Blade shape generation method, device and equipment of gas compressor and storage medium

Abstract

The invention provides a blade shape generation method of a gas compressor, which can be applied to the technical field of artificial intelligence. The blade shape generation method of the gas compressor comprises the steps that expected performance parameters of a blade by a user are input into a shape design model, so that the shape of the blade is designed, and a plurality of geometrical shapes are obtained; inputting the plurality of geometrical shapes into a performance evaluation model, and determining a performance evaluation value of each geometrical shape; screening the plurality of geometric shapes according to the performance evaluation value to obtain at least one target shape conforming to the expected performance parameter; wherein the shape design model and the performance evaluation model are obtained by training sample blade performance and flow field simulation data in a training set, and the flow field simulation data are obtained by analyzing geometric parameters of sample blades and performing numerical simulation. The invention further provides a blade shape generation device and equipment of the gas compressor and a storage medium.

Description

Technical Field

[0001] The present disclosure relates to the field of artificial intelligence technology, and more specifically, to a method, device, equipment, and storage medium for generating a blade shape of a compressor. Background Art

[0002] As one of the core components of an aircraft engine, the compressor / fan has an excellent aerodynamic design that is crucial to the performance of the entire engine. Aerodynamic design, especially the three-dimensional blade shape design, is of great significance to the overall design process of the engine. At present, when using large models to design compressor blades, there are problems such as long design cycles and high requirements on designer experience. Summary of the invention

[0003] In view of this, the present disclosure provides a method, apparatus, device, storage medium and program product for generating a blade shape of a compressor.

[0004] One aspect of the present disclosure provides a method for generating a blade shape for a compressor, comprising: inputting a user's expected performance parameters for a blade into a shape design model to design the shape of the blade to obtain a plurality of geometric shapes; inputting the plurality of geometric shapes into a performance evaluation model to determine a performance evaluation value for each of the geometric shapes; screening the plurality of geometric shapes according to the performance evaluation values ​​to obtain at least one target shape that meets the expected performance parameters; wherein the shape design model and the performance evaluation model are both trained using sample blade performance and flow field simulation data in a training set, and the flow field simulation data is obtained by numerically simulating the geometric parameters of the sample blades after analysis.

[0005] According to an embodiment of the present disclosure, the above-mentioned shape design model includes a diffusion model, a latent space representation and a decoder. The above-mentioned input of the user's expected performance parameters for the blade into the shape design model to design the shape of the above-mentioned blade to obtain multiple geometric shapes, including: inputting the above-mentioned expected performance parameters as constraints into the above-mentioned diffusion model, and outputting the latent space representation of the shape design data of the above-mentioned blade; inputting the latent space representation of the above-mentioned shape design data into the above-mentioned decoder to convert the above-mentioned shape design data into a three-dimensional geometric shape.

[0006] According to an embodiment of the present disclosure, the above-mentioned performance evaluation model includes an encoder and a prediction network. The above-mentioned multiple geometric shapes are respectively input into the performance evaluation model to determine the performance evaluation value of each of the above-mentioned geometric shapes, including: for each of the above-mentioned geometric shapes, the above-mentioned geometric shape is input into the encoder to convert the above-mentioned geometric shape into a latent space representation; the latent space representation obtained by the conversion is input into the above-mentioned prediction network, and the performance evaluation value of the above-mentioned geometric shape is output.

[0007] According to an embodiment of the present disclosure, the above-mentioned method also includes: inputting the above-mentioned target shape into a flow field prediction model for feature extraction, and outputting multiple blade surface parameters predicted for the above-mentioned target shape; based on the above-mentioned blade surface parameters, reconstructing the three-dimensional flow field of the above-mentioned target shape to obtain three-dimensional flow field information, wherein the above-mentioned flow field information includes temperature information and pressure information.

[0008] According to an embodiment of the present disclosure, the above method also includes: integrating the above shape design model, the above performance evaluation model and the above flow field prediction model to construct an integrated design platform, wherein the above integrated design platform is used to visualize the flow field of the three-dimensional flow field information that meets the above expected performance parameters based on the expected performance parameters input by the above user.

[0009] According to an embodiment of the present disclosure, the above-mentioned shape design model and the above-mentioned performance evaluation model are trained in the following manner: determining a sample space according to the acquired geometric parameters of the above-mentioned sample blades; performing Latin hypercube sampling on the geometric parameters in the above-mentioned sample space to obtain a plurality of sample data; performing batch grid processing and single-channel calculation on the above-mentioned sample data using a performance simulation method to obtain the above-mentioned flow field simulation data, wherein the above-mentioned flow field simulation data includes a plurality of sample shapes; performing format conversion on the above-mentioned sample shapes using a point cloud method to obtain shape training data; performing model training using the above-mentioned shape training data and the corresponding sample blade performance to obtain the above-mentioned shape design model and the above-mentioned performance evaluation model.

[0010] According to an embodiment of the present disclosure, the above-mentioned model training using the above-mentioned shape training data and the corresponding sample blade performance includes: performing noise elimination on the above-mentioned shape training data, so as to perform model training using the eliminated data after noise elimination and the corresponding sample blade performance; wherein, in the process of training the above-mentioned shape design model, the shape design experience of the blade and the related alignment strategies are introduced.

[0011] Another aspect of the present disclosure provides a blade shape generating device for a compressor, comprising: a shape design module, used to input the user's expected performance parameters for the blade into a shape design model, so as to design the shape of the above-mentioned blade and obtain a plurality of geometric shapes; a performance evaluation module, used to input the plurality of the above-mentioned geometric shapes into the performance evaluation model, and determine the performance evaluation value of each of the above-mentioned geometric shapes; a shape screening module, used to screen the plurality of the above-mentioned geometric shapes according to the above-mentioned performance evaluation value, and obtain at least one target shape that meets the above-mentioned expected performance parameters; wherein the above-mentioned shape design model and the above-mentioned performance evaluation model are both trained using sample blade performance and flow field simulation data in a training set, and the above-mentioned flow field simulation data is obtained by numerically simulating the geometric parameters of the above-mentioned sample blades after analysis.

[0012] Another aspect of the present disclosure provides an electronic device, including: one or more processors; and a memory for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method described above.

[0013] Another aspect of the present disclosure provides a computer-readable storage medium storing computer-executable instructions, which are used to implement the above method when executed.

[0014] Another aspect of the present disclosure provides a computer program product, the computer program product comprising computer executable instructions, and the instructions are used to implement the above method when executed.

[0015] According to the embodiments of the present disclosure, the shape design model is used to realize the automatic design of the blade, and the performance of the designed geometric shape is evaluated to quickly obtain the desired design result. In the process of model training, a training set including sample blade performance and flow field simulation data is established, so that the model has a higher learning efficiency for features. By designing the blade shape based on the user's expected performance parameters for the blade, and evaluating the performance of the geometric shape during the design process, the design efficiency of the blade is significantly improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The above and other objects, features and advantages of the present disclosure will become more apparent through the following description of the embodiments of the present disclosure with reference to the accompanying drawings, in which:

[0017] Figure 1 An exemplary system architecture to which the method and apparatus for generating blade shapes of a compressor disclosed in the present invention can be applied is schematically shown;

[0018] Figure 2 A flowchart of a method for generating a blade shape for a compressor according to an embodiment of the present disclosure is schematically shown;

[0019] Figure 3 A processing flow chart of a shape design model in a method for generating a blade shape for a compressor according to an embodiment of the present disclosure is schematically shown;

[0020] Figure 4 A processing flow chart of a performance evaluation model in a method for generating a blade shape for a compressor according to an embodiment of the present disclosure is schematically shown;

[0021] Figure 5 A schematic diagram schematically shows a three-dimensional flow field information prediction in a method for generating a blade shape for a compressor according to an embodiment of the present disclosure;

[0022] Figure 6A schematic diagram of an integrated design platform in a method for generating a blade shape for a compressor according to an embodiment of the present disclosure is schematically shown;

[0023] Figure 7 A model training flow chart in a method for generating a blade shape for a compressor according to an embodiment of the present disclosure is schematically shown;

[0024] Figure 8 A block diagram schematically shows a shape generating device according to an embodiment of the present disclosure; and

[0025] Fig. 9 A block diagram of an electronic device suitable for implementing a shape generation method according to an embodiment of the present disclosure is schematically shown. DETAILED DESCRIPTION

[0026] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present disclosure. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present disclosure. However, it is apparent that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present disclosure.

[0027] The terms used herein are only for describing specific embodiments and are not intended to limit the present disclosure. The terms "comprise", "include", etc. used herein indicate the existence of the features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0028] All terms (including technical and scientific terms) used herein have the meanings commonly understood by those skilled in the art unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0029] When using expressions such as "at least one of A, B, and C, etc.", they should generally be interpreted according to the meaning of the expression commonly understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.).

[0030] In the embodiments of the present disclosure, the collection, updating, analysis, processing, use, transmission, provision, disclosure, storage, etc. of the data involved (for example, including but not limited to user personal information) are in compliance with the provisions of relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. In particular, necessary measures are taken for user personal information to prevent illegal access to user personal information data and maintain the security of user personal information and network security.

[0031] In the embodiments of the present disclosure, the user's authorization or consent is obtained before obtaining or collecting the user's personal information.

[0032] As the thrust-to-weight ratio of aircraft engines continues to increase, the load of its core component, the compressor stage, is approaching its limit, leaving the maximum efficiency and optimal design range of existing compressor products, which poses certain challenges to the current compressor design system at typical load levels. The traditional one-dimensional, two-dimensional, and three-dimensional iterative blade design methods are gradually unable to meet the requirements of wide stability margins. At the same time, there is a lack of effective mapping between different blade shapes and specific performance parameters, which also increases the difficulty of design to a certain extent.

[0033] Therefore, improving the traditional design process and developing a fast and effective new model for 3D blade design is of great significance for the design of high-load compressors. At present, generative models represented by large models have swept various fields with their powerful generation capabilities. They learn to abstract the essential laws and probability distribution of data by training on large-scale data sets and generate new data.

[0034] The embodiments of the present disclosure provide a method, device, equipment, storage medium and program product for generating the blade shape of a compressor. The method for generating the blade shape of a compressor includes: inputting the user's expected performance parameters for the blade into a shape design model to design the shape of the blade and obtain multiple geometric shapes; inputting the multiple geometric shapes into a performance evaluation model to determine the performance evaluation value of each geometric shape; screening the multiple geometric shapes according to the performance evaluation value to obtain at least one target shape that meets the expected performance parameters; wherein the shape design model and the performance evaluation model are both trained using the sample blade performance and flow field simulation data in the training set, and the flow field simulation data is obtained by numerically simulating the geometric parameters of the sample blade after analyzing them.

[0035] Figure 1 The exemplary system architecture of the compressor blade shape generation method and device disclosed in the present invention is schematically shown. It should be noted that: Figure 1 What is shown is merely an example of a system architecture to which the embodiments of the present disclosure can be applied, in order to help those skilled in the art understand the technical content of the present disclosure, but it does not mean that the embodiments of the present disclosure cannot be used in other devices, systems, environments or scenarios.

[0036] like Figure 1 As shown, the system architecture 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 is used to provide a medium for a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired and / or wireless communication links, etc.

[0037] The user may use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 through the network 104 to receive or send messages, etc. Various communication client applications may be installed on the first terminal device 101, the second terminal device 102, and the third terminal device 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, and / or social platform software, etc. (only for example).

[0038] The first terminal device 101, the second terminal device 102, and the third terminal device 103 may be various electronic devices having display screens and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, and the like.

[0039] The server 105 may be a server that provides various services, such as a background management server (only as an example) that provides support for websites browsed by users using the first terminal device 101, the second terminal device 102, and the third terminal device 103. The background management server may analyze and process the received data such as user requests, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal device.

[0040] It should be noted that the compressor blade shape generation method provided in the embodiment of the present disclosure can generally be executed by the server 105. Accordingly, the compressor blade shape generation device provided in the embodiment of the present disclosure can generally be set in the server 105. The compressor blade shape generation method provided in the embodiment of the present disclosure can also be executed by a server or server cluster that is different from the server 105 and can communicate with the first terminal device 101, the second terminal device 102, the third terminal device 103 and / or the server 105. Correspondingly, the compressor blade shape generation device provided in the embodiment of the present disclosure can also be set in a server or server cluster that is different from the server 105 and can communicate with the first terminal device 101, the second terminal device 102, the third terminal device 103 and / or the server 105. Alternatively, the compressor blade shape generation method provided in the embodiment of the present disclosure can also be performed by the first terminal device 101, the second terminal device 102 or the third terminal device 103, or can also be performed by other terminal devices different from the first terminal device 101, the second terminal device 102 or the third terminal device 103. Correspondingly, the blade shape generating device of the compressor provided in the embodiment of the present disclosure can also be arranged in the first terminal device 101, the second terminal device 102 or the third terminal device 103, or in other terminal devices different from the first terminal device 101, the second terminal device 102 or the third terminal device 103.

[0041] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0042] Figure 2 A flowchart of a method for generating a blade shape for a compressor according to an embodiment of the present disclosure is schematically shown.

[0043] like Figure 2 As shown, the method includes operations S210 to S230.

[0044] In operation S210 , the user's expected performance parameters for the blade are input into a shape design model to design the shape of the blade to obtain a plurality of geometric shapes.

[0045] In operation S220 , a plurality of geometric shapes are input into a performance evaluation model, and a performance evaluation value of each geometric shape is determined.

[0046] In operation S230 , a plurality of geometric shapes are screened according to the performance evaluation values ​​to obtain at least one target shape that meets the desired performance parameters.

[0047] According to an embodiment of the present disclosure, the user's expected performance parameters for the blade, such as mass flow, isentropic efficiency, and total pressure ratio, are obtained. The total pressure ratio refers to the ratio of the total pressure at the blade outlet to the total pressure at the blade inlet, which is used to evaluate the performance and efficiency of the blade, and the blade is a blade for a high-load compressor. The expected performance parameters are input into the shape design model, and the shape of the blade is designed to obtain multiple geometric shapes. The geometric shape can be represented by a three-dimensional coordinate point (x, y, z).

[0048] According to an embodiment of the present disclosure, a geometric shape is input into a performance evaluation model, and a performance evaluation value of the geometric shape is predicted. The performance evaluation values ​​include flow rate, pressure ratio, and efficiency. By comparing the multiple performance evaluation values ​​output by the performance evaluation model with the expected performance parameters, a geometric shape greater than or equal to the expected performance parameter can be screened out as a target shape. If there is no geometric shape greater than or equal to the expected performance parameter among the multiple performance evaluation values, the shape design model can be used to perform multiple iterations to gradually improve the performance of the blade until a target shape that meets the expected performance parameters is obtained.

[0049] According to the embodiments of the present disclosure, the shape design model and the performance evaluation model are trained using the sample blade performance and flow field simulation data in the training set. Since a large amount of high-fidelity data is required during the training process, it is necessary to construct a training set for the model. Specifically, the construction process involves studying the geometric parameters of the sample blades of the high-load compressor to determine the boundary conditions, parameter variation range, and simulation method required to construct the training set.

[0050] According to the embodiment of the present disclosure, after analyzing the geometric parameters of the sample blade, the performance of the geometric parameters is simulated using a simulation method according to the parameter variation range in the analysis result to obtain flow field simulation data. The geometric parameters include the inlet metal angle, outlet metal angle, chord length, maximum thickness, maximum thickness position, curvature, sweep shape, etc. of three typical sections of the sample blade: the root, the middle and the tip.

[0051] According to the embodiments of the present disclosure, after the training set is constructed, the performance of the sample blade can be used as input for the training of the shape design model to learn the geometric data. On the contrary, the training of the performance evaluation model uses the blade geometric data as input to learn the performance of the sample blade.

[0052] According to the embodiments of the present disclosure, the shape design model is used to realize the automatic design of the blade, and the performance of the designed geometric shape is evaluated to quickly obtain the desired design result. In the process of model training, a training set including sample blade performance and flow field simulation data is established, so that the model has a higher learning efficiency for features. By designing the blade shape based on the user's expected performance parameters for the blade, and evaluating the performance of the geometric shape during the design process, the design efficiency of the blade is significantly improved.

[0053] Figure 3 A processing flow chart of a shape design model in a method for generating a blade shape for a compressor according to an embodiment of the present disclosure is schematically shown.

[0054] According to an embodiment of the present disclosure, the user's expected performance parameters for the blade are input into a shape design model to design the shape of the blade to obtain multiple geometric shapes, including: inputting the expected performance parameters as constraints into a diffusion model to output a latent space representation of the shape design data of the blade; and inputting the latent space representation of the shape design data into a decoder to convert the shape design data into a three-dimensional geometric shape.

[0055] According to an embodiment of the present disclosure, the desired performance parameter 310 is input as a constraint into the diffusion model 320. Since the diffusion model 320 fully learns the distribution characteristics of the data during the training process, the diffusion model 320 can be used to design the shape of the blade to obtain shape design data. Next, the shape design data is converted into a latent space representation 330, which is a compressed data form that contains the information required to generate the blade shape.

[0056] According to an embodiment of the present disclosure, the key features of the shape design data are captured in the latent space representation 330, and the key features are input into the decoder 340. The decoder 340 may be a 3D encoder, which is a neural network that can convert the latent space representation 330 into a blade shape in a three-dimensional space. Finally, the decoder 340 outputs a plurality of three-dimensional geometric shapes 350. By designing the geometric shape of the blade using the shape design model, multiple design schemes can be generated in a short time, effectively shortening the design cycle.

[0057] Figure 4 The flowchart schematically shows a processing flow of a performance evaluation model in a method for generating a blade shape for a compressor according to an embodiment of the present disclosure.

[0058] According to an embodiment of the present disclosure, multiple geometric shapes are respectively input into a performance evaluation model to determine a performance evaluation value of each geometric shape, including: for each geometric shape, inputting the geometric shape into an encoder to convert the geometric shape into a latent space representation; inputting the converted latent space representation into a prediction network, and outputting a performance evaluation value for the geometric shape.

[0059] According to an embodiment of the present disclosure, the geometric shape 410 is input into the encoder 420, wherein the encoder 420 may be a 3D encoder. The encoder 420 is responsible for encoding the geometric shape into a format that can be processed by the machine learning model, such as converting the geometric shape 410 into a feature vector. After encoding into a feature vector, the model may consider additional condition information, such as fluid parameters, which are referred to as "incoming flow conditions 430", which may include environmental factors, operating conditions, or other variables that affect performance.

[0060] According to an embodiment of the present disclosure, the encoded feature vector and the fluid condition 430 are input into the prediction network 440. The prediction network 440 is a machine learning model that learns how to predict performance parameters from the input features. The prediction network 440 outputs key indicators for evaluating the performance of the geometric shape, which may specifically be a performance evaluation value 450. By predicting the performance of the geometric shape during the shape design of the blade, instant feedback can be provided at each stage of the design process, thereby ensuring the optimization and accuracy of the design solution.

[0061] Figure 5 A schematic diagram of three-dimensional flow field information in a method for generating a blade shape for a compressor according to an embodiment of the present disclosure is schematically shown.

[0062] According to an embodiment of the present disclosure, the method for generating a blade shape for a compressor also includes: inputting a target shape into a flow field prediction model for feature extraction, and outputting a plurality of blade surface parameters predicted for the target shape; and reconstructing the three-dimensional flow field of the target shape according to the blade surface parameters to obtain three-dimensional flow field information, wherein the flow field information includes temperature information and pressure information.

[0063] According to the embodiments of the present disclosure, in order to quickly evaluate the target shape, a deep learning prediction technology suitable for the strong nonlinear and unsteady flow field inside the compressor can be constructed, and a study on the rapid reconstruction of the three-dimensional flow field of the compressor blade channel can be carried out based on the deep learning related algorithm. Specifically, the blade surface parameters are predicted by using the flow field prediction model, where the blade surface parameters include physical parameters such as temperature, pressure, and Mach number. By predicting the blade surface parameters, sufficient feedback can be given to the designer.

[0064] According to the embodiments of the present disclosure, the flow field prediction model is trained based on the three-dimensional flow field of the blade channel of the sample blade and the performance of the sample blade. After obtaining the blade surface parameters, numerical simulation can be performed on it to simulate the behavior of the fluid in the blade channel to obtain fluid information. This includes simulating the flow, pressure distribution and temperature change of the fluid.

[0065] According to an embodiment of the present disclosure, after obtaining the fluid information and the flow field information when the blades rotate, the flow field is reconstructed using a deep learning model. A deep learning algorithm is used to model the strong nonlinear and unsteady flow field in the compressor blade channel to ensure accurate capture and reconstruction of complex flow field characteristics. This model can accurately predict the static temperature, static pressure and Mach number at different positions of the blade, and the relative error does not exceed 3%, thereby achieving rapid and accurate reconstruction of the compressor stator flow field. During the reconstruction process, physical knowledge, such as the basic principles of fluid dynamics, can be embedded to improve the accuracy and efficiency of the reconstruction.

[0066] According to the embodiments of the present disclosure, after the reconstruction is completed, the distribution and characteristics of the flow field are intuitively displayed through flow field visualization technology, such as three-dimensional flow field information. Figure 5 As shown, the three-flow flow field animation is displayed in multiple dimensions, including a side view 510 of a compressor blade, a topology optimization concept diagram 520 of a compressor blade, a three-dimensional model 530 of a compressor rotor, and a cross-sectional view 540 of a compressor. By reconstructing the three-dimensional flow field in the blade channel, important information is provided for blade design and performance analysis.

[0067] Figure 6 A schematic diagram of an integrated design platform in a method for generating a blade shape for a compressor according to an embodiment of the present disclosure is schematically shown.

[0068] According to an embodiment of the present disclosure, the method for generating a blade shape for a compressor also includes: integrating a shape design model, a performance evaluation model and a flow field prediction model to construct an integrated design platform, wherein the integrated design platform is used to visualize the three-dimensional flow field information that meets the expected performance parameters based on the expected performance parameters input by the user.

[0069] According to the embodiment of the present disclosure, the shape design model 610, the performance evaluation model 620 and the flow field prediction model 630 are integrated to realize the function that after the user inputs the expected performance parameters, the model generates multiple target shapes in a very short time. And by displaying the three-dimensional flow field information 640 on the integrated design platform 600, the design results are displayed using virtual display technology, which facilitates designers to interact and adjust the design results.

[0070] Figure 7The model training flow chart in the method for generating blade shape of a compressor according to an embodiment of the present disclosure is schematically shown.

[0071] like Figure 7 As shown, the method includes operations S710~S750.

[0072] In operation S710, a sample space is determined according to the acquired geometric parameters of the sample blade.

[0073] In operation S720, Latin hypercube sampling is performed on the geometric parameters in the sample space to obtain a plurality of sample data.

[0074] In operation S730, batch grid processing and single-channel calculation are performed on the sample data using a performance simulation method to obtain flow field simulation data, wherein the flow field simulation data includes a plurality of sample shapes.

[0075] In operation S740, the sample shape is converted into a format by using a point cloud method to obtain shape training data.

[0076] In operation S750, model training is performed using the shape training data and the corresponding sample blade performance to obtain a shape design model and a performance evaluation model.

[0077] According to an embodiment of the present disclosure, the geometric parameters of the acquired sample blades are analyzed to determine the boundary conditions and parameter variation range, such as: inlet total temperature and total pressure and outlet static pressure, etc. The geometric parameters are statistically analyzed to obtain a sample space, and the geometric parameters in the sample space are subjected to Latin hypercube sampling to obtain a plurality of sample data. Specifically, the parameter variation range can be equally divided into several layers, and the number of layers is usually equal to the required number of samples. A sample point is randomly selected in each layer of each variable to ensure that only one sample point is selected in each layer. The sample points of each variable are combined into a sample vector, so that sample data uniformly distributed in the multidimensional sample space is obtained. The above steps are repeated until the required number of sample data is obtained. In order to further improve the randomness of the samples, the order of the generated sample vectors can be disrupted.

[0078] According to the embodiments of the present disclosure, the main advantage of Latin hypercube sampling is that it can cover the sample space more efficiently, especially in the scenario of non-uniform distribution. Through the stratified random sampling method, it can ensure that there are enough samples in different areas of the sample space, avoiding the bias problem that may exist in random sampling.

[0079] According to the embodiment of the present disclosure, after the sample data is determined, the simulation method used in the simulation is verified to check the effectiveness of the simulation prevention. When it is determined that the simulation method is effective, the sample data is batch-grid processed and single-channel calculated to form the geometric shape of the blade under different operating conditions, thereby obtaining a flow field simulation data set.

[0080] According to an embodiment of the present disclosure, before model training, it is necessary to convert the format of the sample shape using a point cloud method. The sample shape is represented by a point cloud method, and modeling is performed using the point cloud method. Modeling using the point cloud method does not rely on a specific parameterization method, and can make the most of previous sample shapes. The training data converted using the point cloud method and the corresponding sample blade performance are used for model training to obtain a shape design model and a performance evaluation model. The sample shape is expressed using point cloud technology, and the model is trained using the coordinate points of the sample shape, so that the training data is not restricted to a single source, thereby increasing the utilization rate of existing data.

[0081] According to an embodiment of the present disclosure, model training is performed using shape training data and corresponding sample blade performance, including: performing noise elimination on the shape training data, so as to perform model training using the eliminated data after noise elimination and the corresponding sample blade performance; wherein, in the process of training the shape design model, the shape design experience of the blade and related alignment strategies are introduced.

[0082] According to the embodiments of the present disclosure, in the process of model training, a corresponding algorithm is introduced to eliminate noise signals in the shape training data to ensure that the generated shape training data is smooth enough. For the training of the shape design model, expert experience and alignment strategy are incorporated into the model training process to ensure that the network generates geometric shapes in a favorable direction. By incorporating expert experience and alignment strategy, it is ensured that the generated geometric shapes do not violate existing physical laws and can make full use of previous design experience.

[0083] Figure 8 The block diagram of a shape generating device according to an embodiment of the present disclosure is schematically shown.

[0084] like Figure 8 As shown, the compressor blade shape generating device 800 includes a shape design module 810 , a performance evaluation module 820 , and a shape screening module 830 .

[0085] The shape design module 810 is used to input the user's expected performance parameters for the blade into the shape design model to design the shape of the blade to obtain multiple geometric shapes.

[0086] The performance evaluation module 820 is used to input multiple geometric shapes into the performance evaluation model to determine the performance evaluation value of each geometric shape.

[0087] The shape screening module 830 is used to screen multiple geometric shapes according to the performance evaluation value to obtain at least one target shape that meets the expected performance parameters.

[0088] Among them, the shape design model and the performance evaluation model are both trained using the sample blade performance and flow field simulation data in the training set. The flow field simulation data is obtained by numerical simulation after analyzing the geometric parameters of the sample blades.

[0089] According to an embodiment of the present disclosure, the shape design module 810 includes a data determination submodule and a shape determination submodule.

[0090] A data determination submodule is used to input the desired performance parameters as constraints into the diffusion model and output a latent space representation of the blade shape design data;

[0091] The shape determination submodule is used to input the latent space representation of the shape design data into the decoder to convert the shape design data into a three-dimensional geometric shape.

[0092] According to an embodiment of the present disclosure, the performance evaluation module 820 includes a shape conversion submodule and an evaluation determination submodule.

[0093] The shape conversion submodule is used to input the geometric shape into the encoder for each geometric shape to convert the geometric shape into a latent space representation.

[0094] The evaluation and determination submodule is used to input the transformed latent space representation into the prediction network and output the performance evaluation value of the geometric shape.

[0095] According to an embodiment of the present disclosure, the compressor blade shape generating device 800 further includes a feature extraction module and a flow field reconstruction module.

[0096] The feature extraction module is used to input the target shape into the flow field prediction model for feature extraction and output a plurality of blade surface parameters predicted for the target shape.

[0097] The flow field reconstruction module is used to reconstruct the three-dimensional flow field of the target shape according to the blade surface parameters to obtain the three-dimensional flow field information, wherein the flow field information includes temperature information and pressure information.

[0098] According to an embodiment of the present disclosure, the blade shape generating device 800 of the compressor also includes a platform construction module, which is used to integrate the shape design model, the performance evaluation model and the flow field prediction model to construct an integrated design platform, wherein the integrated design platform is used to visualize the flow field of three-dimensional flow field information that meets the expected performance parameters according to the expected performance parameters input by the user.

[0099] According to an embodiment of the present disclosure, the compressor blade shape generating device 800 further includes a space determination module, a parameter sampling module, a data simulation module, a format conversion module and a model training module.

[0100] The space determination module is used to determine the sample space according to the acquired geometric parameters of the sample leaves.

[0101] The parameter sampling module is used to perform Latin hypercube sampling on the geometric parameters in the sample space to obtain multiple sample data.

[0102] The data simulation module is used to perform batch grid processing and single-channel calculation on the sample data using a performance simulation method to obtain flow field simulation data, wherein the flow field simulation data includes multiple sample shapes.

[0103] The format conversion module is used to convert the sample shape format using point clouding to obtain shape training data.

[0104] The model training module is used to perform model training using shape training data and corresponding sample blade performance to obtain a shape design model and a performance evaluation model.

[0105] According to an embodiment of the present disclosure, the model training module includes a noise elimination submodule and a strategy introduction submodule.

[0106] The noise elimination submodule is used to eliminate noise from the shape training data so as to use the eliminated data after noise elimination and the corresponding sample blade performance for model training.

[0107] The strategy introduction submodule is used to introduce the blade shape design experience and related alignment strategies during the training of the shape design model.

[0108] According to the embodiments of the present invention, any one or more of the modules, submodules, units, and subunits, or at least part of the functions of any one of them can be implemented in one module. According to the embodiments of the present invention, any one or more of the modules, submodules, units, and subunits can be split into multiple modules for implementation. According to the embodiments of the present invention, any one or more of the modules, submodules, units, and subunits can be at least partially implemented as hardware circuits, such as field programmable gate arrays (FPGAs), programmable logic arrays (PLAs), systems on chips, systems on substrates, systems on packages, application specific integrated circuits (ASICs), or can be implemented by hardware or firmware in any other reasonable way of integrating or packaging circuits, or can be implemented in any one of the three implementation methods of software, hardware, and firmware, or in any appropriate combination of any of them. Alternatively, according to the embodiments of the present invention, one or more of the modules, submodules, units, and subunits can be at least partially implemented as computer program modules, and when the computer program modules are run, the corresponding functions can be performed.

[0109] For example, any multiple of the shape design module 810, the performance evaluation module 820, and the shape screening module 830 can be combined in one module / unit / sub-unit for implementation, or any one of the modules / units / sub-units can be split into multiple modules / units / sub-units. Alternatively, at least part of the functions of one or more of these modules / units / sub-units can be combined with at least part of the functions of other modules / units / sub-units and implemented in one module / unit / sub-unit. According to an embodiment of the present disclosure, at least one of the shape design module 810, the performance evaluation module 820, and the shape screening module 830 can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application specific integrated circuit (ASIC), or can be implemented by hardware or firmware such as any other reasonable way of integrating or packaging the circuit, or implemented in any one of the three implementation methods of software, hardware, and firmware, or in an appropriate combination of any of them. Alternatively, at least one of the shape design module 810 , the performance evaluation module 820 , and the shape screening module 830 may be at least partially implemented as a computer program module, and when the computer program module is executed, the corresponding function may be performed.

[0110] It should be noted that the blade shape generating device part of the compressor in the embodiment of the present disclosure corresponds to the blade shape generating method part of the compressor in the embodiment of the present disclosure. The description of the blade shape generating device part of the compressor specifically refers to the blade shape generating method part of the compressor, which will not be repeated here.

[0111] Fig. 9A block diagram of an electronic device suitable for implementing a shape generation method according to an embodiment of the present disclosure is schematically shown. Fig. 9 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0112] like Fig. 9 As shown, the electronic device 900 according to an embodiment of the present disclosure includes a processor 901, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 902 or a program loaded from a storage part 908 into a random access memory (RAM) 903. The processor 901 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or a related chipset and / or a dedicated microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 901 may also include an onboard memory for caching purposes. The processor 901 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present disclosure.

[0113] In RAM 903, various programs and data required for the operation of electronic device 900 are stored. Processor 901, ROM 902 and RAM 903 are connected to each other via bus 904. Processor 901 performs various operations of the method flow according to the embodiment of the present disclosure by executing the program in ROM 902 and / or RAM 903. It should be noted that the program can also be stored in one or more memories other than ROM 902 and RAM 903. Processor 901 can also perform various operations of the method flow according to the embodiment of the present disclosure by executing the program stored in the one or more memories.

[0114] According to an embodiment of the present disclosure, the electronic device 900 may further include an input / output (I / O) interface 905, which is also connected to the bus 904. The electronic device 900 may further include one or more of the following components connected to the input / output (I / O) interface 905: an input portion 906 including a keyboard, a mouse, etc.; an output portion 907 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage portion 908 including a hard disk, etc.; and a communication portion 909 including a network interface card such as a LAN card, a modem, etc. The communication portion 909 performs communication processing via a network such as the Internet. A drive 910 is also connected to the input / output (I / O) interface 905 as needed. A removable medium 911, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 910 as needed, so that a computer program read therefrom is installed into the storage portion 908 as needed.

[0115] According to an embodiment of the present disclosure, the method flow according to an embodiment of the present disclosure can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a computer-readable storage medium, and the computer program contains a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 909, and / or installed from the removable medium 911. When the computer program is executed by the processor 901, the above-mentioned functions defined in the system of the embodiment of the present disclosure are executed. According to an embodiment of the present disclosure, the system, equipment, device, module, unit, etc. described above can be implemented by a computer program module.

[0116] The present disclosure also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or may exist independently without being assembled into the device / apparatus / system. The above computer-readable storage medium carries one or more programs, and when the above one or more programs are executed, the method according to the embodiment of the present disclosure is implemented.

[0117] According to an embodiment of the present disclosure, the computer-readable storage medium may be a non-volatile computer-readable storage medium. For example, it may include, but is not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, apparatus, or device.

[0118] For example, according to an embodiment of the present disclosure, the computer-readable storage medium may include the ROM 902 and / or the RAM 903 described above and / or one or more memories other than the ROM 902 and the RAM 903 .

[0119] An embodiment of the present disclosure also includes a computer program product, which includes a computer program, and the computer program contains a program code for executing the method provided by the embodiment of the present disclosure. When the computer program product runs on an electronic device, the program code is used to enable the electronic device to implement the compressor blade shape generation method provided by the embodiment of the present disclosure.

[0120] When the computer program is executed by the processor 901, the above functions defined in the system / device of the embodiment of the present disclosure are executed. According to the embodiment of the present disclosure, the system, device, module, unit, etc. described above can be implemented by a computer program module.

[0121] In one embodiment, the computer program may be based on a tangible storage medium such as an optical storage device, a magnetic storage device, etc. In another embodiment, the computer program may also be transmitted and distributed in the form of a signal on a network medium, and downloaded and installed through the communication part 909, and / or installed from a removable medium 911. The program code contained in the computer program may be transmitted using any appropriate network medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0122] According to an embodiment of the present disclosure, the program code for executing the computer program provided by the embodiment of the present disclosure can be written in any combination of one or more programming languages. Specifically, these computing programs can be implemented using high-level process and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, Java, C++, python, "C" language or similar programming languages. The program code can be executed entirely on the user computing device, partially on the user device, partially on the remote computing device, or entirely on the remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (for example, using an Internet service provider to connect through the Internet).

[0123] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram may represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of the code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box may also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart, can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions. It can be understood by those skilled in the art that the features recorded in the various embodiments of the present disclosure can be combined and / or combined in a variety of ways, even if such a combination or combination is not explicitly recorded in the present disclosure. In particular, without departing from the spirit and teaching of the present disclosure, the features described in the various embodiments of the present disclosure may be combined and / or combined in a variety of ways. All of these combinations and / or combinations fall within the scope of the present disclosure.

[0124] The embodiments of the present disclosure are described above. However, these embodiments are only for illustrative purposes and are not intended to limit the scope of the present disclosure. Although the embodiments are described above, this does not mean that the measures in the various embodiments cannot be used in combination to advantage. Without departing from the scope of the present disclosure, those skilled in the art may make a variety of substitutions and modifications, which should all fall within the scope of the present disclosure.

Claims

1. A method for generating a blade shape of a compressor, comprising: Inputting the user's expected performance parameters for the blade into the shape design model to design the shape of the blade to obtain a plurality of geometric shapes; Inputting a plurality of said geometric shapes into a performance evaluation model to determine a performance evaluation value of each said geometric shape; Screening the plurality of geometric shapes according to the performance evaluation value to obtain at least one target shape that meets the desired performance parameter; The shape design model and the performance evaluation model are both obtained by training using sample blade performance and flow field simulation data in a training set. The flow field simulation data is obtained by numerically simulating the geometric parameters of the sample blades after analysis.

2. The method according to claim 1, wherein: The shape design model includes a diffusion model, a latent space representation and a decoder. The user's expected performance parameters for the blade are input into the shape design model to design the shape of the blade to obtain multiple geometric shapes, including: Inputting the desired performance parameters into the diffusion model as constraints, and outputting a latent space representation of the shape design data of the blade; The latent space representation of the shape design data is input into the decoder to convert the shape design data into a three-dimensional geometric shape.

3. The method according to claim 1, wherein: The performance evaluation model includes an encoder and a prediction network, and the plurality of geometric shapes are respectively input into the performance evaluation model to determine the performance evaluation value of each geometric shape, including: For each of the geometric shapes, inputting the geometric shape into an encoder to convert the geometric shape into a latent space representation; The latent space representation obtained by transformation is input into the prediction network, and a performance evaluation value of the geometric shape is output.

4. The method according to claim 1, further comprising: Inputting the target shape into a flow field prediction model to perform feature extraction, and outputting a plurality of blade surface parameters predicted for the target shape; According to the blade surface parameters, the target shape is reconstructed into a three-dimensional flow field to obtain three-dimensional flow field information, wherein the flow field information includes temperature information and pressure information.

5. The method according to claim 4, further comprising: The shape design model, the performance evaluation model and the flow field prediction model are integrated to construct an integrated design platform, wherein the integrated design platform is used to visualize the flow field of the three-dimensional flow field information that meets the expected performance parameters input by the user.

6. The method according to claim 1, wherein: The shape design model and the performance evaluation model are trained in the following manner: Determining a sample space according to the acquired geometric parameters of the sample leaf; Performing Latin hypercube sampling on the geometric parameters in the sample space to obtain a plurality of sample data; Using a performance simulation method to perform batch grid processing and single-channel calculation on the sample data to obtain the flow field simulation data, wherein the flow field simulation data includes a plurality of sample shapes; Converting the sample shape into a point cloud to obtain shape training data; Model training is performed using the shape training data and the corresponding sample blade performance to obtain the shape design model and the performance evaluation model.

7. The method according to claim 6, wherein: The method of performing model training using the shape training data and the corresponding sample blade performance includes: De-noising the shape training data, so as to perform model training using the de-noised data and corresponding sample blade performance; In the process of training the shape design model, the shape design experience of the blade and the related alignment strategy are introduced.

8. A compressor blade shape generating device, comprising: A shape design module, used for inputting the user's expected performance parameters of the blade into the shape design model to design the shape of the blade to obtain a plurality of geometric shapes; A performance evaluation module, used for inputting a plurality of said geometric shapes into a performance evaluation model, and determining a performance evaluation value of each said geometric shape; A shape screening module, used for screening the plurality of geometric shapes according to the performance evaluation value to obtain at least one target shape that meets the expected performance parameters; The shape design model and the performance evaluation model are both obtained by training using sample blade performance and flow field simulation data in a training set. The flow field simulation data is obtained by numerically simulating the geometric parameters of the sample blades after analysis.

9. An electronic device, comprising: one or more processors; a memory for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, enables the processor to implement the method according to any one of claims 1 to 7.

Images