Modeling method and device for aero-engine multi-stage compressor assembly process parameter-component performance correction coefficient

Abstract

The application provides an aero-engine multi-stage compressor assembly process parameter-component performance correction coefficient modeling method and device, and relates to the technical field of data processing.The method comprises the following steps: obtaining assembly process parameters of high-pressure compressors of multiple aero-engines and corresponding component performance correction coefficients;characterizing the component performance correction coefficients as a weighted sum of performance correction coefficient components of each stage of the high-pressure compressors, so as to construct a layered mapping model of the high-pressure compressors;wherein the performance correction coefficient component of each stage is associated with the corresponding assembly process parameter through a preset mapping relationship;jointly identifying the weighted weight parameters in the layered mapping model and the mapping parameters contained in the preset mapping relationship associated with each stage, so as to determine the target mapping relationship from the assembly process parameters to the component performance correction coefficients.Through the method provided by the application, the overfitting problem of high-dimensional modeling under the condition of a small sample is effectively overcome.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method and apparatus for modeling assembly process parameters and component performance correction coefficients for multi-stage compressors in aero-engines. Background Technology

[0002] In the field of aero-engine design and manufacturing, accurately predicting the performance of high-pressure compressors is crucial for ensuring key indicators such as overall engine thrust and fuel consumption.

[0003] Traditional performance modeling primarily relies on component-level mechanistic models based on aerodynamic and thermodynamic principles. While these models possess clear physical meaning, they struggle to accurately characterize performance deviations caused by subtle variations in complex assembly process parameters. Existing technologies either attempt to directly fit data using purely data-driven methods, but in practical engineering, the available engine sample data for training is extremely limited, making such methods prone to overfitting, resulting in poor model generalization ability and low prediction reliability. Alternatively, simplified linear regression methods may be employed, but these fail to characterize the nonlinear and coupled effects of assembly parameters at each stage of a multi-stage compressor on overall performance, leading to insufficient prediction accuracy. Therefore, a mechanism- and data-driven approach is urgently needed to address these issues. Summary of the Invention

[0004] This invention provides a method and apparatus for modeling assembly process parameters and component performance correction coefficients for multi-stage compressors in aero-engines.

[0005] This invention provides a method for modeling assembly process parameters and component performance correction coefficients for multi-stage compressors in aero-engines, comprising the following steps: Obtain the assembly process parameters and corresponding component performance correction coefficients for the high-pressure compressors of multiple aero engines; The component performance correction coefficient is characterized as a weighted sum of the performance correction coefficient components of each stage of the high-pressure compressor, so as to construct a hierarchical mapping model of the high-pressure compressor; wherein, the performance correction coefficient component of each stage is associated with the corresponding assembly process parameters through a preset mapping relationship. The weighted parameters in the hierarchical mapping model and the mapping parameters contained in the preset mapping relationships associated with each level are jointly identified to determine the target mapping relationship from assembly process parameters to component performance correction coefficients.

[0006] According to the present invention, a method for modeling assembly process parameters and component performance correction coefficients for multi-stage compressors of aero-engines is provided. The method for obtaining the assembly process parameters and corresponding component performance correction coefficients of high-pressure compressors for multiple aero-engines includes: Obtain the assembly process parameters of the high-pressure compressors of the multiple aero engines; Based on the test performance data of the multiple aero engines, the component performance correction coefficient of the high-pressure compressor corresponding to each aero engine is determined.

[0007] According to the present invention, a method for modeling assembly process parameters and component performance correction coefficients of a multi-stage compressor for an aero-engine is provided. The method involves characterizing the component performance correction coefficients as a weighted sum of the performance correction coefficient components of each stage of the high-pressure compressor to construct a hierarchical mapping model of the high-pressure compressor. The method includes: For each stage of the high-pressure compressor, a preset mapping relationship is established between the performance correction coefficient component and the corresponding assembly process parameter. The overall component performance correction coefficients of the high-pressure compressor are established as a weighted sum model of the performance correction coefficient components of all stages; wherein, the coefficients in the weighted sum model are the weighting parameters.

[0008] According to the present invention, a method for modeling assembly process parameters and component performance correction coefficients of a multi-stage compressor for an aero-engine is provided, wherein the preset mapping relationship is a linear mapping relationship or a nonlinear mapping relationship based on Taylor expansion.

[0009] According to the present invention, a modeling method for assembly process parameters and component performance correction coefficients of a multi-stage compressor for an aero-engine is provided. The linear mapping relationship based on Taylor expansion is a first-order Taylor expansion model, which is used to characterize the linear relationship between the performance correction coefficient components and the corresponding assembly process parameters.

[0010] According to the present invention, a modeling method for assembly process parameters and component performance correction coefficients of a multi-stage compressor for an aero-engine is provided. The nonlinear mapping relationship based on Taylor expansion is a high-order Taylor expansion model, which is used to characterize the polynomial relationship between the performance correction coefficient components and the corresponding assembly process parameters, which includes high-order terms.

[0011] According to the present invention, a method for modeling assembly process parameters and component performance correction coefficients for a multi-stage compressor of an aero-engine includes the following steps: jointly identifying the weighted parameters in the hierarchical mapping model and the mapping parameters contained in the preset mapping relationships associated with each stage to determine the target mapping relationship from assembly process parameters to component performance correction coefficients. The weighted parameters in the hierarchical mapping model and the mapping parameters contained in the preset mapping relationships associated with each level are jointly identified using the maximum likelihood estimation algorithm to obtain the initial identification result. During the identification process, constraints are introduced to optimize the initial identification results in order to determine the target mapping relationship from assembly process parameters to component performance correction coefficients; wherein, the constraints include equality constraints, inequality constraints, or mixed constraints that simultaneously include equality constraints and inequality constraints.

[0012] According to the present invention, a method for modeling assembly process parameters and component performance correction coefficients of a multi-stage compressor for an aero-engine is provided, wherein the constraints are used to limit the proportional relationship or numerical range between the components of the performance correction coefficients at each stage.

[0013] This invention also provides a modeling device for assembly process parameters and component performance correction coefficients of multi-stage compressors for aero-engines, comprising the following modules: The acquisition module is used to acquire the assembly process parameters and corresponding component performance correction coefficients of the high-pressure compressors of multiple aero engines. A construction module is used to characterize the component performance correction coefficient as a weighted sum of the performance correction coefficient components of each stage of the high-pressure compressor, so as to construct a hierarchical mapping model of the high-pressure compressor; wherein, the performance correction coefficient component of each stage is associated with the corresponding assembly process parameters through a preset mapping relationship; The identification module is used to jointly identify the weighted parameters in the hierarchical mapping model and the mapping parameters contained in the preset mapping relationships associated at each level, so as to determine the target mapping relationship from assembly process parameters to component performance correction coefficients.

[0014] According to the present invention, a modeling device for assembly process parameters and component performance correction coefficients of a multi-stage compressor for an aero-engine is provided, wherein the acquisition module is specifically used for: Obtain the assembly process parameters of the high-pressure compressors of the multiple aero engines; Based on the test performance data of the multiple aero engines, the component performance correction coefficient of the high-pressure compressor corresponding to each aero engine is determined.

[0015] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the modeling method for assembly process parameters and component performance correction coefficients of a multi-stage compressor for aero-engines as described above.

[0016] The present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the modeling method for assembly process parameters and component performance correction coefficients of aero-engine multi-stage compressors as described above.

[0017] The present invention also provides a computer program product, including a computer program, which, when executed by a processor, implements the modeling method for assembly process parameters and component performance correction coefficients of a multi-stage compressor for aero-engines as described above.

[0018] This invention provides a method and apparatus for modeling assembly process parameters and component performance correction coefficients for multi-stage compressors in aero-engines. The method involves acquiring assembly process parameters and corresponding component performance correction coefficients for high-pressure compressors of multiple aero-engines; representing the component performance correction coefficients as a weighted sum of the performance correction coefficient components at each stage of the high-pressure compressor to construct a hierarchical mapping model of the high-pressure compressor; wherein each stage's performance correction coefficient component is associated with its corresponding assembly process parameter through a preset mapping relationship; and jointly identifying the weighted parameters in the hierarchical mapping model and the mapping parameters included in the preset mapping relationships associated with each stage to determine the target mapping relationship from assembly process parameters to component performance correction coefficients. Therefore, this invention embeds prior knowledge at the mechanistic level that the overall performance is synthesized from the performance of each level according to weights. This decomposes the high-dimensional and complex assembly parameter-overall performance mapping problem into multiple low-dimensional and easier-to-learn sub-mapping problems, effectively overcoming the overfitting problem of high-dimensional modeling under small sample conditions. At the same time, the joint identification mechanism ensures the coordinated optimization of each sub-model and the overall weight parameters, so that the final target mapping relationship can not only learn the subtle influence of assembly parameters from limited data more accurately, but also maintain the clarity of the model structure and physical interpretability. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating the modeling method for assembly process parameters and component performance correction coefficients of multi-stage compressors for aero-engines provided by the present invention.

[0021] Figure 2 This is a layered structure diagram of the assembly process parameters-component performance correction coefficient model provided by the present invention.

[0022] Figure 3 This is a schematic diagram of the modeling device for assembly process parameters and component performance correction coefficients of multi-stage compressors for aero-engines provided by the present invention.

[0023] Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0025] The following is combined with Figures 1-4 This invention describes a method and apparatus for modeling assembly process parameters and component performance correction coefficients for a multi-stage compressor in an aero-engine.

[0026] Figure 1 This is a flowchart illustrating the modeling method for assembly process parameters and component performance correction coefficients of multi-stage compressors in aero-engines provided by this invention. Figure 1 As shown, the method includes the following: Step 100: Obtain the assembly process parameters and corresponding component performance correction coefficients for the high-pressure compressors of multiple aero engines.

[0027] It should be noted that the High-Pressure Compressor (HPC) is one of the core components of an aero-engine turbofan. Its structure typically consists of multi-stage rotors (e.g., 8 to 12 stages). Each stage contains a row of rotating blades (movers) and a row of stationary guide vanes (stutterers). The rotating blades perform work on the airflow to achieve pressurization, while the stutterers are used to adjust the airflow direction and improve the compression efficiency of the next stage. The performance of the HPC has a decisive impact on the engine's overall thrust, fuel consumption rate, and operational stability. Even subtle changes in the assembly process parameters of each stage of the rotor can significantly alter its actual operating characteristics by affecting internal flow losses and leakage, leading to dispersion in the overall engine performance. Therefore, accurately modeling and predicting the impact of these assembly parameters on HPC performance is a key technical step in achieving precise control of engine manufacturing assembly quality and accurate performance prediction.

[0028] Step 100 obtains the assembly process parameters and corresponding component performance correction coefficients for the high-pressure compressors of multiple aero engines, including: Step 110: Obtain the assembly process parameters of the high-pressure compressors of the multiple aero engines.

[0029] Step 120: Based on the test performance data of the multiple aero engines, determine the component performance correction coefficient of the high-pressure compressor corresponding to each aero engine.

[0030] Specifically, assembly process parameters refer to the key geometric parameters that are measurable and have tolerances during engine manufacturing and assembly, such as the rotor assembly process parameters in each stage of a high-pressure compressor. Even small changes in these parameters can significantly affect the compressor's aerodynamic efficiency and flow characteristics. Secondly, component performance correction coefficients are used to quantitatively characterize the deviation of actual high-pressure compressor component performance (such as efficiency and flow coefficient) from its design nominal values. These are obtained by inputting the test performance data of the entire engine (such as thrust and exhaust temperature) into a known engine overall mechanism model and then reverse-engineering, representing the unique performance characteristics of that specific engine due to assembly differences.

[0031] Step 200: The component performance correction coefficient is characterized as a weighted sum of the performance correction coefficient components of each stage of the high-pressure compressor, so as to construct a hierarchical mapping model of the high-pressure compressor; wherein, the performance correction coefficient component of each stage is associated with the corresponding assembly process parameters through a preset mapping relationship.

[0032] Specifically, the performance correction coefficients at each level assume that the overall correction coefficient can be decomposed into local contributions originating from each rotor level. A preset mapping relationship defines the mathematical relationship between each level component and its corresponding set of assembly process parameters, which can be a linear or nonlinear function. The weighted sum expresses that the overall correction coefficient is accumulated from each level component with certain weights, where the weighting parameters reflect the relative influence of each level on the overall performance.

[0033] Step 300: Jointly identify the weighted parameters in the hierarchical mapping model and the mapping parameters contained in the preset mapping relationships associated with each level, so as to determine the target mapping relationship from assembly process parameters to component performance correction coefficients.

[0034] Specifically, through joint identification, using sample data from multiple engines, all weight parameters and specific parameters in all preset mapping relationships are simultaneously optimized and solved, thereby obtaining a target mapping relationship that can directly predict the HPC performance correction coefficient based on the HPC assembly process parameters of the new engine.

[0035] In one embodiment, it is assumed that a certain type of high-pressure compressor has M s Stage. For a test-run engine A, measure j assembly process parameters (total M) of each stage rotor. s ×j=L assembly parameters), and simultaneously, the overall efficiency correction coefficient of the high-pressure compressor (e.g., +1.5%) is derived from its test data. This embodiment assumes that this +1.5% overall correction is contributed by stage 1 (+0.2%, component 1), stage 2 (-0.1%, component 2), ... stage M. s Level contributed +0.3% (component M)s The components are synthesized according to specific weights (e.g., a weight vector of [0.1, 0.12, …, 0.08]). The +0.2% component value of level 1 is calculated from the j specific assembly process parameters of level 1 using a pre-defined function (e.g., a first-order linear model). The goal of this embodiment is to collect N such engine samples (each with L assembly parameters and one component performance correction coefficient) and automatically learn through an algorithm: 1) the weight of each level; and 2) the specific function (i.e., mapping relationship) used to calculate the component values ​​from its k assembly process parameters at each level. After learning, for a newly assembled engine B with only its L assembly process parameter values ​​known, the established model can be used to first calculate its performance correction components level by level, then perform a weighted summation, and finally predict the overall performance correction coefficient of its high-pressure compressor, thereby pre-evaluating its performance level.

[0036] The above describes the steps of the modeling method for assembly process parameters and component performance correction coefficients of multi-stage compressors for aero-engines provided by this invention. As can be seen from the above description, the modeling method for assembly process parameters and component performance correction coefficients of multi-stage compressors for aero-engines provided by this invention obtains the assembly process parameters and corresponding component performance correction coefficients of high-pressure compressors from multiple aero-engines; the component performance correction coefficients are represented as a weighted sum of the performance correction coefficient components at each stage of the high-pressure compressor to construct a hierarchical mapping model of the high-pressure compressor; wherein, the performance correction coefficient component at each stage is associated with the corresponding assembly process parameter through a preset mapping relationship; the weighted parameters in the hierarchical mapping model and the mapping parameters included in the preset mapping relationships associated at each stage are jointly identified to determine the target mapping relationship from assembly process parameters to component performance correction coefficients. Therefore, this invention embeds prior knowledge at the mechanistic level that the overall performance is synthesized from the performance of each level according to weights. This decomposes the high-dimensional and complex assembly parameter-overall performance mapping problem into multiple low-dimensional and easier-to-learn sub-mapping problems, effectively overcoming the overfitting problem of high-dimensional modeling under small sample conditions. At the same time, the joint identification mechanism ensures the coordinated optimization of each sub-model and the overall weight parameters, so that the final target mapping relationship can not only learn the subtle influence of assembly parameters from limited data more accurately, but also maintain the clarity of the model structure and physical interpretability.

[0037] Based on the above embodiments, in this embodiment, step 200 characterizes the component performance correction coefficient as a weighted sum of the performance correction coefficient components of each stage of the high-pressure compressor, in order to construct a hierarchical mapping model of the high-pressure compressor, including: Step 210: For each stage of the high-pressure compressor, establish the preset mapping relationship between the performance correction coefficient component and the corresponding assembly process parameter.

[0038] Step 220: Establish the overall component performance correction coefficient of the high-pressure compressor as a weighted sum model of the performance correction coefficient components of all stages; wherein, the coefficients in the weighted sum model are the weighting parameters.

[0039] Specifically, each stage of a high-pressure compressor refers to the sequentially arranged compression units in its structure. Each stage typically contains a set of moving parts and a set of stators, working together to compress the airflow once. The performance correction factor component is a variable defined separately for each stage. It characterizes the change in the local performance of that stage relative to its design nominal value due to deviations in specific assembly process parameters from the design value. A predefined mapping relationship describes how the performance correction factor component of each stage is calculated from a specific set of assembly process parameters for its corresponding stage. This relationship can be a simple linear function (such as a weighted sum) or a complex function containing nonlinear terms (such as square terms or cross terms). The weighted sum model is the core of the hierarchical mapping. It indicates that the overall component performance correction factor of the high-pressure compressor (i.e., the total deviation affecting the overall performance) is not a simple sum of the components at each stage, but rather a weighted sum. The weighting parameter is a coefficient vector corresponding to the stage number. Each weight parameter quantifies the contribution ratio or degree of influence of the corresponding stage's performance correction component on the overall total performance deviation. For example, stages that are more critical to aerodynamic design or more sensitive to assembly process parameters are usually assigned greater weights.

[0040] In one embodiment, the construction process of the assembly process parameter-component performance correction coefficient model is described in detail.

[0041] This embodiment introduces prior mechanistic information: it is assumed that the performance correction coefficient of the entire HPC component is caused by each level of rotor, and therefore can be distributed to each level of rotor according to a certain proportion. The performance correction coefficient components of each level of rotor are determined by the assembly process parameters corresponding to each level of rotor. Based on this prior information, a two-layer model is designed. First, the overall performance correction coefficient of the HPC component can be expressed by the performance correction coefficients of each level of rotor as follows: in w s_k , k =1,…, M s , is the weight of the performance correction coefficients generated by each rotor level when integrated into a whole; M s It is the number of HPC rotor stages; It is the overall performance correction factor for HPC components; It is the first kPerformance correction factors for each stage of the rotor. The relationship between the performance correction factors for each stage of the rotor and the assembly process parameters can be referenced as follows: in These are the assembly process parameters for the k-th stage rotor; It is the vector of performance correction coefficients for all rotors; It is a vector of assembly process parameters for all rotors; This represents the mapping relationship between the assembly process parameters and performance correction coefficients of the k-th stage rotor. The final HPC assembly process parameter-component performance correction coefficient model is as follows: Figure 2 This is a hierarchical structure diagram of the assembly process parameter-component performance correction coefficient model provided by the present invention, as shown below. Figure 2 As shown, the bottom layer of the model, "Input Layer: Assembly Parameters," represents the specific assembly process parameter vectors corresponding to each level (e.g., level 1 to level Ms), denoted as Φ_a1_HPC to Φ_aMs_HPC, such as the assembly process parameter data for each rotor level. The middle layer is the "Hidden Layer: Hierarchical Model," where each node g_1_HPC to g_Ms_HPC represents a preset mapping function. Its function is to receive the assembly parameter vector of the corresponding level as input and calculate the performance correction coefficient components of that level, i.e., θ_s1_HPC to θ_sMs_HPC. These components characterize the independent impact of each level's assembly deviation on its own performance. In the top layer of the model, "Output Layer," the final performance correction coefficient θ_p_HPC of the entire machine is not a simple sum of the components, but is obtained by weighted summation using a set of weighted parameters xw_s1 to xw_sMs, i.e., θ_p_HPC = xw_s1 θ_s1_HPC+ xw_s2 θ_s2_HPC + … + xw_sMs θ_sMs_HPC. This structure clearly embodies the core idea that "overall performance deviation stems from the weighted assembly of contributions at each level," and achieves an interpretable mapping from multi-level, high-dimensional assembly parameters at the bottom layer to a single overall performance index at the top layer.

[0042] The modeling method for assembly process parameters and component performance correction coefficients of multi-stage compressors for aero-engines provided in this embodiment systematically solves the core problem of the difficulty in establishing a high-precision assembly parameter-performance prediction model for high-pressure compressors of aero-engines under small sample conditions through a structured hierarchical mapping modeling method.

[0043] Based on the above embodiments, in this embodiment, step 300 jointly identifies the weighted parameters in the hierarchical mapping model and the mapping parameters included in the preset mapping relationships associated at each level, in order to determine the target mapping relationship from assembly process parameters to component performance correction coefficients, including: Step 310: Using the maximum likelihood estimation algorithm, jointly identify the weighted weight parameters in the hierarchical mapping model and the mapping parameters contained in the preset mapping relationships associated with each level to obtain the initial identification result.

[0044] Step 320: In the identification process, introduce constraints to optimize the initial identification results in order to determine the target mapping relationship from assembly process parameters to component performance correction coefficients; wherein, the constraints include equality constraints, inequality constraints, or mixed constraints that simultaneously include equality constraints and inequality constraints.

[0045] It should be noted that the constraints are used to limit the proportional relationship or numerical range between the performance correction coefficient components at each level.

[0046] Specifically, joint identification refers to treating all unknown parameters in the model—that is, the weighted weight parameters representing the contribution of each level and the mapping parameters defining the mapping relationship between each level (such as the coefficients in a linear model)—as a whole parameter set, and simultaneously estimating them using sample data from multiple engines. Maximum likelihood estimation is a classic parameter estimation method. Its core idea is to find a set of parameter values ​​that maximizes the probability of the model's prediction given the existing sample data, thereby obtaining the initial identification result with the highest matching degree to the training data.

[0047] However, due to the limited number of engine samples and the correlation between assembly parameters, simply relying on data fitting may lead to unrealistic or overfitting results. Therefore, it is necessary to introduce constraints to incorporate prior engineering knowledge into the optimization process to guide and correct the identification direction. Equality constraints can be used to enforce specific mathematical relationships between certain parameters (e.g., limiting the ratio of weights between adjacent levels to a fixed range); inequality constraints are used to restrict the reasonable range of parameter values ​​(e.g., stipulating that all mapping parameters must be positive, or that a certain level of performance correction coefficient component must not exceed a certain physical upper limit); mixed constraints apply both types of restrictions simultaneously. The common purpose of these constraints is to limit the proportional relationship or numerical range between performance correction coefficient components at each level. Essentially, they are used to regularize and correct the initial identification results, thereby ensuring that, under limited data conditions, the final learned target mapping relationship not only fits the data but also conforms to physical laws.

[0048] In one embodiment, the joint identification process of the assembly process parameter-component performance correction coefficient model is specifically described.

[0049] The HPC assembly process parameters and HPC performance correction factor data for N engines are as follows: The superscript 'i' represents the specific value of the HPC assembly process parameter corresponding to the i-th engine. Values ​​of HPC component performance correction factors ; It is a sample vector composed of the performance correction coefficient values ​​of N engine HPC components; It is a sample vector composed of the HPC assembly process parameter values ​​of N engines. During identification, the maximum likelihood estimation algorithm is used. w s_k and g HPCs_k The unknown parameters in parentheses are estimated. During the identification process, to avoid overfitting when there is strong collinearity in the data due to a large number of unknown parameters, constraints can be introduced to prevent excessive differences in correction coefficients between different components. When constraints are introduced, they can take three forms: equality constraints, inequality constraints, and mixed equality / inequality constraints, as follows: (1) (2) (3) The modeling method for assembly process parameters and component performance correction coefficients of multi-stage compressors for aero-engines provided in this embodiment effectively solves the overfitting problem that is prone to occur in complex hierarchical models with small samples and a large number of parameters by introducing a joint identification and constraint optimization mechanism based on maximum likelihood estimation, thereby ensuring the generalization ability and physical rationality of the final mapping relationship.

[0050] Based on the above embodiments, in this embodiment, the preset mapping relationship is a linear mapping relationship or a nonlinear mapping relationship based on Taylor expansion.

[0051] The linear mapping relationship based on Taylor expansion is a first-order Taylor expansion model, used to characterize the linear relationship between the performance correction coefficient components and the corresponding assembly process parameters.

[0052] The nonlinear mapping relationship based on Taylor expansion is a high-order Taylor expansion model, used to characterize the polynomial relationship between the performance correction coefficient components and the corresponding assembly process parameters, which includes high-order terms.

[0053] Specifically, the core idea of ​​the Taylor expansion-based mapping relationship is to expand an unknown complex function into a polynomial near its reference point (e.g., the average or design value of each assembly parameter). When expanding and retaining only the first-order terms, a linear mapping relationship or a first-order Taylor expansion model is obtained. This model assumes that the performance correction coefficient component and the corresponding assembly process parameter have an approximately simple proportional relationship. For example, for every unit increase in a certain assembly process parameter, the performance correction component of that level increases or decreases by a fixed amount approximately linearly. It is simple in form, has few parameters, and is very stable under small samples, making it suitable for scenarios where the parameter variation range is small and the influence is approximately linear. When expanding and retaining the second-order terms, a nonlinear mapping relationship or a higher-order Taylor expansion model is obtained. This model, in addition to including the above linear relationship, also introduces higher-order terms of the assembly parameters (i.e., the square of the parameter itself) and possible cross terms (here limited to a polynomial relationship, usually including cross terms), thus enabling the characterization of more complex nonlinear effects. For example, the performance correction may deteriorate at an accelerating rate as the assembly process parameter increases, or there may be interactive effects between two different assembly process parameters. This polynomial relationship has a stronger fitting ability than the linear model, and can capture the curvature changes in the performance response. It is suitable for situations where the parameter variation range is large or the influencing mechanism has obvious nonlinearity.

[0054] In one embodiment, the specific algorithm design process for identifying the first-order Taylor expansion model of assembly process parameters and component performance correction coefficients is described in detail.

[0055] g HPCs_k The first-order Taylor expansion model is designed as follows: in It is a matrix composed of the parameters to be identified corresponding to the k-th stage rotor. The HPC assembly process parameters - component performance correction coefficient model is as follows: w all It is a vector composed of all parameters to be identified, corresponding to each stage of the rotor. Composed of.

[0056] (1) Without introducing constraints, the update formula for all parameters using the maximum likelihood method is: (2) With equality constraints introduced, the update formula for all parameters using the maximum likelihood method is: The modeling method for assembly process parameters and component performance correction coefficients of multi-stage compressors for aero-engines provided in this embodiment can be specifically constructed as a first-order linear or higher-order nonlinear model based on the Taylor expansion principle by clearly defining the preset mapping relationship. This provides a set of mathematical tools that are both simple and complex, flexible and practical, for the key link of "how the performance correction coefficient components are calculated from the assembly process parameters". This achieves a good balance between the model's expressive power and the learning reliability under limited data conditions.

[0057] The following describes the modeling device for assembly process parameters and component performance correction coefficients of aero-engine multi-stage compressors provided by the present invention. The modeling device for assembly process parameters and component performance correction coefficients of aero-engine multi-stage compressors described below can be referred to in correspondence with the modeling method for assembly process parameters and component performance correction coefficients of aero-engine multi-stage compressors described above.

[0058] Figure 3 This is a schematic diagram of the modeling device for assembly process parameters and component performance correction coefficients of multi-stage compressors for aero-engines provided by the present invention. Figure 3 As shown, the modeling device for assembly process parameters and component performance correction coefficients of multi-stage compressors for aero-engines provided by the present invention includes: The acquisition module 301 is used to acquire the assembly process parameters and corresponding component performance correction coefficients of the high-pressure compressors of multiple aero engines. The construction module 302 is used to characterize the component performance correction coefficient as a weighted sum of the performance correction coefficient components of each stage of the high-pressure compressor, so as to construct a hierarchical mapping model of the high-pressure compressor; wherein, the performance correction coefficient component of each stage is associated with the corresponding assembly process parameters through a preset mapping relationship. The identification module 303 is used to jointly identify the weighted parameters in the hierarchical mapping model and the mapping parameters contained in the preset mapping relationships associated at each level, so as to determine the target mapping relationship from assembly process parameters to component performance correction coefficients.

[0059] The present invention provides a modeling device for assembly process parameters and component performance correction coefficients of multi-stage compressors for aero-engines. This device acquires the assembly process parameters and corresponding component performance correction coefficients of high-pressure compressors from multiple aero-engines. The component performance correction coefficients are represented as a weighted sum of the performance correction coefficient components at each stage of the high-pressure compressor, thereby constructing a hierarchical mapping model of the high-pressure compressor. Each stage's performance correction coefficient component is associated with its corresponding assembly process parameter through a preset mapping relationship. The weighted parameters in the hierarchical mapping model and the mapping parameters included in the preset mapping relationships associated with each stage are jointly identified to determine the target mapping relationship from assembly process parameters to component performance correction coefficients. Therefore, this invention embeds prior knowledge at the mechanistic level that the overall performance is synthesized from the performance of each level according to weights. This decomposes the high-dimensional and complex assembly parameter-overall performance mapping problem into multiple low-dimensional and easier-to-learn sub-mapping problems, effectively overcoming the overfitting problem of high-dimensional modeling under small sample conditions. At the same time, the joint identification mechanism ensures the coordinated optimization of each sub-model and the overall weight parameters, so that the final target mapping relationship can not only learn the subtle influence of assembly parameters from limited data more accurately, but also maintain the clarity of the model structure and physical interpretability.

[0060] Based on the above embodiments, in this embodiment, the acquisition module 301 is specifically used for: Obtain the assembly process parameters of the high-pressure compressors of the multiple aero engines; Based on the test performance data of the multiple aero engines, the component performance correction coefficient of the high-pressure compressor corresponding to each aero engine is determined.

[0061] Based on the above embodiments, in this embodiment, the construction module 302 is specifically used for: For each stage of the high-pressure compressor, a preset mapping relationship is established between the performance correction coefficient component and the corresponding assembly process parameter. The overall component performance correction coefficients of the high-pressure compressor are established as a weighted sum model of the performance correction coefficient components of all stages; wherein, the coefficients in the weighted sum model are the weighting parameters.

[0062] Based on the above embodiments, in this embodiment, the preset mapping relationship is a linear mapping relationship or a nonlinear mapping relationship based on Taylor expansion.

[0063] Based on the above embodiments, in this embodiment, the linear mapping relationship based on Taylor expansion is a first-order Taylor expansion model, which is used to characterize the linear relationship between the performance correction coefficient components and the corresponding assembly process parameters.

[0064] Based on the above embodiments, in this embodiment, the nonlinear mapping relationship based on Taylor expansion is a high-order Taylor expansion model, which is used to characterize the polynomial relationship between the performance correction coefficient components and the corresponding assembly process parameters, which includes high-order terms.

[0065] Based on the above embodiments, in this embodiment, the identification module 303 is specifically used for: The weighted parameters in the hierarchical mapping model and the mapping parameters contained in the preset mapping relationships associated with each level are jointly identified using the maximum likelihood estimation algorithm to obtain the initial identification result. During the identification process, constraints are introduced to optimize the initial identification results in order to determine the target mapping relationship from assembly process parameters to component performance correction coefficients; wherein, the constraints include equality constraints, inequality constraints, or mixed constraints that simultaneously include equality constraints and inequality constraints.

[0066] Based on the above embodiments, in this embodiment, the constraint condition is used to limit the proportional relationship or numerical range between the performance correction coefficient components at each level.

[0067] Figure 4 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 4 As shown, the electronic device can be a robot or other electronic device. This electronic device may include: a processor 410, a communication interface 420, a memory 430, and a communication bus 440. The processor 410, communication interface 420, and memory 430 communicate with each other via the communication bus 440. The processor 410 can call logical instructions from the memory 430 to execute a modeling method for assembly process parameters and component performance correction coefficients of multi-stage compressors for aero-engines, including: Obtain the assembly process parameters and corresponding component performance correction coefficients for the high-pressure compressors of multiple aero engines; The component performance correction coefficient is characterized as a weighted sum of the performance correction coefficient components of each stage of the high-pressure compressor, so as to construct a hierarchical mapping model of the high-pressure compressor; wherein, the performance correction coefficient component of each stage is associated with the corresponding assembly process parameters through a preset mapping relationship. The weighted parameters in the hierarchical mapping model and the mapping parameters contained in the preset mapping relationships associated with each level are jointly identified to determine the target mapping relationship from assembly process parameters to component performance correction coefficients.

[0068] Furthermore, the logical instructions in the aforementioned memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0069] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the modeling method for assembly process parameters and component performance correction coefficients of multi-stage compressors for aero-engines provided by the above methods, including: Obtain the assembly process parameters and corresponding component performance correction coefficients for the high-pressure compressors of multiple aero engines; The component performance correction coefficient is characterized as a weighted sum of the performance correction coefficient components of each stage of the high-pressure compressor, so as to construct a hierarchical mapping model of the high-pressure compressor; wherein, the performance correction coefficient component of each stage is associated with the corresponding assembly process parameters through a preset mapping relationship. The weighted parameters in the hierarchical mapping model and the mapping parameters contained in the preset mapping relationships associated with each level are jointly identified to determine the target mapping relationship from assembly process parameters to component performance correction coefficients.

[0070] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the modeling method for assembly process parameters and component performance correction coefficients of aero-engine multi-stage compressors provided by the methods described above, including: Obtain the assembly process parameters and corresponding component performance correction coefficients for the high-pressure compressors of multiple aero engines; The component performance correction coefficient is characterized as a weighted sum of the performance correction coefficient components of each stage of the high-pressure compressor, so as to construct a hierarchical mapping model of the high-pressure compressor; wherein, the performance correction coefficient component of each stage is associated with the corresponding assembly process parameters through a preset mapping relationship. The weighted parameters in the hierarchical mapping model and the mapping parameters contained in the preset mapping relationships associated with each level are jointly identified to determine the target mapping relationship from assembly process parameters to component performance correction coefficients.

[0071] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0072] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0073] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for modeling assembly process parameters and component performance correction coefficients for multi-stage compressors in aero-engines, characterized in that, include: Obtain the assembly process parameters and corresponding component performance correction coefficients for the high-pressure compressors of multiple aero engines; The component performance correction coefficient is characterized as a weighted sum of the performance correction coefficient components of each stage of the high-pressure compressor, so as to construct a hierarchical mapping model of the high-pressure compressor; wherein, the performance correction coefficient component of each stage is associated with the corresponding assembly process parameters through a preset mapping relationship. The weighted parameters in the hierarchical mapping model and the mapping parameters contained in the preset mapping relationships associated with each level are jointly identified to determine the target mapping relationship from assembly process parameters to component performance correction coefficients.

2. The modeling method for assembly process parameters and component performance correction coefficients of multi-stage compressors for aero-engines according to claim 1, characterized in that, The acquisition of assembly process parameters and corresponding component performance correction coefficients for the high-pressure compressors of multiple aero-engines includes: Obtain the assembly process parameters of the high-pressure compressors of the multiple aero engines; Based on the test performance data of the multiple aero engines, the component performance correction coefficient of the high-pressure compressor corresponding to each aero engine is determined.

3. The modeling method for assembly process parameters and component performance correction coefficients of multi-stage compressors for aero-engines according to claim 1, characterized in that, The step of characterizing the component performance correction coefficient as a weighted sum of the performance correction coefficient components of each stage of the high-pressure compressor to construct a hierarchical mapping model of the high-pressure compressor includes: For each stage of the high-pressure compressor, a preset mapping relationship is established between the performance correction coefficient component and the corresponding assembly process parameter. The overall component performance correction coefficients of the high-pressure compressor are established as a weighted sum model of the performance correction coefficient components of all stages; wherein, the coefficients in the weighted sum model are the weighting parameters.

4. The modeling method for assembly process parameters and component performance correction coefficients of multi-stage compressors for aero-engines according to any one of claims 1-3, characterized in that, The preset mapping relationship is a linear mapping relationship or a nonlinear mapping relationship based on Taylor expansion.

5. The modeling method for assembly process parameters and component performance correction coefficients of multi-stage compressors for aero-engines according to claim 4, characterized in that, The linear mapping relationship based on Taylor expansion is a first-order Taylor expansion model, used to characterize the linear relationship between the performance correction coefficient components and the corresponding assembly process parameters.

6. The modeling method for assembly process parameters and component performance correction coefficients of multi-stage compressors for aero-engines according to claim 4, characterized in that, The nonlinear mapping relationship based on Taylor expansion is a high-order Taylor expansion model, used to characterize the polynomial relationship between the performance correction coefficient components and the corresponding assembly process parameters, which includes high-order terms.

7. The modeling method for assembly process parameters and component performance correction coefficients of multi-stage compressors for aero-engines according to claim 1, characterized in that, The step of jointly identifying the weighted parameters in the hierarchical mapping model and the mapping parameters contained in the preset mapping relationships associated at each level to determine the target mapping relationship from assembly process parameters to component performance correction coefficients includes: The weighted parameters in the hierarchical mapping model and the mapping parameters contained in the preset mapping relationships associated with each level are jointly identified using the maximum likelihood estimation algorithm to obtain the initial identification result. During the identification process, constraints are introduced to optimize the initial identification results in order to determine the target mapping relationship from assembly process parameters to component performance correction coefficients; wherein, the constraints include equality constraints, inequality constraints, or mixed constraints that simultaneously include equality constraints and inequality constraints.

8. The modeling method for assembly process parameters and component performance correction coefficients of multi-stage compressors for aero-engines according to claim 7, characterized in that, The constraints are used to limit the proportional relationship or numerical range between the performance correction coefficient components at each level.

9. A modeling device for assembly process parameters and component performance correction coefficients of a multi-stage compressor for an aero-engine, characterized in that, include: The acquisition module is used to acquire the assembly process parameters and corresponding component performance correction coefficients of the high-pressure compressors of multiple aero engines. A construction module is used to characterize the component performance correction coefficient as a weighted sum of the performance correction coefficient components of each stage of the high-pressure compressor, so as to construct a hierarchical mapping model of the high-pressure compressor; wherein, the performance correction coefficient component of each stage is associated with the corresponding assembly process parameters through a preset mapping relationship; The identification module is used to jointly identify the weighted parameters in the hierarchical mapping model and the mapping parameters contained in the preset mapping relationships associated at each level, so as to determine the target mapping relationship from assembly process parameters to component performance correction coefficients.

10. The modeling device for assembly process parameters and component performance correction coefficients of a multi-stage compressor for aero-engines according to claim 9, characterized in that, The acquisition module is specifically used for: Obtain the assembly process parameters of the high-pressure compressors of the multiple aero engines; Based on the test performance data of the multiple aero engines, the component performance correction coefficient of the high-pressure compressor corresponding to each aero engine is determined.

Images