Design method, system and medium for multi-working-condition heavy-duty gas turbine compressor

Abstract

The application provides a heavy-duty gas turbine compressor design method, system and medium suitable for multiple working conditions. The method comprises the following steps: in the two-dimensional through-flow inverse problem design stage, through a plurality of set blade profile geometric parameters and a preset artificial neural network proxy model, through-flow loss and lag angle are calculated; based on a positive problem analysis file generated by a positive and inverse problem coupling design, two-dimensional through-flow positive problem analysis is carried out, including judging whether the compressor working margin meets the requirements by calculating a stall factor; for the deviation between the matching parameters of each stage of the compressor and the expected values of the through-flow design, the adjustment amount of the blade profile geometric parameters corresponding to the deviation is predicted through through-flow analysis; and the VGVs control law of the compressor under variable load working conditions is optimized and designed by using the positive problem analysis file and a particle swarm optimization algorithm. The method fully utilizes the positive problem analysis file generated in the positive and inverse problem coupling mode, can reduce the number of iterative designs, and improves the aerodynamic design efficiency.

Description

Technical Field

[0001] This application relates to the field of gas turbine compressor design technology, and in particular to a design method, system and medium for heavy-duty gas turbine compressors applicable to multiple operating conditions. Background Technology

[0002] As the power and efficiency parameters of heavy-duty gas turbines continue to increase, the inlet flow rate and stage load requirements of the compressor, one of the three major components of the gas turbine, are also gradually increasing. This trend makes the internal flow of the compressor more complex, and various losses such as tip leakage and wake mixing increase significantly, making the compressor matching design problem more prominent.

[0003] Furthermore, since the compressor's operating efficiency drops significantly under varying load conditions in gas turbines, current gas turbine models all employ multi-row adjustable stationary vanes (VGVs) technology in order to expand the compressor's operating range and improve its efficiency under varying load conditions.

[0004] The compressor design schemes in related technologies cannot set reasonable initial losses for each stage of the blades, resulting in a large number of design iterations, low design efficiency, and an inability to reasonably determine the control law of VGVs under different load conditions. Summary of the Invention

[0005] This application aims to at least partially address one of the technical problems in the related art.

[0006] Therefore, the first objective of this application is to propose a design method for heavy-duty gas turbine compressors applicable to multiple operating conditions. This method makes full use of the forward problem analysis file generated under the forward and backward problem coupling mode, which can reduce the number of iterative design times and improve aerodynamic design efficiency.

[0007] The second objective of this application is to propose a design system for heavy-duty gas turbine compressors applicable to multiple operating conditions.

[0008] The third objective of this application is to provide a non-transitory computer-readable storage medium.

[0009] To achieve the above objectives, the first aspect of this application is to propose a design method for a heavy-duty gas turbine compressor applicable to multiple operating conditions, comprising the following steps:

[0010] In the two-dimensional inverse flow problem design stage of heavy-duty gas turbine compressor, the flow loss and lag angle are calculated by setting multiple blade geometry parameters and pre-built artificial neural network surrogate model to achieve coupled design of forward and inverse problems;

[0011] Based on the positive problem analysis file generated by the positive and negative problem coupling design, a two-dimensional flow positive problem analysis is performed, wherein the two-dimensional flow positive problem analysis includes determining whether the compressor operating margin meets the requirements by calculating the stall factor;

[0012] After completing the two-dimensional flow forward problem analysis, airfoil design, and three-dimensional CFD analysis, the airfoil geometry parameter adjustment amount corresponding to the deviation between the compressor's matching parameters at each stage and the expected values ​​of the flow design is predicted through flow analysis.

[0013] In response to the completion of aerodynamic design and blade frequency tuning, the VGVs control law of the multi-row adjustable stator blades of the compressor under variable load conditions is optimized using the aforementioned forward problem analysis file and particle swarm optimization algorithm, and a compressor design scheme containing the designed VGVs control law is output.

[0014] Optionally, the artificial neural network surrogate model is pre-constructed, including: taking multiple airfoil geometric parameters as input parameters, selecting multiple target parameters from the input parameters based on the compressor operating conditions; generating multiple two-dimensional airfoils based on the multiple target parameters, and obtaining the flow loss and lag angle of the multiple two-dimensional airfoils under corresponding environmental conditions through numerical calculation; and using the multiple target parameters and the corresponding flow loss and lag angle, establishing a surrogate model through an artificial neural network to associate the flow loss and lag angle with the relationship between the multiple target parameters.

[0015] Optionally, the step of determining whether the compressor operating margin meets the requirements by calculating the stall factor includes: determining the current blade grid limit D factor based on the correlation curve between the blade grid limit D factor and the airfoil bend angle; calculating the value of the stall factor based on the blade grid diffusion factor and the current blade grid limit D factor; and evaluating the compressor operating margin based on the value of the stall factor.

[0016] Optionally, the step of predicting the adjustment amount of the blade profile geometry parameters corresponding to the deviation through flow analysis includes: analyzing the blade profile geometry deviation corresponding to the deviation through the positive problem analysis file; using flow analysis tools to predict the adjustment direction and the adjustment amount of the blade profile geometry parameters corresponding to the blade profile geometry deviation; and based on the adjustment direction and the adjustment amount, iteratively performing blade profile adjustment design, three-dimensional CFD analysis, and deviation analysis until the matching parameters at each level meet the requirements.

[0017] Optionally, the optimization design of the VGVs control law for the multi-row adjustable stator vanes of the compressor under variable load conditions includes: initializing the particle swarm using the closing angle of each row of adjustable stator vanes as input variables and the isentropic efficiency at the compressor's common operating point as the optimization objective, wherein different closing angles of the inlet guide vanes are used to reflect different load conditions; calculating the compressor characteristic curve using the forward problem analysis file, and calculating the isentropic efficiency at the compressor's common operating point based on the compressor characteristic curve; determining the isentropic efficiency of each particle based on the isentropic efficiency at the compressor's common operating point, and correcting the isentropic efficiency of each particle using the compressor inlet flow rate to obtain the fitness of each particle; and performing automatic particle flight optimization based on the obtained fitness until the VGVs control law with the highest isentropic efficiency is obtained.

[0018] Optionally, calculating the isentropic efficiency of the compressor common operating point based on the compressor characteristic line includes: calculating the inlet flow rate of the compressor common operating point based on the compressor characteristic line; calculating the pressure ratio of the compressor common operating point based on the inlet flow rate of the compressor common operating point; and obtaining the isentropic efficiency of the compressor common operating point by interpolation of the compressor characteristic line using the inlet flow rate and the pressure ratio of the compressor common operating point.

[0019] Optionally, the isentropic efficiency of each particle can be corrected using the following formula:

[0020]

[0021] in, It is the isentropic efficiency after correction for any particle. It is the isentropic efficiency of any particle determined based on the isentropic efficiency at the common operating point of the compressor. It is the intake airflow at the common operating point of the compressor. It optimizes the intake flow rate at the common operating point of the starting point. It is the intake airflow at the compressor design point. It is the isentropic efficiency at the compressor design point. It optimizes the isentropic efficiency of the common working point at the starting point.

[0022] To achieve the above objectives, a second aspect of this application also proposes a heavy-duty gas turbine compressor design system applicable to multiple operating conditions, comprising the following modules:

[0023] The calculation module is used to calculate the flow loss and lag angle in the two-dimensional inverse flow problem design stage of heavy-duty gas turbine compressors by setting multiple airfoil geometric parameters and a pre-built artificial neural network surrogate model, so as to realize the coupled design of forward and inverse problems.

[0024] The analysis module is used to perform two-dimensional flow positive problem analysis based on the positive problem analysis file generated by the positive and negative problem coupling design. The two-dimensional flow positive problem analysis includes determining whether the compressor operating margin meets the requirements by calculating the stall factor.

[0025] The prediction module is used to predict the adjustment amount of the airfoil geometry parameters corresponding to the deviation between the matching parameters of each stage of the compressor and the expected value of the flow design after completing the two-dimensional flow positive problem analysis, airfoil design and three-dimensional CFD analysis.

[0026] The optimization module is used to optimize the VGVs control law of the multi-row adjustable stator blades of the compressor under variable load conditions in response to the completion of aerodynamic design and blade frequency modulation. It utilizes the positive problem analysis file and particle swarm optimization algorithm to optimize the VGVs control law of the compressor under variable load conditions and outputs a compressor design scheme containing the completed VGVs control law.

[0027] To achieve the above objectives, the third aspect of this application also proposes a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the heavy-duty gas turbine compressor design method applicable to multiple operating conditions as described in any one of the first aspects.

[0028] The technical solutions provided by the embodiments of this application bring at least the following beneficial effects: This application adopts a coupled design mode of positive and negative problems. In the two-dimensional flow design stage, the compressor margin can be quickly evaluated through flow positive problem analysis, and the compressor efficiency can be quickly evaluated through a multi-circular arc blade loss and lag angle surrogate model based on artificial neural networks. This allows for a relatively reasonable aerodynamic layout scheme to be provided in the first round of design, thereby greatly reducing the number of iterative designs and shortening the compressor aerodynamic layout iteration cycle. Furthermore, this application uses flow tool prediction bias for compressor matching adjustment and utilizes flow analysis tools to quickly predict the blade parameter adjustment direction. This transforms the inter-stage matching design iteration from iteration between blade design and time-consuming three-dimensional CFD analysis into flow-based iteration, thereby reducing the number of three-dimensional CFD analyses, greatly accelerating the compressor inter-stage matching design speed, and significantly shortening the compressor matching design iteration cycle. Moreover, this application optimizes the VGVs control law, effectively predicting the optimal VGVs control law of the gas turbine compressor under variable load operating conditions, providing theoretical reference values ​​for actual unit operation and commissioning. This helps reduce reliance on unit operating experience and the risk of human error during gas turbine operation and commissioning, and improves the operating efficiency and reliability of the compressor under non-design conditions.

[0029] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0030] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0031] Figure 1 A flowchart illustrating a heavy-duty gas turbine compressor design method applicable to multiple operating conditions, as proposed in an embodiment of this application;

[0032] Figure 2 This is a schematic diagram of the design process of a compressor in a related embodiment;

[0033] Figure 3 This is a schematic diagram illustrating a specific compressor design process according to an embodiment of this application;

[0034] Figure 4 This is a schematic diagram of the correlation curve between the blade gate limit D factor and the airfoil bend angle proposed in an embodiment of this application;

[0035] Figure 5 This is a schematic diagram of a matching design process based on flow prediction proposed in an embodiment of this application;

[0036] Figure 6 A flowchart illustrating an optimization design method for VGVs control laws proposed in an embodiment of this application;

[0037] Figure 7 This is a schematic diagram of a specific VGVs control law optimization process proposed in an embodiment of this application;

[0038] Figure 8 This is a schematic diagram comparing the highest compressor efficiency before and after VGVs control law optimization according to an embodiment of this application;

[0039] Figure 9 This is a schematic diagram of a heavy-duty gas turbine compressor design system applicable to multiple operating conditions, as proposed in an embodiment of this application. Detailed Implementation

[0040] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0041] It should be noted that, due to the high power and large circulating air flow of heavy-duty gas turbines, multi-stage axial flow compressors are typically used for gas intake and compression in relevant embodiments. However, in the design of multi-stage axial flow compressors, the large number of stages can easily lead to incoordination between stages. Furthermore, because the compressor operates with counter-pressure flow, high-load areas are prone to stall and separation, and the internal flow structure of the compressor is quite complex, making accurate simulation difficult.

[0042] Based on the aforementioned problems, inappropriate selection of interstage parameters in any stage of a multi-stage axial compressor can lead to performance degradation of that stage and even affect the operation of other rows of blades, causing their operating parameters to deviate from the design parameters, thus preventing the overall compressor performance from meeting design goals. In related embodiments, it is difficult to provide reasonable initial loss values ​​and distributions when designing the aerodynamic layout of the compressor. Furthermore, after the initial design of the compressor aerodynamic scheme is completed, a series of structural integrity designs are required, such as adjusting the shroud amount of the blade overlap line, adjusting the frequency tuning design of the blade thickness and chord length, and adjusting the excitation intensity of the inter-row spacing. This necessitates multiple iterations to complete the compressor aerodynamic design scheme, resulting in a long development cycle.

[0043] Besides the design point performance, for the multi-row adjustable stator vane adjustment technology of compressors that optimizes the operating performance under non-design conditions, there is a lack of efficient methods in the relevant embodiments for determining how to find the optimal adjustment law for each specific operating condition during the design phase. It requires a lot of testing and debugging to determine.

[0044] In summary, the design process of the heavy-duty gas turbine compressor in the relevant embodiments mainly suffers from the following three problems: First, in the aerodynamic layout design of the compressor, the given values ​​for the losses of each row of blades rely heavily on engineers' experience. As the compressor flow load index increases, it is difficult to determine a suitable flow loss coefficient at the initial design stage, and consequently, it is difficult to accurately predict the matching conditions and operating margins of each stage of the compressor. Second, the design process involves multiple iterations, such as small iterations of blade profile matching design and large iterations of structural integrity design, resulting in a long development cycle. Third, during the variable operating condition operation of the gas turbine, the control law relies on engineering experience and is difficult to determine efficiently.

[0045] Therefore, this application proposes a design method and system for heavy-duty gas turbine compressors applicable to multiple operating conditions. By considering the geometric and aerodynamic parameters of the compressor in advance at the initial stage of aerodynamic layout design, the number of iterative design cycles can be reduced, R&D efficiency can be improved, and the R&D cycle can be shortened.

[0046] The following description, with reference to the accompanying drawings, illustrates a design method and system for a heavy-duty gas turbine compressor applicable to multiple operating conditions, as proposed in the embodiments of this application.

[0047] Figure 1This is a flowchart illustrating a design method for a heavy-duty gas turbine compressor applicable to multiple operating conditions, as proposed in an embodiment of this application. Figure 1 As shown, the method includes the following steps:

[0048] Step S101: In the two-dimensional inverse flow problem design stage of the heavy-duty gas turbine compressor, the flow loss and lag angle are calculated by setting multiple blade geometry parameters and a pre-built artificial neural network surrogate model to achieve the coupled design of forward and inverse problems.

[0049] It should be noted that, in order to facilitate understanding of the implementation principle of the compressor design method of this application, the compressor design process in the relevant embodiments will be described below first. The compressor design process in the relevant embodiments is as follows: Figure 2 As shown, the initial design involves inputting design parameters such as compressor flow rate, pressure ratio, and efficiency. Then, a one-dimensional centerline design of the compressor is performed based on these parameters, followed by technical feasibility studies. After completing the technical feasibility studies and one-dimensional centerline design, a two-dimensional inverse flow problem is designed. During this design process, the losses of each stage of the blades are manually assigned based on the engineer's experience. The airfoil design is then performed using parameters such as the inlet and outlet airflow angles and Mach numbers of each row of blades calculated from the flow design as boundary conditions. After completing the S1 airfoil design, S1 cascade flow analysis is performed, followed by three-dimensional stacking and design of the airfoil. In this application, S1 and S2 both refer to two-dimensional flow surface calculation and analysis, differing only in the direction of the flow surface cut. S1 involves cutting the blade in half and analyzing the deflection effect of the cut surface airfoil on the airflow direction. S2 involves cutting the compressor into a half-section and analyzing the changes in flow parameters from upstream to downstream and the parameter distribution at different blade heights.

[0050] Subsequently, after completing the airfoil design, a three-dimensional CFD analysis is performed to determine whether the matching parameters of each stage of the compressor meet the expected flow path design. If not, the airfoil design parameters need to be readjusted until the matching parameters meet the design expectations. This airfoil design matching adjustment iteration typically requires more than ten rounds, which is quite time-consuming. The expected flow path design of the compressor mainly involves two aspects: the stator outlet airflow angle (also known as the pre-swirl angle) and the compressor stage pressure ratio. Determining whether the matching parameters of each stage of the compressor meet the expected flow path design means judging whether the pre-swirl angle and pressure ratio distribution of each stage calculated by CFD are consistent with the flow path design results.

[0051] Furthermore, after the matching design is completed, it is necessary to check whether the compressor design point performance meets the design specifications. If not, the compressor flow profile and stage load distribution need to be readjusted, and the two-dimensional inverse flow problem design calculation needs to be restarted. This iteration generally takes about four to five times. After completing the aerodynamic scheme where the design point performance meets the design specifications, the preliminary design of blade strength vibration begins, analyzing the resonant frequencies of each order of each row of blades. Usually, it is necessary to return to the airfoil design stage to adjust the blade thickness and chord length to avoid the resonance range. This iteration is called blade frequency tuning, and it usually takes about four to five rounds. After the preliminary design iteration of blade strength vibration is completed, the detailed design of blade strength vibration (e.g., blade root design) and the control law design of VGVs and anti-surge relief valve under varying compressor operating conditions can be carried out simultaneously (this design process is based on the results of S2 forward flow problem analysis calculations). After completing the detailed design of strength vibration and the control law design, the experimental verification stage can begin. When the control law is not reasonably formulated, multiple rounds of experimental adjustments may be required to successfully find a suitable compressor operating path.

[0052] The design method of this application is... Figure 2 Based on the previous work, several improvements have been made, making full use of the positive problem analysis file generated by the positive and negative problem coupling design pattern during the inverse problem design process.

[0053] Specifically, in combination Figure 3 The design flow shown illustrates the compressor design method of this application. In the two-dimensional flow design of the compressor, this application adopts a coupled positive and negative problem mode, eliminating the reliance on engineer experience to determine the losses of each stage of the compressor blades. Instead, multiple airfoil geometric parameters, such as the chord length, number of blades, and relative thickness of each stage, are pre-defined at the initial design stage. Then, a pre-built artificial neural network surrogate model automatically calculates the losses and lag angles of each stage of the blades.

[0054] It should be noted that the inverse problem in this application refers to designing compressor blade geometry parameters to achieve existing design specifications, such as flow rate and pressure ratio. The forward problem refers to analyzing the achievable performance of existing compressor blade geometry parameters, including performance in terms of flow rate, pressure ratio, and surge margin. The coupled forward and inverse problem design mode refers to considering the influence of geometric parameters in advance during the inverse problem design process, and obtaining the loss lag angle through pre-set geometric parameters and related models.

[0055] One possible approach is to implement coupled forward and reverse problem design during the S2 flow inverse problem design calculation. This involves providing the radial distribution of each stage's pressure ratio and pre-swirl angle, along with parameters such as chord length, consistency, relative blade thickness, tip clearance, and blockage coefficient for each row of blades. A loss model is then used to automatically calculate the compressor's moving blade efficiency and the total pressure recovery coefficient of the stationary blades. Simultaneously, the inlet and outlet geometric angles of each stage of the blades are automatically adjusted iteratively until the radial distribution of the compressor's pressure ratio and pre-swirl angle meets the design expectations. At this point, while obtaining aerodynamic layout parameters such as the inlet and outlet airflow angles and Mach numbers for each row of blades, the airfoil geometric parameters for achieving this aerodynamic layout are also derived—i.e., the flow forward problem analysis file. Therefore, the forward problem analysis file obtained simultaneously with the design facilitates subsequent compressor characteristic curve calculation and analysis, and allows for a preliminary assessment of the aerodynamic layout's operational margin using the compressor stall factor PGK.

[0056] Among them, by Figure 3 As can be seen, this application automatically calculates the loss and lag angle using an artificial neural network surrogate model, replacing the traditional loss and lag angle model. Traditional loss and lag angle models are derived from summarizing and generalizing data from traditional airfoil types such as double-circular arcs. However, since the actual design uses multi-circular arc airfoils (the mid-curve and thickness distribution are constructed using three-segment circular arcs), which differ significantly from traditional airfoils, calculating airfoil loss and lag angle using traditional empirical models is unreliable. Artificial neural networks, on the other hand, possess nonlinear modeling and self-learning capabilities, making them suitable for establishing complex relationships between airfoil loss and lag angle and airfoil geometric and flow parameters. The advantage of the artificial neural network surrogate model in this application compared to the traditional loss and lag angle model is that it can be trained using high-precision data (such as CFD calculation results or experimental results), achieving improved calculation accuracy without increasing computational time.

[0057] In one embodiment of this application, pre-constructing an artificial neural network agent model includes the following steps:

[0058] The first step is to use various blade geometry parameters as input parameters and select multiple target parameters from the input parameters based on the compressor operating conditions.

[0059] For example, various geometric parameters used in blade design, such as inlet geometry angle, bend angle, installation angle, consistency, maximum relative thickness, maximum thickness relative position, angle of attack, inlet Mach number, and axial tightness ratio, are used as model input parameters. Then, the range of values ​​for the above input parameters is evaluated within the compressor operating conditions, and several parameter combinations (i.e., multiple target parameters) are selected from the above input parameters.

[0060] The second step is to generate multiple two-dimensional airfoils based on multiple target parameters, and to obtain the flow loss and lag angle of the multiple two-dimensional airfoils under corresponding environmental conditions through numerical calculation.

[0061] For example, multiple two-dimensional airfoils are generated using a multi-circle arc modeling method based on the selected airfoil geometry parameters. Then, numerical calculations are used to obtain the losses and lag angles of these multiple two-dimensional airfoils under the corresponding inlet Mach number, angle of attack, and axial tightness ratio.

[0062] Among them, Mach number, angle of attack, and axial tightness ratio are the environmental conditions for blade operation, which can be regarded as the input conditions for calculating losses and lag angle in this embodiment.

[0063] In this embodiment, numerical calculation can be performed using a partitioning method, that is, solving the inviscid Euler equation in the mainstream region and solving the boundary layer integral equation in the viscous region near the wall. This method has the advantages of fast convergence speed and high calculation accuracy, and can quickly obtain the loss and lag angle corresponding to multiple combinations of input parameters.

[0064] The third step involves using multiple target parameters and their corresponding flow loss and lag angle to establish a surrogate model through an artificial neural network to correlate the flow loss and lag angle with the multiple target parameters.

[0065] Specifically, by utilizing the nonlinear modeling and self-learning capabilities of artificial neural networks, the loss and lag angle corresponding to different combinations of input parameters obtained in the previous step are used as training data to establish a surrogate model for relating the leaf shape loss and lag angle to the input parameters, so as to facilitate flow design and analysis.

[0066] Step S102: Based on the positive problem analysis file generated by the positive and negative problem coupling design, perform two-dimensional flow positive problem analysis. The two-dimensional flow positive problem analysis includes determining whether the compressor operating margin meets the requirements by calculating the stall factor.

[0067] Specifically, as mentioned above, after completing the two-dimensional inverse flow problem design using a coupled forward and inverse problem model, this application simultaneously generates compressor blade parameters and a forward problem analysis file that meet the design matching conditions. Therefore, the forward problem analysis file can be used directly for forward problem analysis. The forward problem analysis file is a calculation input file containing various compressor blade geometric parameter information.

[0068] In one embodiment of this application, determining whether the compressor operating margin meets the requirements by calculating the stall factor includes: determining the current blade grid limit D factor based on the correlation curve between the blade grid limit D factor and the airfoil bend angle; calculating the value of the stall factor based on the blade grid diffusion factor and the current blade grid limit D factor; and evaluating the compressor operating margin based on the value of the stall factor.

[0069] Specifically, this embodiment uses a stall factor PGK A preliminary assessment is made to determine whether the compressor operating margin meets the requirements. This includes the stall factor. PGKThe calculation formula is as follows:

[0070] PGK = DF / DF SL

[0071] in, DF It is the cascade diffusion factor (D factor for short). DF SL It is the leaf gate-limited D factor. DF SL The value is related to the average bend angle of the blade profile, and the specific correlation curve is as follows: Figure 4 As shown.

[0072] In this embodiment, the leaf angle from the positive problem analysis file can be substituted. Figure 4 The curve shown provides the cascade confinement D-factor under the current design conditions. Then, the determined cascade diffusion factor and cascade confinement D-factor are substituted into the above formula to calculate the stall factor. PGK The value. Furthermore, through the stall factor... PGK When judging the operating margin, the larger the blade stall factor, the higher the load on the blade. When the stall factor reaches 1, the compressor is considered to have reached the near stall point.

[0073] Therefore, this application considers the influence of geometric factors from the initial aerodynamic layout design in the compressor design and makes a preliminary assessment of the compressor margin. After completing the first round of airfoil design and three-dimensional computational fluid dynamics (CFD) analysis, the compressor's efficiency and operating margin parameters are closer to the design specifications. As a result, the number of iterations for adjusting the compressor flow channel profile and stage load distribution will be relatively small, usually no more than once.

[0074] Step S103: After completing the two-dimensional flow forward problem analysis, airfoil design, and three-dimensional CFD analysis, the airfoil geometry parameter adjustment amount corresponding to the deviation between the matching parameters of each stage of the compressor and the expected values ​​of the flow design is predicted through flow analysis.

[0075] Specifically, after completing step S102 and as follows Figure 3 As shown, after completing the blade profile design, 3D design, S1 cascade flow analysis, and 3D CFD analysis according to the design method in the relevant embodiments, matching design is performed, that is, determining whether the matching parameters of each stage of the compressor meet the expected flow design. This application uses flow analysis examples to quickly evaluate the matching deviation and the adjustment amount of the blade profile geometry parameters, thereby reducing the number of 3D CFD iteration calculations.

[0076] It should be noted that the matching design method in the relevant embodiments mainly involves continuously comparing and adjusting the airfoil geometry parameters through CFD calculations. For example, conventional inter-stage matching adjustment uses iterative methods for airfoil design and three-dimensional CFD analysis of each stage of the compressor. Due to the coupling between each stage of the compressor, the adjustment is difficult and requires many iterations. In some other embodiments, a stage-by-stage matching method is also used. Specifically, the inlet stage is designed first to ensure that the inlet stage flow-pressure ratio meets the design requirements; based on the inlet stage, a second stage is superimposed to ensure that the first two stages meet the flow design requirements; then, matching is performed stage by stage. As the number of superimposed stages increases, the computation time increases, and the total number of iterations and computation time are basically equivalent to the conventional inter-stage matching adjustment method described above.

[0077] This application focuses on compressor matching adjustment based on deviation prediction using flow analysis tools. In one embodiment of this application, the adjustment amount of the airfoil geometry parameters corresponding to the deviation is predicted through flow analysis, including: analyzing the airfoil geometry deviation corresponding to the deviation using a forward problem analysis file; using flow analysis tools to predict the adjustment direction and amount of the airfoil geometry parameters corresponding to the airfoil geometry deviation; and based on the adjustment direction and amount, iteratively performing airfoil adjustment design, three-dimensional CFD analysis, and deviation analysis until the matching parameters at each stage meet the requirements.

[0078] Specifically, in this embodiment, after discovering a deviation between the matching results of the 3D CFD calculation and the expected flow design, instead of directly adjusting the airfoil geometry based on experience, it analyzes the airfoil geometry deviation corresponding to the current matching deviation using the forward problem analysis file generated during the flow inverse problem design calculation process. Then, flow analysis tools are used to quickly predict the direction and amount of airfoil parameter adjustment, transforming the inter-stage matching design iteration from iteration between airfoil design and time-consuming 3D CFD analysis into an iteration primarily focused on flow. Therefore, this embodiment can reduce the number of 3D CFD analyses, accelerate the inter-stage matching design speed, and typically complete the matching design iteration in one round.

[0079] As one possible approach, when using flow analysis tools to quickly predict the direction and amount of airfoil parameter adjustment, a small geometric adjustment amount is first set, and then a large number of trial calculations are performed using flow analysis tools. The adjustment amount is changed according to the calculation results. Taking advantage of the fast calculation characteristics of flow analysis tools (which can usually output a calculation result in a few seconds), the adjustment amounts of all geometric parameters are calculated.

[0080] The specific matching design process in this embodiment is as follows: Figure 5As shown, after the airfoil design and 3D CFD calculation analysis are completed, a deviation analysis of inter-stage matching and design requirements is performed, including comparing the deviation between the actual pressure ratio distribution of each stage of the compressor and the design pressure ratio distribution of the flow path. Then, inter-stage matching design is iteratively performed. Using the S2 flow path forward problem analysis example, the airfoil geometry parameters in the flow path example are adjusted to ensure that the compressor stage matching results obtained from the flow path forward problem analysis are consistent with the 3D CFD calculation results. It should be noted that the change in flow path airfoil parameters before and after adjustment is the amount of adjustment required for the airfoil geometry parameters in the actual airfoil design process. Thus, the airfoil geometry adjustment amount is predicted through flow path analysis. After obtaining the airfoil geometry adjustment amount, the matching design deviation can be eliminated through one round of airfoil design iteration.

[0081] Step S104: In response to the completion of aerodynamic design and blade frequency modulation, the VGVs control law of the multi-row adjustable stator blades of the compressor under variable load conditions is optimized using forward problem analysis file and particle swarm optimization algorithm, and the compressor design scheme containing the designed VGVs control law is output.

[0082] Specifically, in completing the matching design in step S103, and as... Figure 3 After completing the design point performance design and blade frequency tuning, this application further utilizes the S2 forward flow problem analysis example and particle swarm optimization algorithm to optimize the VGVs control law of the compressor under variable load conditions. The specific implementation methods for subsequent aerodynamic design and blade frequency tuning can be found in the relevant embodiments and will not be elaborated here.

[0083] It should be noted that when a gas turbine operates under varying load conditions, the compressor inlet airflow needs to be reduced, thereby reducing the turbine load. This reduction is achieved by closing the angles of the compressor inlet adjustable guide vanes (IGV) and each row of adjustable stationary vanes (VGV), thus decreasing the compressor inlet flow area. In this case, it is necessary to determine the control law of VGVs (the aforementioned inlet adjustable guide vanes and each row of adjustable stationary vanes are collectively referred to as VGVs) under different loads. Although multi-row adjustable stationary vane adjustment technology for compressors has been applied in various models, there is a lack of efficient methods for determining the optimal adjustment law for each specific operating condition during the design phase. Extensive commissioning and testing are usually required to determine this optimal law.

[0084] Since different inlet flow rates correspond to different gas turbine loads, and the compressor inlet flow rate depends on IGV angle control, different load operating conditions can be represented by selecting different IGV angles. Based on this, this embodiment uses the closing angle of each row of VGVs as input variables and optimizes the VGVs control law under different IGV angles using particle swarm optimization.

[0085] The fundamental idea behind particle swarm optimization (PSO) is information sharing among the population. Each solution to the optimization problem is reduced to the spatial state of a single "particle," and all particles "fly" and search in a multi-dimensional space. Each particle has a fitness value determined by a fitness calculation function to judge the quality of its current position. Simultaneously, each particle is endowed with a memory function, capable of remembering the optimal position it has searched for. Each particle has a velocity that determines its flight distance and direction, dynamically adjusted based on its own flight experience and that of its companions.

[0086] To more clearly illustrate the specific implementation process of the optimized design of the VGVs control law of the compressor under variable load conditions in this application, the following description uses a design method proposed in one embodiment of this application as an example. Figure 6 This is a flowchart of an optimization design method for VGVs control law proposed in an embodiment of this application, as shown below. Figure 6 As shown, the method includes the following steps:

[0087] Step S601: Initialize the particle swarm using the closing angle of each row of adjustable stator vanes as the input variable and the isentropic efficiency at the common operating point of the compressor as the optimization objective.

[0088] The different closing angles of the inlet guide vanes are used to reflect different load conditions.

[0089] Specifically, this embodiment uses the closing angle of each row of VGVs as the input variable and the isentropic efficiency at the compressor's common operating point as the optimization objective. The particle swarm optimization method is used to optimize the VGVs control law under different IGV angles. During optimization, the particle swarm optimization is first initialized based on the aforementioned determined input variables and optimization objectives.

[0090] As an example, based on the operating conditions targeted by the current control law design, an IGV angle is set, and the initial value of the angles of the other rows of adjustable stator blades is set to zero degrees as the optimization starting point. Then, the dimensional space is determined according to the number of rows of adjustable stator blades in the VGVs. For example, if there are D rows of adjustable stator blades, the optimization is carried out in D-dimensional space, and the number of particles can be 20 to 40.

[0091] Step S602: Calculate the compressor characteristic curve using the positive problem analysis file, and calculate the isentropic efficiency of the compressor's common operating point based on the compressor characteristic curve.

[0092] Specifically, to more clearly illustrate the specific implementation flow of the optimized design method in the embodiments of this application, the following is combined with... Figure 7The specific process is explained below. After initializing the particle swarm, the initial angle of each particle is randomly generated within a limited angle range during the first optimization process. Then, the compressor characteristic curve is calculated using a flow path program, and the compressor common operating point is calculated based on the compressor characteristic curve.

[0093] In other words, the compressor characteristic curve in this embodiment is calculated using the S2 forward problem analysis example. Under different VGVs control laws, the characteristic curves of the compressor flow rate-pressure ratio and flow efficiency under the control law are calculated using the aforementioned forward problem analysis file. Then, the operating point is obtained using the compressor common operating point prediction method proposed in this embodiment, and the compressor isentropic efficiency at the operating point is calculated.

[0094] One possible implementation is to calculate the isentropic efficiency of the compressor common operating point based on the compressor characteristic line, including: calculating the intake flow rate of the compressor common operating point based on the compressor characteristic line; calculating the pressure ratio of the compressor common operating point based on the intake flow rate of the compressor common operating point; and obtaining the isentropic efficiency of the compressor common operating point by interpolation of the compressor characteristic line using the intake flow rate and pressure ratio of the compressor common operating point.

[0095] Specifically, in actual calculations, the average flow rate near the stall point and near the blockage point of the compressor characteristic curve can be used as the inlet flow rate at the compressor's common operating point. m match Then, the compressor's common operating point is calculated using the following formula:

[0096]

[0097] in, P match It is the common operating point pressure ratio of the compressor. P des It is the compressor design point pressure ratio. m des It is the compressor design point intake flow rate. P des and m des This can be obtained from the design calculations of the above-mentioned inverse flow problem.

[0098] Then, after calculating the inlet flow rate and pressure ratio at the common operating point of the compressor, the isentropic efficiency of the compressor at that point can be obtained by interpolation of the compressor characteristic curve.

[0099] Step S603: Determine the isentropic efficiency of each particle based on the isentropic efficiency of the compressor's common operating point, and correct the isentropic efficiency of each particle using the compressor inlet flow rate to obtain the fitness of each particle.

[0100] Specifically, the isentropic efficiency of each particle is first determined based on the isentropic efficiency of the compressor's common operating point. Since the current isentropic efficiency is obtained through calculation, the isentropic efficiency of each particle needs to be corrected by the inlet (intake) flow rate and then used as the fitness of each calculation point (particle).

[0101] As one possible implementation, the isentropic efficiency of each particle is corrected using the following formula:

[0102]

[0103] in, It is the isentropic efficiency after correction for any particle. It is the isentropic efficiency of any particle determined based on the isentropic efficiency at the common operating point of the compressor. It is the intake airflow at the common operating point of the compressor. It optimizes the intake flow rate at the common operating point of the starting point. It is the intake airflow at the compressor design point. It is the isentropic efficiency at the compressor design point. It optimizes the isentropic efficiency of the common working point at the starting point.

[0104] The optimization starting point is when the VGVs angle of each row is zero degrees. As shown in the formula above, this application corrects the isentropic efficiency obtained by direct interpolation of the characteristic curve by using the intake flow rates at the common operating point and design point.

[0105] Step S604: Based on the obtained fitness, perform automatic particle flight optimization until the VGVs control law with the highest isentropic efficiency is obtained.

[0106] Specifically, in this embodiment, the isentropic efficiency calculation result is corrected by the compressor inlet flow rate. The corrected isentropic efficiency for each particle is used as the fitness of each calculation point (particle) in the particle swarm optimization algorithm. Then, the position and velocity of the corresponding particle are updated based on the fitness, and the algorithm is further optimized according to... Figure 7 The method shown is used for iterative looping, and the compressor variable load control law with the highest isentropic efficiency is obtained by automatically searching for the optimal particle flight path.

[0107] As an example, this application also compares the highest isentropic efficiency of compressors under different load conditions under different control laws. Figure 8 It can be seen that the efficiency is lowest when VGVs control is not used (i.e., single IGV control), and the efficiency is improved when the linear control law in related technologies is used. However, the compressor efficiency is the highest when the VGVs control law optimized in this application is used.

[0108] Furthermore, such as Figure 3As shown, the output of the completed compressor design scheme (including the various design results mentioned above) can be used for subsequent experimental verification, which can reduce the number of rounds of experimental verification.

[0109] In summary, the heavy-duty gas turbine compressor design method applicable to multiple operating conditions in this application adopts a coupled positive and negative problem design mode. In the two-dimensional flow design stage, compressor margin can be quickly assessed through flow forward problem analysis, and compressor efficiency can be quickly assessed through a multi-circular-arc blade loss and lag angle surrogate model based on artificial neural networks. This allows for a relatively reasonable aerodynamic layout scheme to be provided in the first round of design, thereby significantly reducing the number of iterative designs and shortening the compressor aerodynamic layout iteration cycle. Furthermore, this method adjusts compressor matching based on flow tool prediction deviations and uses flow analysis tools to quickly predict the direction of blade parameter adjustment. It transforms the inter-stage matching design iteration from iteration between blade design and time-consuming three-dimensional CFD analysis into flow-driven iteration, thereby reducing the number of three-dimensional CFD analyses, greatly accelerating the compressor inter-stage matching design speed, and significantly shortening the compressor matching design iteration cycle. Moreover, this method optimizes the VGVs control law, effectively predicting the optimal VGVs control law of the gas turbine compressor under variable load operating conditions, providing theoretical reference values ​​for actual unit operation and commissioning. This helps reduce reliance on unit operating experience and the risk of human error during gas turbine operation and commissioning, and improves the operating efficiency and reliability of the compressor under non-design conditions.

[0110] To achieve the above embodiments, this application also proposes a heavy-duty gas turbine compressor design system applicable to multiple operating conditions. Figure 9 This is a schematic diagram of a heavy-duty gas turbine compressor design system applicable to multiple operating conditions, as proposed in an embodiment of this application. Figure 9 As shown, the system includes:

[0111] The calculation module 100 is used to calculate the flow loss and lag angle in the two-dimensional inverse flow problem design stage of heavy-duty gas turbine compressor by setting multiple blade geometry parameters and a pre-built artificial neural network surrogate model, so as to realize the coupled design of forward and inverse problems.

[0112] Analysis module 200 is used to perform two-dimensional flow forward problem analysis based on the forward problem analysis file generated by the forward and reverse problem coupling design. The two-dimensional flow forward problem analysis includes determining whether the compressor operating margin meets the requirements by calculating the stall factor.

[0113] The prediction module 300 is used to predict the adjustment amount of the blade geometry parameters corresponding to the deviation between the matching parameters of each stage of the compressor and the expected values ​​of the flow design after completing the two-dimensional flow forward problem analysis, blade design and three-dimensional CFD analysis.

[0114] The optimization module 400 is used to optimize the VGVs control law of the compressor with multiple rows of adjustable stator blades under variable load conditions in response to the completion of aerodynamic design and blade frequency modulation. It uses forward problem analysis files and particle swarm optimization algorithm to optimize the design of the compressor VGVs control law with multiple rows of adjustable stator blades under variable load conditions, and outputs a compressor design scheme containing the completed VGVs control law.

[0115] Optionally, in one embodiment of this application, the calculation module 100 is specifically used for: taking multiple airfoil geometric parameters as input parameters, selecting multiple target parameters from the input parameters based on the compressor operating conditions; generating multiple two-dimensional airfoils based on the multiple target parameters, and obtaining the flow loss and lag angle of the multiple two-dimensional airfoils under corresponding environmental conditions through numerical calculation; and using the multiple target parameters and the corresponding flow loss and lag angle, establishing a surrogate model through an artificial neural network to associate the flow loss and lag angle with the relationship between the multiple target parameters and the multiple target parameters.

[0116] It should be noted that the foregoing explanation of the embodiment of the design method for heavy-duty gas turbine compressor applicable to multiple operating conditions also applies to the system of this embodiment, and will not be repeated here.

[0117] In summary, the heavy-duty gas turbine compressor design system applicable to multiple operating conditions in this application embodiment makes full use of the forward problem analysis file generated under the forward and reverse problem coupling mode, which can reduce the number of iterative design times and improve aerodynamic design efficiency.

[0118] To implement the above embodiments, this application also proposes an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to execute the heavy-duty gas turbine compressor design method applicable to multiple operating conditions as described in any of the first aspects above.

[0119] To implement the above embodiments, this application also proposes a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the heavy-duty gas turbine compressor design method applicable to multiple operating conditions as described in any one of the first aspect embodiments above.

[0120] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0121] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0122] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0123] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0124] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0125] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0126] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0127] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. A method for designing a heavy duty gas turbine compressor suitable for multiple operating conditions, characterized in that, The method comprises the following steps: In the two-dimensional through-flow inverse problem design stage of the heavy-duty gas turbine compressor, through setting multiple blade profile geometric parameters and a pre-constructed artificial neural network proxy model, through-flow loss and lag angle are calculated to realize the coupling design of the forward problem and the inverse problem; Based on the forward problem analysis file generated by the coupling design of the forward problem and the inverse problem, two-dimensional through-flow forward problem analysis is performed, wherein the two-dimensional through-flow forward problem analysis comprises judging whether the compressor operating margin meets the requirements by calculating a stall factor; After completing the two-dimensional through-flow forward problem analysis, blade profile design and three-dimensional CFD analysis, for the deviation between the matching parameters of each stage of the compressor and the expected values of the through-flow design, the adjustment amount of the blade profile geometric parameters corresponding to the deviation is predicted through through-flow analysis; In response to completing the aerodynamic design and blade frequency adjustment, the VGVs control law of the multiple rows of adjustable stators of the compressor under variable load conditions is optimized and designed by using the forward problem analysis file and a particle swarm optimization algorithm, and a compressor design scheme containing the designed VGVs control law is output, wherein the optimization and design of the VGVs control law of the multiple rows of adjustable stators of the compressor under variable load conditions comprises: taking the closing angle of each row of adjustable stators as an input variable and taking the isentropic efficiency of the common operating point of the compressor as an optimization target, particle swarm initialization is performed, wherein different closing angles of the inlet guide vanes are used to reflect different load conditions; the characteristic line of the compressor is calculated through the forward problem analysis file, and the isentropic efficiency of the common operating point of the compressor is calculated based on the characteristic line of the compressor; the isentropic efficiency of each particle is determined according to the isentropic efficiency of the common operating point of the compressor, and the isentropic efficiency of each particle is corrected by the inlet flow rate of the compressor to obtain the fitness of each particle; the particle automatic flight optimization is performed based on the obtained fitness until the VGVs control law with the highest isentropic efficiency is obtained; the isentropic efficiency of the common operating point of the compressor is calculated based on the characteristic line of the compressor, comprising: calculating the inlet flow rate of the common operating point of the compressor based on the characteristic line of the compressor; calculating the pressure ratio of the common operating point of the compressor based on the inlet flow rate of the common operating point of the compressor; obtaining the isentropic efficiency of the common operating point of the compressor through the characteristic line interpolation by using the inlet flow rate and the pressure ratio of the common operating point of the compressor; wherein the isentropic efficiency of each particle is corrected by the following formula: wherein, is the isentropic efficiency of any particle corrected for the particle, is the isentropic efficiency of any particle determined from the isentropic efficiency of the compressor common operating point, is the mass flow rate of the compressor common operating point, is the mass flow rate of the common operating point of the optimization starting point, is the mass flow rate of the compressor design point, is the isentropic efficiency of the compressor design point, is the isentropic efficiency of the common operating point of the optimization starting point.

2. The method of claim 1, wherein, The pre-constructed artificial neural network proxy model comprises: Taking multiple blade profile geometric parameters as input parameters, multiple target parameters are selected from the input parameters based on the operating conditions of the compressor; Based on the multiple target parameters, multiple two-dimensional blade profiles are generated, and through numerical calculation, the through-flow loss and lag angle of the multiple two-dimensional blade profiles under corresponding environmental conditions are obtained; By using the multiple target parameters and the corresponding through-flow loss and lag angle, an artificial neural network is used to establish a proxy model for correlating the relationship between the through-flow loss and lag angle and the multiple target parameters.

3. The method of claim 1, wherein, The judgment of whether the compressor operating margin meets the requirements by calculating the stall factor comprises: determine the current cascade limit D factor according to the correlation curve between the cascade limit D factor and the bending angle of the blade profile; calculate the value of the stall factor according to the cascade diffusion factor and the current cascade limit D factor, and evaluate the compressor operating margin according to the value of the stall factor.

4. The method of claim 1, wherein, The adjustment amount of the blade profile geometric parameter corresponding to the deviation is predicted through through-flow analysis, including: analyzing the blade profile geometric deviation corresponding to the deviation through the forward problem analysis file; using a through-flow analysis tool, predicting the adjustment direction and the adjustment amount of the blade profile geometric parameter corresponding to the blade profile geometric deviation; based on the adjustment direction and the adjustment amount, cyclically performing blade profile adjustment design, three-dimensional CFD analysis and deviation analysis until the matching parameters of each stage meet the requirements.

5. A heavy duty gas turbine compressor design system suitable for multiple operating conditions, characterized by, The method comprises the following modules: a calculation module, configured to calculate through-flow loss and lag angle through a plurality of blade profile geometric parameters set and a pre-constructed artificial neural network proxy model in a two-dimensional through-flow inverse problem design stage of a heavy-duty gas turbine compressor, so as to realize forward problem and inverse problem coupling design; an analysis module, configured to perform two-dimensional through-flow forward problem analysis based on a forward problem analysis file generated by the forward problem and inverse problem coupling design, wherein the two-dimensional through-flow forward problem analysis comprises judging whether the compressor operating margin meets the requirements by calculating a stall factor; a prediction module, configured to, after completing the two-dimensional through-flow forward problem analysis, blade profile design and three-dimensional CFD analysis, predict an adjustment amount of a blade profile geometric parameter corresponding to a deviation between matching parameters of each stage of the compressor and expected values of through-flow design through through-flow analysis. An optimization module is configured to, in response to completion of the aerodynamic design and the blade frequency tuning, optimize design of a control law of VGVs of multiple rows of adjustable stators of the compressor under variable load conditions by using the forward problem analysis file and a particle swarm optimization algorithm, and output a compressor design scheme including the designed VGV control law; wherein the optimization module is specifically configured to: take the closing angle of each row of adjustable stators as an input variable and take the isentropic efficiency at a common working point of the compressor as an optimization target, and perform particle swarm initialization, wherein different closing angles of the inlet guide vanes are used to reflect different load conditions; calculate a compressor characteristic line by using the forward problem analysis file, and calculate the isentropic efficiency at the common working point of the compressor based on the compressor characteristic line; determine the isentropic efficiency of each particle according to the isentropic efficiency at the common working point of the compressor, and correct the isentropic efficiency of each particle by using the inlet flow rate of the compressor to obtain the fitness of each particle; perform automatic flight optimization of the particles based on the obtained fitness until the VGV control law with the highest isentropic efficiency is obtained; the calculation of the isentropic efficiency at the common working point of the compressor based on the compressor characteristic line comprises: calculating the inlet flow rate at the common working point of the compressor based on the compressor characteristic line; calculating the pressure ratio at the common working point of the compressor based on the inlet flow rate at the common working point of the compressor; and obtaining the isentropic efficiency at the common working point of the compressor by interpolation of the compressor characteristic line by using the inlet flow rate and the pressure ratio at the common working point of the compressor; wherein the isentropic efficiency of each particle is corrected by the following formula: wherein, is the isentropic efficiency of any particle corrected for particle, is the isentropic efficiency of any particle determined from the isentropic efficiency of the compressor common operating point, is the mass flow rate of the compressor common operating point, is the mass flow rate of the common operating point of the optimization starting point, is the mass flow rate of the compressor design point, is the isentropic efficiency of the compressor design point, is the isentropic efficiency of the common operating point of the optimization starting point.

6. The system of claim 5, wherein, The calculation module is specifically configured to: select multiple target parameters from the input parameters based on the operating conditions of the compressor; generate multiple two-dimensional blade profiles based on the multiple target parameters, and obtain the throughflow loss and the lag angle of the multiple two-dimensional blade profiles under corresponding environmental conditions by numerical calculation; establish an agent model for associating the throughflow loss and the lag angle with the relationship between the multiple target parameters by using the multiple target parameters and the corresponding throughflow loss and lag angle.

7. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program, when executed by a processor, implements the design method for the compressor of the heavy-duty gas turbine suitable for multiple conditions according to any one of claims 1-4.

Images