2 stream surface design scheme through the following steps, including:

[0019] Compare the flow rate, efficiency, and surge margin in the one-dimensional design scheme with the target flow rate, target efficiency, and target surge margin in the design objective;

[0020] If the flow rate, efficiency, and surge margin in the one-dimensional design scheme do not meet the target flow rate, target efficiency, and target surge margin in the design objective, then change the number of stages and inlet pre-whirl of the one-dimensional design scheme, and adjust the distribution of the axial velocity, reaction degree, target loading coefficient, and elementary parameters along the flow path;

[0021] If the flow rate, efficiency, and surge margin in the one-dimensional design scheme meet the target flow rate, target efficiency, and target surge margin in the design objective, then perform radial work distribution and reaction degree design of the compressor on the basis of the one-dimensional design scheme to determine a two-dimensional S2 stream surface design scheme;

[0022] According to the two-dimensional S2 stream surface design scheme, perform a forward problem simulation of the two-dimensional S2 stream surface to determine the performance parameters of the two-dimensional S2 stream surface design scheme and the aerodynamic and thermodynamic parameters after circumferential averaging of the compressor.

[0023] In a possible implementation manner, determine an initial three-dimensional hot condition through-flow design scheme through the following method, including:

[0024] Compare the flow rate, efficiency, and surge margin in the two-dimensional S2 stream surface design scheme with the target flow rate, target efficiency, and target surge margin in the design objective;

[0025] If the flow rate, efficiency, and surge margin in the two-dimensional S2 stream surface design scheme do not meet the target flow rate, target efficiency, and target surge margin in the design objective, then change the way of step-by-step radial work distribution in the two-dimensional S2 stream surface design scheme;

[0026] If the flow rate, efficiency, and surge margin in the two-dimensional S2 flow surface design scheme meet the target flow rate, target efficiency, and target surge margin in the design objectives, then a two-dimensional airfoil is selected based on the two-dimensional S2 flow surface design scheme, and the three-dimensional blade is determined by stacking the two-dimensional airfoils;

[0027] Based on the three-dimensional blade, an initial three-dimensional hot flow path design scheme is determined, and a forward problem simulation is performed on the initial three-dimensional hot flow path design scheme to obtain multiple initial three-dimensional performance parameters in the initial three-dimensional hot flow path design scheme.

[0028] In a possible implementation manner, the to-be-designed compressor is a multi-stage axial flow compressor. The steps to determine multiple one-dimensional correction coefficients of the one-dimensional design scheme and multiple two-dimensional S2 flow surface correction coefficients of the two-dimensional S2 flow surface design scheme include:

[0029] Input multiple initial three-dimensional performance parameters in the initial three-dimensional hot flow path design scheme, multiple corresponding one-dimensional performance parameters in the one-dimensional design scheme, and multiple corresponding two-dimensional S2 flow surface performance parameters in the two-dimensional S2 flow surface design scheme into a pre-established dimensionality reduction data transfer model;

[0030] For each stage of the to-be-designed compressor, substitute an initial three-dimensional performance parameter corresponding to the initial three-dimensional hot flow path design scheme of this stage into the empirical formula of one-dimensional simulation, and combine the relevant design and performance parameters of one-dimensional design to determine the reference one-dimensional correction coefficient corresponding to this one-dimensional performance parameter of this stage;

[0031] For each stage of the to-be-designed compressor, substitute an initial three-dimensional performance parameter in the initial three-dimensional hot flow path design scheme of this stage into the empirical formula of two-dimensional S2 flow surface simulation, and combine the relevant design and performance parameters of two-dimensional S2 flow surface design to determine the reference two-dimensional S2 flow surface correction coefficient corresponding to the two-dimensional S2 flow surface performance parameter of this stage;

[0032] Based on the reference one-dimensional correction coefficient and the reference two-dimensional S2 flow surface correction coefficient generated by dimensionality reduction of the performance parameters of the initial three-dimensional hot flow path design scheme, train the dimensionality reduction data transfer model, and output multiple one-dimensional correction coefficients in the forward problem simulation of the one-dimensional design scheme and multiple two-dimensional S2 flow surface correction coefficients in the forward problem simulation of the two-dimensional S2 flow surface design scheme by the dimensionality reduction data transfer model.

[0033] In a possible implementation manner, the target three-dimensional hot flow path design scheme is determined through the following steps:

[0034] For the rotational speed and pressure ratio of each stage, a one-dimensional correction coefficient in the one-dimensional forward problem simulation is determined based on the dimension-reduced data transfer model corresponding to this stage;

[0035] For the rotational speed and pressure ratio of each stage, a two-dimensional S2 stream surface correction coefficient in the two-dimensional S2 stream surface forward problem simulation is determined based on the dimension-reduced data transfer model corresponding to this stage;

[0036] Change the design parameters in the one-dimensional design scheme to determine a target one-dimensional design scheme. In the one-dimensional forward problem simulation, use the one-dimensional correction coefficient generated by the dimension-reduced data transfer model;

[0037] Change the stage-by-stage radial work distribution method of the two-dimensional S2 stream surface design scheme to determine a target two-dimensional S2 stream surface design scheme. In the two-dimensional S2 stream surface forward problem simulation, use the two-dimensional S2 stream surface correction coefficient generated by the dimension-reduced data transfer model;

[0038] Based on the target two-dimensional S2 stream surface design scheme, select a target two-dimensional blade profile for stacking to generate a target three-dimensional blade, and determine a target three-dimensional hot flow path design scheme of the compressor based on the target three-dimensional blade.

[0039] In a possible implementation manner, the design method further includes:

[0040] If any performance in the operating conditions performance of the target three-dimensional hot flow path design scheme is inconsistent with the corresponding operating conditions performance in the target two-dimensional S2 stream surface design scheme and the corresponding operating conditions performance in the target one-dimensional design scheme, perform dimension reduction transfer iterative processing on the initial three-dimensional hot flow path design scheme until the operating conditions performance of the target three-dimensional hot flow path design scheme is corresponding and consistent with the operating conditions performance of the target two-dimensional S2 stream surface design scheme and the operating conditions performance of the target one-dimensional design scheme.

[0041] The embodiment of the present application also provides a design device for a multi-stage axial flow compressor of a gas turbine. The design device includes:

[0042] A first dimension elevation design module, configured to sequentially determine a one-dimensional design scheme, a two-dimensional S2 stream surface design scheme, and an initial three-dimensional hot flow path design scheme for the to-be-designed compressor in the initial dimension elevation design process according to the design target of the to-be-designed compressor;

[0043] A dimensionality reduction data transfer module is configured to input multiple initial three-dimensional performance parameters in the initial three-dimensional hot flow design scheme, multiple corresponding one-dimensional performance parameters in the one-dimensional design scheme, and multiple corresponding two-dimensional S2 flow surface performance parameters in the two-dimensional S2 flow surface design scheme into a pre-established dimensionality reduction data transfer model, and output multiple one-dimensional correction coefficients in the one-dimensional direct problem simulation process and multiple two-dimensional S2 flow surface correction coefficients in the two-dimensional S2 flow surface direct problem simulation process; wherein, the one-dimensional correction coefficients and the two-dimensional S2 flow surface correction coefficients are obtained by comparing parameters after dimensionality reduction processing of multiple initial three-dimensional performance parameters in the initial three-dimensional hot flow design scheme.

[0044] A second dimensionality increase design module is configured to determine a target three-dimensional hot flow design scheme based on multiple of the one-dimensional correction coefficients and multiple of the two-dimensional S2 flow surface correction coefficients, and detect whether the operating conditions performance of the target three-dimensional hot flow design scheme reaches the design target.

[0045] A structure design module is configured to, if the operating conditions performance of the target three-dimensional hot flow design scheme reaches the design target, determine the cold-state geometric model and structure design of the compressor according to the target three-dimensional hot flow design scheme, and obtain a compressor that meets the design target.

[0046] An embodiment of the present application further provides an electronic device, including: a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the electronic device runs, the processor communicates with the memory through the bus. When the machine-readable instructions are run by the processor, the steps of the design method of the gas turbine multi-stage axial flow compressor as described above are executed.

[0047] An embodiment of the present application further provides a computer-readable storage medium. A computer program is stored on the computer-readable storage medium. When the computer program is run by a processor, the steps of the design method of the gas turbine multi-stage axial flow compressor as described above are executed.

[0048] The present application provides a design method and device for a multi-stage axial flow compressor of a gas turbine. The design method includes: according to the design objectives of the compressor to be designed, sequentially determining a one-dimensional design scheme, a two-dimensional S2 stream surface design scheme, and an initial three-dimensional hot condition through-flow design scheme for the compressor to be designed during the initial dimensionality elevation design process; inputting multiple initial three-dimensional performance parameters in the initial three-dimensional hot condition through-flow design scheme, multiple corresponding one-dimensional performance parameters in the one-dimensional design scheme, and multiple corresponding two-dimensional S2 stream surface performance parameters in the two-dimensional S2 stream surface design scheme into a pre-established dimensionality reduction data transfer model, and outputting multiple one-dimensional correction coefficients during the one-dimensional direct problem simulation process and multiple two-dimensional S2 stream surface correction coefficients during the two-dimensional S2 stream surface direct problem simulation process; wherein, the one-dimensional correction coefficients and the two-dimensional S2 stream surface correction coefficients are obtained by comparing parameters after dimensionality reduction processing of multiple initial three-dimensional performance parameters in the initial three-dimensional hot condition through-flow design scheme; based on multiple one-dimensional correction coefficients and multiple two-dimensional S2 stream surface correction coefficients, determining a target three-dimensional hot condition through-flow design scheme, and detecting whether the operating conditions performance of the target three-dimensional hot condition through-flow design scheme reaches the design objectives; if the operating conditions performance of the target three-dimensional hot condition through-flow design scheme reaches the design objectives, then determining the cold state geometric model and structural design of the compressor according to the target three-dimensional hot condition through-flow design scheme, and obtaining a compressor that meets the design objectives.

[0049] In this way, a dimensionality reduction data transfer model is obtained by performing dimensionality reduction processing on performance parameters between the initial three-dimensional hot condition through-flow design scheme and the one-dimensional design scheme, and between the initial three-dimensional hot condition through-flow design scheme and the two-dimensional S2 stream surface design scheme. The target three-dimensional hot condition through-flow design scheme is quickly and accurately determined by using multiple one-dimensional correction coefficients and multiple two-dimensional S2 stream surface correction coefficients generated by the dimensionality reduction data transfer model. The cold state geometric model and structural design of the compressor are determined by using the target three-dimensional hot condition through-flow design scheme, and a compressor that meets the design objectives is obtained, thereby ensuring the consistency of the multi-dimensional design results and improving the accuracy of the design results.

[0050] To make the above objectives, features, and advantages of the present application more obvious and understandable, the following specifically gives preferred embodiments and, in conjunction with the accompanying drawings, makes a detailed description as follows. Description of the Drawings

[0051] To more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present application and should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can also be obtained based on these drawings without creative efforts.

[0052] Figure 1 Flow chart of a design method for a multi-stage axial flow compressor of a gas turbine provided by an embodiment of the present application;

[0053] Figure 2 Schematic diagram of the simulation result of a one-dimensional design scheme in a design method for a multi-stage axial flow compressor of a gas turbine provided by an embodiment of the present application;

[0054] Figure 3 Data dimensionality reduction transfer flow chart of a design method for a multi-stage axial flow compressor of a gas turbine provided by an embodiment of the present application;

[0055] Figure 4 Schematic diagram of the simulation result of a target two-dimensional S2 stream surface design scheme in a design method for a multi-stage axial flow compressor of a gas turbine provided by an embodiment of the present application;

[0056] Figure 5 Schematic diagram of multi-dimensional simulation results in a design method for a multi-stage axial flow compressor of a gas turbine provided by an embodiment of the present application;

[0057] Figure 6 Schematic diagram of the compressor structure design in a design method for a multi-stage axial flow compressor of a gas turbine provided by an embodiment of the present application;

[0058] Figure 7 One of the schematic diagrams of the structure of a design device for a multi-stage axial flow compressor of a gas turbine provided by an embodiment of the present application;

[0059] Figure 8 Another schematic diagram of the structure of a design device for a multi-stage axial flow compressor of a gas turbine provided by an embodiment of the present application;

[0060] Figure 9 Schematic diagram of the structure of an electronic device provided by an embodiment of the present application.

[0061] Icon: 1 - Inlet cylinder; 2 - Inlet adjustable guide vane; 3 - Compressor cylinder; 4 - Low-pressure air extraction chamber; 5 - Air extraction pipe; 6 - High-pressure air extraction chamber; 7 - Combustion and compression cylinder; 8 - Diffuser; 9 - Thrust bearing; 10 - Radial bearing; 11 - Moving blade; 12 - Static blade; 13 - Moving impeller disc; 14 - Static vane ring; 15 - Main shaft; 700 - Design device; 710 - First dimension elevation design module; 720 - Dimensionality reduction data transfer module; 730 - Second dimension elevation design module; 740 - Structure design module; 750 - Iterative processing module; 900 - Electronic device; 910 - Processor; 920 - Memory; 930 - Bus. Detailed implementation manners

[0062] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of this application. It should be understood that the accompanying drawings in this application are only for the purposes of illustration and description, and are not used to limit the protection scope of this application. Additionally, it should be understood that the schematic drawings are not drawn to scale. The flowcharts used in this application illustrate the operations implemented according to some embodiments of this application. It should be understood that the operations in the flowchart may not be implemented in sequence, and steps without a logical context relationship may be reversed or implemented simultaneously. Furthermore, those skilled in the art can add one or more other operations to the flowchart or remove one or more operations from the flowchart under the guidance of the content of this application.

[0063] In addition, the described embodiments are only a part of the embodiments of this application, rather than all of the embodiments. The components of the embodiments of this application usually described and illustrated in the accompanying drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of this application claimed, but merely represents the selected embodiments of this application. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative efforts fall within the protection scope of this application.

[0064] To enable those skilled in the art to use the content of this application, in combination with the specific application scenario of "designing a compressor", the following implementation manners are given. For those skilled in the art, without departing from the spirit and scope of this application, the general principles defined here can be applied to other embodiments and application scenarios.

[0065] The following methods, devices, electronic devices, or computer-readable storage media in the embodiments of this application can be applied to any scenario that requires "designing a compressor". The embodiments of this application do not limit the specific application scenario. Any solution that uses the design method and device of a multi-stage axial flow compressor of a gas turbine provided in the embodiments of this application falls within the protection scope of this application.

[0066] First, an introduction to the applicable application scenarios of this application is provided. This application can be applied to the technical field of compressor design.

[0067] The compressor is one of the three core components of a gas turbine, and its successful development is the first milestone on the R & D route of gas turbines. With the continuous development of the world's heavy industry level, multi-stage axial compressors of gas turbines are constantly evolving towards high efficiency, high load, and high surge margin. How to quickly and accurately obtain the aerodynamic design scheme of the compressor and even establish a multi-dimensional design system for the compressor in a serialized and modularized manner is a technical problem that must be solved in the face of future gas turbine technological revolutions and industrial transformations.

[0068] It has been found through research that at the present stage, the aerodynamic design method based on the traditional compressor design process has been relatively mature. However, the low-dimensional simulation tools that rely on empirical regression seriously limit the result consistency in the upscaling design process of the compressor and the accuracy of low-dimensional design, resulting in over-reliance on three-dimensional simulations in the design and prolonging the design cycle. Therefore, how to quickly and accurately design the compressor according to the design objectives has become an urgent problem to be solved.

[0069] Based on this, the embodiments of the present application provide a design method for a multi-stage axial compressor of a gas turbine. The core idea of this method lies in the downscaling data transfer between high-dimensional and low-dimensional simulation tools. During the initial upscaling design process, the low-dimensional simulation tools need to avoid using radial models with relatively weak robustness, such as radial mixing and end region blockage. After completing the one-dimensional inverse problem design and forward problem simulation, two-dimensional S2 stream surface inverse problem design and forward problem simulation, three-dimensional blade modeling design, and three-dimensional single-stage forward problem simulation, the performance parameters in the three-dimensional simulation results are extracted, and a downscaling data transfer model is obtained through training a neural network. When performing the upscaling design again, the low-dimensional simulation tools will combine the obtained downscaling data to correct the correction coefficients in the one-dimensional simulation and the correction coefficients in the two-dimensional S2 stream surface simulation, give a more accurate performance evaluation, and provide relatively accurate boundary conditions for the three-dimensional single-stage simulation. Through multiple iterations, a compressor that meets the design objectives can be obtained. The present invention improves the result consistency between multi-dimensional simulations through downscaling data transfer, and thus, relying on the rapidity of low-dimensional simulations, shortens the design iteration cycle while ensuring the accuracy of simulations.

[0070] Please refer to Figure 1 , Figure 1 which is a flowchart of a design method for a multi-stage axial compressor of a gas turbine provided by the embodiments of the present application. As Figure 1 shown, the design method provided by the embodiments of the present application includes:

[0071] S101: According to the design objectives of the compressor to be designed, during the initial upscaling design process, determine in sequence a one-dimensional design scheme, a two-dimensional S2 stream surface design scheme, and an initial three-dimensional hot condition through-flow design scheme for the compressor to be designed.

[0072] In this step, according to the design objectives of the compressor to be designed, in the initial dimensionality-up design process, a one-dimensional design scheme, a two-dimensional S2 stream surface design scheme, and an initial three-dimensional hot flow design scheme for the compressor to be designed are determined successively.

[0073] Here, the design objectives are the requirements for the compressor design scheme. For example, if the compressor uses natural aspiration, the working medium is air, the target is to achieve a total pressure ratio of 23, the design point efficiency is not less than 88%, the surge margin is not less than 20%, etc. Please refer to Table 1.

[0074] Table 1 Design Objectives of Multistage Axial Compressor

[0075] Design point flow rate 166 kg / s Design point pressure ratio 23 Design point efficiency ≮88％ Surge margin at design speed ≮20％ Design speed 6000 RPM

[0076] Here, the one-dimensional design scheme is a design scheme for determining parameters such as axial velocity, reaction degree, axial clearance, and target load coefficient along the flow path according to the design objectives.

[0077] Here, the two-dimensional S2 stream surface design scheme is to determine the radial work distribution method and inlet pre-whirl for each stage based on the already determined one-dimensional design scheme.

[0078] Here, the initial three-dimensional hot flow design scheme is to select appropriate classical two-dimensional blade profiles based on the already determined two-dimensional S2 stream surface design scheme and stack them to generate three-dimensional rotor blades and stator blades.

[0079] Furthermore, the one-dimensional design scheme is determined through the following steps, including:

[0080] 1): Based on the design objectives, determine multiple one-dimensional design parameters for the compressor to be designed.

[0081] Among them, multiple one-dimensional design parameters for the compressor to be designed are determined according to the design objectives.

[0082] Here, the multiple one-dimensional design parameters include target parameters such as flow rate, pressure ratio, efficiency, and number of stages, boundary conditions such as total pressure and total temperature at the inlet, overall parameters such as rotational speed and tip tangential velocity of the first-stage rotor blade, the planned two-dimensional blade profile to be used, and other one-dimensional design parameters such as the step-by-step planning of parameters such as axial velocity, reaction degree, axial clearance, and target load coefficient along the flow path.

[0083] 2): Based on the multiple one-dimensional design parameters, perform one-dimensional design at the mean radius of the compressor to determine the one-dimensional design scheme.

[0084] Among them, the one-dimensional design scheme is obtained through one-dimensional inverse problem design at the mean radius for a preset number of times according to the multiple one-dimensional design parameters.

[0085] Among them, the one-dimensional design scheme is a design scheme for the compressor to be designed for the one-dimensional design parameters.

[0086] 3): Perform a forward problem simulation based on the elementary stage basic parameters and meridional channel parameters in the one-dimensional design scheme to determine the performance parameters of the one-dimensional design scheme and the aerodynamic and thermodynamic parameters at the mean radius of the compressor.

[0087] Among them, perform a forward problem simulation based on the elementary stage basic parameters and meridional channel parameters in the one-dimensional design scheme to obtain the performance parameters of the one-dimensional design scheme and the aerodynamic and thermodynamic parameters at the mean radius of the compressor. Among them, the elementary stage basic parameters and meridional channel parameters include other elementary stage basic parameters such as the inlet and outlet geometric angles, leading edge radius, size and position of the maximum thickness of each row of blades, and other meridional channel parameters such as the mean radius, hub ratio, aspect ratio, etc.

[0088] Among them, the performance parameters and aerodynamic and thermodynamic parameters include other performance parameters such as the diffusion factor, De Haller number, static pressure rise coefficient, incidence angle, trailing angle of each row of blades, the loading coefficient and reaction degree of each stage, and the efficiency and surge margin of the whole machine.

[0089] Here, during the initial dimension elevation design process, a default one-dimensional correction coefficient needs to be adopted.

[0090] Further, please refer to Figure 2 , Figure 2 which is a schematic diagram of the simulation results of the one-dimensional design scheme in a design method of a multi-stage axial flow compressor of a gas turbine provided by an embodiment of the present application. As Figure 2 shown, the elementary stage basic parameters and performance parameters stage by stage are given in Figure 2 : the reaction degree, loading coefficient, flow coefficient, and the diffusion factors of the rotor blade and stator blade.

[0091] Further, determine the two-dimensional S2 stream surface design scheme through the following steps, including:

[0092] (1): Compare the performance indicators such as flow rate, efficiency, and surge margin in the one-dimensional design scheme with the performance indicators such as target flow rate, target efficiency, and target surge margin in the design target.

[0093] Here, it also includes: simultaneously compare the performance parameters such as the diffusion factor, loading coefficient, static pressure rise coefficient, and De Haller number in the one-dimensional design scheme with the design limit values.

[0094] Here, the design rotational speed efficiency, pressure ratio, flow rate, surge margin, and the efficiency, pressure, flow rate, and surge margin under off-design conditions in the one-dimensional design scheme are obtained by analyzing the performance parameters of six characteristic lines, which are: 100% rated speed with the inlet adjustable guide vane fully open, 95.9% rated speed with the inlet adjustable guide vane fully open (corresponding to the design rotational speed at an ambient temperature of 40°C), 105.7% rated speed with the inlet adjustable guide vane fully open (corresponding to the design rotational speed at an ambient temperature of -15°C), 100% rated speed with the inlet adjustable guide vane fully closed, 70% rated speed with the inlet adjustable guide vane fully closed, and 50% rated speed with the inlet adjustable guide vane fully closed.

[0095] Here, the design limit values refer to the empirical limit values of performance parameters such as the diffusion factor, loading coefficient, static pressure rise coefficient, and De Haller number. For example, the diffusion factor is not greater than 0.45, the static pressure rise coefficient is not greater than 0.45, the De Haller number is not less than 0.70, and the loading coefficient of each stage is controlled at about 0.35 at most. The diffusion factor of the last-stage stator blade is not greater than 0.60. Please refer to Table 2.

[0096] Table 2 Design Limits of Multistage Axial Compressor

[0097] Height of exit guide vane ≥40 mm Exit axial Mach number ≯0.2 Limit value of stage loading coefficient ≯0.35 Limit value of diffusion factor ≯0.45 Limit value of De Haller number ≮0.70 Limit value of static pressure rise coefficient ≯0.45

[0098] (2) If the flow rate, efficiency, and surge margin in the one-dimensional design scheme do not meet the target flow rate, target efficiency, and target surge margin in the design objective, then change the number of stages and inlet pre-whirl of the one-dimensional design scheme, and adjust the axial velocity, degree of reaction, target loading coefficient, etc., as well as the distribution of elemental parameters along the flow path.

[0099] Among them, if the performance indicators such as the flow rate in the one-dimensional design scheme are inconsistent with the target performance indicators such as the target flow rate in the design objective, or the performance indicators such as the efficiency and surge margin in the one-dimensional design scheme are inferior to the target efficiency and target surge margin in the design objective, or the performance parameters such as the diffusion factor, loading coefficient, and static pressure rise coefficient in the one-dimensional design scheme are greater than the design limit values, then change the design parameters such as the number of stages and inlet pre-whirl of the one-dimensional design scheme, and adjust the distribution of elemental parameters such as the axial velocity, degree of reaction, and target loading coefficient along the flow path.

[0100] Here, the main adjustment contents are the axial velocity, degree of reaction, and target loading coefficient distributed along the flow path. If the diffusion factor or static pressure rise coefficient of the middle and rear stages exceeds the design limit value, then reduce the target loading coefficient of the middle and rear stages, or increase the axial velocity of the middle and rear stages. If the surge margin at the design rotational speed is inferior to the target surge margin, then reduce the target loading coefficient of the surge-induced stage. If the diffusion factor or static pressure rise coefficient of the rotor blade exceeds the design limit value, then reduce the degree of reaction; if the diffusion factor or static pressure rise coefficient of the stator blade exceeds the design limit value, then increase the degree of reaction, etc.

[0101] (3) If the flow rate, efficiency, and surge margin in the one-dimensional design scheme meet the target flow rate, target efficiency, and target surge margin in the design objectives, then based on the one-dimensional design scheme, the work distribution and reaction degree design in the radial direction of the compressor are carried out to determine the two-dimensional S2 stream surface design scheme.

[0102] Among them, if other performance indicators such as the flow rate in the one-dimensional design scheme correspond to and are consistent with other performance indicators such as the target flow rate in the design objectives, and performance indicators such as the efficiency and surge margin in the one-dimensional design scheme are better than the target efficiency and target surge margin in the design objectives, and performance parameters such as the diffusion factor, loading coefficient, and static pressure rise coefficient in the one-dimensional design scheme are less than the design limit values, then based on the one-dimensional design scheme, the work distribution and reaction degree design in the radial direction of the compressor are carried out to determine the two-dimensional S2 stream surface design scheme.

[0103] For example, if the flow rate of the one-dimensional design scheme is consistent with the target flow rate in the design objectives, the surge margin in the one-dimensional design scheme is better than the target surge margin in the design objectives, and the efficiency in the one-dimensional design scheme is better than the target efficiency in the design objectives.

[0104] At the same time, the diffusion factor in the one-dimensional design scheme is not greater than the set limit value of 0.45, the loading coefficient is not greater than the set limit value of 0.35, the static pressure rise coefficient is not greater than the set limit value of 0.45, and the De Haller number is not less than the set limit value of 0.70.

[0105] Among them, the two-dimensional S2 stream surface design based on the one-dimensional design scheme is to use the aerodynamic and thermodynamic parameters at the mean radius in the forward problem simulation results of the one-dimensional design scheme as the boundary conditions for the two-dimensional S2 stream surface design and the design input parameters at the mean diameter.

[0106] (4) According to the two-dimensional S2 stream surface design scheme, a forward problem simulation of the two-dimensional S2 stream surface is carried out to determine the performance parameters of the two-dimensional S2 stream surface design scheme and the aerodynamic and thermodynamic parameters averaged in the circumferential direction of the compressor.

[0107] Among them, the performance parameters and aerodynamic and thermodynamic parameters include the radial distributions of the diffusion factor, De Haller number, static pressure rise coefficient, incidence angle, and trailing angle of each row of blades, the radial distributions of the loading coefficient and reaction degree of each stage, and other performance parameters such as the efficiency and surge margin of the whole machine.

[0108] Here, in the initial dimension elevation design process, the default two-dimensional S2 stream surface correction coefficient needs to be adopted. Try to avoid using radial models with weak robustness such as radial mixing and end region blockage. Only the two-dimensional blade profile loss model is used in the loss model.

[0109] Further, an initial three-dimensional hot condition flow path design scheme is determined through the following methods, including:

[0110] a: Comparing the flow rate, efficiency, and surge margin in the two-dimensional S2 stream surface design scheme with the target flow rate, target efficiency, and target surge margin in the design objectives.

[0111] Here, it also includes: simultaneously comparing the radial distributions of performance parameters such as the diffusion factor, loading coefficient, static pressure rise coefficient, and De Haller number in the two-dimensional S2 stream surface design scheme with the design limit values.

[0112] Among them, the design rotational speed efficiency, pressure ratio, flow rate, surge margin, and the efficiency, pressure, flow rate, and surge margin under off-design conditions in the two-dimensional S2 stream surface design scheme are obtained by analyzing the performance parameters of six characteristic lines. These six characteristic lines are: 100% rated speed with the inlet adjustable guide vane fully open, 95.9% rated speed with the inlet adjustable guide vane fully open (corresponding to the design rotational speed at an ambient temperature of 40°C), 105.7% rated speed with the inlet adjustable guide vane fully open (corresponding to the design rotational speed at an ambient temperature of -15°C), 100% rated speed with the inlet adjustable guide vane fully closed, 70% rated speed with the inlet adjustable guide vane fully closed, and 50% rated speed with the inlet adjustable guide vane fully closed.

[0113] b: If the flow rate, efficiency, and surge margin in the two-dimensional S2 stream surface design scheme do not meet the target flow rate, target efficiency, and target surge margin in the design objectives, then change the step-by-step radial work distribution method in the two-dimensional S2 stream surface design scheme.

[0114] Among them, if other performance indicators such as the flow rate in the two-dimensional S2 stream surface design scheme are inconsistent with other performance indicators such as the target flow rate in the design objectives, or performance indicators such as the efficiency and surge margin in the two-dimensional S2 stream surface design scheme are inferior to the corresponding target indicators in the design objectives, or the radial distributions of performance parameters such as the diffusion factor, loading coefficient, and static pressure rise coefficient in the two-dimensional S2 stream surface design scheme are greater than the design limit values, then change the step-by-step radial work distribution method in the two-dimensional S2 stream surface design scheme.

[0115] Here, the radial work distribution method can be changed to various methods such as equal work distribution and equal reaction degree distribution.

[0116] c: If the flow rate, efficiency, and surge margin in the two-dimensional S2 stream surface design scheme meet the target flow rate, target efficiency, and target surge margin in the design objectives, then select a two-dimensional blade profile based on the two-dimensional S2 stream surface design scheme, and stack the two-dimensional blade profiles to determine the three-dimensional blades.

[0117] Among them, if other performance indicators such as the flow rate in the two-dimensional S2 flow surface design scheme correspond to the target flow rate and other performance indicators in the design target, and performance indicators such as efficiency and surge margin in the two-dimensional S2 flow surface design scheme are better than the corresponding target indicators in the design target, and the radial distributions of performance parameters such as diffusion factor, loading coefficient, and static pressure rise coefficient in the two-dimensional S2 flow surface design scheme are less than the design limit values, then a two-dimensional airfoil is selected based on the two-dimensional S2 flow surface design scheme, and the three-dimensional blade is determined by stacking the two-dimensional airfoils.

[0118] Among them, the two-dimensional airfoil is a standard two-dimensional airfoil generated based on conditions such as Mach number, Reynolds number, and inlet and outlet airflow angles in the two-dimensional S2 flow surface design scheme.

[0119] Among them, the three-dimensional blade is obtained by stacking on the basis of the two-dimensional airfoil.

[0120] d: Based on the three-dimensional blade, an initial three-dimensional hot flow path design scheme is determined, and a forward problem simulation is performed on the initial three-dimensional hot flow path design scheme to obtain multiple initial three-dimensional performance parameters in the initial three-dimensional hot flow path design scheme.

[0121] Among them, three-dimensional numerical simulation technology (CFD) is used to predict the performance of the three-dimensional flow field of the three-dimensional blade. The simulation results of the two-dimensional S2 flow surface design scheme are used as the inlet and outlet boundary conditions, and the stage performance of the compressor is predicted by step-by-step simulation. The inlet boundary conditions are total temperature, total pressure, and airflow angle direction, and the outlet boundary condition is the outlet back pressure. The step-by-step three-dimensional simulation will be carried out based on multiple working conditions, including at least the following seven working condition points: the design working condition with the inlet adjustable guide vane fully open (the pressure ratio is consistent with the target pressure ratio), the near-surge working condition at the design speed with the inlet adjustable guide vane fully open (the back pressure is 115% of the design working condition), the near-blockage working condition at the design speed with the inlet adjustable guide vane fully open (the back pressure is 85% of the design working condition), the common working condition at a low speed (95.9% of the rated speed) with the inlet adjustable guide vane fully open (the pressure ratio is consistent with the target pressure ratio at 95.9% speed), the near-surge working condition at a low speed (95.9% of the rated speed) with the inlet adjustable guide vane fully open (the back pressure is 115% of the common working condition at 95.9% speed), the common working condition at a super speed (105.7% of the rated speed) with the inlet adjustable guide vane fully open (the pressure ratio is consistent with the target pressure ratio at 105.7% speed), the near-surge working condition at a low speed (70% of the rated speed) with the inlet adjustable guide vane fully closed (the back pressure is 115% of the common working condition at 70% speed). After the simulation is completed, the three-dimensional simulation results are parametrically analyzed to obtain the performance parameters of the initial three-dimensional hot flow path design scheme and the detailed aerodynamic and thermodynamic parameters in the flow field.

[0122] Among them, the performance parameters and detailed aerodynamic and thermodynamic parameters in the flow field include the flow direction positions of each stage, pressure ratio, trailing angles of moving blades and stationary blades, flow rate, load coefficient, efficiency, and other performance parameters and aerodynamic and thermodynamic parameters.

[0123] S102: Input multiple initial three-dimensional performance parameters in the initial three-dimensional hot flow design scheme, multiple corresponding one-dimensional performance parameters in the one-dimensional design scheme, and multiple corresponding two-dimensional S2 stream surface performance parameters in the two-dimensional S2 stream surface design scheme into a pre-established dimensionality reduction data transfer model, and output multiple one-dimensional correction coefficients in the one-dimensional direct problem simulation process and multiple two-dimensional S2 stream surface correction coefficients in the two-dimensional S2 stream surface direct problem simulation process; among them, the one-dimensional correction coefficients and the two-dimensional S2 stream surface correction coefficients are obtained by comparing parameters after dimensionality reduction processing of multiple initial three-dimensional performance parameters in the initial three-dimensional hot flow design scheme.

[0124] In this step, input multiple initial three-dimensional performance parameters in the initial three-dimensional hot flow design scheme, multiple corresponding one-dimensional performance parameters in the one-dimensional design scheme, and multiple corresponding two-dimensional S2 stream surface performance parameters in the two-dimensional S2 stream surface design scheme into a pre-established dimensionality reduction data transfer model, and output multiple one-dimensional correction coefficients in the one-dimensional direct problem simulation process and multiple two-dimensional S2 stream surface correction coefficients in the two-dimensional S2 stream surface direct problem simulation process.

[0125] Among them, the dimensionality reduction data transfer model adopts a neural network model, and the training and test samples are obtained by performing data dimensionality reduction on multiple initial three-dimensional performance parameters in the initial three-dimensional hot flow design scheme, multiple one-dimensional performance parameters in the one-dimensional design scheme, and multiple two-dimensional S2 stream surface performance parameters in the two-dimensional S2 stream surface design scheme.

[0126] Among them, the one-dimensional correction coefficients and two-dimensional S2 stream surface correction coefficients are generated by the dimensionality reduction data transfer model.

[0127] Here, the one-dimensional correction coefficients may include one-dimensional trailing angle correction coefficients, flow rate correction coefficients, load correction coefficients, efficiency correction coefficients, etc.

[0128] Here, the two-dimensional S2 stream surface correction coefficients may include two-dimensional S2 stream surface stationary blade loss / moving blade efficiency correction coefficients, trailing angle correction coefficients, axial velocity correction coefficients, etc.

[0129] Here, the initial three-dimensional performance parameters include the relative reduced speed of the compressor to be designed in the initial three-dimensional hot flow design scheme, as well as the flow direction positions of each stage, pressure ratio, flow rate, loss / efficiency, trailing angle, blockage factor, and other performance parameters.

[0130] Here, the one-dimensional performance parameters include the relative reduced speed of the compressor to be designed in the one-dimensional design scheme, as well as the parameters at the mean diameter of the flow direction position, pressure ratio, flow rate, loss / efficiency, and stagger angle at each stage, and other performance parameters such as the blockage factor.

[0131] Here, the two-dimensional S2 stream surface performance parameters include the relative reduced speed of the compressor to be designed in the two-dimensional S2 stream surface design scheme, as well as the radial profiles of the flow direction position, pressure ratio, flow rate, loss / efficiency, and stagger angle at each stage, and other performance parameters such as the blockage factor.

[0132] Among them, the dimensionality reduction data transfer model is used to establish a neural network model after dimensionality reduction processing of the performance parameters in the initial three-dimensional hot flow design scheme, one-dimensional design scheme, and two-dimensional S2 stream surface design scheme.

[0133] Among them, multiple performance parameters in the initial three-dimensional hot flow design scheme, corresponding performance parameters in the one-dimensional design scheme and two-dimensional S2 stream surface design scheme, and parameters such as rotational speed and flow direction position are input into the pre-established dimensionality reduction data transfer model. The model quality is improved through training, and multiple one-dimensional correction coefficients in the one-dimensional forward problem simulation and multiple two-dimensional S2 stream surface correction coefficients in the two-dimensional S2 stream surface forward problem simulation required for the re-dimensioning design are generated by calling the model.

[0134] Further, please refer to Figure 3 , Figure 3 , which is the data dimensionality reduction transfer flow chart of a design method for a multi-stage axial flow compressor of a gas turbine provided by an embodiment of the present application. As Figure 3 shown, determining multiple one-dimensional correction coefficients for the one-dimensional design scheme and multiple two-dimensional S2 stream surface correction coefficients for the two-dimensional S2 stream surface design scheme through the following steps includes:

[0135] S301: Input multiple initial three-dimensional performance parameters in the initial three-dimensional hot flow design scheme, corresponding multiple one-dimensional performance parameters in the one-dimensional design scheme, and corresponding multiple two-dimensional S2 stream surface performance parameters in the two-dimensional S2 stream surface design scheme into the pre-established dimensionality reduction data transfer model.

[0136] Here, the dimensionality reduction data transfer model is obtained through training of an artificial neural network.

[0137] Here, multiple initial three-dimensional performance parameters in the initial three-dimensional hot flow design scheme, corresponding multiple one-dimensional performance parameters in the one-dimensional design scheme, and corresponding multiple two-dimensional S2 stream surface performance parameters in the two-dimensional S2 stream surface design scheme will be processed by the dimensionality reduction data transfer model into reference one-dimensional correction coefficients and reference two-dimensional S2 stream surface correction coefficients as the input of the neural network model.

[0138] S302: For each stage of the compressor to be designed, substitute an initial three-dimensional performance parameter corresponding to the initial three-dimensional through-flow design scheme of this stage into the empirical formula of one-dimensional simulation, and combine relevant design and performance parameters of one-dimensional design to determine the reference one-dimensional correction coefficient corresponding to this one-dimensional performance parameter of this stage.

[0139] For the reference one-dimensional correction coefficient being the reference stagger angle correction coefficient, determine the reference stagger angle correction coefficient in the one-dimensional dimensionality reduction data transfer model through the following formula:

[0140]

[0141] where, x f is the relative position of the maximum camber of the airfoil in the one-dimensional design scheme, B 2k is the geometric outlet angle in the one-dimensional design scheme, b t is the solidity in the one-dimensional design scheme, θ is the airfoil camber angle in the one-dimensional design scheme, K 1ref is the reference stagger angle correction coefficient in the one-dimensional dimensionality reduction data transfer model, and δ is the stagger angle in the initial three-dimensional design scheme. During the one-dimensional direct problem simulation, δ in the above formula is changed to the stagger angle of the rotor or stator blades in the one-dimensional design scheme, and the reference correction coefficient K 1ref is changed to the correction coefficient K1 generated by the dimensionality reduction data transfer model.

[0142] For the reference one-dimensional correction coefficient being the reference flow rate correction coefficient, determine the reference flow rate correction coefficient in the one-dimensional dimensionality reduction data transfer model through the following formula.

[0143] q = K 2ref ·q0;

[0144] where, q is the flow coefficient at the inlet of the current stage, q0 is the flow coefficient at the outlet of the previous stage (if the current stage is the first stage, it is the flow coefficient at the inlet of the compressor in the initial three-dimensional design scheme), and K 2ref is the reference flow rate correction coefficient. During the one-dimensional direct problem simulation, if the current stage is the first stage, q0 in the above formula is changed to the flow coefficient at the inlet of the compressor in the one-dimensional design scheme, and the reference correction coefficient K 2ref is changed to the correction coefficient K2 generated by the dimensionality reduction data transfer model.

[0145] For the reference one-dimensional correction coefficient being the reference load correction coefficient, determine the reference load correction coefficient in the one-dimensional dimensionality reduction data transfer model through the following formula.

[0146] H Z0 = H T0 ·K 3ref ·U K1 ;

[0147] Among them, H Z0 is the actual machining amount of the moving blade in the initial three-dimensional hot flow design scheme, and H T0 is the theoretical machining amount of the moving blade in the one-dimensional design scheme, and K 3ref is the reference load correction coefficient in the one-dimensional dimensionality reduction data transfer model, and U K1 is the circumferential velocity at the inlet of the moving blade in the one-dimensional design scheme. During the one-dimensional direct problem simulation, H in the above formula Z0 is changed to the actual machining amount of the moving blade in the one-dimensional design scheme, and the reference correction coefficient K 3ref is changed to the correction coefficient K3 generated by the dimensionality reduction data transfer model.

[0148] Regarding the reference one-dimensional correction coefficient as the reference efficiency correction coefficient, the reference efficiency correction coefficient in the one-dimensional dimensionality reduction data transfer model is determined by the following formula:

[0149] η o = [η k ·τ + η a ·(1 - τ)]·K 4ref ;

[0150] Among them, η0 is the compressor stage efficiency in the initial three-dimensional hot flow design scheme, η k and η a are the efficiencies of the moving blade and the stationary blade respectively, τ is the degree of reaction, and K 4ref is the reference efficiency correction coefficient in the one-dimensional dimensionality reduction data transfer model. During the one-dimensional direct problem simulation, η0 in the above formula is changed to the compressor stage efficiency in the one-dimensional design scheme, and the reference correction coefficient K 4ref is changed to the correction coefficient K4 generated by the dimensionality reduction data transfer model.

[0151] S303: For each stage of the compressor to be designed, substitute an initial three-dimensional performance parameter in the initial three-dimensional hot flow design scheme of this stage into the empirical formula of the two-dimensional S2 stream surface simulation, and combine the relevant design and performance parameters of the two-dimensional S2 stream surface design to determine the reference two-dimensional S2 stream surface correction coefficient corresponding to the two-dimensional S2 stream surface performance parameter of this stage.

[0152] Regarding the reference two-dimensional S2 stream surface correction coefficient as the reference stationary blade loss / moving blade efficiency correction coefficient, the reference stationary blade loss / moving blade efficiency correction coefficient in the two-dimensional S2 stream surface dimensionality reduction data transfer model is determined by the following formula:

[0153] η k，j = η k0，j ·K 5ref，j , ω a，j = ω a0，j ·K 6ref，j ;

[0154] Among them, η k is the moving blade efficiency in the initial three-dimensional hot condition through-flow design scheme, η k0 is the moving blade efficiency in the two-dimensional S2 stream surface design scheme, K 5ref is the correction coefficient for the reference two-dimensional S2 stream surface moving blade efficiency, j represents the radial flow tube number, and K 5ref，j is a matrix composed of j elements; ω a is the stationary blade loss in the initial three-dimensional hot condition through-flow design scheme, ω a0 is the stationary blade loss in the two-dimensional S2 stream surface design scheme, K 6ref is the correction coefficient for the reference two-dimensional S2 stream surface stationary blade loss, j also represents the radial flow tube number, so K 6ref，j is a matrix composed of j elements. During the two-dimensional S2 stream surface direct problem simulation, the product of η k0 and the two-dimensional S2 stream surface correction coefficient η k is used as the moving blade efficiency in the output two-dimensional S2 stream surface design scheme; the product of ω a0 and the two-dimensional S2 stream surface correction coefficient ω a is used as the stationary blade loss in the output two-dimensional S2 stream surface design scheme, and the reference correction coefficient K 5ref is changed to the correction coefficient K5 generated by the dimensionality reduction data transfer model, and the reference correction coefficient K 6ref is changed to the correction coefficient K6 generated by the dimensionality reduction data transfer model.

[0155] Regarding the reference two-dimensional S2 stream surface correction coefficient as the reference stagger angle correction coefficient, the reference stagger angle correction coefficient in the two-dimensional S2 stream surface dimensionality reduction data transfer model is determined through the following formula:

[0156] δ k，j = δ k0，j ·K 7ref，j , δ a，j = δ a0，j ·K 8ref，j ;

[0157] Among them, δ k and δ a represent the stagger angles of the moving blade and the stationary blade in the initial three-dimensional hot condition through-flow design scheme respectively, δ k0 and δ a0 are the stagger angles of the moving blade and the stationary blade in the two-dimensional S2 stream surface design scheme, K 7ref and K 8ref are the reference two-dimensional S2 stream surface stagger angle correction coefficients respectively, j also represents the radial flow tube number, K 7ref，j and K 8ref，j are also matrices composed of j elements. During the two-dimensional S2 stream surface direct problem simulation, δk0 and δ a0 The product δ with the correction coefficient of the two-dimensional S2 stream surface k and δ a As the leading and trailing angles of the moving and stationary blades in the output two-dimensional S2 stream surface design scheme, referring to the correction coefficient K 7ref Changed to the correction coefficient K7 generated by the dimensionality reduction data transfer model, referring to the correction coefficient K 8ref Changed to the correction coefficient K8 generated by the dimensionality reduction data transfer model.

[0158] For the reference two-dimensional S2 stream surface correction coefficient being the reference axial velocity correction coefficient, the reference axial velocity correction coefficient in the two-dimensional S2 stream surface dimensionality reduction data transfer model is determined by the following formula:

[0159] V zk，j = V zk0，j ·K 9ref，j , V za，j = V za0，j ·K 10ref，j :

[0160] Wherein, V zk and V za are respectively the axial velocities at the inlet of the moving blade and the inlet of the stationary blade in the initial three-dimensional hot flow path design scheme, V zk0 and V za0 represent the axial velocities at the inlet of the moving blade and the inlet of the stationary blade in the two-dimensional S2 stream surface design scheme, j represents the radial flow tube number, K 9ref and K 10ref are the reference two-dimensional S2 stream surface axial velocity correction coefficients, K 9refj and K 10refj are respectively matrices composed of j elements, mainly used to introduce the blockage in the end region or the displacement thickness of the annulus wall boundary layer. During the two-dimensional S2 stream surface direct problem simulation, V zk0 and V za0 The products V zk and V za with the two-dimensional S2 stream surface correction coefficient are used as the axial velocities at the inlet of the moving blade and the inlet of the stationary blade in the output two-dimensional S2 stream surface design scheme, referring to the correction coefficient K 9ref Changed to the correction coefficient K9 generated by the dimensionality reduction data transfer model, referring to the correction coefficient K 10ref Changed to the correction coefficient K 10 .

[0161] S304: Based on the reference one-repair positive coefficient and the reference two-dimensional S2 flow surface correction coefficient generated by dimensionality reduction of the performance parameters of the initial three-dimensional hot flow design scheme, the dimensionality reduction data transfer model is trained, and the dimensionality reduction data transfer model outputs multiple one-repair positive coefficients in the one-dimensional design scheme positive problem simulation process and multiple two-dimensional S2 flow surface correction coefficients in the two-dimensional S2 flow surface design scheme positive problem simulation process.

[0162] Here, all the reference-maintenance positive coefficients and reference two-dimensional S2 flow surface correction coefficients generated by dimensionality reduction of the performance parameters of the initial three-dimensional hot flow design scheme are imported into the neural network model, and the performance parameters, speed, flow direction position, reference-maintenance positive coefficients and reference two-dimensional S2 flow surface correction coefficients of the compressor are used as model inputs, and the one-maintenance positive coefficient and the two-dimensional S2 flow surface correction coefficient are used as model outputs. The neural network model is trained to obtain a dimensionality reduction data transfer model based on a neural network.

[0163] For the one-dimensional dimension reduction data transfer model from three-dimensional to one-dimensional, the model form is:

[0164] [K1, K2, K3, K4] = f ANN (n c , π i , row i , K 1ref , K 2ref , K 3ref , K 4ref );

[0165] Among them, n c is the relative reduced speed, π i is the stage pressure ratio, row i is the flow position of the compressor stage, corresponding to the i-th stage of the compressor. Reference correction factor K 1ref ~K 4ref They are matrices with 1 row and i columns, and each number represents the reference lag angle, reference flow, reference load and reference efficiency correction coefficient of the corresponding stage. The correction coefficients K1~K4 are also matrices with 1 row and i columns, and each number represents the lag angle, flow, load and efficiency correction coefficient of the corresponding stage.

[0166] The two-dimensional S2 flow surface dimension reduction data transfer model for three-dimensional transmission to two-dimensional S2 flow surface has the following model form:

[0167] [K5, K6, K7, K8, K9, K 10 ] = f ANN (n c , π i , row i , K 5ref , K 6ref , K7ref , K 8ref , K 9ref , K 10ref );

[0168] Among them, n c is the relative reduced rotational speed, π i is the stage pressure ratio, row i is the flow direction position where the compressor stage is located, corresponding to the i-th stage of the compressor. The reference correction coefficients K 5ref ~K 1ref0 are matrices of the j-th row and i-th column respectively. j represents the radial flow tube number. Therefore, each column in the matrix represents the reference two-dimensional S2 flow surface correction coefficient of the corresponding stage. The correction coefficients K5~K 10 are also matrices of the j-th row and i-th column. j represents the radial flow tube number. Therefore, each column in the matrix represents the two-dimensional S2 flow surface correction coefficient of the corresponding stage.

[0169] Here, it will be determined whether to increase the number of operating conditions for three-dimensional simulation to supplement the training set of the model according to the training quality of the artificial neural network. For example, if the coefficient of determination R2 of the trained model is less than 0.9, representative three-dimensional simulation operating conditions should be appropriately increased as training samples. At the same time, it is necessary to ensure that at least 10% of the samples are reserved as the test set.

[0170] Here, the one-dimensional correction coefficients output by the one-dimensional dimensionality reduction data transfer model include the stagger angle correction coefficient, flow rate correction coefficient, load correction coefficient, efficiency correction coefficient, etc. The two-dimensional S2 flow surface correction coefficients output by the two-dimensional S2 flow surface dimensionality reduction data transfer model include the two-dimensional S2 flow surface stator loss / rotor efficiency correction coefficient, stagger angle correction coefficient, and axial velocity correction coefficient, etc.

[0171] S103: Based on the dimensionality reduction data transfer model described above, determine the target three-dimensional hot flow path design scheme, and detect whether the operating condition performance of the target three-dimensional hot flow path design scheme reaches the design target.

[0172] In this step, according to the multiple one-dimensional correction coefficients and multiple two-dimensional S2 flow surface correction coefficients determined by the dimensionality reduction data transfer model, determine the target three-dimensional hot flow path design scheme. After determining the target three-dimensional hot flow path design scheme, detect whether the operating condition performance of the target three-dimensional hot flow path design scheme reaches the design target.

[0173] Here, for the rotational speed and pressure ratio of each stage, based on the dimensionality reduction data transfer model corresponding to this stage, determine the one-dimensional correction coefficients in the one-dimensional direct problem simulation process.

[0174] Here, for the rotational speed and pressure ratio of each stage, a two-dimensional S2 stream surface correction coefficient in the two-dimensional S2 stream surface direct problem simulation is determined based on the reduced-dimension data transfer model corresponding to this stage.

[0175] The target three-dimensional hot flow path scheme is determined based on multiple one-dimensional correction coefficients generated by the reduced-dimension data transfer model and multiple two-dimensional S2 stream surface correction coefficients generated by the reduced-dimension data transfer model.

[0176] Further, through the following steps, a target three-dimensional hot flow path design scheme is determined based on multiple one-dimensional correction coefficients and multiple two-dimensional S2 stream surface correction coefficients:

[0177] i: Change the design parameters in the one-dimensional design scheme to determine a target one-dimensional design scheme. During the one-dimensional direct problem simulation, the one-dimensional correction coefficients generated by the reduced-dimension data transfer model are used.

[0178] Here, for each one-dimensional correction coefficient of each stage, based on the one-dimensional correction coefficient corresponding to this stage, the default one-dimensional correction coefficient corresponding to this stage and this correction coefficient used in the one-dimensional direct problem simulation during the initial dimension elevation design process is changed.

[0179] Among them, based on the one-dimensional correction coefficients determined during the one-dimensional direct problem simulation, the number of stages, inlet pre-whirl and other design parameters in the one-dimensional design scheme are changed, and the distributions of elementary parameters such as axial velocity, reaction degree, and target load coefficient along the flow path are adjusted to determine a target one-dimensional design scheme.

[0180] Among them, based on the multiple one-dimensional correction coefficients determined, a target one-dimensional design scheme is determined. The determination process of the target one-dimensional design scheme is the same as that of the one-dimensional design scheme, and this part will not be elaborated here.

[0181] ii: Change the stage-by-stage radial work distribution method in the two-dimensional S2 stream surface design scheme to determine a target two-dimensional S2 stream surface design scheme. During the two-dimensional S2 stream surface direct problem simulation, the two-dimensional S2 stream surface correction coefficients generated by the reduced-dimension data transfer model are used.

[0182] Here, for each two-dimensional S2 stream surface correction coefficient of each stage, based on the two-dimensional S2 stream surface correction coefficient corresponding to this stage, the default two-dimensional S2 stream surface correction coefficient corresponding to this stage and this correction coefficient used in the two-dimensional S2 stream surface direct problem simulation during the initial dimension elevation design process is changed.

[0183] Among them, based on the multiple two-dimensional S2 stream surface correction coefficients determined, a target two-dimensional S2 stream surface design scheme is determined. The determination process of the target two-dimensional S2 stream surface design scheme is the same as that of the two-dimensional S2 stream surface design scheme, and this part will not be elaborated here.

[0184] iii: Based on the target two-dimensional S2 stream surface design scheme, select a target two-dimensional blade profile for stacking to generate a target three-dimensional blade, and determine the target three-dimensional hot flow design scheme of the compressor based on the target three-dimensional blade.

[0185] Among them, based on the target two-dimensional S2 stream surface design scheme, re-select a suitable two-dimensional blade profile, stack and generate a three-dimensional blade, and determine the target three-dimensional hot flow design scheme of the compressor. The determination process of the target three-dimensional hot flow design scheme is consistent with the determination process of the initial three-dimensional hot flow design scheme, and this part will not be elaborated here.

[0186] Further, please refer to Figure 4 , Figure 4 , which is a schematic diagram of the simulation results of the target two-dimensional S2 stream surface design scheme in a design method for a multi-stage axial compressor of a gas turbine provided in an embodiment of the present application. As Figure 4 shown, the hollow circles are the simulation results of the forward problem of the two-dimensional S2 stream surface, and the solid lines are the simulation results of the three-dimensional forward problem. At the design point, the two-dimensional radial air flow angle distribution has a high consistency with the three-dimensional simulation results.

[0187] Further, please refer to Figure 5 , Figure 5 , which is a schematic diagram of the multi-dimensional simulation results in a design method for a multi-stage axial compressor of a gas turbine provided in an embodiment of the present application. As Figure 5 shown, after dimensionality reduction processing, the multi-dimensional simulation results of the compressor characteristic line have a high consistency.

[0188] S104: If the operating condition performance of the target three-dimensional hot flow design scheme reaches the design target, determine the cold geometric model and structural design of the compressor according to the target three-dimensional hot flow design scheme to obtain a compressor that meets the design target.

[0189] Here, the design target is the requirement of the compressor design scheme, such as the design point efficiency of the compressor, the surge margin at the design speed, etc.

[0190] Further, based on the target three-dimensional hot flow design scheme, through cold-hot state conversion and structural design, determine a three-dimensional cold blade model, including:

[0191] Detect whether the combined stress of the blade bending stress and tensile stress of the three-dimensional blade in the three-dimensional cold blade model is less than the ratio of the yield limit to the safety factor and the tensile stress in the three-dimensional blade is less than the ratio of the yield limit to the safety factor.

[0192] If the combined stress of the blade bending stress and the tensile stress in the three-dimensional cold-state blade model in the target three-dimensional hot-state flow design scheme is greater than the ratio of the yield limit to the safety factor or the tensile stress is greater than the ratio of the yield limit to the safety factor, then adjust the geometric model in the three-dimensional hot-state flow design scheme until the combined stress of the blade bending stress and the tensile stress in the three-dimensional cold-state blade model is less than the ratio of the yield limit to the safety factor and the tensile stress is less than the ratio of the yield limit to the safety factor, so as to obtain a compressor that meets the design objective.

[0193] Here, detect whether the combined stress of the blade bending stress and the tensile stress in the three-dimensional blade is less than the ratio of the yield limit to the safety factor and whether the tensile stress in the three-dimensional blade is less than the ratio of the yield limit to the safety factor. When the combined stress of the blade bending stress and the tensile stress in the three-dimensional blade in the target three-dimensional hot-state flow design scheme is greater than the ratio of the yield limit to the safety factor or the tensile stress is greater than the ratio of the yield limit to the safety factor, perform frequency modulation, shroud amount adjustment, etc. on the three-dimensional blade in the target three-dimensional hot-state flow design scheme until the combined stress of the blade bending stress and the tensile stress in the three-dimensional blade in the target three-dimensional hot-state flow design scheme is less than the ratio of the yield limit to the safety factor and the tensile stress is less than the ratio of the yield limit to the safety factor, so as to obtain a compressor that meets the design objective.

[0194] In a specific embodiment, please refer to Figure 6 , Figure 6 which is a schematic structural design diagram of a compressor in a design method of a multi-stage axial-flow compressor of a gas turbine provided by an embodiment of the present application. As Figure 6As shown in the figure, according to the design objectives, a target three-dimensional hot-state through-flow design scheme is determined, and the compressor is subjected to cold-hot state conversion and structural design based on the target three-dimensional hot-state through-flow design scheme. The radial intake method is adopted, and an intake volute is designed at the compressor inlet. A row of inlet adjustable guide vanes 2 is arranged on the inlet cylinder 1 at the compressor inlet. Two bleed ports, low-pressure and high-pressure, are arranged between stages, located at the outlets of the sixth-stage and eleventh-stage stator vanes respectively. Through the full-circumference bleed grooves, the mainstream fluid is introduced into the annular low-pressure bleed chamber 4 and high-pressure bleed chamber 6, and then through the circumferentially evenly distributed bleed pipes 5, the high-pressure gas is released into the atmosphere or reused. The cylinder adopts a split-plane structure form, and the flange surface between the compressor cylinder 3 and the combustion-pressure cylinder 7 is arranged at the low-pressure bleed chamber 4. The impeller disk 13 of the compressor rotor is installed on the main shaft 15 by shrink fitting. There are axial dovetail-type blade root grooves and circumferential "T"-type moving blade fixing grooves on the disk, and "T"-type positioning keys are installed in the moving blade fixing grooves to prevent the axial movement of the compressor moving blades 11. After the stator blades 12 are fixed by inner and outer shroud bands, they are welded to form a complete circle of stator blade ring 14, and then the stator blade ring is cut into upper and lower half rings and screwed into the stator blade grooves of the cylinder. The compressor is designed for axial exhaust, and the high-temperature and high-pressure gas at the compressor outlet is discharged into the large chamber of the combustion-pressure cylinder 7 through the diffuser 8.

[0195] Further, the design method further includes: if any performance in the operating conditions performance of the target three-dimensional hot-state through-flow design scheme is inconsistent with the corresponding operating conditions performance in the target two-dimensional S2 stream surface design scheme and the corresponding operating conditions performance in the target one-dimensional design scheme, then perform a dimensionality reduction transfer iteration process on the initial three-dimensional hot-state through-flow design scheme until the operating conditions performance of the target three-dimensional hot-state through-flow design scheme, the operating conditions performance of the target two-dimensional S2 stream surface design scheme, and the operating conditions performance of the target one-dimensional design scheme are all correspondingly consistent.

[0196] Here, if the design point pressure ratio and the airflow deflection angle in the target three-dimensional hot-state through-flow design scheme deviate greatly from the forward problem simulation results of the target one-dimensional design scheme or the target two-dimensional S2 stream surface design scheme, then appropriately adjust the inlet and outlet geometric angles of the three-dimensional blades to improve the inter-stage matching. If it is still difficult to improve the inter-stage matching, then change the radial work distribution method and start iterative design from the design process of the target two-dimensional S2 stream surface design scheme; or adjust the distribution of elemental parameters such as circumferential velocity, reaction degree, and target load coefficient along the process and start iterative design from the design process of the target one-dimensional design scheme. Stop the iterative process until the three-dimensional through-flow simulation results are better than the compressor design objectives.

[0197] In a specific embodiment, the design objective of the compressor to be designed is that the compressor uses natural aspiration, the working medium is air, the target total pressure ratio is 23, the design point efficiency is not less than 88%, and the surge margin is not less than 20%. According to the design objective, a one-dimensional design scheme for parameters such as the axial velocity, degree of reaction, axial clearance, and target loading coefficient along the flow path is determined. It is judged whether the one-dimensional design scheme meets the design objective. If it meets, based on the one-dimensional design scheme, the radial work distribution method and the inlet pre-whirl at each stage are determined, and a two-dimensional S2 flow surface design scheme for the inlet and outlet geometric angles of the moving blades and stationary blades at each stage along the radial direction is obtained. It is judged whether the two-dimensional S2 flow surface design scheme meets the design objective. If it meets, an initial three-dimensional hot flow path design scheme is carried out based on the two-dimensional S2 flow surface design scheme. Data dimensionality reduction processing is performed on the performance parameters of the initial three-dimensional hot flow path design scheme to obtain a dimensionality reduction data transfer model based on a neural network. Using multiple one-dimensional correction coefficients and multiple two-dimensional S2 flow surface correction coefficients generated by the dimensionality reduction data transfer model, the design is dimensionally ascended again to obtain a target three-dimensional hot flow path design scheme. It is judged whether the target three-dimensional hot flow path design scheme meets the design objective. If it meets, the cold-state geometric model and structural design of the compressor are determined according to the three-dimensional hot flow path design scheme, and a compressor meeting the design objective is obtained.

[0198] The present application provides a design method for a multi-stage axial flow compressor of a gas turbine. The design method includes: according to the design objective of the compressor to be designed, successively determining a one-dimensional design scheme, a two-dimensional S2 flow surface design scheme, and an initial three-dimensional hot flow path design scheme for the compressor to be designed during the initial dimensionality ascension design process; inputting multiple initial three-dimensional performance parameters in the initial three-dimensional hot flow path design scheme, multiple corresponding one-dimensional performance parameters in the one-dimensional design scheme, and multiple corresponding two-dimensional S2 flow surface performance parameters in the two-dimensional S2 flow surface design scheme into a pre-established dimensionality reduction data transfer model, and outputting multiple one-dimensional correction coefficients in the one-dimensional direct problem simulation process and multiple two-dimensional S2 flow surface correction coefficients in the two-dimensional S2 flow surface direct problem simulation process; wherein, the one-dimensional correction coefficients and the two-dimensional S2 flow surface correction coefficients are obtained by comparing parameters after dimensionality reduction processing of multiple initial three-dimensional performance parameters in the initial three-dimensional hot flow path design scheme; based on multiple one-dimensional correction coefficients and multiple two-dimensional S2 flow surface correction coefficients, a target three-dimensional hot flow path design scheme is determined, and it is detected whether the operating conditions performance of the target three-dimensional hot flow path design scheme reaches the design objective. If the operating conditions performance of the target three-dimensional hot flow path design scheme reaches the design objective, the cold-state geometric model and structural design of the compressor are determined according to the target three-dimensional hot flow path design scheme, and a compressor meeting the design objective is obtained.

[0199] In this way, during the initial dimensionality-up design process, the low-dimensional simulation tool needs to avoid using radial models with relatively weak robustness, such as radial mixing and end-region blockage. After completing the one-dimensional inverse problem design, the one-dimensional forward problem simulation, the two-dimensional S2 stream surface inverse problem design, the two-dimensional forward problem simulation, the three-dimensional blade shaping design, and the three-dimensional single-stage forward problem simulation, the performance parameters in the three-dimensional simulation results are extracted, and a dimensionality reduction data transfer model is obtained through training a neural network. During the next dimensionality-up design, the low-dimensional simulation tool will combine the obtained dimensionality reduction data, correct the correction coefficients in the one-dimensional simulation and the correction coefficients in the two-dimensional S2 stream surface simulation, give a more accurate performance evaluation, and provide relatively accurate boundary conditions for the three-dimensional single-stage simulation. Through multiple iterations, a compressor that meets the design objectives can be obtained. The present invention improves the result consistency between multi-dimensional simulations through dimensionality reduction data transfer, and thus, relying on the rapidity of low-dimensional simulations, shortens the design iteration cycle while ensuring the simulation accuracy.

[0200] Please refer to Figure 7 、 Figure 8 , Figure 7 which is one of the structural schematic diagrams of a design device for a multi-stage axial-flow compressor of a gas turbine provided by an embodiment of the present application. Figure 8 which is the second of the structural schematic diagrams of a design device for a multi-stage axial-flow compressor of a gas turbine provided by an embodiment of the present application. As shown in Figure 7 the design device 700 includes:

[0201] A first dimensionality-up design module 710, configured to sequentially determine a one-dimensional design scheme, a two-dimensional S2 stream surface design scheme, and an initial three-dimensional hot-state through-flow design scheme for the compressor to be designed during the initial dimensionality-up design process according to the design objectives of the compressor to be designed;

[0202] A dimensionality reduction data transfer module 720, configured to input multiple initial three-dimensional performance parameters in the initial three-dimensional hot-state through-flow design scheme, multiple corresponding one-dimensional performance parameters in the one-dimensional design scheme, and multiple corresponding two-dimensional S2 stream surface performance parameters in the two-dimensional S2 stream surface design scheme into a pre-established dimensionality reduction data transfer model, and output multiple one-dimensional correction coefficients during the one-dimensional forward problem simulation and multiple two-dimensional S2 stream surface correction coefficients during the two-dimensional S2 stream surface forward problem simulation; wherein, the one-dimensional correction coefficients and the two-dimensional S2 stream surface correction coefficients are obtained by comparing parameters after dimensionality reduction processing of multiple initial three-dimensional performance parameters in the initial three-dimensional hot-state through-flow design scheme;

[0203] A second dimensionality-up design module 730, configured to determine a target three-dimensional hot-state through-flow design scheme based on multiple one-dimensional correction coefficients and multiple two-dimensional S2 stream surface correction coefficients, and detect whether the operating conditions performance of the target three-dimensional hot-state through-flow design scheme reaches the design objectives;

[0204] A structural design module 740, which is configured to, if the operating performance of the target three-dimensional hot-state flow-through design scheme reaches the design target, determine the cold-state geometric model and structural design of the compressor according to the target three-dimensional hot-state flow-through design scheme, so as to obtain a compressor that meets the design target.

[0205] Further, when the structural design module 740 is used to determine the cold-state geometric model and structural design of the compressor according to the target three-dimensional hot-state flow-through design scheme to obtain a compressor that meets the design target, the structural design module 740 is further configured to:

[0206] Based on the target three-dimensional hot-state flow-through design scheme, determine a three-dimensional cold-state blade model through cold-hot state conversion and structural design;

[0207] Detect whether the combined stress of the blade bending stress and the tensile stress of the three-dimensional blades in the three-dimensional cold-state blade model is less than the ratio of the yield limit to the safety factor and whether the tensile stress in the three-dimensional blades is less than the ratio of the yield limit to the safety factor;

[0208] If the combined stress of the blade bending stress and the tensile stress of the three-dimensional cold-state blade model in the target three-dimensional hot-state flow-through design scheme is greater than the ratio of the yield limit to the safety factor or the tensile stress is greater than the ratio of the yield limit to the safety factor, adjust the geometric model in the three-dimensional hot-state flow-through design scheme until the combined stress of the blade bending stress and the tensile stress of the three-dimensional cold-state blade model is less than the ratio of the yield limit to the safety factor and the tensile stress is less than the ratio of the yield limit to the safety factor, so as to obtain a compressor that meets the design target.

[0209] Further, the first dimension-elevating design module 710 is configured to determine a one-dimensional design scheme through the following steps:

[0210] Based on the design target, determine a plurality of one-dimensional design parameters for the compressor to be designed;

[0211] Based on the plurality of one-dimensional design parameters, perform one-dimensional design at the average radius of the compressor to determine a one-dimensional design scheme;

[0212] According to the basic parameters of the elementary stage and the meridian flow path parameters in the one-dimensional design scheme, perform a forward problem simulation to determine the performance parameters of the one-dimensional design scheme and the aerodynamic thermodynamic parameters at the average radius of the compressor.

[0213] Further, the first dimension-elevating design module 710 is configured to determine a two-dimensional S2 stream surface design scheme through the following steps, including:

[0214] Compare the flow rate, efficiency, and surge margin in the one-dimensional design solution with the target flow rate, target efficiency, and target surge margin in the design objective;

[0215] If the flow rate, efficiency, and surge margin in the one-dimensional design solution do not meet the target flow rate, target efficiency, and target surge margin in the design objective, then change the number of stages and inlet pre-whirl of the one-dimensional design solution, and adjust the distribution of axial velocity, reaction degree, target loading coefficient, and elementary parameters along the flow path;

[0216] If the flow rate, efficiency, and surge margin in the one-dimensional design solution meet the target flow rate, target efficiency, and target surge margin in the design objective, then perform work distribution and reaction degree design in the radial direction of the compressor based on the one-dimensional design solution to determine a two-dimensional S2 stream surface design solution;

[0217] Conduct a forward problem simulation of the two-dimensional S2 stream surface according to the two-dimensional S2 stream surface design solution to determine the performance parameters of the two-dimensional S2 stream surface design solution and the aerodynamic and thermodynamic parameters averaged in the circumferential direction of the compressor.

[0218] Furthermore, the first dimension elevation design module 710 is used to determine an initial three-dimensional hot condition through-flow design solution in the following manner, including:

[0219] Compare the flow rate, efficiency, and surge margin in the two-dimensional S2 stream surface design solution with the target flow rate, target efficiency, and target surge margin in the design objective;

[0220] If the flow rate, efficiency, and surge margin in the two-dimensional S2 stream surface design solution do not meet the target flow rate, target efficiency, and target surge margin in the design objective, then change the stage-by-stage radial work distribution method in the two-dimensional S2 stream surface design solution;

[0221] If the flow rate, efficiency, and surge margin in the two-dimensional S2 stream surface design solution meet the target flow rate, target efficiency, and target surge margin in the design objective, then select a two-dimensional blade profile based on the two-dimensional S2 stream surface design solution, and stack the two-dimensional blade profiles to determine a three-dimensional blade;

[0222] Based on the three-dimensional blade, determine an initial three-dimensional hot condition through-flow design solution, and conduct a forward problem simulation of the initial three-dimensional hot condition through-flow design solution to obtain multiple initial three-dimensional performance parameters in the initial three-dimensional hot condition through-flow design solution.

[0223] Further, when the dimensionality reduction data transfer module 720 is used for the compressor to be designed which is a multi-stage axial flow compressor, the following steps are used to determine multiple one-dimensional correction coefficients for the one-dimensional design scheme and multiple two-dimensional S2 stream surface correction coefficients for the two-dimensional S2 stream surface design scheme:

[0224] Input multiple initial three-dimensional performance parameters in the initial three-dimensional hot flow design scheme, multiple corresponding one-dimensional performance parameters in the one-dimensional design scheme, and multiple corresponding two-dimensional S2 stream surface performance parameters in the two-dimensional S2 stream surface design scheme into the pre-established dimensionality reduction data transfer model;

[0225] For each stage of the compressor to be designed, substitute an initial three-dimensional performance parameter corresponding to the initial three-dimensional hot flow design scheme of this stage into the empirical formula of one-dimensional simulation, and combine relevant design and performance parameters of one-dimensional design to determine the reference one-dimensional correction coefficient corresponding to this one-dimensional performance parameter of this stage;

[0226] For each stage of the compressor to be designed, substitute an initial three-dimensional performance parameter in the initial three-dimensional hot flow design scheme of this stage into the empirical formula of two-dimensional S2 stream surface simulation, and combine relevant design and performance parameters of two-dimensional S2 stream surface design to determine the reference two-dimensional S2 stream surface correction coefficient corresponding to the two-dimensional S2 stream surface performance parameter of this stage.

[0227] Based on the reference one-dimensional correction coefficients and reference two-dimensional S2 stream surface correction coefficients generated by dimensionality reduction of the performance parameters of the initial three-dimensional hot flow design scheme, train the dimensionality reduction data transfer model, and the dimensionality reduction data transfer model outputs multiple one-dimensional correction coefficients in the forward problem simulation process of the one-dimensional design scheme and multiple two-dimensional S2 stream surface correction coefficients in the forward problem simulation process of the two-dimensional S2 stream surface design scheme.

[0228] Further, the second dimensionality increase design module 730 is used to determine the target three-dimensional hot flow design scheme through the following steps:

[0229] For the rotational speed and pressure ratio of each stage, determine the one-dimensional correction coefficients in the one-dimensional forward problem simulation based on the dimensionality reduction data transfer model corresponding to this stage;

[0230] For the rotational speed and pressure ratio of each stage, determine the two-dimensional S2 stream surface correction coefficients in the two-dimensional S2 stream surface forward problem simulation based on the dimensionality reduction data transfer model corresponding to this stage;

[0231] Change the design parameters in the one-dimensional design scheme to determine the target one-dimensional design scheme. In the one-dimensional forward problem simulation, use the one-dimensional correction coefficients generated by the dimensionality reduction data transfer model;

[0232] Change the step-by-step radial work distribution method of the two-dimensional S2 flow surface design scheme to determine the target two-dimensional S2 flow surface design scheme. During the simulation process of the two-dimensional S2 flow surface direct problem, use the two-dimensional S2 flow surface correction coefficient generated by the dimensionality reduction data transfer model;

[0233] Based on the target two-dimensional S2 flow surface design scheme, select the target two-dimensional blade profiles for stacking to generate the target three-dimensional blade, and determine the target three-dimensional hot flow path design scheme of the compressor based on the target three-dimensional blade.

[0234] Further, as Figure 8 shown, the design device 700 further includes an iterative processing module 750, and the iterative processing module 750 is used for:

[0235] If any performance in the operating conditions performance of the target three-dimensional hot flow path design scheme is inconsistent with the corresponding operating conditions performance in the target two-dimensional S2 flow surface design scheme and the corresponding operating conditions performance in the target one-dimensional design scheme, perform dimensionality reduction transfer iterative processing on the initial three-dimensional hot flow path design scheme until the operating conditions performance of the target three-dimensional hot flow path design scheme is corresponding and consistent with the operating conditions performance of the target two-dimensional S2 flow surface design scheme and the operating conditions performance of the target one-dimensional design scheme.

[0236] The present application provides a design device for a multi-stage axial flow compressor of a gas turbine. The first dimensionality increase design module is used to sequentially determine a one-dimensional design scheme, a two-dimensional S2 flow surface design scheme, and an initial three-dimensional hot flow path design scheme for the compressor to be designed during the initial dimensionality increase design process; the dimensionality reduction data transfer module is used to input multiple initial three-dimensional performance parameters in the initial three-dimensional hot flow path design scheme, multiple corresponding one-dimensional performance parameters in the one-dimensional design scheme, and multiple corresponding two-dimensional S2 flow surface performance parameters in the two-dimensional S2 flow surface design scheme into a pre-established dimensionality reduction data transfer model, and output multiple one-dimensional correction coefficients during the one-dimensional direct problem simulation process and multiple two-dimensional S2 flow surface correction coefficients during the two-dimensional S2 flow surface direct problem simulation process; wherein, the one-dimensional correction coefficients and the two-dimensional S2 flow surface correction coefficients are obtained through parameter comparison after dimensionality reduction processing of multiple initial three-dimensional performance parameters in the initial three-dimensional hot flow path design scheme; the second dimensionality increase design module is used to determine a target three-dimensional hot flow path design scheme based on multiple one-dimensional correction coefficients and multiple two-dimensional S2 flow surface correction coefficients, and detect whether the operating conditions performance of the target three-dimensional hot flow path design scheme reaches the design target; the structure design module is used to, if the operating conditions performance of the target three-dimensional hot flow path design scheme reaches the design target, determine the cold-state geometric model and structure design of the compressor according to the target three-dimensional hot flow path design scheme to obtain a compressor that meets the design target.

[0237] In this way, during the initial dimensionality-up design process, the low-dimensional simulation tool needs to avoid using radial models with relatively weak robustness, such as radial mixing and end-region blockage. After completing the one-dimensional inverse problem design and forward problem simulation, two-dimensional S2 stream surface inverse problem design and forward problem simulation, three-dimensional blade modeling design, and three-dimensional single-stage forward problem simulation, the performance parameters in the three-dimensional simulation results are extracted, and a dimensionality reduction data transfer model is obtained through training a neural network. During the subsequent dimensionality-up design, the low-dimensional simulation tool will combine the obtained dimensionality reduction data, correct the correction coefficients in the one-dimensional simulation and the correction coefficients in the two-dimensional S2 stream surface simulation, give a more accurate performance evaluation, and provide relatively accurate boundary conditions for the three-dimensional single-stage simulation. Through multiple iterations, a compressor that meets the design objectives can be obtained. The present invention improves the result consistency between multi-dimensional simulations through dimensionality reduction data transfer, and thus, relying on the rapidity of low-dimensional simulations, shortens the design iteration cycle while ensuring the accuracy of the simulations.

[0238] Please refer to Figure 9 , Figure 9 which is a schematic structural diagram of an electronic device provided by an embodiment of the present application. As Figure 9 shown in

[0239] , the electronic device 900 includes a processor 910, a memory 920, and a bus 930. Figure 1 and Figure 3 shown, the memory 920 stores machine-readable instructions executable by the processor 910. When the electronic device 900 runs, the processor 910 communicates with the memory 920 through the bus 930. When the machine-readable instructions are executed by the processor 910, the steps of a design method for a multi-stage axial-flow compressor of a gas turbine as described in the method embodiments above

[0240] can be executed. For the specific implementation manners, reference can be made to the method embodiments, which will not be elaborated herein. Figure 1 and Figure 3 shown can be executed. For the specific implementation manners, reference can be made to the method embodiments, which will not be elaborated herein.

[0241] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments, which will not be elaborated herein.

[0242] In several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there may be other division methods in actual implementation. For another example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some communication interfaces, and the indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0243] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0244] In addition, in each embodiment of this application, the functional units can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0245] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a non-volatile computer-readable storage medium executable by a processor. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or this part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of this application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks, or optical discs that can store program codes.

[0246] Finally, it should be noted that the above-described embodiments are only specific embodiments of the present application, which are used to illustrate the technical solutions of the present application, rather than limiting it. The protection scope of the present application is not limited thereto. Although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that any person skilled in the art within the technical scope disclosed by the present application can still modify the technical solutions described in the foregoing embodiments, or can easily think of changes, or make equivalent replacements for some of the technical features; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A design method for a multi-stage axial compressor of a gas turbine, characterized in that, The design method includes: According to the design objectives of the compressor to be designed, in the initial dimensionality-up design process, determine in sequence the one-dimensional design scheme, two-dimensional S2 stream surface design scheme, and initial three-dimensional hot-state flow path design scheme for the compressor to be designed; Input multiple initial three-dimensional performance parameters in the initial three-dimensional hot-state flow path design scheme, multiple corresponding one-dimensional performance parameters in the one-dimensional design scheme, and multiple corresponding two-dimensional S2 stream surface performance parameters in the two-dimensional S2 stream surface design scheme into the pre-established dimensionality-reduction data transfer model, and output multiple one-dimensional correction coefficients in the one-dimensional direct problem simulation process and multiple two-dimensional S2 stream surface correction coefficients in the two-dimensional S2 stream surface direct problem simulation process; wherein, the one-dimensional correction coefficients and the two-dimensional S2 stream surface correction coefficients are obtained through parameter comparison after dimensionality reduction processing of multiple initial three-dimensional performance parameters in the initial three-dimensional hot-state flow path design scheme; Based on multiple one-dimensional correction coefficients and multiple two-dimensional S2 stream surface correction coefficients, determine the target three-dimensional hot-state flow path design scheme, and detect whether the operating conditions performance of the target three-dimensional hot-state flow path design scheme reaches the design objectives; If the operating conditions performance of the target three-dimensional hot-state flow path design scheme reaches the design objectives, then determine the cold-state geometric model and structural design of the compressor according to the target three-dimensional hot-state flow path design scheme, and obtain a compressor that meets the design objectives; The compressor to be designed is a multi-stage axial-flow compressor. The steps to determine multiple one-dimensional correction coefficients of the one-dimensional design scheme and multiple two-dimensional S2 stream surface correction coefficients of the two-dimensional S2 stream surface design scheme include: Input multiple initial three-dimensional performance parameters in the initial three-dimensional hot-state flow path design scheme, multiple corresponding one-dimensional performance parameters in the one-dimensional design scheme, and multiple corresponding two-dimensional S2 stream surface performance parameters in the two-dimensional S2 stream surface design scheme into the pre-established dimensionality-reduction data transfer model; For each stage of the compressor to be designed, substitute a corresponding initial three-dimensional performance parameter in the initial three-dimensional hot-state flow path design scheme of this stage into the empirical formula of one-dimensional simulation, and combine the relevant design and performance parameters of one-dimensional design to determine the reference one-dimensional correction coefficient corresponding to this one-dimensional performance parameter of this stage; For each stage of the compressor to be designed, substitute an initial three-dimensional performance parameter in the initial three-dimensional hot-state flow path design scheme of this stage into the empirical formula of two-dimensional S2 stream surface simulation, and combine the relevant design and performance parameters of two-dimensional S2 stream surface design to determine the reference two-dimensional S2 stream surface correction coefficient corresponding to the two-dimensional S2 stream surface performance parameter of this stage; Based on the reference one-dimensional correction coefficients and reference two-dimensional S2 stream surface correction coefficients generated by dimensionality reduction of the performance parameters of the initial three-dimensional hot-state flow path design scheme, train the dimensionality-reduction data transfer model, and output multiple one-dimensional correction coefficients in the one-dimensional design scheme direct problem simulation process and multiple two-dimensional S2 stream surface correction coefficients in the two-dimensional S2 stream surface design scheme direct problem simulation process by the dimensionality-reduction data transfer model; Determine the target three-dimensional hot-state flow path design scheme through the following steps: For each stage of rotational speed and pressure ratio, a one-dimensional correction coefficient in the one-dimensional direct problem simulation is determined based on the reduced-dimension data transfer model corresponding to this stage; For each stage of rotational speed and pressure ratio, a two-dimensional S2 stream surface correction coefficient in the two-dimensional S2 stream surface direct problem simulation is determined based on the reduced-dimension data transfer model corresponding to this stage; Change the design parameters in the one-dimensional design scheme to determine the target one-dimensional design scheme. During the one-dimensional direct problem simulation, use the one-dimensional correction coefficient generated by the reduced-dimension data transfer model; Change the stage-by-stage radial work distribution method of the two-dimensional S2 stream surface design scheme to determine the target two-dimensional S2 stream surface design scheme. During the two-dimensional S2 stream surface direct problem simulation, use the two-dimensional S2 stream surface correction coefficient generated by the reduced-dimension data transfer model; Based on the target two-dimensional S2 stream surface design scheme, select the target two-dimensional airfoil for stacking to generate the target three-dimensional blade, and determine the target three-dimensional hot flow design scheme of the compressor based on the target three-dimensional blade; The design method further includes: If any performance in the operating conditions performance of the target three-dimensional hot flow design scheme is inconsistent with the corresponding operating conditions performance in the target two-dimensional S2 stream surface design scheme and the corresponding operating conditions performance in the target one-dimensional design scheme, perform reduced-dimension transfer iterative processing on the initial three-dimensional hot flow design scheme until the operating conditions performance of the target three-dimensional hot flow design scheme is in correspondence with the operating conditions performance of the target two-dimensional S2 stream surface design scheme and the operating conditions performance of the target one-dimensional design scheme.

2. The design method according to claim 1, characterized in that, Determining the cold-state geometric model and structural design of the compressor according to the target three-dimensional hot flow design scheme to obtain a compressor meeting the design target includes: Based on the target three-dimensional hot flow design scheme, determine the three-dimensional cold-state blade model through cold-hot state conversion and structural design; Detect whether the combined stress of the blade bending stress and tensile stress of the three-dimensional blades in the three-dimensional cold-state blade model is less than the ratio of the yield limit to the safety factor and whether the tensile stress in the three-dimensional blades is less than the ratio of the yield limit to the safety factor; If the combined stress of the blade bending stress and tensile stress of the three-dimensional cold-state blade model in the target three-dimensional hot flow design scheme is greater than the ratio of the yield limit to the safety factor or the tensile stress is greater than the ratio of the yield limit to the safety factor, adjust the geometric model in the three-dimensional hot flow design scheme until the combined stress of the blade bending stress and tensile stress of the three-dimensional cold-state blade model is less than the ratio of the yield limit to the safety factor and the tensile stress is less than the ratio of the yield limit to the safety factor, to obtain a compressor meeting the design target.

3. The design method according to claim 1, characterized in that, The one-dimensional design scheme is determined through the following steps, including: Based on the design target, determine multiple one-dimensional design parameters for the compressor to be designed; Based on the multiple one-dimensional design parameters, perform one-dimensional design at the mean radius of the compressor to determine the one-dimensional design scheme; Perform a forward problem simulation based on the elementary stage basic parameters and meridional flow path parameters in the one-dimensional design scheme to determine the performance parameters of the one-dimensional design scheme and the aerothermodynamic parameters at the mean radius of the compressor.

4. The design method according to claim 1, characterized in that, Determine the two-dimensional S2 stream surface design scheme through the following steps, including: Compare the flow rate, efficiency, and surge margin in the one-dimensional design scheme with the target flow rate, target efficiency, and target surge margin in the design objective; If the flow rate, efficiency, and surge margin in the one-dimensional design scheme do not meet the target flow rate, target efficiency, and target surge margin in the design objective, then change the number of stages and inlet pre-whirl of the one-dimensional design scheme, and adjust the axial velocity, reaction degree, target loading coefficient, and the distribution of elementary parameters along the flow path; If the flow rate, efficiency, and surge margin in the one-dimensional design scheme meet the target flow rate, target efficiency, and target surge margin in the design objective, then perform work distribution and reaction degree design in the radial direction of the compressor based on the one-dimensional design scheme to determine the two-dimensional S2 stream surface design scheme; Perform a forward problem simulation of the two-dimensional S2 stream surface according to the two-dimensional S2 stream surface design scheme to determine the performance parameters of the two-dimensional S2 stream surface design scheme and the aerothermodynamic parameters averaged in the circumferential direction of the compressor.

5. The design method according to claim 1, characterized in that, Determine the initial three-dimensional hot condition through-flow design scheme in the following way, including: Compare the flow rate, efficiency, and surge margin in the two-dimensional S2 stream surface design scheme with the target flow rate, target efficiency, and target surge margin in the design objective; If the flow rate, efficiency, and surge margin in the two-dimensional S2 stream surface design scheme do not meet the target flow rate, target efficiency, and target surge margin in the design objective, then change the stage-by-stage radial work distribution method in the two-dimensional S2 stream surface design scheme; If the flow rate, efficiency, and surge margin in the two-dimensional S2 stream surface design scheme meet the target flow rate, target efficiency, and target surge margin in the design objective, then select two-dimensional blade profiles based on the two-dimensional S2 stream surface design scheme, and stack the two-dimensional blade profiles to determine three-dimensional blades; Based on the three-dimensional blades, determine the initial three-dimensional hot condition through-flow design scheme, and perform a forward problem simulation on the initial three-dimensional hot condition through-flow design scheme to obtain multiple initial three-dimensional performance parameters in the initial three-dimensional hot condition through-flow design scheme.

6. A design device for a multi-stage axial flow compressor of a gas turbine, characterized in that, The design device includes: A first dimension elevation design module for sequentially determining a one-dimensional design scheme, a two-dimensional S2 stream surface design scheme, and an initial three-dimensional hot condition through-flow design scheme for the compressor to be designed during the initial dimension elevation design process according to the design objective of the compressor to be designed; The dimensionality reduction data transfer module is used to input multiple initial three-dimensional performance parameters in the initial three-dimensional hot flow design scheme, multiple corresponding one-dimensional performance parameters in the one-dimensional design scheme, and multiple corresponding two-dimensional S2 flow surface performance parameters in the two-dimensional S2 flow surface design scheme into a pre-established dimensionality reduction data transfer model, and output multiple one-dimensional correction coefficients in the one-dimensional direct problem simulation process and multiple two-dimensional S2 flow surface correction coefficients in the two-dimensional S2 flow surface direct problem simulation process; wherein, the one-dimensional correction coefficients and the two-dimensional S2 flow surface correction coefficients are obtained by comparing parameters after dimensionality reduction processing of multiple initial three-dimensional performance parameters in the initial three-dimensional hot flow design scheme. The second dimensionality increase design module is used to determine a target three-dimensional hot flow design scheme based on multiple one-dimensional correction coefficients and multiple two-dimensional S2 flow surface correction coefficients, and detect whether the operating conditions performance of the target three-dimensional hot flow design scheme reaches the design target. The structure design module is used to, if the operating conditions performance of the target three-dimensional hot flow design scheme reaches the design target, determine the cold-state geometric model and structural design of the compressor according to the target three-dimensional hot flow design scheme, and obtain a compressor that meets the design target. When the dimensionality reduction data transfer module is used for the compressor to be designed as a multi-stage axial flow compressor and determines multiple one-dimensional correction coefficients of the one-dimensional design scheme and multiple two-dimensional S2 flow surface correction coefficients of the two-dimensional S2 flow surface design scheme through the following steps, the dimensionality reduction data transfer module is specifically used for: Input multiple initial three-dimensional performance parameters in the initial three-dimensional hot flow design scheme, multiple corresponding one-dimensional performance parameters in the one-dimensional design scheme, and multiple corresponding two-dimensional S2 flow surface performance parameters in the two-dimensional S2 flow surface design scheme into a pre-established dimensionality reduction data transfer model. For each stage of the compressor to be designed, substitute one initial three-dimensional performance parameter corresponding to the initial three-dimensional hot flow design scheme of this stage into the empirical formula of one-dimensional simulation, and combine the relevant design and performance parameters of one-dimensional design to determine the reference one-dimensional correction coefficient corresponding to this one-dimensional performance parameter of this stage. For each stage of the compressor to be designed, substitute one initial three-dimensional performance parameter in the initial three-dimensional hot flow design scheme of this stage into the empirical formula of two-dimensional S2 flow surface simulation, and combine the relevant design and performance parameters of two-dimensional S2 flow surface design to determine the reference two-dimensional S2 flow surface correction coefficient corresponding to the two-dimensional S2 flow surface performance parameter of this stage. Based on the reference one-dimensional correction coefficients and reference two-dimensional S2 flow surface correction coefficients generated by dimensionality reduction of the performance parameters of the initial three-dimensional hot flow design scheme, train the dimensionality reduction data transfer model, and output multiple one-dimensional correction coefficients in the one-dimensional direct problem simulation process of the one-dimensional design scheme and multiple two-dimensional S2 flow surface correction coefficients in the two-dimensional S2 flow surface direct problem simulation process by the dimensionality reduction data transfer model. The second dimensionality increase design module determines the target three-dimensional hot flow design scheme through the following steps: For each stage of rotational speed and pressure ratio, a one-dimensional correction coefficient in the one-dimensional direct problem simulation is determined based on the reduced-dimensional data transfer model corresponding to this stage; For each stage of rotational speed and pressure ratio, a two-dimensional S2 stream surface correction coefficient in the two-dimensional S2 stream surface direct problem simulation is determined based on the reduced-dimensional data transfer model corresponding to this stage; The design parameters in the one-dimensional design scheme are changed to determine a target one-dimensional design scheme. In the one-dimensional direct problem simulation, the one-dimensional correction coefficient generated by the reduced-dimensional data transfer model is adopted; The stage-by-stage radial work distribution method of the two-dimensional S2 stream surface design scheme is changed to determine a target two-dimensional S2 stream surface design scheme. In the two-dimensional S2 stream surface direct problem simulation, the two-dimensional S2 stream surface correction coefficient generated by the reduced-dimensional data transfer model is adopted; Based on the target two-dimensional S2 stream surface design scheme, a target two-dimensional airfoil is selected for stacking to generate a target three-dimensional blade, and a target three-dimensional hot flow path design scheme of the compressor is determined based on the target three-dimensional blade; The design device further includes an iterative processing module, and the iterative processing module is used for: If any performance in the operating conditions performance of the target three-dimensional hot flow path design scheme is inconsistent with the corresponding operating conditions performance in the target two-dimensional S2 stream surface design scheme and the corresponding operating conditions performance in the target one-dimensional design scheme, then a reduced-dimensional transfer iterative processing is performed on the initial three-dimensional hot flow path design scheme until the operating conditions performance of the target three-dimensional hot flow path design scheme is corresponding consistent with the operating conditions performance of the target two-dimensional S2 stream surface design scheme and the operating conditions performance of the target one-dimensional design scheme.

7. An electronic device, characterized in that, Including: A processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the electronic device runs, the processor communicates with the memory through the bus. When the machine-readable instructions are run by the processor, the steps of a design method of a multi-stage axial flow compressor of a gas turbine as described in any one of claims 1 to 5 are executed.

Images