A method and device for determining a blade deformation simulation parameter and an electronic device

Abstract

The present disclosure provides a method and device for determining a blade deformation simulation parameter and an electronic device. The method comprises: generating a deformation driving model of a blade model based on blade modeling information, the deformation driving model comprising a plurality of sub-models, each sub-model being used to drive the blade model to produce deformation in an activated state; optimizing activation information of the deformation driving model using a Bayesian optimization algorithm to determine target activation information of the deformation driving model corresponding to a target blade geometry; and determining the target activation information of the deformation driving model corresponding to the target blade geometry as the blade deformation simulation parameter. The method provided by the present disclosure can not only realize forward design of the blade producing target deformation, but also drive the blade to actively produce deformation under different engine operating conditions by designing target activation information of different blade geometries to obtain corresponding blade geometries, so that the engine can have good aerodynamic performance in a wide speed range.

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, and in particular to a method, apparatus and electronic device for determining blade deformation simulation parameters. Background Technology

[0002] Aero engines generate thrust by compressing and expanding gases, with nearly 90% of the thrust coming from the high-speed rotation of the fan blades at the front of the engine. Under actual operating conditions, the fan blades are subjected to complex stresses, which alters their profile, leading to reduced performance or even breakage, thus causing engine performance degradation and impacting flight safety.

[0003] In related technologies, the fan blades are first modeled as curved surfaces and then imported into simulation software such as ANSYS or related simulation platforms. Next, the fan blade model is meshed, and then finite element analysis is performed to evaluate the deformation of the fan blades. However, existing technologies cannot achieve forward design of the expected deformation of the fan blades, nor can they configure the parameters to achieve the expected deformation. Summary of the Invention

[0004] According to one aspect of this disclosure, a method for determining blade deformation simulation parameters is provided, the method comprising:

[0005] A deformation-driven model for generating a blade model based on blade modeling information is provided. The deformation-driven model includes multiple sub-models, each of which is used to drive the blade model to produce deformation when it is in an active state.

[0006] The activation information of the deformation-driven model is optimized using the Bayesian optimization algorithm to determine the target activation information of the deformation-driven model corresponding to the target blade geometry. The activation information of the deformation-driven model includes the activation parameters of at least one sub-model that is in an active state.

[0007] The target activation information of the deformation-driven model corresponding to the target blade geometry is determined as the blade deformation simulation parameters.

[0008] According to another aspect of this disclosure, an apparatus for determining blade deformation simulation parameters is provided, the apparatus comprising:

[0009] The generation module is used to generate a deformation-driven model of the blade model based on the blade modeling information. The deformation-driven model includes multiple sub-models, and each sub-model is used to drive the blade model to produce deformation when it is in an active state.

[0010] The determination module is used to optimize the activation information of the deformation driving model using a Bayesian optimization algorithm, and to determine the target activation information of the deformation driving model corresponding to the target blade geometry. The activation information of the deformation driving model includes the activation parameters of at least one sub-model that is in an active state.

[0011] The determination module is also used to determine the target activation information of the deformation driving model corresponding to the target blade geometry as blade deformation simulation parameters.

[0012] According to another aspect of this disclosure, an electronic device is provided, comprising:

[0013] Processor; and,

[0014] Memory for stored programs;

[0015] The program includes instructions that, when executed by a processor, cause the processor to perform the method described according to exemplary embodiments of the present disclosure.

[0016] According to another aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are used to cause a computer to perform the method described according to an exemplary embodiment of this disclosure.

[0017] One or more technical solutions provided in the exemplary embodiments of this disclosure can generate a deformation-driven model of the blade model based on blade modeling information. The deformation-driven model includes multiple sub-models, each of which, when activated, can drive the blade model to deform. Based on this, the exemplary embodiments of this disclosure can utilize a Bayesian optimization algorithm to optimize the activation information of the deformation-driven model, determining the target activation information of the deformation-driven model corresponding to the target blade geometry. The activation information of the deformation-driven model includes activation parameters of at least one sub-model in an activated state. Furthermore, the target activation information of the deformation-driven model corresponding to the target blade geometry is determined as the blade deformation simulation parameter. Therefore, the method of the exemplary embodiments of this disclosure can determine the target activation information of the blade in the target blade geometry through the deformation-driven model. Using this target activation information as the blade deformation simulation parameter can drive the blade to achieve the target deformation corresponding to the target blade geometry, realizing a forward design for the blade to generate the target deformation. Simultaneously, the method of the exemplary embodiments of this disclosure can also design target activation information for different blade geometries to drive the blade to actively deform under different engine operating conditions to obtain the corresponding blade geometry, thereby enabling the engine to have good aerodynamic performance over a wide speed range and improving working efficiency. Attached Figure Description

[0018] Further details, features, and advantages of this disclosure are disclosed in the following description of exemplary embodiments in conjunction with the accompanying drawings, in which:

[0019] Figure 1 A flowchart illustrating a method for determining blade deformation simulation parameters according to an exemplary embodiment of the present disclosure is shown;

[0020] Figure 2A cross-sectional schematic diagram of a deformable blade model according to an exemplary embodiment of the present disclosure is shown;

[0021] Figure 3 An optimization flowchart of activation information for a deformation-driven model according to an exemplary embodiment of the present disclosure is shown;

[0022] Figure 4 A flowchart illustrating the construction process of the proxy model of an exemplary embodiment of this disclosure is shown;

[0023] Figure 5 An optimized flowchart of the acquisition function of an exemplary embodiment of this disclosure is shown;

[0024] Figure 6 A schematic block diagram of a device for determining blade deformation simulation parameters according to an exemplary embodiment of the present disclosure is shown.

[0025] Figure 7 A schematic block diagram of a chip according to an exemplary embodiment of the present disclosure is shown;

[0026] Figure 8 A structural block diagram of an exemplary electronic device that can be used to implement embodiments of the present disclosure is shown. Detailed Implementation

[0027] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0028] It should be understood that the steps described in the method embodiments of this disclosure may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of this disclosure is not limited in this respect.

[0029] The term "comprising" and its variations as used herein are open-ended, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Definitions of other terms will be given in the description below. It should be noted that the concepts of "first", "second", etc., used in this disclosure are only used to distinguish different devices, modules, or units, and are not intended to limit the order of functions performed by these devices, modules, or units or their interdependencies.

[0030] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0031] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0032] Before introducing the embodiments of this disclosure, the relevant terms involved in the embodiments of this disclosure are first defined as follows:

[0033] Finite Element Analysis (FEA) is a method of simulating real physical systems (geometry and load conditions) using mathematical approximations. It uses simple, interacting elements (i.e., units) to approximate an infinite number of unknowns in a real system with a finite number of unknowns. FEA offers high computational accuracy and can adapt to various complex shapes, making it an effective engineering analysis tool. Commonly used FEA software includes ANSYS, ABAQUS, and MSC.

[0034] Latin hypercube sampling (LHS) is a method for approximating random sampling from a multivariate parameter distribution. It belongs to stratified sampling techniques and is commonly used in computer experiments or Monte Carlo integration. By maximizing the stratification of each marginal distribution, LHS ensures full coverage of the range of each variable, achieving efficient sampling within the distribution interval of the variable.

[0035] Bayesian optimization is a black-box optimization algorithm used to solve the extremum problem of a function whose expression is unknown. It predicts the probability distribution of the function value at any point based on the function values ​​at a set of sampled points, achieved through Gaussian process regression. A sampling function is constructed based on the Gaussian process regression results to measure the extent to which each point is worth exploring. The extremum of the sampling function is then solved to determine the next sampling point. Finally, the extrema of this set of sampled points are returned as the extrema of the function.

[0036] Acquisition function: is the expectation of the posterior function, used as a heuristic method to select the next extreme value evaluation point.

[0037] In related technologies, conventional techniques such as adjustable stator blades are insufficient to support the demands of future aero-engines, such as high flow rate across the entire speed range and simultaneous adjustment of bypass ratio / pressure ratio. To address this bottleneck, the concept of a controllable deformable fan has been proposed. This concept fundamentally overcomes the constraint of non-adjustable blade aerodynamic geometry, enabling decoupled adjustment of flow capacity and pressure capacity, further improving the limit of aerodynamic matching adjustment. Current research has demonstrated the structural feasibility of deformable blades using piezoelectric materials to drive blade deformation. However, conventional finite element analysis methods are only used to evaluate the blade deformation of controllable deformable fans, but are not suitable for achieving the forward design of controllable deformable fans that yields the desired deformation.

[0038] To address the aforementioned issues, this exemplary embodiment provides a method for determining blade deformation simulation parameters. This method not only enables forward design of the blade to generate target deformation, but also allows the blade to actively generate deformation under different engine operating conditions by designing target activation information for different blade geometries, thereby obtaining the corresponding blade geometry. This enables the engine to have good aerodynamic performance over a wide speed range.

[0039] The method for determining blade deformation simulation parameters according to an exemplary embodiment of this disclosure can be executed by an electronic device equipped with one or more modeling software programs, which can be stored in a computer-readable storage medium. The modeling software includes, but is not limited to, one or more of the following: UG, CAD, BIM (Building Information Modeling), SPCS, PKPM-PC, Revit, Navisworks, Bentley Navigator, Tekla Structures, and ArchiCAD. The electronic device can include, but is not limited to, desktop computers, laptops, tablets, etc.

[0040] Figure 1 A flowchart illustrating a method for determining blade deformation simulation parameters according to an exemplary embodiment of this disclosure is shown. Figure 1 As shown, the method for determining blade deformation simulation parameters in an exemplary embodiment of this disclosure may include:

[0041] Step 101: Generate a deformation-driven model of the blade model based on the blade modeling information. The deformation-driven model includes multiple sub-models, and each sub-model is used to drive the blade model to produce deformation when it is in an active state.

[0042] The aforementioned blade modeling information can be the three-dimensional modeling information of the blade model corresponding to the blade, and the aforementioned deformation driving model can be the three-dimensional model that drives the blade model to deform. That is, in practical applications, the deformation driving model is a three-dimensional model of the piezoelectric material arranged on the blade surface, and the blade can be a fan blade located at the front of an engine. An exemplary embodiment of this disclosure can arrange piezoelectric material on the blade surface and use the piezoelectric material to drive the blade to deform. For example, the piezoelectric material can be a piezoelectric fiber composite material (MFC). This piezoelectric material can deform under the action of an external electric field, thereby causing the blade to deform. A blade with piezoelectric material arranged on its surface is called a deformable blade.

[0043] Figure 2 A schematic cross-sectional view of a deformable blade model according to an exemplary embodiment of this disclosure is shown. Figure 2 As shown, in the cross-section of the deformable blade model, the deformable blade model 200 includes a blade model 201 and a deformation driving model 202. The deformation driving model 202 is located on the surface of the blade model 201. Under the action of an external electric field, the deformation driving model 202 can drive the blade model 201 to deform.

[0044] The aforementioned deformation-driven model can include multiple sub-models, each of which can be a blade deformation-driven unit. The model parameters of each sub-model can include at least its dimensional parameters, and may also include its characteristic parameters. The characteristic parameters of a sub-model can be functional parameters defining whether it can be used to drive the blade model to produce deformation. Based on these characteristic parameters, the state parameters (whether it is in an active state) and the driving parameters of an active sub-model can be determined. When a sub-model is in an active state, it is used to drive the blade model to produce deformation.

[0045] For example, an exemplary embodiment of this disclosure can utilize the ANSYS birth and death element method to select whether to activate a sub-model during the calculation process by defining the "birth" or "death" state of a sub-model. When the sub-model is in a "birth" state, it is in an active state and can be used to drive the corresponding part of the blade model to produce deformation; when the sub-model is in a "death" state, it is in an inactive state and cannot drive the corresponding part of the blade model to produce deformation.

[0046] As can be seen, the exemplary embodiments of this disclosure can construct a deformation-driven model that matches the blade model, and drive the blade model to deform through the deformation-driven model.

[0047] Step 102: Optimize the activation information of the deformation-driven model using the Bayesian optimization algorithm to determine the target activation information of the deformation-driven model corresponding to the target blade geometry. The activation information of the deformation-driven model includes the activation parameters of at least one sub-model that is in an active state.

[0048] The activation information of the aforementioned deformation-driven model can be used to drive the blade model to generate blade deformation geometry corresponding to the activation information. The activation information of the deformation-driven model can include activation parameters of at least one sub-model that is in an active state. When a sub-model is active, piezoelectric material is arranged at the corresponding position on the blade surface, and the piezoelectric material drives the blade to deform. The number of active sub-models can be one or more. When there is only one active sub-model, it can drive the entire blade model to deform; when there are multiple active sub-models, they can collectively drive the blade model to produce deformations with different distributions.

[0049] In the method of the exemplary embodiments of this disclosure, the activation parameters of each sub-model in the active state may include: distribution parameters of the active sub-model and driving parameters of the active sub-model. The distribution parameters of the active sub-model can be used to determine the position of the active sub-model in the deformation driving model. The distribution parameters of the active sub-model can be represented by three-dimensional coordinates or by model number. The driving parameters of the active sub-model can be the parameters that enable the sub-model to drive the blade model to deform. Therefore, the exemplary embodiments of this disclosure can define the activation parameters of each sub-model in the active state by the distribution parameters and driving parameters of the active sub-model, thereby determining the activation information of the deformation driving model based on the activation parameters of multiple active sub-models.

[0050] For example, regardless of whether a sub-model is in an active or inactive state, the activation parameter of each sub-model can be represented by a feature vector pr, where pr = (state, voltage), and r is the distribution parameter of the sub-model represented by its model number, where r is an integer greater than or equal to 0 and less than or equal to N, and N represents the total number of sub-models included in the deformation driving model; state represents the life / death state of the r-th sub-model, taking a value of 0 or 1. When state is 0, the sub-model is inactive; when state is 1, the sub-model is active; voltage represents the driving parameter of the r-th sub-model. Therefore, when the state of the r-th sub-model is 1, the voltage of the r-th sub-model can be adjusted to drive different degrees of deformation in the corresponding parts of the blade model. It is evident that the method of this exemplary embodiment can drive different degrees of deformation in the corresponding parts of the blade model by adjusting the life / death state and driving parameters of each sub-model, which is efficient and simple to operate, thereby improving the efficiency of obtaining blade deformation simulation parameters.

[0051] Bayesian optimization algorithms offer several advantages: parameter updates utilize a Gaussian process, incorporating prior parameter information and continuously updating the prior; they involve fewer iterations and are computationally faster; and they remain robust to non-convex problems. Therefore, the exemplary method of this disclosure, when optimizing the activation information of a deformation-driven model using Bayesian optimization to determine the target activation information of the deformation-driven model corresponding to the target blade geometry, can reduce overexpansion and overexploration to a certain extent, accelerate optimization efficiency, reduce computational resource waste, and achieve better optimization results.

[0052] Step 103: Determine the target activation information of the deformation driving model corresponding to the target blade geometry as the blade deformation simulation parameter. In this exemplary embodiment, the target activation information of the deformation driving model corresponding to the target blade geometry is determined, and this target activation information is used as the blade deformation simulation parameter. The blade model is then driven to generate the target deformation corresponding to the target blade geometry using this blade deformation simulation parameter. Based on this, when the blade deformation simulation parameter is applied to an engine blade, it can drive the blade to generate the target deformation.

[0053] As can be seen, the method of the exemplary embodiment of this disclosure can determine the target activation information of the blade in the target blade geometry through the deformation driving model. The target activation information can be used as the blade deformation simulation parameter to drive the blade to achieve the target deformation corresponding to the target blade geometry, thus realizing the forward design of the blade to generate target deformation. At the same time, the method of the exemplary embodiment of this disclosure can also drive the blade to actively generate deformation under different engine operating conditions by designing the target activation information of different blade geometries to obtain the corresponding blade geometry, thereby enabling the engine to have good aerodynamic performance in a wide speed range and improve working efficiency.

[0054] In an alternative embodiment, the objective function of the Bayesian optimization algorithm in this disclosure may include: a similarity parameter between the blade geometry generated by the deformation-driven model and the target blade geometry.

[0055] The aforementioned similarity parameters can be used to evaluate the similarity between the blade geometry generated by the deformation-driven model and the target blade geometry. The higher the similarity between the blade geometry generated by the deformation-driven model and the target blade geometry, the higher the matching degree between the activation information of the deformation-driven model and the target activation information of the deformation-driven model that generates the target blade geometry. In this case, the blade geometry generated by the blade model driven by the activation information of the deformation-driven model is closer to the target blade geometry.

[0056] For example, in the method of the exemplary embodiments of this disclosure, the above similarity parameters can satisfy:

[0057]

[0058] Where F is the sum of the Euclidean distances of the arcs in each blade height section between the blade geometry generated by the deformation-driven model and the target blade geometry. Let be the three-dimensional coordinates of the k-th discrete point on the i-th blade height section of the target blade geometry. The coordinates of the kth discrete point on the i-th blade height section of the blade geometry generated by the deformation-driven model are given. m represents the number of discrete points on each blade height section, and n represents the number of blade height sections.

[0059] Understandably, when the similarity parameter satisfies the sum of the Euclidean distances of the arcs in each blade height section between the blade geometry generated by the deformation-driven model and the target blade geometry, the similarity between the two geometrys can be evaluated based on the magnitude of this sum. A smaller sum of the Euclidean distances in each blade height section indicates a higher similarity between the two geometrys.

[0060] As can be seen, the method of the exemplary embodiment of this disclosure can determine the target activation information of the deformation-driven model based on the optimization objective function of the Bayesian optimization algorithm, so that the blade model generates the target blade geometry under the drive of the target activation information of the deformation-driven model.

[0061] In one alternative approach, Figure 3 A flowchart illustrating the optimization of activation information for a deformation-driven model according to an exemplary embodiment of this disclosure is shown. Figure 3 As shown in the exemplary embodiment of this disclosure, optimizing the activation information of the deformation-driven model using the Bayesian optimization algorithm may include:

[0062] Step 301: Construct a proxy model based on initial samples of multiple blade geometry models. Each initial sample includes activation information of the deformation-driven model and similarity parameters between the blade geometry corresponding to the activation information and the target blade geometry.

[0063] The aforementioned surrogate model can include a prior probability model and an observation model. The prior probability model can be a probability distribution expressing the confidence level in an uncertain quantity before considering certain factors; it can be a hypothetical prior. The observation model can describe the mechanism of observed data generation, i.e., the likelihood distribution. The surrogate model is used to substitute for an unknown objective function. Starting from a hypothetical prior, it iteratively increases the amount of information and corrects the prior to obtain a more accurate surrogate model.

[0064] The activation information of the aforementioned deformation-driven model can be randomly determined. For example, an exemplary embodiment of this disclosure can determine the feature vectors of the activation parameters of N sub-models included in the deformation-driven model. Once the feature vectors of the activation parameters of all N sub-models are determined, the activation information of the deformation-driven model of the initial sample can be obtained. If the number of blade geometric models is M, the feature vector of the activation information of the deformation-driven model of the j-th blade geometric model can be expressed as: P j =(p1,p2,…,p N-1 ,p N ), j is an integer greater than or equal to 0 and less than or equal to M, p1 is the feature vector of the activation parameters of the first sub-model included in the deformation driving model of the j-th blade geometry model, p2 is the feature vector of the activation parameters of the second sub-model included in the deformation driving model of the j-th blade geometry model, p N-1 The deformation-driving model of the j-th blade geometry includes the feature vector of the activation parameters of the (N-1)-th sub-model, p N The deformation-driven model of the j-th blade geometry includes the eigenvectors of the activation parameters of the N-th sub-model.

[0065] Based on this, the method of the exemplary embodiments of this disclosure can determine the similarity parameters between the blade geometry and the target blade geometry corresponding to the activation information of M blade geometric models, and construct a surrogate model using the M blade geometric models as initial samples.

[0066] Step 302: Based on the surrogate model, solve for the maximum value of the acquisition function, determine the maximum similarity parameter between the blade geometry and the target blade geometry, and the activation information of the deformation-driven model corresponding to the maximum similarity parameter.

[0067] The aforementioned acquisition function is a heuristic method for selecting the next optimal evaluation point. It determines an optimal solution under the current sample exploration, namely, the maximum similarity parameter between the blade geometry and the target blade geometry. It should be understood that this optimal solution can be a global optimum or a local optimum, depending on the specific application scenario; no specific limitation is made here.

[0068] Therefore, the method of the exemplary embodiments of this disclosure can solve for the maximum value of the acquisition function based on the surrogate model, determine the maximum similarity parameter between the blade geometry and the target blade geometry, and then determine the activation information of the deformation driving model corresponding to the maximum similarity parameter by using the maximum similarity parameter between the blade geometry and the target blade geometry.

[0069] Step 303: Determine whether the maximum similarity parameter between the blade geometry determined based on the optimization objective function and the target blade geometry satisfies the iteration termination condition.

[0070] If the maximum similarity parameter between the blade geometry determined based on the optimization objective function and the target blade geometry satisfies the iteration termination condition, it indicates that the maximum similarity parameter is the optimal solution under the current sample exploration. In this case, proceed to step 304. If the maximum similarity parameter between the blade geometry determined based on the optimization objective function and the target blade geometry does not satisfy the iteration termination condition, it indicates that the maximum similarity parameter is not the optimal solution under the current sample exploration. In this case, proceed to step 305.

[0071] For example, the iteration termination condition may include the maximum similarity parameter between the blade geometry and the target blade geometry being less than or equal to a first preset threshold. The first preset threshold may be an empirical value determined based on experience, or a custom value determined based on actual conditions.

[0072] At this point, the method of the exemplary embodiment of this disclosure can determine whether to continue the next iteration based on whether the maximum similarity parameter between the blade geometry and the target blade geometry obtained in the current iteration satisfies the iteration termination condition. When the maximum similarity parameter between the blade geometry and the target blade geometry satisfies the iteration termination condition, the iteration is terminated; when the maximum similarity parameter between the blade geometry and the target blade geometry does not satisfy the iteration termination condition, the next iteration is continued.

[0073] For example, the iteration termination condition may include the difference between the maximum similarity parameter of the blade geometry obtained in two adjacent iterations and the target blade geometry being less than or equal to a second preset threshold. The second preset threshold may be an empirical value determined based on experience, or a custom value determined based on actual conditions.

[0074] At this point, the method of this exemplary embodiment can determine whether to continue the next iteration based on whether the difference between the maximum similarity parameter between the blade geometry and the target blade geometry obtained in the current iteration and the maximum similarity parameter between the blade geometry and the target blade geometry obtained in the previous iteration satisfies the iteration termination condition. The iteration terminates when the difference between the maximum similarity parameter between the blade geometry and the target blade geometry obtained in the current iteration and the maximum similarity parameter between the blade geometry and the target blade geometry obtained in the previous iteration satisfies the iteration termination condition; the next iteration continues when the difference between the maximum similarity parameter between the blade geometry and the target blade geometry obtained in the current iteration and the maximum similarity parameter between the blade geometry and the target blade geometry obtained in the previous iteration does not satisfy the iteration termination condition.

[0075] Step 304: If the maximum similarity parameter between the blade geometry determined based on the optimization objective function and the target blade geometry satisfies the iteration termination condition, the target activation information of the deformation-driven model is determined as the activation information of the deformation-driven model corresponding to the maximum similarity parameter.

[0076] If the maximum similarity parameter between the blade geometry determined based on the optimization objective function and the target blade geometry satisfies the iteration termination condition, it indicates that the maximum similarity parameter is the optimal solution under the current sample exploration. At this time, the target activation information of the deformation-driven model is determined to be the activation information of the deformation-driven model corresponding to the maximum similarity parameter.

[0077] Step 305: If the maximum similarity parameter between the blade geometry and the target blade geometry determined based on the optimization objective function does not meet the iteration termination condition, update the initial sample based on the maximum similarity parameter between the blade geometry and the target blade geometry and the activation information of the deformation-driven model corresponding to the maximum similarity parameter.

[0078] If the maximum similarity parameter between the blade geometry determined based on the optimization objective function and the target blade geometry does not meet the iteration termination condition, it means that the maximum similarity parameter is not the optimal solution under the current sample exploration. At this time, the maximum similarity parameter between the blade geometry and the target blade geometry and the activation information of the deformation-driven model corresponding to the maximum similarity parameter are taken as a new blade geometry model. The initial sample in step 301 is updated based on the new blade geometry model, and then the next iteration is executed.

[0079] As can be seen, in the method of the exemplary embodiments of this disclosure, the surrogate model can obtain a new blade geometry model after each iteration, continuously increase the amount of information and correct the prior through the new blade geometry model to obtain a more accurate surrogate model, so that in the last iteration, the target activation information of the deformation driving model is determined based on the acquisition function as the activation information of the deformation driving model corresponding to the maximized similarity parameter. When the activation information of the deformation driving model corresponding to the maximized similarity parameter is applied to the deformation driving model, the blade model can be driven to generate the target deformation, thus realizing the forward design of the blade model generating the target deformation.

[0080] In one alternative approach, Figure 4 A flowchart illustrating the construction process of an agent model according to an exemplary embodiment of this disclosure is shown. Figure 4 As shown in the exemplary embodiments of this disclosure, constructing a proxy model based on initial samples of multiple blade geometric models may include:

[0081] Step 401: Determine the activation information of multiple deformation-driven models based on the hierarchical sampling method.

[0082] The stratified sampling method described above can sample from the distribution range of independent variables, ensuring that the obtained blade geometry model fully covers the range of each variable, thereby improving the sampling efficiency of the blade geometry model. This stratified sampling method can be Latin hypercube sampling (LHS), a method for approximating random sampling from a multivariate parameter distribution, commonly used in computer experiments or Monte Carlo integration.

[0083] For example, for the j-th blade geometric model, an exemplary embodiment of this disclosure can use the feature vectors of the activation parameters of the sub-models as the independent variable space, and determine the feature vectors of the activation parameters of the N sub-models included in the deformation-driven model based on the Latin hypercube sampling method, thereby determining the feature vector of the activation information of the deformation-driven model of the blade geometric model as P. j =(p1,p2,…,p N-1 ,p N Based on this, an exemplary embodiment of the present disclosure can determine the feature vector of the activation information of the deformation-driven model of M blade geometric models.

[0084] Step 402: Determine the blade geometry corresponding to each activation information based on the finite element method. The aforementioned finite element method can be achieved by using finite element analysis software to solve for the blade geometry corresponding to the activation information of the deformation-driven model.

[0085] For example, an exemplary embodiment of this disclosure can input the feature vector of the activation information of the deformation driving model of each blade geometry model into finite element analysis software. After solving the problem, the finite element analysis software outputs the blade geometry corresponding to the feature vector of the activation information of the deformation driving model of the blade geometry model. Based on this, an exemplary embodiment of this disclosure can determine the blade geometry corresponding to the feature vector of the activation information of the deformation driving model of M corresponding blade geometry models.

[0086] Step 403: Construct a surrogate model using the activation information of multiple deformation-driven models and the similarity parameters between the corresponding blade geometry and the target blade geometry as initial samples.

[0087] An exemplary embodiment of this disclosure can, after determining the blade geometry corresponding to the feature vectors of the activation information of the deformation-driven models of the M corresponding blade geometric models in step 402, determine the similarity parameters between the blade geometry corresponding to the feature vectors of the activation information of the deformation-driven models of the M corresponding blade geometric models and the target blade geometry based on the optimization objective function. Then, the feature vectors of the activation information of the deformation-driven models of the M blade geometric models determined in step 401 and the similarity parameters between the blade geometry corresponding to the feature vectors of the activation information of the deformation-driven models of the M corresponding blade geometric models and the target blade geometry are used as M blade geometric models. A surrogate model is constructed using these M blade geometric models as initial samples.

[0088] As can be seen, the method of the exemplary embodiment of this disclosure uses a stratified sampling method to determine multiple blade geometric models as initial samples to construct a surrogate model. This can ensure that each blade geometric model included in the surrogate model has full coverage of every variable range, thereby improving the sampling efficiency of the blade geometric model and ensuring the accuracy of the optimal solution determined based on multiple blade geometric models.

[0089] In one alternative approach, Figure 5 An optimized flowchart of the acquisition function of an exemplary embodiment of this disclosure is shown. Figure 5 As shown in the exemplary embodiment of this disclosure, determining the maximum similarity parameter between the blade geometry and the target blade geometry, and the activation information of the deformation-driven model corresponding to the maximum similarity parameter, based on solving the maximum value of the acquisition function using a surrogate model, may include:

[0090] Step 501: Determine the predicted similarity parameters between the blade geometry and the target blade geometry based on the surrogate model.

[0091] The predicted similarity parameters between the blade geometry and the target blade geometry can be the predicted values ​​of the surrogate model. In essence, they are the posterior mean of the surrogate model in the exploration space determined by the current sample. Therefore, the predicted similarity parameters are not necessarily the optimal solution for the similarity parameters between the blade geometry and the target blade geometry in the exploration space determined by the current sample. Further exploration is needed to determine the optimal solution for the similarity parameters between the blade geometry and the target blade geometry in the exploration space determined by the current sample.

[0092] Step 502: Determine the maximum value of the acquisition function based on the predicted similarity parameter between the blade geometry and the target blade geometry. The maximum similarity parameter between the blade geometry and the target blade geometry is the maximum value of the acquisition function.

[0093] The aforementioned acquisition function can be used to determine the optimal solution for the similarity parameters between the blade geometry and the target blade geometry within the exploration space defined by the current sample. This acquisition function can be an upper confidence interval acquisition function, which may include the predicted value of the surrogate model and the posterior probability determined by the surrogate model distribution. The role of the posterior probability determined by the surrogate model distribution is to increase the explorability of the optimal solution within the exploration space defined by the current sample. In this case, an exemplary embodiment of this disclosure can determine the optimal solution for the similarity parameters between the blade geometry and the target blade geometry based on the acquisition function.

[0094] For example, in the exemplary embodiments of this disclosure, the acquisition function A(P) of the Bayesian optimization algorithm can satisfy:

[0095]

[0096] in, σ(P) is the similarity parameter between the blade geometry and the target blade geometry; σ(P) is a Gaussian distribution function with zero mean, and λ represents the weight.

[0097] The aforementioned data acquisition functions are essentially upper confidence interval data acquisition functions (such as the UCB data acquisition function), characterized by clear meaning, good stability, and suitability for most optimization tasks. The UCB data acquisition function is essentially the upper bound of a certain confidence interval of the target distribution. Different values ​​of λ correspond to different confidence intervals, where λ is a user-defined parameter greater than zero. For example, when λ = 1.96, the UCB function represents the upper bound of the 95% confidence interval of the target distribution.

[0098] λσ(P) represents the posterior probability determined by the surrogate model distribution, σ(P) represents the standard deviation of the posterior mean of the surrogate model, and λ, when representing the weight, can be used to balance the posterior mean and the standard deviation of the surrogate model. The magnitude of λ represents the trade-off between two objectives in the optimization process: coarsely exploring global extrema and precisely developing local extrema. When a smaller value is taken, the optimization strategy leans towards coarse exploration, with stronger global optimization ability; when a larger value is taken, the optimization strategy leans towards precise development, with stronger local optimization ability.

[0099] For example, an exemplary embodiment of this disclosure can refit a curve with the activation information of the deformation-driven model as the independent variable and the maximum similarity parameter between the blade geometry and the target blade geometry as the dependent variable based on the acquisition function. In this case, the maximum similarity parameter between the blade geometry and the target blade geometry can be determined to be the maximum value of the acquisition function.

[0100] Step 503: Determine the activation information of the deformation-driven model corresponding to the maximum similarity parameter based on the maximum similarity parameter between the blade geometry and the target blade geometry.

[0101] For example, an exemplary embodiment of this disclosure can refit a curve with the activation information of the deformation-driven model as the independent variable and the maximum similarity parameter between the blade geometry and the target blade geometry as the dependent variable based on the acquisition function, and determine the activation information of the deformation-driven model corresponding to the maximum similarity parameter from the curve as the activation information of the deformation-driven model corresponding to the maximum value of the acquisition function.

[0102] As can be seen, the method of the exemplary embodiment of this disclosure solves the maximization of the acquisition function based on the surrogate model in each iteration, so that the acquisition function can determine the maximum similarity parameter between the blade geometry and the target blade geometry under the current sample exploration. The newly added blade geometry model determined based on the maximum similarity parameter between the blade geometry and the target blade geometry can continuously increase the amount of information iteratively, thereby obtaining target activation information that can drive the blade model to produce target deformation.

[0103] One or more technical solutions provided in the exemplary embodiments of this disclosure can generate a deformation-driven model of the blade model based on blade modeling information. The deformation-driven model includes multiple sub-models, each of which, when activated, can drive the blade model to deform. Based on this, the exemplary embodiments of this disclosure can utilize a Bayesian optimization algorithm to optimize the activation information of the deformation-driven model, determining the target activation information of the deformation-driven model corresponding to the target blade geometry. The activation information of the deformation-driven model includes activation parameters of at least one sub-model in an activated state. Furthermore, the target activation information of the deformation-driven model corresponding to the target blade geometry is determined as the blade deformation simulation parameter. Therefore, the method of the exemplary embodiments of this disclosure can determine the target activation information of the blade in the target blade geometry through the deformation-driven model. Using this target activation information as the blade deformation simulation parameter can drive the blade to achieve the target deformation corresponding to the target blade geometry, realizing a forward design for the blade to generate the target deformation. Simultaneously, the method of the exemplary embodiments of this disclosure can also design target activation information for different blade geometries to drive the blade to actively deform under different engine operating conditions to obtain the corresponding blade geometry, thereby enabling the engine to have good aerodynamic performance over a wide speed range and improving working efficiency.

[0104] The foregoing mainly describes the solutions provided by the embodiments of this disclosure. It is understood that, in order to achieve the above functions, the electronic device includes hardware structures and / or software modules corresponding to the execution of each function. Those skilled in the art should readily recognize that, based on the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein, this disclosure can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this disclosure.

[0105] This disclosure embodiment can divide the electronic device into functional units according to the above method example. For example, each function can be divided into a separate functional module, or two or more functions can be integrated into one processing module. The integrated module can be implemented in hardware or as a software functional module. It should be noted that the module division in this disclosure embodiment is illustrative and only represents one logical functional division; other division methods may be used in actual implementation.

[0106] In the case of dividing each functional module according to its corresponding functions, an exemplary embodiment of this disclosure provides a device for determining blade deformation simulation parameters. This device for determining blade deformation simulation parameters can be an electronic device or a chip applied to an electronic device. Figure 6A schematic block diagram of a device for determining blade deformation simulation parameters according to an exemplary embodiment of the present disclosure is shown. Figure 6 As shown, the device 600 includes:

[0107] The generation module 601 is used to generate a deformation driving model of the blade model based on the blade modeling information. The deformation driving model includes multiple sub-models, and each sub-model is used to drive the blade model to produce deformation when it is in an active state.

[0108] The determination module 602 is used to optimize the activation information of the deformation driving model using a Bayesian optimization algorithm, and determine the target activation information of the deformation driving model corresponding to the target blade geometry. The activation information of the deformation driving model includes the activation parameters of at least one sub-model that is in an active state.

[0109] The determining module 602 is also used to determine the target activation information of the deformation driving model corresponding to the target blade geometry as blade deformation simulation parameters.

[0110] As one possible implementation, the activation parameters of each sub-model in the active state include: the distribution parameters of the active sub-model and the driving parameters of the active sub-model.

[0111] As one possible implementation, the objective function of the Bayesian optimization algorithm includes: the similarity parameter between the blade geometry generated by the deformation-driven model and the target blade geometry.

[0112] As one possible implementation, the similarity parameter satisfies:

[0113]

[0114] Where F(P) is the sum of the Euclidean distances of the arcs in each blade height section between the blade geometry generated by the deformation-driven model and the target blade geometry. Let be the three-dimensional coordinates of the k-th discrete point on the i-th blade height section of the target blade geometry. The coordinates of the kth discrete point on the i-th blade height section of the blade geometry generated by the deformation-driven model are given. m represents the number of discrete points on each blade height section, and n represents the number of blade height sections.

[0115] As one possible implementation, the determining module 602 is also used to construct a surrogate model based on initial samples of multiple blade geometry models. Each initial sample includes activation information of the deformation-driven model and similarity parameters between the blade geometry and the target blade geometry corresponding to the activation information. Based on the surrogate model, the maximum value of the acquisition function is solved to determine the maximum similarity parameter between the blade geometry and the target blade geometry and the activation information of the deformation-driven model corresponding to the maximum similarity parameter. If the maximum similarity parameter between the blade geometry and the target blade geometry determined based on the optimization objective function satisfies the iteration termination condition, the target activation information of the deformation-driven model is determined to be the activation information of the deformation-driven model corresponding to the maximum similarity parameter. Otherwise, the initial samples are updated based on the maximum similarity parameter between the blade geometry and the target blade geometry and the activation information of the deformation-driven model corresponding to the maximum similarity parameter.

[0116] As one possible implementation, the determination module 602 is also used to determine the activation information of multiple deformation-driven models based on a hierarchical sampling method; determine the blade geometry corresponding to each activation information based on the finite element method; and construct a surrogate model using the activation information of multiple deformation-driven models and the similarity parameters between the blade geometry corresponding to the activation information and the target blade geometry as initial samples.

[0117] As one possible implementation, the determining module 602 is also used to determine the predicted similarity parameters between the blade geometry and the target blade geometry based on the surrogate model; determine the maximum value of the acquisition function based on the predicted similarity parameters between the blade geometry and the target blade geometry, wherein the maximum similarity parameter between the blade geometry and the target blade geometry is the maximum value of the acquisition function; and determine the activation information of the deformation driving model corresponding to the maximum similarity parameter based on the maximum similarity parameter between the blade geometry and the target blade geometry.

[0118] As one possible implementation, the acquisition function is an upper confidence interval acquisition function, which includes the predicted value of the surrogate model and the posterior probability determined by the distribution of the surrogate model.

[0119] As one possible implementation, the iteration termination condition includes the maximum similarity parameter between the blade geometry and the target blade geometry being less than or equal to a first preset threshold.

[0120] The iteration termination condition includes the difference between the maximum similarity parameter of the blade geometry obtained in two adjacent iterations and the target blade geometry being less than or equal to a second preset threshold.

[0121] As one possible implementation, the model parameters of each sub-model include at least the size parameters of the sub-model, and the model parameters of each sub-model also include the characteristic parameters of the sub-model.

[0122] As one possible implementation, the acquisition function A(P) of the Bayesian optimization algorithm satisfies:

[0123]

[0124] in, σ(P) is the similarity parameter between the blade geometry and the target blade geometry; σ(P) is a Gaussian distribution function with zero mean, and λ represents the weight.

[0125] Figure 7 A schematic block diagram of a chip according to an exemplary embodiment of this disclosure is shown. (As follows) Figure 7 As shown, the chip 700 includes one or more (including two) processors 701 and a communication interface 702. The communication interface 702 can support the server in performing the data transmission and reception steps in the above method, and the processor 701 can support the server in performing the data processing steps in the above method.

[0126] Optional, such as Figure 7 As shown, the chip 700 also includes a memory 703, which may include read-only memory and random access memory, and provides operation instructions and data to the processor. A portion of the memory may also include non-volatile random access memory (NVRAM).

[0127] In some implementations, such as Figure 7 As shown, processor 701 executes corresponding operations by calling operation instructions stored in memory (which may be stored in the operating system). Processor 701 controls the processing operations of any terminal device; processor can also be called a central processing unit (CPU). Memory 703 may include read-only memory and random access memory, and provides instructions and data to processor 701. A portion of memory 703 may also include NVRAM. For example, in applications, memory, communication interfaces, and other components are coupled together via a bus system, which may include, in addition to a data bus, a power bus, a control bus, and a status signal bus, etc. However, for clarity, in... Figure 7 The general designated all buses as Bus System 704.

[0128] The methods disclosed in the embodiments of this disclosure can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above methods can be completed by integrated logic circuits in the processor's hardware or by instructions in software form. The processor can be a general-purpose processor, a digital signal processor (DSP), an ASIC, a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this disclosure. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this disclosure can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. This storage medium is located in memory; the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above methods.

[0129] Exemplary embodiments of this disclosure also provide an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor. The memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to cause the electronic device to perform a method according to an embodiment of this disclosure.

[0130] Exemplary embodiments of this disclosure also provide a non-transitory computer-readable storage medium storing a computer program, wherein the computer program, when executed by a computer's processor, is used to cause the computer to perform a method according to embodiments of this disclosure.

[0131] Exemplary embodiments of this disclosure also provide a computer program product, including a computer program, wherein, when executed by a processor of a computer, the computer program is used to cause the computer to perform a method according to an embodiment of this disclosure.

[0132] refer to Figure 8The present invention describes a structural block diagram of an electronic device 800 that can serve as a server or client of the present disclosure, which is an example of a hardware device that can be applied to various aspects of the present disclosure. The electronic device is intended to represent various forms of digital electronic computer devices, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0133] like Figure 8 As shown, the electronic device 800 includes a computing unit 801, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 802 or a computer program loaded from a storage unit 808 into a random access memory (RAM) 803. The RAM 803 may also store various programs and data required for the operation of the device 800. The computing unit 801, ROM 802, and RAM 803 are interconnected via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0134] Multiple components in electronic device 800 are connected to I / O interface 805, including: input unit 806, output unit 807, storage unit 808, and communication unit 809. Input unit 806 can be any type of device capable of inputting information to electronic device 800. Input unit 806 can receive input digital or character information and generate key signal inputs related to user settings and / or function control of electronic device. Output unit 807 can be any type of device capable of presenting information and may include, but is not limited to, a display, speaker, video / audio output terminal, vibrator, and / or printer. Storage unit 808 may include, but is not limited to, disks and optical discs. Communication unit 809 allows electronic device 800 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks, and may include, but is not limited to, modems, network cards, infrared communication devices, wireless communication transceivers, and / or chipsets, such as Bluetooth™ devices, WiFi devices, WiMax devices, cellular communication devices, and / or the like.

[0135] like Figure 8As shown, computing unit 801 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of computing unit 801 include, but are not limited to, central processing unit (CPU), graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Computing unit 801 performs the various methods and processes described above. For example, in some embodiments, the methods of exemplary embodiments of this disclosure can be implemented as computer software programs tangibly contained in a machine-readable medium, such as storage unit 808. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 800 via ROM 802 and / or communication unit 809. In some embodiments, computing unit 801 can be configured to perform methods by any other suitable means (e.g., by means of firmware).

[0136] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0137] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0138] As used in this disclosure, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, device, and / or apparatus (e.g., disk, optical disk, memory, programmable logic device (PLD)) for providing machine instructions and / or data to a programmable processor, including machine-readable media that receive machine instructions as machine-readable signals. The term "machine-readable signal" refers to any signal for providing machine instructions and / or data to a programmable processor.

[0139] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0140] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0141] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other.

[0142] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer programs or instructions. When the computer program or instructions are loaded and executed on a computer, the processes or functions described in the embodiments of this disclosure are performed, in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, a terminal, a user equipment, or other programmable device. The computer program or instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another. For example, the computer program or instructions can be transferred from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center integrating one or more available media. The available medium can be a magnetic medium, such as a floppy disk, hard disk, or magnetic tape; it can also be an optical medium, such as a digital video disc (DVD); or it can be a semiconductor medium, such as a solid-state drive (SSD).

[0143] Although this disclosure has been described in conjunction with specific features and embodiments, it will be apparent that various modifications and combinations can be made therein without departing from the spirit and scope of this disclosure. Accordingly, this specification and drawings are merely exemplary illustrations of the disclosure as defined by the appended claims and are to be considered as covering any and all modifications, variations, combinations, or equivalents within the scope of this disclosure. It is obvious that those skilled in the art can make various alterations and modifications to this disclosure without departing from its spirit and scope. Thus, this disclosure is also intended to include any such modifications and modifications that fall within the scope of the claims of this disclosure and their equivalents.

Claims

1. A method for determining parameters for simulating blade deformation, characterized in that, The method includes: A deformation-driven model for generating a blade model based on blade modeling information is provided. The deformation-driven model includes multiple sub-models, and each sub-model is used to drive the blade model to deform when it is in an active state. The activation information of the deformation-driven model is optimized using a Bayesian optimization algorithm to determine the target activation information of the deformation-driven model corresponding to the target blade geometry. The activation information of the deformation-driven model includes the activation parameters of at least one sub-model that is in an active state. The target activation information of the deformation driving model corresponding to the target blade geometry is determined as the blade deformation simulation parameters; The optimization of the activation information of the deformation-driven model using the Bayesian optimization algorithm includes: A proxy model is constructed based on initial samples of multiple blade geometry models. Each initial sample includes activation information of the deformation-driven model and similarity parameters between the blade geometry and the target blade geometry corresponding to the activation information. Based on the surrogate model, the maximum value of the acquisition function is solved, and the maximum similarity parameter between the blade geometry and the target blade geometry and the activation information of the deformation driving model corresponding to the maximum similarity parameter are determined. If the maximum similarity parameter between the blade geometry and the target blade geometry determined based on the optimization objective function satisfies the iteration termination condition, the target activation information of the deformation-driven model is determined to be the activation information of the deformation-driven model corresponding to the maximum similarity parameter. Otherwise, the initial sample is updated based on the maximum similarity parameter between the blade geometry and the target blade geometry and the activation information of the deformation-driven model corresponding to the maximum similarity parameter.

2. The method according to claim 1, characterized in that, The activation parameters of each sub-model in the active state include: the distribution parameters of the active sub-model and the driving parameters for activating the sub-model.

3. The method according to claim 1, characterized in that, The objective function of the Bayesian optimization algorithm includes: the similarity parameter between the blade geometry generated by the deformation-driven model and the target blade geometry.

4. The method according to claim 3, characterized in that, The similarity parameter satisfies: Where F(P) is the sum of the Euclidean distances of the arcs in each blade height section between the blade geometry generated by the deformation-driven model and the target blade geometry. Let be the three-dimensional coordinates of the k-th discrete point on the i-th blade height section of the target blade geometry. The coordinates of the kth discrete point on the i-th blade height section of the blade geometry generated by the deformation-driven model are given. m represents the number of discrete points on each blade height section, and n represents the number of blade height sections.

5. The method according to claim 1, characterized in that, The initial sample construction of the surrogate model based on multiple blade geometric models includes: Activation information of multiple deformation-driven models is determined based on a stratified sampling method; The blade geometry corresponding to each activation information is determined using the finite element method. Using the activation information of multiple deformation-driven models and the similarity parameters between the blade geometry and the target blade geometry corresponding to the activation information as initial samples, a surrogate model is constructed.

6. The method according to claim 1, characterized in that, The step of determining the maximum similarity parameter between the blade geometry and the target blade geometry, and the activation information of the deformation-driven model corresponding to the maximum similarity parameter, based on solving the acquisition function using the surrogate model, includes: The predicted similarity parameters between the blade geometry and the target blade geometry are determined based on the surrogate model. The maximum value of the acquisition function is determined based on the predicted similarity parameter between the blade geometry and the target blade geometry, and the maximum similarity parameter between the blade geometry and the target blade geometry is the maximum value of the acquisition function. The activation information of the deformation-driven model corresponding to the maximum similarity parameter between the blade geometry and the target blade geometry is determined.

7. The method according to claim 1, characterized in that, The acquisition function is an upper confidence interval acquisition function, which includes the predicted value of the surrogate model and the posterior probability determined by the distribution of the surrogate model.

8. The method according to claim 1, characterized in that, The iteration termination condition includes that the maximum similarity parameter between the blade geometry and the target blade geometry is less than or equal to a first preset threshold; or, The iteration termination condition includes the difference between the maximum similarity parameter of the blade geometry and the target blade geometry obtained in two adjacent iterations being less than or equal to a second preset threshold.

9. The method according to claim 1, characterized in that, The model parameters of each sub-model include at least the size parameters of the sub-model, and the model parameters of each sub-model also include the characteristic parameters of the sub-model.

10. The method according to claim 1, characterized in that, The acquisition function A(P) of the Bayesian optimization algorithm satisfies: in, This is the similarity parameter between the blade geometry and the target blade geometry; Let λ be a Gaussian distribution function with a mean of zero, and let λ represent the weights.

11. A device for determining blade deformation simulation parameters, characterized in that, The device includes: The generation module is used to generate a deformation driving model of the blade model based on the blade modeling information. The deformation driving model includes multiple sub-models, and each sub-model is used to drive the blade model to produce deformation when it is in an active state. The determination module is used to optimize the activation information of the deformation driving model using a Bayesian optimization algorithm, and to determine the target activation information of the deformation driving model corresponding to the target blade geometry. The activation information of the deformation driving model includes the activation parameters of at least one sub-model that is in an active state. The determining module is also used to determine the target activation information of the deformation driving model corresponding to the target blade geometry as blade deformation simulation parameters; The determining module is further configured to construct a surrogate model based on initial samples of multiple blade geometry models. Each initial sample includes activation information of the deformation-driven model and a similarity parameter between the blade geometry and the target blade geometry corresponding to the activation information. Based on the surrogate model, the maximum value of the acquisition function is calculated to determine the maximum similarity parameter between the blade geometry and the target blade geometry and the activation information of the deformation-driven model corresponding to the maximum similarity parameter. If the maximum similarity parameter between the blade geometry and the target blade geometry determined based on the optimization objective function satisfies the iteration termination condition, the target activation information of the deformation-driven model is determined to be the activation information of the deformation-driven model corresponding to the maximum similarity parameter. Otherwise, the initial sample is updated based on the maximum similarity parameter between the blade geometry and the target blade geometry and the activation information of the deformation-driven model corresponding to the maximum similarity parameter.

12. An electronic device, characterized in that, include: processor; as well as, Memory for stored programs; The program includes instructions that, when executed by the processor, cause the processor to perform the method according to any one of claims 1 to 10.

13. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to perform the method according to any one of claims 1 to 10.

Images