Gas turbine combustion chamber outlet temperature distribution rapid prediction method and related equipment

Abstract

The invention relates to the technical field of combustion system optimization, and discloses a gas turbine combustion chamber outlet temperature distribution rapid prediction method and related equipment. The method comprises the following steps: collecting combustion chamber outlet temperature distribution data obtained through a CFD experiment mode; performing main mode extraction on the combustion chamber outlet temperature distribution data based on intrinsic orthogonal decomposition to obtain a POD coefficient matrix and a POD mode matrix; performing statistical fitting on the POD coefficient matrix by using a deep neural network, and establishing a POD coefficient prediction model; constructing an outlet temperature prediction model based on a parameterized reduced-order model based on the POD coefficient prediction model and the POD modal matrix; and utilizing the outlet temperature prediction model to perform rapid simulation prediction on the outlet temperature distribution of the combustion chamber under the combustion parameters to be optimized. According to the method, the calculation cost and time consumption of engine combustion chamber simulation can be effectively reduced, meanwhile, high simulation precision is kept, and powerful support is provided for optimization design of an advanced combustion chamber.

Description

Technical Field

[0001] The present invention relates to the technical field of combustion system optimization, and in particular to a method for quickly predicting outlet temperature distribution of a gas turbine combustion chamber and related equipment. Background Art

[0002] Traditional combustion system optimization mainly relies on high-fidelity simulation and experimental testing. However, with the rapid development of big data and machine learning technologies, data-driven design optimization methods can build proxy models with low computational cost and low accuracy loss based on existing data, accelerate the design process of advanced gas turbines, and have received widespread attention.

[0003] Under the background of global carbon dioxide and nitrogen oxide emission restrictions, traditional hydrocarbon fuel combustion technology faces severe challenges. However, hydrogen-blended gas engine technology has gradually become an important development direction in the field of gas turbines due to its low emissions and high efficiency.

[0004] However, the complexity of hydrogen combustion and the large parameter space pose challenges to the design of the combustion chamber of the hydrogen-blended gas engine. Traditional design optimization methods, such as high-fidelity simulation and experimental testing, require huge costs to traverse all parameter ranges, which not only consumes time and resources, but also often cannot meet the real-time needs of industrial production. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a method for quickly predicting the outlet temperature distribution of a gas turbine combustion chamber and related equipment in view of the problems existing in the prior art.

[0006] In a first aspect, the present invention provides a method for quickly predicting the temperature distribution at the outlet of a gas turbine combustion chamber, comprising:

[0007] Collect the combustion chamber outlet temperature distribution data obtained through CFD experiments;

[0008] Based on the intrinsic orthogonal decomposition, the main mode of the combustion chamber outlet temperature distribution data is extracted to obtain the POD coefficient matrix and POD mode matrix.

[0009] The POD coefficient matrix is ​​statistically fitted using a deep neural network to establish a POD coefficient prediction model;

[0010] Based on the POD coefficient prediction model and POD modal matrix, an outlet temperature prediction model based on parameterized reduced-order model is constructed;

[0011] The outlet temperature prediction model is used to quickly simulate and predict the combustion chamber outlet temperature distribution under the combustion parameters to be optimized.

[0012] In a second aspect, the present invention provides a system for quickly predicting the outlet temperature distribution of a gas turbine combustion chamber, which is used in the method for quickly predicting the outlet temperature distribution of a gas turbine combustion chamber provided by the above technical solution. The system comprises:

[0013] A data acquisition module is used to collect the combustion chamber outlet temperature distribution data obtained through CFD experiments;

[0014] A dimension reduction processing module is used to extract the main mode of the combustion chamber outlet temperature distribution data based on intrinsic orthogonal decomposition to obtain the POD coefficient matrix and POD mode matrix;

[0015] The model building module is used to perform statistical fitting on the POD coefficient matrix using a deep neural network to establish a POD coefficient prediction model; based on the POD coefficient prediction model and the POD modal matrix, an outlet temperature prediction model based on a parameterized reduced-order model is constructed;

[0016] The simulation prediction module is used to use the outlet temperature prediction model to quickly simulate and predict the combustion chamber outlet temperature distribution under the combustion parameters to be optimized.

[0017] In a third aspect, the present invention provides a computer-readable storage medium, comprising instructions, which, when executed on a computer, enable the computer to execute a method for rapidly predicting the outlet temperature distribution of a gas turbine combustion chamber as provided in the above technical solution.

[0018] In a fourth aspect, the present invention provides an electronic device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, a method for rapidly predicting the outlet temperature distribution of a gas turbine combustion chamber is implemented as provided in the above technical solution.

[0019] In a fifth aspect, the present invention further provides a computer program product, the computer program product comprising computer instructions, the computer instructions being stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the method for rapidly predicting the temperature distribution at the outlet of a gas turbine combustion chamber provided by the above scheme.

[0020] The beneficial effects of the present invention are as follows: the present invention performs dimensionality reduction processing on the combustion chamber outlet temperature distribution data through the intrinsic orthogonal decomposition POD method, and establishes a POD coefficient prediction model using a deep neural network. When prediction is made based on new combustion parameters, the POD modal data part is fixed, and only the POD coefficients with differences are predicted, and the common part does not need to be repeatedly calculated, thereby achieving a replacement for high-fidelity numerical simulation with a shorter design cycle and lower computing consumption, and studying the influence of multiple design parameters on the combustion chamber performance in a wide range of parameter space.

[0021] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description or learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 A flowchart of a method for quickly predicting the outlet temperature distribution of a gas turbine combustion chamber is shown as an exemplary embodiment of the present application;

[0023] Figure 2 for Figure 1 A flowchart of step S5 in the illustrated embodiment in an exemplary embodiment;

[0024] Figure 3 A flow chart of a method for quickly predicting the outlet temperature distribution of a gas turbine combustion chamber according to another exemplary embodiment of the present application;

[0025] Figure 4 A block diagram of a system for rapid prediction of gas turbine combustor outlet temperature distribution according to an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0026] The following describes the embodiments of the present disclosure through specific examples, and those skilled in the art can easily understand other advantages and effects of the present disclosure from the contents disclosed in this specification. Obviously, the described embodiments are only part of the embodiments of the present disclosure, rather than all of the embodiments. The present disclosure can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present disclosure. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present disclosure.

[0027] It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described herein may be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on the present disclosure, it should be understood by those skilled in the art that an aspect described herein may be implemented independently of any other aspect, and two or more of these aspects may be combined in various ways. For example, any number of aspects described herein may be used to implement the device and / or practice the method. In addition, other structures and / or functionalities other than one or more of the aspects described herein may be used to implement this device and / or practice this method.

[0028] Aiming at the defects of high-fidelity numerical simulation and experimental testing, the present invention provides a method for quickly predicting the outlet temperature distribution of a gas turbine combustion chamber to predict the outlet temperature of the combustion chamber. Compared with the prior art, the present invention can effectively reduce the computational cost and time consumption of engine combustion chamber simulation while maintaining a high simulation accuracy, providing strong support for the optimal design of advanced combustion chambers.

[0029] The solution provided in the embodiments of the present invention can be executed by any electronic device, for example, a terminal device, or jointly executed by a terminal device and a server. The server can be an independent server, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms, etc., and is not limited here.

[0030] Figure 1 The flowchart of a method for quickly predicting the outlet temperature distribution of a gas turbine combustion chamber is shown as an exemplary embodiment of the present application. Figure 1 As shown, in an exemplary embodiment, the rapid prediction of the gas turbine combustion chamber outlet temperature distribution may include steps S1 to S5, which are described in detail as follows:

[0031] S1, collect the combustion chamber outlet temperature distribution data obtained by CFD experiment;

[0032] S2, extract the main mode of the combustion chamber outlet temperature distribution data based on the intrinsic orthogonal decomposition to obtain the POD coefficient matrix and POD mode matrix;

[0033] S3, using deep neural network to statistically fit the POD coefficient matrix and establish a POD coefficient prediction model;

[0034] S4, constructing an outlet temperature prediction model based on a parameterized reduced-order model based on the POD coefficient prediction model and the POD modal matrix;

[0035] S5, using the outlet temperature prediction model to perform a rapid simulation prediction of the combustion chamber outlet temperature distribution under the combustion parameters to be optimized.

[0036] The embodiment of the present invention performs dimensionality reduction processing on the combustion chamber outlet temperature distribution data through the intrinsic orthogonal decomposition POD method, and establishes a POD coefficient prediction model using a deep neural network. When making predictions based on new combustion parameters, the POD modal data part is fixed, and only the POD coefficients with differences are predicted, and the common part does not need to be calculated repeatedly, thereby achieving a replacement for high-fidelity numerical simulation with a shorter design cycle and lower computing consumption, and studying the influence of multiple design parameters on the combustion chamber performance in a wide range of parameter space.

[0037] The embodiments of the present invention can effectively reduce the computational cost and time consumption of engine combustion chamber simulation (compared to high-fidelity numerical simulation, the computational time is shortened by more than 4 orders of magnitude), while maintaining a high simulation accuracy, providing strong support for the optimal design of advanced combustion chambers.

[0038] Optionally, Figure 1 In an exemplary embodiment, step S1 in the illustrated embodiment may specifically include steps S11 to S12. The combustion chamber outlet temperature distribution data is obtained through the above steps, which are described in detail as follows:

[0039] S11, collecting flow field data under different parameters in the gas turbine combustion chamber obtained by CFD numerical simulation.

[0040] It should be noted that CFD (Computational Fluid Dynamics) has developed along with the development of computer technology and numerical computing technology. Simply put, CFD is equivalent to "virtually" conducting experiments on computers to simulate actual fluid flow conditions. Its basic principle is to numerically solve the differential equations that control fluid flow, and obtain the discrete distribution of the flow field of fluid flow in a continuous area, thereby approximately simulating the fluid flow condition.

[0041] Flow field data is a set of sequential, large, fast, and continuous data sequences, usually generated continuously by thousands of data sources. These data are usually sent in the form of data records and are small in size (about a few thousand bytes). The characteristic of flow field data is that it grows infinitely over time and is suitable for dynamic data processing and analysis. Process data generally includes gas temperature, flow rate, pressure, and gas composition.

[0042] S12, extracting combustion chamber outlet temperature distribution data from the flow field data and constructing a temperature distribution matrix.

[0043] According to the geometric shape of the combustion chamber outlet and the distribution of temperature data, determine the dimension of the temperature distribution matrix, simplify the shape of the combustion chamber outlet into a plane, and determine the number of rows and columns of the matrix according to the resolution of the temperature data. Fill the temperature data collected in the flow field into the temperature distribution matrix, map the data points to the corresponding positions of the matrix, and perform data smoothing or interpolation processing as needed. After the construction is completed, ensure that it can accurately reflect the temperature distribution of the combustion chamber outlet.

[0044] The embodiment of the present invention facilitates the collection of combustion chamber outlet temperature distribution data by constructing a temperature distribution matrix, realizes graphical display of the temperature distribution matrix, and makes the temperature distribution of the combustion chamber outlet clear at a glance. The temperature difference at the combustion chamber outlet can be quantified and numerical data can be provided. This facilitates the subsequent prediction of the temperature distribution of the combustion chamber outlet.

[0045] Optionally, Figure 1 In an exemplary embodiment, step S2 in the illustrated embodiment may specifically include steps S21 to S22. Through the above steps, a POD coefficient matrix and a POD modal matrix are obtained, which are described in detail as follows:

[0046] S21, using intrinsic orthogonal decomposition to obtain the eigenvalues ​​and eigenvectors of the temperature distribution matrix; where different eigenvalues ​​correspond to different POD modes; the size of different eigenvalues ​​represents the strength of the influence of different flow conditions on the flow, and then the POD mode corresponding to each eigenvalue is obtained.

[0047] S22, retaining the first preset number of POD modes whose eigenvalues ​​are arranged from large to small as main modes, constructing a POD mode matrix using the main modes, and obtaining the corresponding POD coefficient matrix.

[0048] The present invention implements retaining a preset number of POD modes with the largest eigenvalues, constructing a POD modal matrix that stores the most important information, and obtaining POD coefficients corresponding to each working condition, thereby obtaining a POD coefficient matrix.

[0049] In an embodiment of the present invention, POD is used to extract the main dynamic features (i.e., POD mode) in the training set. This process generates a low-dimensional subspace, and the training set data is mapped to the subspace, thereby obtaining the corresponding subspace coordinates (i.e., POD coefficients).

[0050] For a given data set, the proper orthogonal decomposition (POD) method constructs a set of orthogonal basis functions (i.e., POD modes) so that the L2 error when the original data is mapped to this set of basis functions is minimized. The L2 error refers to the sum of squares of the difference between the target value and the estimated value in the least squares method. The basis function is selected based on the size of the corresponding eigenvalue, where the POD mode corresponding to the largest eigenvalue represents the most significant feature in the data. By retaining the first preset number r of POD modes with larger eigenvalues, the most important information in the data can be recovered with minimal uncertainty.

[0051] Map the training data at design point i to the POD modal matrix Where m is the number of grid points corresponding to each design point, the POD coefficient vector or coordinate associated with the subspace can be obtained The physical field is further reconstructed using the formula:

[0052]

[0053] Among them, φ j represents the jth POD mode in the POD mode matrix, represents the i-th coefficient of the j-th POD mode in the POD coefficients.

[0054] The above formula can be regarded as a parameterized reduced-order model ROM obtained by training with the original data in the training set. In order to develop the parameterized reduced-order model ROM, it is necessary to predict the coefficient vector (β(new)) at the unknown design point (x(new)) based on the existing information in the training set. The prediction of the POD coefficient at the new design point can realize the rapid calculation of the spatial field.

[0055] Optionally, Figure 1 In an exemplary embodiment, step S3 in the illustrated embodiment may specifically include steps S31 to S33. Through the above steps, a POD coefficient prediction model is obtained, which is described in detail as follows:

[0056] S31, obtaining the combustion parameters corresponding to each POD coefficient in the POD coefficient matrix;

[0057] S32, constructing a training set based on the combustion parameters and the corresponding POD coefficients;

[0058] S33, training a deep neural network based on the training set to obtain a POD coefficient prediction model.

[0059] It should be noted that the deep neural network defines a mapping function β p (x) = f(x; θ), where β p (x) is used to represent the coefficient vector at the design point x.p (x) Information learning parameter θ, θ is not a single value, θ includes weights and bias {W (i) , b (i) , i=1,2,…l}. θ is iteratively optimized using the training data set and an appropriate loss function. By learning the θ obtained, the new x(new) can be input to infer β(new) to obtain the POD coefficient matrix. The POD coefficient matrix and the POD mode can calculate the combustion chamber outlet temperature. The loss function is used to evaluate the degree of deviation between the predicted value and the actual value on the entire training set. The embodiment of the present invention adopts the mean square error loss function. In order to minimize the loss function, the Adam optimization algorithm is used to update the parameter θ until the specified number of iterations or the set loss function threshold is reached. The activation function adopts the ReLU function. The deep neural network DNN has the ability to effectively capture highly nonlinear features.

[0060] See also Figure 2 , Figure 2 for Figure 1 Step S5 in the illustrated embodiment is a flow chart in an exemplary embodiment. Figure 2 As shown, step S5 may specifically include steps S51 to S52. The above steps are used to implement rapid simulation prediction of the combustion chamber outlet temperature distribution under the combustion parameters to be optimized, which is described in detail as follows:

[0061] S51, predicting a new POD coefficient corresponding to the combustion parameter to be optimized based on the POD coefficient prediction model, and updating the POD coefficient matrix based on the new POD coefficient.

[0062] The combustion parameters to be optimized may be a set of combustion parameters in a pre-constructed set of candidate combustion parameters, and the combustion parameters in the set of candidate combustion parameters cover a parameter range required for combustion system optimization.

[0063] S52, reconstructing the flow field based on the updated POD coefficient matrix and POD mode matrix, and realizing rapid simulation prediction of the combustion chamber outlet temperature distribution under the combustion parameters to be optimized.

[0064] The flow field is reconstructed based on the updated POD coefficient matrix and the POD mode matrix to realize the rapid simulation prediction of the combustion chamber outlet temperature distribution under the combustion parameters to be optimized. Specifically, it may include: multiplying the updated POD coefficient matrix and the POD mode matrix to realize the rapid simulation prediction of the combustion chamber outlet temperature distribution under the combustion parameters to be optimized.

[0065] It should be noted that the parameterized reduced-order model (ROM) is used to simulate the flexible fuel combustion process under different hydrogen blending ratios to achieve accurate prediction of physical fields such as velocity, temperature and main components. The parameterized reduced-order model ROM combines a mapping-based dimensionality reduction method, a machine learning-based regression model and a Latin hypercube design method. The parameterized reduced-order model ROM developed in the embodiment of the present invention is data-driven and can replace high-fidelity numerical simulation with a shorter design cycle and lower computational consumption, and study the influence of multiple design parameters on the performance of the combustion chamber in a wide range of parameter space.

[0066] The embodiment of the present invention performs dimensionality reduction processing on the combustion chamber outlet temperature distribution data through the intrinsic orthogonal decomposition POD method, and establishes a POD coefficient prediction model using a deep neural network. When making predictions based on new combustion parameters, the POD modal data part is fixed, and only the POD coefficients with differences are predicted, and the common part does not need to be calculated repeatedly, thereby achieving a replacement for high-fidelity numerical simulation with a shorter design cycle and lower computing consumption, and studying the influence of multiple design parameters on the combustion chamber performance in a wide range of parameter space.

[0067] Optionally, in an exemplary embodiment, the rapid prediction of the gas turbine combustion chamber outlet temperature distribution further includes: testing the outlet temperature prediction model under different test sets to verify its accuracy and reliability.

[0068] The present invention is described in detail below with reference to a specific example. Figure 3 The flowchart of the method for quickly predicting the temperature distribution at the outlet of the gas turbine combustion chamber is shown in another exemplary embodiment of the present application. Figure 3 As shown in FIG. 1 , the rapid prediction method for the gas turbine combustion chamber outlet temperature distribution includes:

[0069] 1. Collect flow field data under different parameters in the gas turbine combustion chamber obtained through CFD numerical simulation.

[0070] Specifically, a design space of parameterized input is designed, and combustion parameters include equivalence ratio and hydrogen blending ratio, etc. The flow field data under different parameters are designed through DoE (Design of Experiment) experimental design method. The embodiment of the present invention designs 20 design points to obtain 20 simulation instances.

[0071] 2. Extract the combustion chamber outlet temperature distribution data from the flow field data and construct the temperature distribution matrix.

[0072] 3. Use eigenorthogonal decomposition to obtain the eigenvalues ​​and eigenvectors of the temperature distribution matrix.

[0073] 4. Keep the first preset number of POD modes whose eigenvalues ​​are arranged from large to small as the main modes, use the main modes to construct the POD mode matrix, and obtain the corresponding POD coefficient matrix.

[0074] 5. Use deep neural network to perform statistical fitting on the POD coefficient matrix and establish a POD coefficient prediction model.

[0075] 6. Predict the new POD coefficient corresponding to the combustion parameters to be optimized based on the POD coefficient prediction model, and update the POD coefficient matrix based on the new POD coefficient.

[0076] 7. Based on the updated POD coefficient matrix and POD modal matrix, the flow field is reconstructed to achieve rapid simulation and prediction of the combustion chamber outlet temperature distribution under the combustion parameters to be optimized.

[0077] It should be noted that the attached Figure 3 The POD-based parameterized ROM (reduced-order model) includes a POD coefficient prediction model and a POD modal matrix. When making predictions based on new combustion parameters, the POD modal data part is fixed, and only the different POD coefficients are predicted. The common part does not need to be calculated repeatedly, thereby achieving a replacement for high-fidelity numerical simulation with a shorter design cycle and lower computational consumption, and studying the influence of multiple design parameters on the combustion chamber performance in a wide range of parameter space.

[0078] Figure 4 The block diagram of a gas turbine combustion chamber outlet temperature distribution rapid prediction system is shown in an exemplary embodiment of the present application. Figure 4 As shown, in an exemplary embodiment, the gas turbine combustor outlet temperature distribution rapid prediction system 40 is used to implement the gas turbine combustor outlet temperature distribution rapid prediction method provided in any of the above embodiments.

[0079] The system may include a data acquisition module 41, a dimensionality reduction processing module 42, a model building module 43 and a simulation prediction module 44, which are described in detail as follows:

[0080] The data acquisition module 41 is used to collect the combustion chamber outlet temperature distribution data obtained by CFD experiment;

[0081] The dimension reduction processing module 42 is used to extract the main mode of the combustion chamber outlet temperature distribution data based on the intrinsic orthogonal decomposition to obtain the POD coefficient matrix and the POD mode matrix;

[0082] The model building module 43 is used to perform statistical fitting on the POD coefficient matrix using a deep neural network to establish a POD coefficient prediction model; based on the POD coefficient prediction model and the POD modal matrix, an outlet temperature prediction model based on a parameterized reduced-order model is constructed;

[0083] The simulation prediction module 44 is used to use the outlet temperature prediction model to perform a rapid simulation prediction of the combustion chamber outlet temperature distribution under the combustion parameters to be optimized.

[0084] An embodiment of the present invention further provides a computer-readable storage medium, comprising instructions, which, when executed on a computer, enable the computer to execute a method for rapidly predicting the outlet temperature distribution of a gas turbine combustion chamber as provided in any of the above embodiments.

[0085] An embodiment of the present invention also provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, a method for rapidly predicting the outlet temperature distribution of a gas turbine combustion chamber as provided in any of the above embodiments is implemented.

[0086] The embodiment of the present invention also provides a computer program product, which includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the gas turbine combustion chamber outlet temperature distribution fast prediction method provided in the above various optional embodiments.

[0087] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0088] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0089] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present invention.

[0090] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0091] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0092] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for quickly predicting the temperature distribution at the outlet of a gas turbine combustion chamber, characterized in that: include: Collect the combustion chamber outlet temperature distribution data obtained through CFD experiments; Based on the intrinsic orthogonal decomposition, the main mode of the combustion chamber outlet temperature distribution data is extracted to obtain the POD coefficient matrix and the POD mode matrix; Using a deep neural network to perform statistical fitting on the POD coefficient matrix and establish a POD coefficient prediction model; Constructing an outlet temperature prediction model based on a parameterized reduced-order model based on the POD coefficient prediction model and the POD modal matrix; The outlet temperature prediction model is used to perform rapid simulation prediction of the combustion chamber outlet temperature distribution under the combustion parameters to be optimized.

2. The method according to claim 1, characterized in that The collecting of combustion chamber outlet temperature distribution data obtained by CFD experiment includes: Collect the flow field data under different parameters in the gas turbine combustion chamber obtained through CFD numerical simulation; The combustion chamber outlet temperature distribution data is extracted from the flow field data to construct a temperature distribution matrix.

3. The method according to claim 2, characterized in that The main mode extraction of the combustion chamber outlet temperature distribution data based on the intrinsic orthogonal decomposition is performed to obtain the POD coefficient matrix and the POD mode matrix, including: The eigenvalues ​​and eigenvectors of the temperature distribution matrix are obtained by using intrinsic orthogonal decomposition; wherein different eigenvalues ​​correspond to different POD modes; The first preset number of POD modes whose eigenvalues ​​are arranged from large to small are retained as main modes, a POD mode matrix is ​​constructed using the main modes, and a corresponding POD coefficient matrix is ​​obtained.

4. The method according to claim 1, characterized in that: The method of using a deep neural network to statistically fit the POD coefficient matrix and establish a POD coefficient prediction model includes: Obtain the combustion parameters corresponding to each POD coefficient in the POD coefficient matrix; Constructing a training set based on the combustion parameters and corresponding POD coefficients; A deep neural network is trained based on the training set to obtain a POD coefficient prediction model.

5. The method according to any one of claims 1 to 4, characterized in that: The method of using the outlet temperature prediction model to perform a rapid simulation prediction of the combustion chamber outlet temperature distribution under the combustion parameters to be optimized includes: Predicting a new POD coefficient corresponding to the combustion parameter to be optimized based on the POD coefficient prediction model, and updating the POD coefficient matrix based on the new POD coefficient; The flow field is reconstructed based on the updated POD coefficient matrix and the POD modal matrix to achieve rapid simulation prediction of the combustion chamber outlet temperature distribution under the combustion parameters to be optimized.

6. The method according to claim 5, characterized in that The flow field reconstruction is performed based on the updated POD coefficient matrix and the POD modal matrix to achieve rapid simulation prediction of the combustion chamber outlet temperature distribution under the combustion parameters to be optimized, including: The updated POD coefficient matrix and the POD mode matrix are multiplied to achieve rapid simulation prediction of the combustion chamber outlet temperature distribution under the combustion parameters to be optimized.

7. The method according to claim 5, characterized in that The combustion parameters to be optimized are a group of combustion parameters in a pre-constructed set of candidate combustion parameters, and the combustion parameters in the set of candidate combustion parameters cover a parameter range required for combustion system optimization.

8. A rapid prediction system for gas turbine combustion chamber outlet temperature distribution, characterized in that: For implementing the method described in any one of claims 1 to 7, the system comprises: A data acquisition module is used to collect the combustion chamber outlet temperature distribution data obtained through CFD experiments; A dimensionality reduction processing module is used to extract the main mode of the combustion chamber outlet temperature distribution data based on intrinsic orthogonal decomposition to obtain a POD coefficient matrix and a POD mode matrix; A model building module, used to perform statistical fitting on the POD coefficient matrix using a deep neural network to establish a POD coefficient prediction model; and to construct an outlet temperature prediction model based on a parameterized reduced-order model based on the POD coefficient prediction model and the POD modal matrix; The simulation prediction module is used to use the outlet temperature prediction model to perform rapid simulation prediction of the combustion chamber outlet temperature distribution under the combustion parameters to be optimized.

9. A computer-readable storage medium comprising instructions, characterized in that: When the instructions are executed on a computer, the computer is caused to execute the method for rapidly predicting the gas turbine combustion chamber outlet temperature distribution according to any one of claims 1 to 7.

10. An electronic device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method for rapidly predicting the outlet temperature distribution of a gas turbine combustion chamber as described in any one of claims 1 to 7 when executing the program.

Images