A data assimilation method for turbine blade film cooling flow
By constructing the state matrix and covariance matrix of the prediction set through data assimilation methods, and combining Kalman gain and relaxation factor, the problem of insufficient simulation accuracy of air film cooling flow of turbine blades is solved, thereby improving simulation accuracy and design reliability.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- AECC HUNAN AVIATION POWERPLANT RES INST
- Filing Date
- 2026-03-20
- Publication Date
- 2026-06-23
Smart Images
Figure CN122263306A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of aero-engine technology, and more specifically to a data assimilation method applicable to film cooling flow in turbine blades. Background Technology
[0002] In engine turbine blade testing, the high-temperature, high-pressure, and high-velocity operating environment of turbine blades presents two major limitations to physical testing of film cooling performance: limited testing conditions and unobservable limitations. This restricts the analysis of blade performance, resulting in data acquisition only from a limited number of measurement points and under limited operating conditions, failing to provide a complete flow field reference for turbine blade cooling design. Against this backdrop, data simulation overcomes physical limitations to predict cooling performance under untested conditions and in observational blind spots. Adapting data simulation to the characteristics of turbine blade film cooling flow and improving its accuracy becomes a crucial issue. Summary of the Invention
[0003] This invention provides a data assimilation method suitable for film cooling flow of turbine blades, in order to solve the problem of how to adapt to the characteristics of film cooling flow of turbine blades and improve the accuracy of data simulation.
[0004] In a first aspect, the present invention provides a data assimilation method suitable for film cooling flow of turbine blades, the method comprising: State vector samples are generated based on the data of air film cooling flow of turbine blades, and a prediction set state matrix is constructed. Construct a prediction set covariance matrix based on the prediction set state matrix, solve the Kalman gain using the prediction set covariance matrix, introduce a relaxation factor to adjust the Kalman gain, and use the adjusted Kalman gain to update the state variables. The simulation prediction values are extracted from the updated state variables. If the error between the simulation prediction values and the actual experimental observation data meets the preset error conditions, the data assimilation results are output.
[0005] This invention constructs a prediction set covariance matrix based on the constructed prediction set state matrix to characterize the correlation and dispersion of state variables, solves the Kalman gain to fuse observation information and model predictions, introduces a relaxation factor to regulate the Kalman gain and improve iterative stability, and effectively improves the data simulation accuracy of turbine blade film cooling flow by fusing experimental observation data and simulation prediction values to correct prediction bias, thereby enhancing the accuracy and reliability of turbine blade film cooling design.
[0006] In one alternative implementation, constructing the prediction set state matrix includes: The Spalare-Allmaras model constant combination was generated using the Latin hypercube sampling method; Numerical simulations were performed on the constants of the Spalare-Allmaras model to obtain the predicted state variables; By concatenating the Spalare-Allmaras model constants and predicted state variables, a state vector is obtained. All state vectors are then used to construct a predicted set state matrix.
[0007] This invention generates a combination of Spalare-Allmaras model constants using the Latin hypersolution method, achieving uniform and independent sampling of the multidimensional parameter space within the range of model constant values. Through data simulation, the Spalare-Allmaras model constants are converted into predicted state variables, and the two are concatenated to form a state vector. A predicted set state matrix is constructed to reflect the degree of dispersion of model constants and flow field state variables and the correlation between various dimensions.
[0008] In one optional implementation, constructing a prediction set covariance matrix based on the prediction set state matrix includes: Calculate the average value of the state vectors, and calculate the deviation between each state vector and the average value of the state vectors. Calculate the transpose of the deviation, and multiply the transpose of the deviation by the deviation to obtain the deviation product matrix; Accumulate the product matrix of the biases of all samples to construct the covariance matrix of the prediction set.
[0009] This invention improves the accuracy of the prediction set covariance matrix by considering the state vectors of all samples and the average value of the state vectors when constructing the prediction covariance matrix, thus avoiding statistical bias caused by single sample bias.
[0010] In one alternative implementation, the prediction set covariance matrix is as follows:
[0011] in, P To predict the set covariance matrix, l The number of members in the set. i For serial number, For the first i A state vector, The average value of the state vector. T This is the transpose of a vector.
[0012] In one alternative implementation, the Kalman gain is solved by combining the prediction set covariance matrix, including: Construct a mapping matrix from the state space to the observation space; Multiply the mapping matrix, the prediction set covariance matrix, and the transpose of the mapping matrix in sequence, add the multiplication result to the perturbation covariance matrix to obtain the intermediate matrix, and then take the inverse matrix of the intermediate matrix. The Kalman gain is determined by multiplying the covariance matrix of the prediction set, the transpose of the mapping matrix, and the inverse matrix.
[0013] This invention achieves matching between high-dimensional prediction and low-dimensional observation by constructing a mapping matrix, and calculates the Kalman gain by fusing multi-dimensional matrices to integrate observation information and model prediction, thereby achieving high-precision simulation.
[0014] In one alternative implementation, the Kalman gain is calculated as follows:
[0015] in, K For Kalman gain, R Let covariance be the perturbation. H This is the mapping matrix from the state space to the observation space.
[0016] In one alternative implementation, the method further includes: If the error between the simulation prediction and the actual experimental observation data does not meet the preset error condition, the process returns to the step of constructing the prediction set covariance matrix based on the prediction set state matrix for iterative updating.
[0017] This invention corrects model constant deviations through an iterative mechanism, reduces the discrepancy between simulation predictions and experimental observations, and improves the simulation accuracy of film cooling flow.
[0018] In a second aspect, the present invention provides a data assimilation device suitable for film cooling flow of turbine blades, the device comprising: A module is built to generate state vector samples based on data from the air film cooling flow of turbine blades and to construct a prediction set state matrix. The update module is used to construct the prediction set covariance matrix based on the prediction set state matrix, solve the Kalman gain by combining the prediction set covariance matrix, introduce a relaxation factor to adjust the Kalman gain, and use the adjusted Kalman gain to update the state variables. The output module is used to extract simulation prediction values from the updated state variables. If the error between the simulation prediction values and the actual experimental observation data meets the preset error conditions, the data assimilation results are output.
[0019] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the data assimilation method applicable to the film cooling flow of turbine blades described in the first aspect or any corresponding embodiment thereof.
[0020] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to perform the data assimilation method applicable to film cooling flow of turbine blades as described in the first aspect or any corresponding embodiment thereof. Attached Figure Description
[0021] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.
[0022] Figure 1 This is a schematic flowchart of a data assimilation method for film cooling flow of turbine blades according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the computational domain for assimilating cooling blade film jet data according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the EnKF data assimilation algorithm for turbulence model constants according to an embodiment of the present invention; Figure 4 This is a schematic diagram of numerical simulation results according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the data assimilation result according to an embodiment of the present invention; Figure 6 This is a structural block diagram of a data assimilation device for film cooling flow of turbine blades according to an embodiment of the present invention; Figure 7 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation
[0023] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0024] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.
[0025] In related technologies, numerical simulation models based on the RANS (Reynolds Average Navier-Stokes) model are constructed based on a semi-theoretical and semi-empirical approach. Most of the model constants are derived from simple experimental calibrations such as plate tests. However, the calibration scenario of plate tests is a regular and simple basic turbulence, while the air film cooling flow on the surface of aero-engine turbine blades is a complex and multi-factor coupled special turbulence. Therefore, conventional numerical simulation models based on the RANS model cannot be applied to the prediction of air film cooling flow on the surface of turbine blades.
[0026] This invention provides a data assimilation method suitable for film cooling flow of turbine blades. Based on the Ensemble Kalman Filter (EnKF) algorithm, the default parameters of the Spalare-Allmaras turbulence model in numerical simulation are adjusted and optimized to improve the simulation accuracy of film jet flow of cooling blades.
[0027] According to an embodiment of the present invention, a data assimilation method for turbine blade film cooling flow is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.
[0028] This embodiment provides a data assimilation method suitable for film cooling flow of turbine blades. Figure 1 This is a flowchart of a data assimilation method for film cooling flow of turbine blades according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps: Step S101: Generate state vector samples based on the data of air film cooling flow of turbine blades, and construct a prediction set state matrix.
[0029] In this embodiment of the invention, physical quantity data of the true characteristics of the air film cooling flow of turbine blades are collected, and state vector samples are generated based on the air film cooling flow data of turbine blades. The state vector consists of model constants and numerical simulation predicted state quantities. The state vector samples are structured to construct a prediction set state matrix.
[0030] Step S102: Construct the prediction set covariance matrix based on the prediction set state matrix, solve the Kalman gain by combining the prediction set covariance matrix, introduce a relaxation factor to adjust the Kalman gain, and update the state variables using the adjusted Kalman gain.
[0031] In this embodiment of the invention, the discreteness and interdimensional correlation of the state vector samples are quantified, and a prediction set covariance matrix is constructed based on the state vectors in the prediction set state matrix. The Kalman gain is then calculated by combining the prediction set covariance matrix to determine the allocation ratio of observation information and model prediction in state updates, thereby fusing observation information and model prediction.
[0032] To avoid overcorrection, deviation of model parameters from physical meaning, and oscillation of simulation results when directly using the original Kalman gain in the turbine blade film cooling scenario, a relaxation factor is introduced to adjust the Kalman gain, thereby enhancing the stability of the calculation. The adjusted Kalman gain is then used to update the state variables.
[0033] Step S103: Extract simulation prediction values from the updated state variables. If the error between the simulation prediction values and the actual experimental observation data meets the preset error conditions, output the data assimilation results.
[0034] In this embodiment of the invention, the simulation prediction value corresponding to the experimental observation dimension and physical quantity is extracted from the updated state variable. The simulation prediction value is compared with the actual experimental observation data to quantify the deviation between the simulation prediction value and the actual experimental observation data. When the error between the simulation prediction value and the actual experimental observation data meets the preset error condition, the data assimilation result is output.
[0035] The data assimilation method for turbine blade film cooling flow provided in this embodiment constructs a prediction set covariance matrix based on the constructed prediction set state matrix to characterize the correlation and dispersion of state variables, solves the Kalman gain to fuse observation information and model predictions, introduces a relaxation factor to regulate the Kalman gain and improve iterative stability, and effectively improves the data simulation accuracy of turbine blade film cooling flow by fusing experimental observation data and simulation prediction values, thereby enhancing the accuracy and reliability of turbine blade film cooling design.
[0036] This embodiment provides a data assimilation method suitable for film cooling flow of turbine blades, the process of which includes the following steps: Step S201: Generate state vector samples based on the data of air film cooling flow of turbine blades, and construct a prediction set state matrix.
[0037] Specifically, step S201 includes: Step S2011: Generate the Spalare-Allmaras model constant combination using the Latin hypercube sampling method; Step S2012: Perform numerical simulation on the constants of the Spalare-Allmaras model to obtain the predicted state variables; Step S2013: Concatenate the Spalare-Allmaras model constants and predicted state variables to obtain a state vector, and construct a prediction set state matrix from all state vectors.
[0038] In embodiments of the present invention, such as Figure 2 As shown, Figure 2 The computational domain for assimilating the film jet data of the cooling blades. Figure 2 (a) is a cylindrical structure. Figure 2 (b) is a dune-shaped structure, and the Spalare-Allmaras model is used for all of them. The basic settings for numerical simulation of the Spalare-Allmaras model are shown in Table 1. A unified and suitable basic settings for numerical simulation are formulated for the two typical film venting structures of aero-engine turbine blades, namely cylindrical and dune-shaped.
[0039] Table 1
[0040] For mesh setup, a 1.38 million hexahedral mesh was used for the cylindrical structure, and a 4.33 million polyhedral mesh was used for the dune-shaped structure. The working fluids used were water and Rhodamine B, with the Rhodamine B concentration set according to experimental conditions. Rhodamine B served as a tracer to effectively match the concentration field in the simulation with experimental observations. Regarding boundary conditions, the main flow inlet was set as a velocity inlet, with its velocity distribution set according to experimental data. The jet inlet was set as a uniform velocity inlet boundary condition, and three velocity ratios (0.4, 0.8, and 1.2) between the jet and the main flow were set. The 1.2 velocity ratio was the core condition for data assimilation, used to characterize the optimized velocity field simulation after data assimilation and further determine the scalability of the velocity and concentration fields under different velocity ratios. The main flow outlet was a pressure outlet, with the pressure set to atmospheric pressure to match the experimental environment.
[0041] The Spalare-Allmaras model constant set was generated using the Latin hypercube sampling method. For example, 100 different combinations of model constants were generated. The Spalare-Allmaras model constants and their range of variation are shown in Table 2. The optimization parameter results obtained by data assimilation of the two structures are shown in Table 3.
[0042] Table 2
[0043] Table 3
[0044] Under different model constants, the predicted state variables are calculated through numerical simulation. A state vector is then formed by combining the Spalare-Allmaras model constants and the predicted state variables. , ,in, u To predict state variables, These are model constants. T This is the transpose of a vector.
[0045] Will N The state vectors are used to construct the prediction set state matrix. X for: .
[0046] The Spalare-Allmaras model constant combination is generated by the Latin hypersolution method. Uniform and independent sampling of the multidimensional parameter space is achieved within the value range of the model constants. The Spalare-Allmaras model constants are converted into predicted state variables through data simulation. The two are concatenated to form a state vector and a predicted set state matrix is constructed to reflect the degree of dispersion of the model constants and flow field state variables and the correlation between various dimensions.
[0047] Step S202: Construct the prediction set covariance matrix based on the prediction set state matrix, solve the Kalman gain by combining the prediction set covariance matrix, introduce a relaxation factor to adjust the Kalman gain, and update the state variables using the adjusted Kalman gain.
[0048] Specifically, in step S202 above, constructing the prediction set covariance matrix based on the prediction set state matrix includes: Step S2021: Calculate the average value of the state vectors, and calculate the deviation between each state vector and the average value of the state vectors. Step S2022: Calculate the transpose of the deviation, and multiply the transpose of the deviation with the deviation to obtain the deviation product matrix; Step S2023: Accumulate the product matrix of the biases of all samples to construct the covariance matrix of the prediction set.
[0049] In this embodiment of the invention, the average value of the state vector is calculated. And calculate the deviation between the state vector and the average value of the state vector. Calculate the transpose of the deviation Then multiply it by the original bias to obtain the bias product matrix. This matrix reflects the autocorrelation and cross-correlation of each dimension of a single sample. l The product matrix of the biases of each sample is summed, and combined with the number of samples, a covariance matrix of the prediction set is constructed.
[0050] The covariance matrix of the prediction set is as follows:
[0051] in, P To predict the set covariance matrix, lThe number of members in the set. i For serial number, For the first i A state vector, The average value of the state vector. T This is the transpose of a vector.
[0052] By considering the state vectors of all samples and their average value when constructing the prediction covariance matrix, statistical bias caused by single-sample bias is avoided, thereby improving the accuracy of the prediction set covariance matrix.
[0053] Specifically, the step S202 above, which involves solving for the Kalman gain by combining the prediction set covariance matrix, includes: Step S2024: Construct the mapping matrix from the state space to the observation space; Step S2025: Multiply the mapping matrix, the prediction set covariance matrix, and the transpose of the mapping matrix in sequence, add the multiplication result to the perturbation covariance matrix to obtain the intermediate matrix, and then obtain the inverse matrix of the intermediate matrix. Step S2026: The product of the covariance matrix of the prediction set, the transpose of the mapping matrix, and the inverse matrix is determined as the Kalman gain.
[0054] In this embodiment of the invention, a mapping matrix from the state space to the observation space is constructed. H For the mapping matrix H Perform the transpose operation to get H T , mapping matrix H Predicting the covariance matrix of the set P The transpose of the mapping matrix H T Perform matrix multiplication, and then multiply the result by the covariance matrix of the perturbation. R Add them together to get the intermediate matrix, then invert the intermediate matrix: .
[0055] Predict the set covariance matrix P The transpose of the mapping matrix H T Inverse matrix Perform matrix multiplication to obtain the Kalman gain.
[0056] The Kalman gain calculation formula is as follows:
[0057] in, K For Kalman gain, R Let be the covariance matrix of the perturbation, and assume that the perturbation follows a Gaussian distribution. H This is the mapping matrix from the state space to the observation space.
[0058] By constructing a mapping matrix, high-dimensional predictions are matched with low-dimensional observations. The Kalman gain is calculated by fusing multi-dimensional matrices to integrate observation information and model predictions, thereby achieving high-precision simulation.
[0059] In some alternative implementations, the method further includes: Step Sa: Update the state variables based on the Kalman gain.
[0060] In this embodiment of the invention, the updated state variables are as follows:
[0061] in, x a For the updated state variables, y exp These are experimental observations. w i This is a perturbation based on experimental uncertainty.
[0062] Specifically, in step S202 above, the state variable is updated using the adjusted Kalman gain, and the updated state variable is as follows:
[0063] in, It is a relaxation factor.
[0064] Step S203: Extract simulation prediction values from the updated state variables. If the error between the simulation prediction values and the actual experimental observation data meets the preset error conditions, output the data assimilation results.
[0065] In some alternative implementations, the method further includes: In step S204, if the error between the simulation prediction value and the actual experimental observation data does not meet the preset error condition, return to step S202 for iterative update.
[0066] In this embodiment of the invention, the state vector of the previous round is used as the initial value. Substituting it into the following iterative formula, the state vector of the next round is calculated to complete the correction of the state vector. The iterative formula is as follows:
[0067] in, For the first j+1 The iteration of the ... i A state vector, j For the number of iterations, For the first j The iteration of the ... i A state vector.
[0068] In each iteration, the simulated predicted value is extracted from the state vector through the mapping matrix, the error between the simulated predicted value and the actual experimental observation data is calculated, and it is checked whether the error meets the preset error condition. If the error between the simulated predicted value and the actual experimental observation data does not meet the preset error condition, the covariance matrix of the prediction set is updated, and steps S202 to S204 are repeated until the error between the simulated predicted value and the actual experimental observation data meets the preset error condition. Finally, the set of constants with the smallest error is selected as the optimal model constants.
[0069] The data assimilation method for film cooling flow of turbine blades provided in this embodiment corrects the model constant deviation through an iterative mechanism, reduces the deviation between simulation predictions and experimental observations, and improves the simulation accuracy of film cooling flow.
[0070] like Figure 3 As shown, Figure 3 The flowchart of the EnKF data assimilation algorithm for turbulence model constants is shown. The process is divided into three stages: prediction, analysis, and confirmation. In the prediction stage, the state matrix of the prediction set is constructed. In the analysis stage, the covariance matrix of the prediction set is constructed, the Kalman gain is calculated, and the set members are updated. In the confirmation stage, if the algorithm converges, the optimized state parameters are output and the process ends.
[0071] For example, taking the speed ratio of 1.2 as an example, the numerical simulation results under its optimized parameters are as follows: Figure 4 As shown, the simulation speed is compared with experimental measurements and default parameters. Figure 5 As shown, after data assimilation, the width of the jet core region increases, and the vertical velocity at the jet exit decreases, improving the prediction accuracy of flow separation and jet penetration depth in the inclined transverse jet through the circular aperture. Using model constants optimized based on a velocity ratio of 1.2, tests have shown that the model can still reduce the overestimation of flow separation by the Spalare-Allmaras default model under other operating conditions, and the simulation results are closer to the experimental measurements.
[0072] The optimized parameters based on the cylindrical structure were used in the numerical simulation of the dune-shaped structure, and the results were compared with those of the dune-shaped structure based on its own optimized parameters. The predicted flow field results of the two sets of optimized parameters are almost identical, both correcting the narrower jet and overestimation of flow separation in the predicted flow field of the Spalare-Allmaras model under the default parameters.
[0073] The heat transfer process based on film cooling is similar to the concentration mixing process of inclined transverse jets, both driven by similar scalar transport equations. Therefore, a new constant, the turbulent Schmidt number Sc, is introduced to modify the mass transport equation. This parameter has a default value of 0.7, and the sampling variation factor is 4. The concentration field is corrected using a data assimilation method. After supplementing the correction with Sc=0.32, the predicted turbulent scalar mixing is appropriately enhanced, better reflecting the actual situation, and the prediction effect is significantly improved.
[0074] The flow field characteristics of film jets with different structures were calculated at different velocity ratios. Data assimilation techniques were used to correct the model constants of the velocity and concentration fields of transversely inclined flat plate jets with different structures, so as to improve the prediction accuracy of numerical simulation.
[0075] This embodiment also provides a data assimilation device suitable for turbine blade film cooling flow. This device is used to implement the above embodiments and preferred embodiments, and details already described will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that performs a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.
[0076] This embodiment provides a data assimilation device suitable for film cooling flow of turbine blades, such as... Figure 6 As shown, it includes: Module 601 is used to generate state vector samples based on the data of air film cooling flow of turbine blades and to construct a prediction set state matrix; The update module 602 is used to construct the prediction set covariance matrix based on the prediction set state matrix, solve the Kalman gain by combining the prediction set covariance matrix, introduce a relaxation factor to adjust the Kalman gain, and update the state variables using the adjusted Kalman gain. The output module 603 is used to extract the simulation prediction value from the updated state variable. If the error between the simulation prediction value and the actual experimental observation data meets the preset error condition, the data assimilation result is output.
[0077] In some alternative implementations, the construction module 601 includes: A constant generation unit is used to generate combinations of Spalare-Allmaras model constants using the Latin hypercube sampling method. The numerical simulation unit is used to numerically simulate the constants of the Spalare-Allmaras model to obtain the predicted state variables; The first building unit is used to concatenate the Spalare-Allmaras model constants and predicted state variables to obtain a state vector, and then construct the predicted set state matrix from all the state vectors.
[0078] In some alternative implementations, the update module 602 includes: The first calculation unit is used to calculate the average value of the state vector and the deviation between each state vector and the average value of the state vector. The first calculation unit is used to calculate the transpose of the deviation, and multiply the transpose of the deviation by the deviation to obtain the deviation product matrix; The second building unit is used to accumulate the product matrix of the biases of all samples to construct the covariance matrix of the prediction set.
[0079] In some alternative implementations, the prediction set covariance matrix is as follows:
[0080] in, P To predict the set covariance matrix, l The number of members in the set. i For serial number, For the first i A state vector, The average value of the state vector. T This is the transpose of a vector.
[0081] In some alternative implementations, the update module 602 further includes: The third building unit is used to construct the mapping matrix from the state space to the observation space; The multiplication unit is used to multiply the mapping matrix, the prediction set covariance matrix, and the transpose of the mapping matrix in sequence, add the multiplication result to the perturbation covariance matrix to obtain the intermediate matrix, and then obtain the inverse matrix of the intermediate matrix. The determination unit is used to determine the Kalman gain by multiplying the covariance matrix of the prediction set, the transpose of the mapping matrix, and the inverse matrix.
[0082] In some alternative implementations, the Kalman gain is calculated as follows:
[0083] in, K For Kalman gain, R Let covariance be the perturbation. H This is the mapping matrix from the state space to the observation space.
[0084] In some alternative embodiments, the device further includes: The iterative update unit is used to return to the step of constructing the covariance matrix of the prediction set based on the state matrix of the prediction set for iterative update if the error between the simulation prediction value and the actual experimental observation data does not meet the preset error condition.
[0085] The data assimilation device for turbine blade film cooling flow provided in this embodiment of the invention can execute the data assimilation method for turbine blade film cooling flow provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the method. Further functional descriptions of the above modules and units are the same as in the corresponding embodiments described above, and will not be repeated here.
[0086] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.
[0087] The following is a detailed reference. Figure 7 This diagram illustrates a suitable structural schematic for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 701, which can perform various appropriate actions and processes based on a program stored in read-only memory (ROM) 702 or a program loaded from memory 708 into random access memory (RAM) 703. The RAM 703 also stores various programs and data required for the operation of the electronic device. The processor 701, ROM 702, and RAM 703 are interconnected via a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.
[0088] Typically, the following devices can be connected to I / O interface 705: input devices 706 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 707 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 708 including, for example, magnetic tapes, hard disks, etc.; and communication devices 709. Communication device 709 allows electronic devices to exchange data via wireless or wired communication with other devices. Although Figure 7 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.
[0089] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 709, or installed from a memory 708, or installed from a ROM 702. When the computer program is executed by the processor 701, it performs the functions defined in the data assimilation method for turbine blade film cooling flow according to embodiments of the present invention.
[0090] Figure 7 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.
[0091] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the data assimilation method for turbine blade film cooling flow shown in the above embodiments is implemented.
[0092] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.
[0093] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and all such modifications and variations fall within the scope defined by the appended invention.
Claims
1. A data assimilation method suitable for film cooling flow of turbine blades, characterized in that, The method includes: State vector samples are generated based on the data of air film cooling flow of turbine blades, and a prediction set state matrix is constructed. Construct a prediction set covariance matrix based on the prediction set state matrix, solve the Kalman gain using the prediction set covariance matrix, introduce a relaxation factor to adjust the Kalman gain, and use the adjusted Kalman gain to update the state variables. The simulation prediction value is extracted from the updated state variable. If the error between the simulation prediction value and the actual experimental observation data meets the preset error condition, the data assimilation result is output.
2. The method according to claim 1, characterized in that, The construction of the prediction set state matrix includes: The Spalare-Allmaras model constant combination was generated using the Latin hypercube sampling method; Numerical simulations were performed on the constants of the Spalare-Allmaras model to obtain the predicted state variables; By concatenating the Spalare-Allmaras model constants and predicted state variables, a state vector is obtained. All state vectors are then used to construct a predicted set state matrix.
3. The method according to claim 2, characterized in that, The step of constructing the prediction set covariance matrix based on the prediction set state matrix includes: Calculate the average value of the state vectors, and calculate the deviation between each state vector and the average value of the state vectors; Calculate the transpose of the deviation, and multiply the transpose of the deviation and the deviation to obtain the deviation product matrix; Accumulate the product matrix of the biases of all samples to construct the covariance matrix of the prediction set.
4. The method according to claim 1 or 3, characterized in that, The covariance matrix of the prediction set is as follows: in, P To predict the set covariance matrix, l The number of members in the set. i For serial number, For the first i A state vector, The average value of the state vector. T This is the transpose of a vector.
5. The method according to claim 4, characterized in that, The step of solving for the Kalman gain by combining the covariance matrix of the prediction set includes: Construct a mapping matrix from the state space to the observation space; Multiply the mapping matrix, the prediction set covariance matrix, and the transpose of the mapping matrix in sequence, add the multiplication result to the perturbation covariance matrix to obtain the intermediate matrix, and then obtain the inverse matrix of the intermediate matrix. The product of the covariance matrix of the prediction set, the transpose of the mapping matrix, and the inverse matrix is determined as the Kalman gain.
6. The method according to claim 5, characterized in that, The Kalman gain calculation formula is as follows: in, K For Kalman gain, R Let be the covariance matrix of the perturbation. H This is the mapping matrix from the state space to the observation space.
7. The method according to claim 1, characterized in that, The method further includes: If the error between the simulated predicted value and the actual experimental observation data does not meet the preset error condition, then return to the step of constructing the prediction set covariance matrix based on the prediction set state matrix for iterative update.
8. A data assimilation device suitable for film cooling flow of turbine blades, characterized in that, The device includes: A module is built to generate state vector samples based on data from the air film cooling flow of turbine blades and to construct a prediction set state matrix. The update module is used to construct a prediction set covariance matrix based on the prediction set state matrix, solve the Kalman gain by combining the prediction set covariance matrix, introduce a relaxation factor to adjust the Kalman gain, and update the state variables using the adjusted Kalman gain. The output module is used to extract simulation prediction values from the updated state variables. If the error between the simulation prediction values and the actual experimental observation data meets the preset error conditions, the data assimilation result is output.
9. An electronic device, characterized in that, include: A memory and a processor are communicatively connected, the memory storing computer instructions, and the processor executing the computer instructions to perform the data assimilation method for film cooling flow of turbine blades as described in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing a computer to perform the data assimilation method applicable to film cooling flow of turbine blades as described in any one of claims 1 to 7.