A method and system for measuring the throat area of ​​a gas turbine disc guide based on ELM.

By using an ELM-based method and a coordinate measuring machine and extreme learning machine model, the throat area of ​​a gas turbine guide vane is automatically measured, which solves the problems of low efficiency and poor repeatability in the existing technology and achieves efficient and accurate throat area measurement.

CN123111172APending Publication Date: 2026-11-17AECC AVIATION POWER CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611153759.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-07-31
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing methods for measuring the throat area of ​​gas turbine guide vanes are inefficient, have poor repeatability, are greatly affected by manual operation, and lack an effective way to convert three-dimensional coordinate data into throat area.

Method used

An ELM-based approach is adopted to acquire the three-dimensional coordinate data of throat feature points through a coordinate measuring machine, establish sample correspondence, train an extreme learning machine model, and realize the automated measurement of throat area.

Benefits of technology

This improves the efficiency and repeatability of measuring the throat area of ​​gas turbine guide vanes, reduces the impact of manual operation, and ensures the accuracy and traceability of measurement results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123111172A_ABST
    Figure CN123111172A_ABST
Patent Text Reader

Abstract

This invention discloses a method and system for measuring the throat area of ​​a gas turbine guide vane based on ELM (Extreme Learning Machine), belonging to the field of gas turbine measurement technology. The method acquires data samples from multiple throat windows, uses a coordinate measuring machine to collect the coordinates of throat feature points on adjacent guide vane blades and upper and lower edge plates, and uses the throat area measured by a dedicated measuring tool as the output label. An extreme learning machine model is trained using a training set and validated and saved using a validation set. Each throat window of the gas turbine guide vane under test is numbered, and three-dimensional coordinate data is acquired in numerical order and input into the model to obtain the corresponding throat area. This invention enables automatic measurement of the throat area of ​​a gas turbine guide vane without specifying the theoretical position of the minimum throat cross-section, improving measurement repeatability and detection efficiency.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of gas turbine measurement technology, specifically relating to a method and system for measuring the throat area of ​​a gas turbine disc guide based on ELM. Background Technology

[0002] A gas turbine guide vane typically includes a guide vane casing, guide vane blades, and mounting rings. Two adjacent guide vane blades, together with the upper and lower edge plates, form an airflow channel. The minimum cross-sectional area of ​​this airflow channel is called the throat area. The throat area determines the flow capacity of the guide vane channel and affects the gas turbine's flow rate, thrust, and speed. Therefore, it is necessary to measure the throat area of ​​each throat window in the guide vane.

[0003] Current throat area measurement typically involves inspectors using specialized measuring tools to measure the upper base, lower base, and height of an approximate trapezoidal cross-section, and then calculating the throat area based on this. However, gas turbine disc guides are large in size, have numerous throat windows, and limited measurement space. Manually placing and adjusting the specialized measuring tools is difficult and inefficient.

[0004] Furthermore, the guide design drawings typically do not explicitly specify the theoretical location of the minimum throat cross-section. Inspectors need to locate the measurement position based on experience, gauge posture, and contact conditions. Different inspectors have varying degrees of operational force and positional judgment, which can lead to different results even when the same inspector performs repeated measurements on the same throat window, resulting in poor measurement repeatability and reproducibility.

[0005] While coordinate measuring machines (CMMs) can measure the three-dimensional coordinates of points on the guide vanes and upper and lower edge plates, they cannot directly determine the actual minimum cross-section and its area based solely on the measured three-dimensional coordinates when the theoretical location of the minimum throat cross-section is unknown. Furthermore, existing technologies lack an effective method for converting the three-dimensional coordinate data of multiple points near the throat into the corresponding throat area.

[0006] For a whole-disc guide, multiple throat windows are continuously distributed along the circumference. Existing manual measurement methods are prone to problems such as inconsistent measurement order, incorrect correspondence between measurement data and throat windows, missed measurements or repeated measurements, making it difficult to achieve continuous automatic measurement of each throat window and traceability of measurement results. Summary of the Invention

[0007] The technical problem to be solved by this invention is to address the shortcomings of the prior art by providing an ELM-based method and system for measuring the throat area of ​​a gas turbine guide plate. This method involves collecting the coordinates of three-dimensional feature points near the throat cross-section and establishing a sample correspondence between these coordinates and the throat area measured by a dedicated measuring instrument. An extreme learning machine model is then trained to establish a mapping relationship between the three-dimensional feature point coordinates and the throat area. This enables automated and repeatable measurement of each throat window of the gas turbine guide plate, addressing the technical problems of unclear theoretical positions of the minimum throat cross-section, difficulty in directly converting existing coordinate measuring machine data into throat area, and the significant impact of manual measurement on operator error, poor measurement repeatability, and low overall plate detection efficiency.

[0008] The present invention adopts the following technical solution: A method for measuring the throat area of ​​a gas turbine disc guide based on ELM includes the following steps: S1. For multiple gas turbine guideway throat windows, based on the structural characteristics of the guideway and the dimensions of the special measuring tool, determine multiple throat feature points near the throat cross section; use a coordinate measuring machine to obtain the three-dimensional coordinate data of each throat feature point on the upper and lower edge plates of two adjacent guideway blades, and use a special measuring tool to obtain the corresponding throat area; establish a sample correspondence relationship between the three-dimensional coordinate data and the corresponding throat area to form a data sample. S2. Divide the data samples into a training set and a validation set; S3. Construct an extreme learning machine model, using the three-dimensional coordinate data in the training set as the model input and the corresponding throat area as the model output, set the number of hidden layer nodes, randomly generate the input weights and hidden layer node biases from the input layer to the hidden layer, and solve the output weights from the hidden layer to the output layer based on the hidden layer output matrix and the corresponding throat area to obtain the throat area measurement model. S4. Validate the throat area measurement model using the validation set, and save the throat area measurement model when the validation result meets the preset accuracy conditions. S5. Number each throat window of the gas turbine guide to be tested, and use the coordinate measuring machine to obtain the three-dimensional coordinate data of the throat feature points corresponding to each throat window according to the numbering order. S6. Input the three-dimensional coordinate data of the throat feature points corresponding to each throat window into the saved throat area measurement model to obtain the throat area measurement results corresponding to each throat window number.

[0009] Furthermore, the two adjacent guide vanes are respectively the first guide vane and the second guide vane; the throat feature points include the blade basin profile measurement points of the first guide vane and the blade back profile measurement points of the second guide vane located on three radially arranged measurement sections, as well as the flow channel wall measurement points located on the upper edge plate and the lower edge plate.

[0010] Furthermore, the three blade basin profile measurement points of the first guide vane and the three blade back profile measurement points of the second guide vane constitute throat width feature points; the flow channel wall measurement points located on the upper edge plate and the flow channel wall measurement points located on the lower edge plate constitute throat height feature points.

[0011] Furthermore, the coordinate values ​​of each of the throat feature points in the corresponding three-dimensional coordinate system are measured respectively, and the coordinate values ​​of each of the throat feature points constitute the three-dimensional coordinate data of the corresponding throat window.

[0012] Furthermore, the extreme learning machine model is a single hidden layer feedforward neural network, and the hidden layer output matrix is ​​determined based on the input weights, the hidden layer node biases, the three-dimensional coordinate data, and the activation functions of the hidden layer nodes.

[0013] Furthermore, the output weights are obtained by solving the Moore-Penrose generalized inverse of the hidden layer output matrix and the throat area corresponding to each data sample in the training set.

[0014] Furthermore, the three-dimensional coordinate data from the validation set are input into the throat area measurement model to obtain the corresponding validation area. The throat area measurement model is then validated based on the measurement deviation between the validation area and the throat area measured by the corresponding dedicated measuring instrument, as well as the accuracy of the validation set.

[0015] Furthermore, in step S5, when observing the gas turbine guide in the forward direction, the throat windows of the gas turbine guide are numbered sequentially in the counterclockwise direction, and the corresponding three-dimensional coordinate data are recorded according to the numbering order of each throat window.

[0016] Furthermore, the complete throat window formed by the first guide vane under test and the adjacent guide vane located to the left of the first guide vane under test when observing the gas turbine guide vane in the forward direction is defined as the first throat window, and the remaining throat windows of the gas turbine guide vane under test are numbered sequentially in a counterclockwise direction starting from the first throat window.

[0017] Secondly, embodiments of the present invention provide an ELM-based system for measuring the throat area of ​​a gas turbine disk guide, including a coordinate measuring machine, a special measuring tool, and a computer device; The coordinate measuring machine is used to measure multiple throat feature points determined near the throat section of the throat window of the gas turbine guide based on the structural characteristics of the gas turbine guide and the dimensions of the special measuring tool. The throat feature points are located on the two adjacent guide blades that make up the throat window, as well as the upper and lower edge plates, so as to obtain the three-dimensional coordinate data of the throat feature points. The special measuring instrument is used to measure the throat area of ​​the throat window; The computer device includes a processor and a memory, the memory storing a computer program, and the processor executing the computer program to cause the computer device to: The three-dimensional coordinate data and the corresponding throat area are obtained. A sample correspondence relationship is established between the three-dimensional coordinate data and the corresponding throat area to form a data sample. The data sample is then divided into a training set and a validation set. An extreme learning machine model is constructed, with the three-dimensional coordinate data in the training set as the model input and the corresponding throat area as the model output. The number of hidden layer nodes is set, and the input weights and hidden layer node biases from the input layer to the hidden layer are randomly generated. The output weights from the hidden layer to the output layer are solved according to the hidden layer output matrix and the corresponding throat area to obtain the throat area measurement model. The throat area measurement model is validated using the validation set, and the throat area measurement model is saved when the validation results meet the preset accuracy conditions. Each throat window of the gas turbine guide under test is numbered, and the three-dimensional coordinate data of the throat feature points corresponding to each throat window are obtained by the coordinate measuring machine according to the numbering order. The three-dimensional coordinate data is then input into the saved throat area measurement model to obtain the throat area measurement result corresponding to each throat window number.

[0018] Compared with the prior art, the present invention has at least the following beneficial effects: An ELM-based method for measuring the throat area of ​​a gas turbine disk guide is proposed. This method involves setting multiple throat feature points near the throat cross-section, collecting 3D coordinate data of adjacent blades and upper and lower edge plates using a coordinate measuring machine (CMM), and using the throat area measured with a dedicated measuring tool as the output label. This establishes a sample correspondence between the spatial information of the measurement points and the actual throat area, solving the problems of the design drawing not providing the theoretical location of the minimum throat cross-section and the difficulty in directly converting CMM measurement results into throat area. Furthermore, an ELM is used to randomly generate input weights and hidden layer node biases, and the output weights are solved based on the hidden layer output matrix. This avoids the iterative parameter adjustments required by traditional neural networks, enabling rapid fitting of the mapping relationship between 3D coordinate data and throat area. The model is validated using a validation set, and models that meet the accuracy requirements are saved, which helps to filter out models with insufficient predictive performance. Numbering and sequentially measuring each throat window of the disk guide establishes a correspondence between window numbers, measurement point data, and area results, thereby reducing the differences in force, posture, and experience in manual measurement, and improving measurement repeatability, detection efficiency, and result traceability.

[0019] Furthermore, measurement points of the blade basin profile of the first guide vane and the blade back profile of the second guide vane are collected at three radially arranged measurement sections, and measurement points of the flow channel walls of the upper and lower edge plates are also collected. This allows for the acquisition of spatial morphological information of the throat channel from multiple radial positions. This avoids the problem that a single section measurement point cannot fully reflect the three-dimensional structure of the throat, and provides more complete input features for the model.

[0020] Furthermore, the measurement points of the leaf basin profile and the leaf back profile are used as throat width feature points, and the measurement points of the upper and lower margins are used as throat height feature points, respectively characterizing the width and height changes of the throat cross-section. This helps to correlate the input features of the extreme learning machine with the geometric influencing factors of the throat area, improving the relevance of the model fitting.

[0021] Furthermore, the coordinates of each throat feature point in the corresponding three-dimensional coordinate system are obtained, and the coordinates are used to construct the three-dimensional coordinate data of the throat window, converting the scattered spatial measurement points into numerical inputs that the model can process. This realizes the conversion of physical blade structural information into machine learning input data, providing a data foundation for subsequent model training and area prediction.

[0022] Furthermore, a single hidden layer feedforward neural network is used as the extreme learning machine model, and the hidden layer output matrix is ​​determined based on the input weights, hidden layer node biases, three-dimensional coordinate data and activation functions. The nonlinear mapping capability is used to learn the relationship between the measurement points of complex leaf shapes and the throat area, without the need to establish an accurate analytical geometric model in advance.

[0023] Furthermore, by solving for the output weights using the Moore-Penrose generalized inverse of the hidden layer output matrix and the throat area corresponding to the training samples, the parameters from the hidden layer to the output layer can be directly obtained. This reduces the need for repeated iterative training, improves model training efficiency, and reduces instability caused by improper selection of iteration parameters.

[0024] Furthermore, the deviation of the validation area from the area measured by the dedicated measuring instrument is used as the accuracy criterion for a single sample. When the measurement deviation meets the requirement of 1%, the measurement is accurate; when the validation set accuracy reaches 98%, the trained extreme learning machine model can be used.

[0025] Furthermore, when observing along the flight path, each throat window is sequentially numbered counterclockwise, and the three-dimensional coordinate data is recorded in numerical order, which unifies the measurement sequence of the entire guide. This setting can reduce missed measurements, repeated measurements, or mismatches between measurement data and window positions, and facilitates the management and traceability of test results.

[0026] Furthermore, by using the complete throat window formed by the first guide vane under test and its adjacent left-side vane as the first throat window, and numbering sequentially from this point, the starting reference for the entire measurement can be clearly defined. The purpose is to eliminate inconsistencies in the understanding of the numbering starting point among different testing personnel and improve the consistency of the measurement process.

[0027] In summary, this invention constructs supervised samples using coordinate measuring machine data and the area of ​​a dedicated measuring tool, and uses an extreme learning machine to fit the mapping relationship between throat feature points and area, solving the problem of difficulty in directly calculating the area when the position of the minimum throat cross section is unknown. Furthermore, it achieves automatic measurement of the entire guide through window numbering, improving detection efficiency, repeatability, and result traceability.

[0028] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0029] Figure 1 This is a flowchart of the ELM-based method for measuring the throat area of ​​a gas turbine disc guide in an embodiment of the present invention. Figure 2 This is a schematic diagram of the throat area measurement of the gas turbine guide airflow passage according to an embodiment of the present invention; Figure 3 This is a schematic diagram showing the arrangement of characteristic points of the throat width of adjacent guide vanes in an embodiment of the present invention; Figure 4 This is a schematic diagram showing the angular numbering of the throat window of the guide tube of the ring gas turbine according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the Extreme Learning Machine network structure used in an embodiment of the present invention; Figure 6A schematic diagram of a computer device provided in an embodiment of the present invention; Figure 7 This is a block diagram of a chip provided according to an embodiment of the present invention.

[0030] Among them, 60. Computer equipment; 61. Processor; 62. Memory; 63. Computer program; 600. Electronic device; 610. Processing unit; 620. Storage unit; 6201. Random access memory unit; 6202. Cache memory unit; 6203. Read-only memory unit; 6204. Program / utility; 6205. Program module; 630. Bus; 640. Display unit; 650. Input / output interface; 660. Network adapter; 700. External device. Detailed Implementation

[0031] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0032] In the description of this invention, it should be understood that the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0033] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0034] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes such combinations. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. Additionally, the character " / " in this invention generally indicates that the preceding and following objects have an "or" relationship.

[0035] It should be understood that although terms such as first, second, third, etc., may be used in the embodiments of the present invention to describe the preset range, these preset ranges should not be limited to these terms. These terms are only used to distinguish the preset ranges from one another. For example, without departing from the scope of the embodiments of the present invention, the first preset range may also be referred to as the second preset range, and similarly, the second preset range may also be referred to as the first preset range.

[0036] Depending on the context, the word "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection." Similarly, depending on the context, the phrase "if determination" or "if detection (of the stated condition or event)" can be interpreted as "when determination," "in response to determination," "when detection (of the stated condition or event)," or "in response to detection (of the stated condition or event)."

[0037] The accompanying drawings illustrate various structural schematic diagrams according to embodiments disclosed in this invention. These drawings are not to scale, and some details have been enlarged for clarity, and some details may have been omitted. The shapes of the various regions and layers shown in the drawings, as well as their relative sizes and positional relationships, are merely exemplary and may deviate from reality due to manufacturing tolerances or technical limitations. Furthermore, those skilled in the art can design regions / layers with different shapes, sizes, and relative positions as needed.

[0038] This invention relates to the process control of the detection of the throat area of ​​the guide vane of a gas turbine, specifically to a method and system for measuring the throat area of ​​the entire guide vane of a gas turbine based on ELM, belonging to the field of gas turbine measurement.

[0039] As one of the core components of a gas turbine, the guide vane mainly consists of a guide vane casing, guide vane blades, and mounting rings. An airflow channel is formed between adjacent guide vane blades, and the minimum cross-sectional area of ​​this channel is called the throat area. The throat area determines the flow capacity of the guide vane channel and directly affects factors such as the gas turbine's flow rate, thrust, and speed. Therefore, it is necessary to measure the throat area of ​​the guide vane blades.

[0040] Traditional measurement methods typically involve inspectors using specialized measuring tools to measure three parameters—the upper base, lower base, and height—of an approximately trapezoidal throat cross-section, then manually calculating the throat area. However, due to the large size of gas turbine guide vanes, manual measurement is difficult. Furthermore, variations in the force applied by inspectors when operating the measuring tools and their perception of the minimum cross-section can easily lead to inconsistent measurement results. Sometimes, even the same inspector may produce different results when measuring the same throat area. Therefore, traditional throat area measurements have poor repeatability and are quite challenging.

[0041] This invention places the guide on a coordinate measuring machine and uses a coordinate measuring program to perform point measurement; after the coordinate measuring machine point measurement is completed, the measurement points are fitted using an extreme learning machine to obtain the throat area value of each throat window.

[0042] The artificial intelligence algorithm used in this invention is the Extreme Learning Machine (ELM).

[0043] I. Overall Measurement Process Please see Figure 1 The present invention provides a method for measuring the throat area of ​​a gas turbine disc guide based on ELM, comprising the following steps: Since the guide design drawing does not provide the theoretical point of the minimum throat window, the throat area cannot be directly calculated by marking points with a coordinate measuring machine. Therefore, this invention employs an extreme learning machine (ELM) for algorithm fitting. Using an ELM for algorithm fitting requires collecting corresponding data samples.

[0044] This embodiment first collects 1000 throat area data samples. Two measurement methods are used for each throat window: coordinate measuring machine (CMM) measurement and measurement using a dedicated measuring instrument.

[0045] Three-coordinate measuring machine is used to obtain the three-dimensional coordinate data of each measuring point near the throat window; a special measuring tool is used to measure and obtain the throat area of ​​the corresponding throat window.

[0046] A correspondence was established between the data obtained from coordinate measuring machine point measurement and the throat area measured using a special measuring tool, and the collected data samples were used for extreme learning machine fitting.

[0047] If the throat area value obtained by fitting the coordinate measuring machine (CMM) data is consistent with the result obtained by using a special measuring instrument, it means that the method of fitting the throat area by combining CMM with a limit learning machine can replace the original method of measuring with a special measuring instrument.

[0048] Figure 1 The process shown includes collecting data samples, determining appropriate measurement points based on the guide structure, performing coordinate measuring machine (CMM) point measurement, measuring with a special measuring tool, using the collected data samples for artificial intelligence fitting, labeling the data, dividing the data into training and validation sets, training the extreme learning machine (ELM) model, testing the ELM model, saving the trained model, outputting the detection results, and evaluating the detection results.

[0049] II. Setting up throat feature points and coordinate measuring machine measurement Since the theoretical location of the minimum cross-section is not given in the guide design drawing, this embodiment sets the measurement point near the throat section based on the structural characteristics of the guide and the size of the special measuring tool, and then uses a coordinate measuring machine to perform point measurement.

[0050] Please see Figure 2 This is a schematic diagram of the throat area measurement of the airflow passage of the gas turbine guide vane according to an embodiment of the present invention. Two adjacent guide vane blades, along with the upper and lower edge plates, together form the airflow passage.

[0051] For ease of description, two adjacent guide vanes will be referred to as the first guide vane and the second guide vane, respectively.

[0052] Three different measurement sections are set along the radial direction of the gas turbine guide vane. Points E1, E2, and E3 are measurement points on the blade profile of the first guide vane on the three different measurement sections set along the radial direction, respectively, and are represented by point E in the figure.

[0053] Points G1, G2, and G3 are measurement points on the blade profile of the second guide vane on the three different measurement sections mentioned above, and are represented by point G in the figure.

[0054] Points H1 and H2 are measurement points on the flow channel wall located on the upper and lower edge plates, respectively.

[0055] Among them, points G1, G2, G3, E1, E2 and E3 are the feature points of the throat width to be measured; points H1 and H2 are the feature points of the throat height to be measured.

[0056] Figure 2 The diagram illustrates the positional relationships between the upper and lower edge plates, measurement points H1 and H2, and different measurement sections arranged radially. By selecting measurement points on different radially arranged measurement sections, measurement data at different locations near the throat can be obtained.

[0057] Please see Figure 3 This is a schematic diagram of the arrangement of characteristic points of the throat width of adjacent guide vanes in an embodiment of the present invention.

[0058] Figure 3 The relative positions of two adjacent guide vanes within a measurement section are shown. The blade basin profile of the first guide vane is positioned opposite to the blade back profile of the second guide vane, forming an airflow channel between them.

[0059] Measurement point E is set on the blade basin profile of the first guide vane, and measurement point G is set on the blade back profile of the second guide vane. Corresponding to the three different measurement sections set radially, measurement points E are E1, E2, and E3, and measurement points G are G1, G2, and G3, respectively.

[0060] The coordinate values ​​of the above-mentioned throat feature points are measured in different three-dimensional coordinate systems, and the three-dimensional coordinate data of the corresponding throat window are formed from the coordinate values ​​of each throat feature point.

[0061] III. Measurement Methods for the Complete Ring Guide For a complete ring guide, there are X throat windows along the circumference. The X measurement windows are numbered sequentially from 1 to X in the counterclockwise direction when observed in the forward direction.

[0062] Among them, the No. 1 throat window is a complete window of a single blade assembly, which is composed of the first measuring blade and its left-side blade when observed in the forward direction.

[0063] After determining the No. 1 throat window, record the original measurement data corresponding to each throat window in sequence according to the numbering order of the measurement windows.

[0064] Please see Figure 4 This is a schematic diagram of the guide's angular orientation according to an embodiment of the present invention. The figure shows multiple throat windows distributed circumferentially along the entire ring guide, and is used to indicate the window numbering direction when viewed in the forward heading.

[0065] The coordinate measuring machine measures the throat feature points near each throat window in sequence according to their numbering, and records the corresponding three-dimensional coordinate data.

[0066] By numbering the throat windows of the entire guide vane, the measurement data of each throat window and the throat area measurement results can be correlated, realizing the measurement of the throat area of ​​the entire guide vane, rather than being limited to the measurement of a single throat area.

[0067] IV. Extreme Learning Machine Model This invention uses an extreme learning machine to fit the mapping relationship between the three-dimensional coordinate data of throat feature points obtained by coordinate measuring machine and the throat area measured by a special measuring instrument.

[0068] The Extreme Learning Machine (SLFN) was first proposed by Professor Huang Guangbin in 2004. SLFN is a single-hidden-layer feedforward neural network algorithm. During training, input parameters are randomly generated, eliminating the need for iterative adjustments. The neural network can calculate output weights simply by pre-setting the number of hidden layer nodes. The network structure of an SLFN is as follows: Figure 5 As shown.

[0069] For the sample size is dataset , , ,in, For the first Input data for each data sample For the first Output data for each data sample.

[0070] In this embodiment, the input data is the coordinate data of the throat feature points measured by a coordinate measuring machine, and the output data is the corresponding throat area measured by a special measuring tool.

[0071] If a single hidden layer feedforward neural network (SLFN) has When there are one hidden node, the description is as follows:

[0072] in, , and For input parameters, Indicates the hidden layer The connection weights between each neuron and the output layer It is the activation function of hidden layer neurons.

[0073] Hidden layer output matrix Represented as:

[0074]

[0075]

[0076]

[0077] Because the input learning parameters are generated randomly, no adjustment is needed. Hidden layer and output connection weights By solving the equation, we obtain:

[0078] in, It is the hidden layer output matrix Moore-Penrose generalized inverse.

[0079] Therefore, based on the three-dimensional coordinate data of the throat feature points obtained by the coordinate measuring machine and the corresponding throat area measured by the special measuring tool, an extreme learning machine model for outputting the throat area is trained.

[0080] V. Training and Validation Sets The collected data samples are divided into training set and validation set.

[0081] The training set is used to train the Extreme Learning Machine (ELM) model. The 3D coordinate data of each data sample in the training set is used as the input to the ELM model, and the throat area measured by the corresponding specialized measuring instrument is used as the output of the ELM model.

[0082] The number of hidden layer nodes in the extreme learning machine model is set, input parameters are randomly generated, and the connection weights from the hidden layer to the output layer are solved based on the hidden layer output matrix and the throat area corresponding to each data sample in the training set, thus obtaining the throat area measurement model.

[0083] The validation set is used to test the trained throat area measurement model. The three-dimensional coordinate data from the validation set is input into the trained throat area measurement model to obtain the corresponding throat area fitting value, which is then compared with the throat area measured by a dedicated measuring instrument.

[0084] When the measurement deviation meets the requirement of 1%, the corresponding measurement result is considered accurate; when the validation set accuracy reaches 98%, the trained extreme learning machine model is considered usable, and the trained model is saved.

[0085] This embodiment does not further limit the specific calculation method of the measurement bias and validation set accuracy.

[0086] VI. Measurement of the throat area of ​​the guide vane of the gas turbine under test After the extreme learning machine model is trained and meets the verification requirements, the gas turbine guide under test is placed on the coordinate measuring machine.

[0087] Following the aforementioned numbering method for the entire ring guide, each throat window of the gas turbine guide under test is numbered.

[0088] The coordinate measuring machine acquires the three-dimensional coordinate data of the throat feature points corresponding to each throat window in sequence according to the numbering order.

[0089] By inputting the three-dimensional coordinate data corresponding to each throat window into the saved extreme learning machine model, the throat area value of the corresponding throat window can be obtained by fitting.

[0090] By establishing a correspondence between the throat area measurement results output by the extreme learning machine model and the corresponding throat window number, the throat area measurement results of each throat window of the whole gas turbine guide are obtained.

[0091] After the Extreme Learning Machine (ELM) model training is completed, the subsequent throat area measurement can be carried out by combining a coordinate measuring machine (CMM) and an ELM, eliminating the need for inspectors to manually measure each throat area one by one using specialized measuring tools.

[0092] In another embodiment of the present invention, an ELM-based throat area measurement system for gas turbine disc guides is provided, comprising a coordinate measuring machine, a special measuring tool, and a computer device.

[0093] A coordinate measuring machine (CMM) is used to measure multiple throat feature points located near the throat section of the gas turbine guide vane window, based on the structural characteristics of the gas turbine guide vane and the dimensions of a dedicated measuring instrument. These throat feature points are located on two adjacent guide vane blades forming the throat window, as well as on the upper and lower edge plates, and are used to obtain the three-dimensional coordinate data of these throat feature points.

[0094] A specialized measuring instrument is used to measure the throat area of ​​the throat window. During the model training and validation phases, the throat area measured by the specialized measuring instrument serves as the target output and calibration data corresponding to the three-dimensional coordinate data.

[0095] A computer device includes a processor and a memory, in which a computer program is stored, and the processor executes the computer program to cause the computer device to: Acquire the three-dimensional coordinate data measured by the coordinate measuring machine and the throat area measured by the special measuring tool, establish the sample correspondence between the two, and form a data sample; The data samples are divided into training and validation sets; Construct an extreme learning machine model, take the three-dimensional coordinate data in the training set as input and the corresponding throat area as output, solve the output weights of the extreme learning machine model, and obtain the throat area measurement model; The throat area measurement model was validated using a validation set, and the model was saved when the validation set accuracy reached 98%. Each throat window of the gas turbine guide under test is numbered, and the three-dimensional coordinate data of each throat window is obtained by a coordinate measuring machine in the order of numbering. The three-dimensional coordinate data is then input into the saved throat area measurement model to obtain the corresponding throat area measurement result.

[0096] This invention provides a terminal device comprising a processor and a memory. The memory stores a computer program, which includes program instructions. The processor executes the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, graphics processing units (GPUs), tensor processing units (TPUs), digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions to achieve a corresponding method flow or corresponding function. The processor described in this embodiment can be used for the operation of an ELM-based method for measuring the throat area of ​​a gas turbine's overall guide vane, including: For multiple gas turbine guideway throat windows, based on the guideway structural characteristics and the dimensions of the specialized measuring instrument, multiple throat feature points are determined near the throat cross-section. A coordinate measuring machine is used to acquire the three-dimensional coordinate data of each throat feature point on adjacent guideway blades, upper and lower edge plates, and the corresponding throat area is obtained using the specialized measuring instrument. A sample correspondence is established between the three-dimensional coordinate data and the corresponding throat area to form a data sample. The data sample is divided into a training set and a validation set. An extreme learning machine model is constructed, using the three-dimensional coordinate data from the training set as the model input and the corresponding throat area as the model output. The number of hidden layer nodes is set, and the input from the input layer to the hidden layer is randomly generated. The weights and hidden layer node biases are calculated, and the output weights from the hidden layer to the output layer are obtained based on the hidden layer output matrix and the corresponding throat area to obtain the throat area measurement model. The throat area measurement model is verified using the validation set, and the throat area measurement model is saved when the verification result meets the preset accuracy conditions. Each throat window of the gas turbine guide under test is numbered, and the three-dimensional coordinate data of the throat feature points corresponding to each throat window is obtained using the coordinate measuring machine according to the numbering order. The three-dimensional coordinate data of the throat feature points corresponding to each throat window are input into the saved throat area measurement model to obtain the throat area measurement result corresponding to each throat window number.

[0097] Please see Figure 6The terminal device is a computer device. In this embodiment, the computer device 60 includes a processor 61, a memory 62, and a computer program 63 stored in the memory 62 and executable on the processor 61. When executed by the processor 61, the computer program 63 implements the ELM-based method for measuring the throat area of ​​a gas turbine disc guide in this embodiment. To avoid repetition, these details are not elaborated here. Alternatively, when executed by the processor 61, the computer program 63 implements the functions of each model / unit in the ELM-based gas turbine disc guide throat area measurement system of this embodiment. To avoid repetition, these details are not elaborated here.

[0098] Computer device 60 can be a desktop computer, laptop, handheld computer, cloud server, or other computing device. Computer device 60 may include, but is not limited to, a processor 61 and a memory 62. Those skilled in the art will understand that... Figure 6 This is merely an example of computer device 60 and does not constitute a limitation on computer device 60. It may include more or fewer components than shown, or combine certain components, or different components. For example, computer device may also include input / output devices, network access devices, buses, etc.

[0099] The processor 61 may be a Central Processing Unit (CPU), or other general-purpose processors, graphics processing units (GPUs), tensor processing units (TPUs), digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0100] The memory 62 can be an internal storage unit of the computer device 60, such as a hard disk or memory of the computer device 60. The memory 62 can also be an external storage device of the computer device 60, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc. equipped on the computer device 60.

[0101] Furthermore, the memory 62 may include both internal storage units of the computer device 60 and external storage devices. The memory 62 is used to store computer programs and other programs and data required by the computer device. The memory 62 can also be used to temporarily store data that has been output or will be output.

[0102] Please see Figure 7 The terminal device is an electronic device 600, which is manifested in the form of a general-purpose computing device. The components of the electronic device may include, but are not limited to: at least one processing unit 610, at least one storage unit 620, a bus 630 connecting different platform components (including storage unit 620 and processing unit 610), a display unit 640, etc.

[0103] The storage unit stores program code, which can be executed by the processing unit 610 to perform the steps described in the method section of this specification according to various exemplary embodiments of the present invention. For example, the processing unit 610 can perform actions such as... Figure 1 The steps are shown in the figure.

[0104] Storage unit 620 may include readable media in the form of volatile storage units, such as random access memory (RAM) 6201 and / or cache memory 6202, and may further include read-only memory (ROM) 6203.

[0105] Storage unit 620 may also include a program / utility 6204 having a set (at least one) program module 6205, such program module 6205 including but not limited to: operating system, one or more application programs, other program modules and program data, each or some combination of these examples may include an implementation of a network environment.

[0106] Bus 630 can represent one or more of several types of bus structures, including a memory cell bus or memory cell controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus using any of the multiple bus structures.

[0107] Electronic device 600 can also communicate with one or more external devices 700 (e.g., keyboard, pointing device, Bluetooth device, etc.), and with one or more devices that enable a user to interact with electronic device 600, and / or with any device that enables electronic device 600 to communicate with one or more other computing devices (e.g., router, modem). This communication can be performed via input / output interface 650. Furthermore, electronic device 600 can also communicate with one or more networks (e.g., local area network, wide area network, and / or public network, such as the Internet) via network adapter 660. Network adapter 660 can communicate with other modules of electronic device 600 via bus 630. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with electronic device 600, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage platforms.

[0108] This invention also provides a storage medium, specifically a computer-readable storage medium, which is a memory device in a terminal device for storing programs and data. It is understood that the computer-readable storage medium here can include both built-in storage media in the terminal device and extended storage media supported by the terminal device; it can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. The computer-readable storage medium provides storage space that stores the terminal's operating system. Furthermore, the storage space also stores one or more instructions suitable for loading and execution by a processor, which can be one or more computer programs (including program code). More specific examples of the computer-readable storage medium include: an electrical connection with one or more wires, a portable disk, a hard disk, random access memory, read-only memory, erasable programmable read-only memory, optical fiber, portable compact disk read-only memory, optical storage device, magnetic storage device, or any suitable combination thereof.

[0109] Computer-readable storage media also include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A readable storage medium can also be any readable medium other than a readable storage medium that can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the readable storage medium can be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, radio frequency, etc., or any suitable combination thereof.

[0110] Program code for performing the operations of this invention can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java and C++, and conventional procedural programming languages ​​such as C or similar languages. The program code can execute entirely on the user's computing device, partially on the user's device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0111] One or more instructions stored in a computer-readable storage medium can be loaded and executed by a processor to implement the corresponding steps of the ELM-based gas turbine whole-disk guide throat area measurement method in the above embodiments; one or more instructions in the computer-readable storage medium are loaded by the processor and executed as follows: For multiple gas turbine guideway throat windows, based on the guideway structural characteristics and the dimensions of the specialized measuring instrument, multiple throat feature points are determined near the throat cross-section. A coordinate measuring machine is used to acquire the three-dimensional coordinate data of each throat feature point on adjacent guideway blades, upper and lower edge plates, and the corresponding throat area is obtained using the specialized measuring instrument. A sample correspondence is established between the three-dimensional coordinate data and the corresponding throat area to form a data sample. The data sample is divided into a training set and a validation set. An extreme learning machine model is constructed, using the three-dimensional coordinate data from the training set as the model input and the corresponding throat area as the model output. The number of hidden layer nodes is set, and the input from the input layer to the hidden layer is randomly generated. The weights and hidden layer node biases are calculated, and the output weights from the hidden layer to the output layer are obtained based on the hidden layer output matrix and the corresponding throat area to obtain the throat area measurement model. The throat area measurement model is verified using the validation set, and the throat area measurement model is saved when the verification result meets the preset accuracy conditions. Each throat window of the gas turbine guide under test is numbered, and the three-dimensional coordinate data of the throat feature points corresponding to each throat window is obtained using the coordinate measuring machine according to the numbering order. The three-dimensional coordinate data of the throat feature points corresponding to each throat window are input into the saved throat area measurement model to obtain the throat area measurement result corresponding to each throat window number.

[0112] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0113] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0114] This invention addresses situations where the design drawings of gas turbine guide vanes do not provide the theoretical cross-sectional area of ​​the blade throat or the theoretical location of the minimum throat cross-section. Based on the structural characteristics of the guide vane and the dimensions of specialized measuring instruments, this invention sets multiple throat feature points near the throat cross-section. A coordinate measuring machine (CMM) is used to acquire the three-dimensional coordinate data of each throat feature point on adjacent guide vane blades and the upper and lower edge plates. A specialist then uses specialized measuring instruments to measure the corresponding throat area. By establishing a sample correspondence between the three-dimensional coordinate data and the measurement results from the specialized measuring instruments, an extreme learning machine model is trained. This model learns the mapping relationship between the coordinates of the throat feature points and the throat area, thus avoiding the limitation of having to accurately determine the theoretical minimum cross-section location before performing geometric calculations.

[0115] After the Extreme Learning Machine (ELM) model is trained and validated, the qualified model is saved. For the subsequent gas turbine guide vanes to be tested, only the three-dimensional coordinate data of each throat window needs to be collected using a coordinate measuring machine at predetermined measurement points. The three-dimensional coordinate data is then input into the saved ELM model to obtain the corresponding throat area. This eliminates the need for inspectors to manually measure each throat window with specialized measuring tools, reducing manual operation steps and inspection time, lowering manual inspection costs, and facilitating engineering applications during production testing.

[0116] Traditional manual measurements are easily affected by the magnitude of force applied by the inspector, the placement of the measuring instrument, and differences in the judgment of the minimum cross-section. Even if the same inspector repeatedly measures the same throat window, different measurement results may occur. This invention uses a coordinate measuring machine to collect the coordinates of throat feature points according to a predetermined measurement procedure, and processes the measurement data according to a unified mapping relationship using a stored limit learning machine model. This helps to reduce the impact of human operation differences on the results, enabling the same throat window to obtain measurement results with better repeatability and consistency under the same measurement conditions.

[0117] This invention sequentially numbers each throat window of the whole-disc guide in a counter-clockwise direction when observed from the forward direction, and collects three-dimensional coordinate data, inputs the model, and outputs the throat area according to the numbering order, so that a clear correspondence is established between the throat window number, the original measurement data, and the area measurement results. This method enables continuous measurement of multiple throat windows of the whole-disc guide, rather than detecting only a single throat window, and helps to reduce missed measurements, repeated measurements, or mismatches between measurement results and window positions.

[0118] Furthermore, this invention does not rely on analytical geometric calculations of simple regular cross-sections, but instead utilizes multi-position three-dimensional feature points that can characterize the throat width and height as input to the limit learning machine. Therefore, it is more suitable for throat structures with complex blade profiles, limited measurement space, and difficult-to-determine minimum cross-section locations. By establishing the sample correspondence between coordinate measuring machine data and measurement results from specialized measuring instruments, an area fitting model suitable for specific guide structures can be built based on existing measurement results.

[0119] For window areas in other blade-like components that are difficult to measure directly, the same technical approach can be adopted. This involves setting feature points near the window that reflect its spatial shape, acquiring three-dimensional coordinate data using a coordinate measuring machine, and using the window area obtained through existing measurement methods as the corresponding output to train an extreme learning machine model to achieve a fitted measurement of the window area. Therefore, this invention is not only applicable to the throat area detection of gas turbine disc guide vanes, but can also be extended to other blade-like components with complex curved surfaces and difficult-to-measure window structures.

[0120] In summary, this invention presents an ELM-based method and system for measuring the throat area of ​​a gas turbine guide vane. Addressing the issues of the lack of theoretical location of the minimum throat cross-section in the gas turbine guide vane design drawings, the difficulty in directly converting coordinate measuring machine (CMM) data into throat area, and the significant influence of operating force, tool posture, and operator experience on manual measurement using specialized measuring tools, this invention proposes a throat area measurement scheme based on Extreme Learning Machine (ELM). By setting measurement points near the throat cross-section for the blade basin profile, blade back profile, and upper and lower edge plates, a CMM is used to acquire the three-dimensional coordinate data of the throat feature points. The throat area measured by the specialized measuring tool is used as the corresponding output to train the ELM model, thereby establishing a mapping relationship between the three-dimensional coordinate data and the throat area. This invention eliminates the need to directly determine the theoretical location of the actual minimum throat cross-section and the need to establish an analytical area calculation model for complex blade profiles. By uniformly numbering and sequentially measuring each throat window of the entire guide, the throat area results can be matched one-to-one with the window position, reducing missed measurements, duplicate measurements, and data mismatches, improving detection efficiency, measurement repeatability, and result traceability, and reducing manual inspection costs. It is suitable for engineering batch inspection and can also be extended to other blade-type parts where the window area is difficult to measure directly.

[0121] The above content is only for illustrating the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made to the technical solution based on the technical concept proposed in this invention shall fall within the scope of protection of the claims of this invention.

Claims

1. An ELM-based method for measuring the throat area of a gas turbine whole disk guide, characterized in that, Includes the following steps: S1. For multiple gas turbine guideway throat windows, based on the structural characteristics of the guideway and the dimensions of the special measuring tool, determine multiple throat feature points near the throat cross section; use a coordinate measuring machine to obtain the three-dimensional coordinate data of each throat feature point on the upper and lower edge plates of two adjacent guideway blades, and use a special measuring tool to obtain the corresponding throat area; establish a sample correspondence relationship between the three-dimensional coordinate data and the corresponding throat area to form a data sample. S2. Divide the data samples into a training set and a validation set; S3. Construct an extreme learning machine model, using the three-dimensional coordinate data in the training set as the model input and the corresponding throat area as the model output, set the number of hidden layer nodes, randomly generate the input weights and hidden layer node biases from the input layer to the hidden layer, and solve the output weights from the hidden layer to the output layer based on the hidden layer output matrix and the corresponding throat area to obtain the throat area measurement model. S4. Validate the throat area measurement model using the validation set, and save the throat area measurement model when the validation result meets the preset accuracy conditions. S5. Number each throat window of the gas turbine guide to be tested, and use the coordinate measuring machine to obtain the three-dimensional coordinate data of the throat feature points corresponding to each throat window according to the numbering order. S6. Input the three-dimensional coordinate data of the throat feature points corresponding to each throat window into the saved throat area measurement model to obtain the throat area measurement results corresponding to each throat window number.

2. The ELM-based gas turbine whole disk guide throat area measurement method of claim 1, wherein, The two adjacent guide vanes are respectively the first guide vane and the second guide vane; the throat feature points include the blade basin profile measurement point of the first guide vane and the blade back profile measurement point of the second guide vane, which are respectively located on three measurement sections arranged radially, as well as the flow channel wall measurement points located on the upper edge plate and the lower edge plate.

3. The ELM-based gas turbine whole disk guide throat area measurement method of claim 2, wherein, The three blade basin profile measurement points of the first guide vane and the three blade back profile measurement points of the second guide vane constitute the throat width feature points; the flow channel wall measurement points located on the upper edge plate and the flow channel wall measurement points located on the lower edge plate constitute the throat height feature points.

4. The ELM-based gas turbine whole disk guide throat area measurement method of claim 2, wherein, The coordinate values ​​of each throat feature point in the corresponding three-dimensional coordinate system are measured respectively, and the coordinate values ​​of each throat feature point are used to construct the three-dimensional coordinate data of the corresponding throat window.

5. The ELM-based gas turbine whole disk guide throat area measurement method of claim 1, wherein, The extreme learning machine model is a single hidden layer feedforward neural network, and the hidden layer output matrix is ​​determined based on the input weights, the hidden layer node biases, the three-dimensional coordinate data, and the activation functions of the hidden layer nodes.

6. The ELM-based gas turbine whole disk guide throat area measurement method of claim 5, wherein, The output weights are obtained by solving the Moore-Penrose generalized inverse of the hidden layer output matrix and the throat area corresponding to each data sample in the training set.

7. The ELM-based gas turbine whole disk guide throat area measurement method of claim 1, wherein, In step S4, the three-dimensional coordinate data of the verification set is input into the throat area measurement model to obtain the corresponding verification area. The throat area measurement model is then verified based on the measurement deviation between the verification area and the throat area measured by the corresponding special measuring instrument, as well as the accuracy of the verification set.

8. The ELM-based gas turbine whole disk guide throat area measurement method of claim 1, wherein, In step S5, when observing the gas turbine guide in the forward direction, the throat windows of the gas turbine guide are numbered sequentially in the counterclockwise direction, and the corresponding three-dimensional coordinate data are recorded according to the numbering order of each throat window.

9. The ELM-based gas turbine bladed disk guide throat area measurement method of claim 8, wherein, The complete throat window, consisting of the first guide vane under test and the adjacent guide vane located to the left of the first guide vane under test when observing the gas turbine guide vane along the forward direction, is defined as the first throat window. Starting from the first throat window, the remaining throat windows of the gas turbine guide vane under test are numbered sequentially in a counterclockwise direction.

10. An ELM-based gas turbine whole disk guider throat area measurement system, characterized in that, This includes coordinate measuring machines, specialized measuring tools, and computer equipment; The coordinate measuring machine is used to measure multiple throat feature points determined near the throat section of the throat window of the gas turbine guide based on the structural characteristics of the gas turbine guide and the dimensions of the special measuring tool. The throat feature points are located on the two adjacent guide blades that make up the throat window, as well as the upper and lower edge plates, so as to obtain the three-dimensional coordinate data of the throat feature points. The special measuring instrument is used to measure the throat area of ​​the throat window; The computer device includes a processor and a memory, the memory storing a computer program, and the processor executing the computer program to cause the computer device to: The three-dimensional coordinate data and the corresponding throat area are obtained. A sample correspondence relationship is established between the three-dimensional coordinate data and the corresponding throat area to form a data sample. The data sample is then divided into a training set and a validation set. An extreme learning machine model is constructed, with the three-dimensional coordinate data in the training set as the model input and the corresponding throat area as the model output. The number of hidden layer nodes is set, and the input weights and hidden layer node biases from the input layer to the hidden layer are randomly generated. The output weights from the hidden layer to the output layer are solved according to the hidden layer output matrix and the corresponding throat area to obtain the throat area measurement model. The throat area measurement model is validated using the validation set, and the throat area measurement model is saved when the validation results meet the preset accuracy conditions. Each throat window of the gas turbine guide under test is numbered, and the three-dimensional coordinate data of the throat feature points corresponding to each throat window are obtained by the coordinate measuring machine according to the numbering order. The three-dimensional coordinate data is then input into the saved throat area measurement model to obtain the throat area measurement result corresponding to each throat window number.