Gas turbine blade adaptive clamping fixing method, device, equipment and medium

Abstract

This disclosure discloses an adaptive clamping and fixing method, apparatus, device, and medium for gas turbine blades. One specific embodiment of the method includes: performing model detection processing on the gas turbine blade to be processed, located in the detection area at the clamp slide rail; driving a positioning clamp module corresponding to the blade profile version information to move along the slide rail connected to the detection area at the clamp slide rail to a preset receiving position where the gas turbine blade to be processed is located; controlling each of the at least one adjustable positioning and fastening devices to perform gap detection processing on the gas turbine blade to be processed in its working area; and controlling each of the at least one adjustable support devices to perform height support adaptive compensation processing on the gas turbine blade to be processed in its working area. This embodiment shortens the time required for changing the fixing clamp and simplifies the fixing operation.

Description

Technical Field

[0001] The embodiments disclosed herein relate to the field of computer technology, and specifically to a method, apparatus, device, and medium for adaptive clamping and fixing of gas turbine blades. Background Technology

[0002] In the research, development, trial production, or maintenance of gas turbine blades, situations often arise where "the same blade has different blade shapes." For example, for the same type of blade, there may be multiple design iterations or performance-tuned versions designed for different operating conditions. These versions of blades may have the same macroscopic installation reference, but the curvature and position of the blade profile (blade head, blade back) differ. Adaptive clamping and fixing of gas turbine blades is a technique for clamping and fixing the gas turbine blades to be processed. Currently, the common method for clamping and fixing gas turbine blades is as follows: a dedicated rigid positioning fixture module is designed and manufactured for each blade shape version. When processing different versions of blades, the entire positioning fixture module is manually disassembled and replaced to achieve version switching, and the positioning fixture module is recalibrated. After the positioning fixture module is switched, the fit between the positioning fixture module and the blade is manually checked and repeatedly adjusted using feeler gauges or dial indicators to fix the gas turbine blade to be processed.

[0003] However, when using the above method to clamp and fix gas turbine blades, the following technical problems often arise: The manual disassembly and replacement of the entire set of fixing fixtures to accommodate different blade shapes is time-consuming. Furthermore, after each changeover, the positioning accuracy of the fixing fixtures needs to be recalibrated and re-aligned. This requires manual inspection and repeated adjustments using feeler gauges or dial indicators to check the fit between the fixture positioning surface and the blade, making the fixing process quite cumbersome. Additionally, the actual gap between the support positioning device and the blade profile can easily become a gap due to individual variations in the blank or batch differences, resulting in a clearance between the support positioning device and the blade.

[0004] The information disclosed in this background section is only intended to enhance the understanding of the background of the inventive concept, and therefore may contain information that does not form prior art known to those skilled in the art. Summary of the Invention

[0005] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0006] Some embodiments of this disclosure provide adaptive clamping and fixing methods, apparatuses, electronic devices, and computer-readable media for gas turbine blades to solve one or more of the technical problems mentioned in the background section above.

[0007] In a first aspect, some embodiments of this disclosure provide an adaptive clamping and fixing method for gas turbine blades. The method includes: performing model detection processing on a gas turbine blade to be processed, located in a detection area at a clamp slide rail, to obtain blade shape version information; driving a positioning clamp module corresponding to the blade shape version information to move along a slide rail connected to the detection area at the clamp slide rail to a preset receiving position where the gas turbine blade to be processed is located, to receive the gas turbine blade to be processed, wherein the positioning clamp module includes at least one adjustable positioning fastening device and at least one adjustable support device; for each of the at least one adjustable positioning fastening devices, controlling the adjustable positioning fastening device to perform a gap detection processing on the gas turbine blade to be processed in its working area to lock the blade in its working area; for each of the at least one adjustable support device, controlling the adjustable support device to perform height support adaptive compensation processing on the gas turbine blade to be processed in its working area to support the blade in its working area.

[0008] Secondly, some embodiments of this disclosure provide an adaptive clamping and fixing device for gas turbine blades. The device includes: a model detection unit configured to perform model detection processing on the gas turbine blade to be processed, which is disposed in the detection area of ​​the clamp slide rail, to obtain blade shape version information; a drive unit configured to drive a positioning clamp module corresponding to the blade shape version information to move along the slide rail connected to the detection area of ​​the clamp slide rail to a preset receiving position where the gas turbine blade to be processed is located, so as to receive the gas turbine blade to be processed, wherein the positioning clamp module includes at least one adjustable positioning fastening device and at least one adjustable support device; a first control unit configured to control each of the at least one adjustable positioning fastening device to perform a gap detection processing on the gas turbine blade to be processed in its working area, so as to lock the blade in its working area; and a second control unit configured to control each of the at least one adjustable support device to perform height support adaptive compensation processing on the gas turbine blade to be processed in its working area, so as to support the blade in its working area.

[0009] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation of the first aspect above.

[0010] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method described in any of the implementations of the first aspect above.

[0011] The above-described embodiments of this disclosure have the following beneficial effects: the adaptive clamping and fixing method for gas turbine blades according to some embodiments of this disclosure improves and shortens the time required for changing the type of fixing fixture and simplifies the fixing operation. Specifically, the reason why changing the type of fixing fixture is time-consuming and the fixing operation is relatively cumbersome is that manually disassembling and replacing the entire set of fixing fixtures to adapt to different blade shapes is time-consuming, and after each change, the positioning accuracy of the fixing fixture needs to be recalibrated. Manually using feeler gauges or dial indicators to check the fit between the fixture positioning surface and the blade is required, and repeated adjustments are made, making the fixing operation cumbersome. At the same time, the actual gap between the support positioning device and the blade profile is easily caused by individual differences in the blank or batch differences, resulting in a gap between the support positioning device and the blade. Based on this, the adaptive clamping and fixing method for gas turbine blades according to some embodiments of this disclosure first performs model detection processing on the gas turbine blade to be processed, which is set in the detection area at the fixture slide rail, to obtain the blade shape version information. Thus, the blade shape version can be automatically identified and the blade shape version information obtained. Next, the positioning fixture module corresponding to the aforementioned blade version information is moved along the slide rail connected to the detection area at the fixture slide rail to the preset receiving station where the gas turbine blade to be processed is located, to receive the gas turbine blade to be processed. The positioning fixture module includes at least one adjustable positioning fastening device and at least one adjustable support device. Thus, the positioning fixture module adapted to the gas turbine blade to be processed can be switched to the corresponding preset station via the slide rail, eliminating the need for manual disassembly and replacement of the entire fixture set, saving the process of recalibrating after a changeover, shortening the changeover time, and enabling rapid switching of positioning fixture modules for multiple blade versions without the need for recalibrating the positioning accuracy after each changeover. Next, for each of the at least one adjustable positioning fastening device, the device is controlled to perform a gap detection process on the gas turbine blade to be processed in its working area to lock the blade in its working area. An adjustable positioning and fastening device is used to detect gaps in the working area and lock the blade, replacing manual inspection and repeated adjustments using feeler gauges or dial indicators. Subsequently, for each of the at least one adjustable support device, the device is controlled to provide adaptive height compensation for the gas turbine blade to be processed in its working area. This automatically detects the support gap between the support device and the blade bottom surface, and automatically compensates for support height deviations caused by individual or batch differences in the blank, ensuring reliable contact between the adjustable support device and the blade surface.Because of the rapid switching between model detection and positioning fixture modules, the compatibility of fixture changes is ensured, eliminating the need for manual disassembly and replacement. This also eliminates the need for recalibrating positioning accuracy after a change in fixture type. An adjustable positioning and fastening device detects gaps in the working area and locks the blade, replacing manual inspection and repeated adjustments using feeler gauges or dial indicators. This shortens the time required for fixture changes and simplifies the fixing operation. Simultaneously, the system automatically detects the support gap between the support device and the blade bottom surface and automatically compensates for support height deviations caused by individual or batch differences in the blank, ensuring a reliable fit between the adjustable support device and the blade surface. Attached Figure Description

[0012] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0013] Figure 1 This is a flowchart of some embodiments of the adaptive clamping and fixing method for gas turbine blades according to the present disclosure; Figure 2 This is a schematic diagram of the structure of some embodiments of the adaptive clamping and fixing device for gas turbine blades according to the present disclosure; Figure 3 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure; Figure 4 This is a structural schematic diagram of some embodiments of the positioning fixture module according to the present disclosure. Detailed Implementation

[0014] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0015] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0016] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0017] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0018] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0019] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0020] Figure 1 A flowchart 100 is shown, illustrating some embodiments of the adaptive clamping and fixing method for gas turbine blades according to this disclosure. The adaptive clamping and fixing method for gas turbine blades includes the following steps: Step 101: Perform model detection processing on the gas turbine blades to be processed, which are set in the detection area at the fixture slide rail, to obtain blade shape version information.

[0021] In some embodiments, the execution subject (e.g., a computing device) of the adaptive clamping and fixing method for gas turbine blades can perform model detection processing on the gas turbine blade to be processed, which is located in the detection area of ​​the clamp slide rail, to obtain blade shape version information. The detection area at the clamp slide rail can be a preset area at the end of the slide rail used to place and identify the model of the gas turbine blade to be processed. The gas turbine blade to be processed can be a gas turbine blade blank or semi-finished product that has not yet completed all machining processes and needs to be clamped and fixed. The gas turbine blade to be processed can be suspended in the detection area of ​​the clamp slide rail by an external material conveying device. The external material conveying device can be an industrial robot or a robotic arm. The blade shape version information refers to the identification information used to distinguish different design iteration versions or different performance fine-tuning versions of the same model of gas turbine blade. The blade shape version information can be identified by characters, representing one of the following: version number, version name, or version code.

[0022] In some optional implementations of certain embodiments, the aforementioned execution entity can perform model detection processing on the gas turbine blade to be processed, which is located in the detection area at the fixture slide rail, through the following steps to obtain blade shape version information: The first step involves acquiring images of a predetermined area of ​​the gas turbine blade to be processed, resulting in a blade identification image. This predetermined area contains pre-etched or pre-pasted character markings. The predetermined area refers to the region on the surface of the gas turbine blade to be processed where character markings have been pre-processed or pasted. These character markings can be strings composed of numbers, letters, or combinations thereof. In practice, the executing entity can use an image acquisition device (e.g., a camera) to acquire images of the predetermined area of ​​the gas turbine blade to be processed as the blade identification image.

[0023] The second step is to process the above leaf identification images to obtain leaf shape version information.

[0024] In addressing the technical problems mentioned in the background section using the above-described technical solutions, and considering the application scenario—an automated production line for mixed-line machining of multiple versions of gas turbine blades—the use of optical character recognition (OCR) to obtain blade shape version information from pre-etched or pasted character markings on the blade surface often presents the following technical challenges: When the pre-etched character markings on the blade surface experience localized wear and corrosion due to long-term use, resulting in broken character strokes, or when slight oil contamination reduces the contrast between the character area and the background, directly processing the blade marking image will yield incorrect character recognition results. Incorrect character recognition results lead to incorrect blade shape version information, causing the positioning module to switch to a preset receiving station that does not match the actual blade shape, resulting in blade clamping failure or increased blade loosening and vibration during machining. The following requirements are necessary for this application scenario: The blade shape version identification of gas turbine blades needs to adapt to common industrial conditions such as wear, corrosion, and oil contamination on the blade surface markings due to long-term use. This ensures accurate identification of blade shape version information even with localized defects in the markings, avoiding misjudgments due to poor marking conditions. This, in turn, guarantees that the positioning module is accurately switched to the preset receiving station that matches the actual blade shape. To address these technical challenges, we have decided to adopt the following solution: In some optional implementations of certain embodiments, the aforementioned execution entity can perform identification processing on the aforementioned leaf identification image through the following steps to obtain leaf shape version information: The first step is to perform grayscale conversion and noise reduction on the aforementioned leaf identification image to obtain a preprocessed leaf identification image. In practice, the execution entity can first perform grayscale conversion on the aforementioned leaf identification image, converting the three channels of the color image into a single grayscale channel using a weighted average. Then, Gaussian filtering is applied to the grayscale image to suppress high-frequency random noise in the image, resulting in the preprocessed leaf identification image. The aforementioned preprocessed leaf identification image can refer to the grayscale image obtained after grayscale conversion and noise reduction of the leaf identification image.

[0025] The second step involves performing distortion correction processing on the preprocessed blade identification image to obtain a distortion-corrected blade identification image. In practice, firstly, the executing entity can obtain the camera intrinsic parameter matrix and distortion coefficients of the image acquisition device. Then, based on the camera intrinsic parameter matrix and distortion coefficients, coordinate transformation and interpolation reconstruction are performed on each pixel in the preprocessed blade identification image to eliminate radial and tangential distortion, resulting in the distortion-corrected blade identification image. Here, the camera intrinsic parameter matrix refers to the parameter matrix describing the camera's focal length and principal point position. The distortion coefficients refer to the mathematical parameters describing the radial and tangential distortion of the camera lens. The distortion-corrected blade identification image can be a blade identification image that has undergone the above distortion correction processing, eliminating perspective distortion and lens distortion.

[0026] The third step involves performing connected component detection on the aforementioned distortion-corrected blade identifier image to obtain a sequence of candidate character regions. In practice, connected component detection technology can be used to perform connected component detection on the aforementioned distortion-corrected blade identifier image to obtain information on each connected region. Each connected region information is represented by the vertex coordinates of the rectangle representing the connected region. Then, the connected region information is sorted from smallest to largest according to the horizontal coordinates of the center point of the region represented by the connected region information, resulting in a sequence of connected region information that serves as the sequence of candidate character regions.

[0027] Fourth, for each character candidate region in the above character candidate region information sequence, perform the following steps: The first sub-step involves performing stroke breakage detection processing on the region image corresponding to the aforementioned candidate character region information in the distortion-corrected blade identification image, obtaining a stroke breakage detection score. Here, the aforementioned region image refers to the local sub-image in the distortion-corrected blade identification image defined by the bounding box of the aforementioned candidate character region information. In practice, the executing entity can perform skeleton extraction processing on the aforementioned region image to obtain the skeleton lines of the character strokes in the aforementioned region image. The skeleton extraction processing can employ the Zhang-Suen thinning algorithm. Then, endpoint detection processing is performed on the aforementioned skeleton lines to identify all endpoints in the aforementioned skeleton lines. Next, the number of endpoint pairs satisfying preset pairing conditions is calculated. The preset pairing conditions can be that the Euclidean distance between the endpoints is less than a preset endpoint distance threshold and the angle between the extension directions of the two endpoints is less than a preset direction angle threshold. Finally, based on the number of endpoint pairs that satisfy the aforementioned preset pairing conditions, a stroke breakage detection score is generated. A higher number of endpoint pairs indicates a more severe stroke breakage and a higher score. For example, the executing entity can assign a first preset score (e.g., 0) to the stroke breakage detection score corresponding to the region image where the number of endpoint pairs satisfying the aforementioned preset pairing conditions is less than or equal to a first threshold (e.g., 3). Then, the executing entity can assign a second preset score (e.g., 5) to the stroke breakage detection score corresponding to the region image where the number of endpoint pairs satisfying the aforementioned preset pairing conditions is greater than the first threshold. The aforementioned Euclidean distance refers to the straight-line distance between two pixels on the image plane. The aforementioned endpoint extension direction refers to the local tangent direction of the skeleton calculated by tracing back a preset number of pixels along the skeleton line from the endpoint.

[0028] The second sub-step involves performing stroke repair processing on the region corresponding to the candidate character area information in the distortion-corrected blade identification image, in response to determining that the stroke breakage detection score is greater than a preset score threshold. For example, the preset score threshold can be 4. In practice, the executing entity can perform connection and completion processing on the character strokes in the region corresponding to the candidate character area information in the distortion-corrected blade identification image through morphological operations.

[0029] The fourth step is to identify the distorted leaf identification image after stroke restoration as the leaf identification image to be identified.

[0030] The fifth step involves inputting the aforementioned leaf identification image into a pre-trained identification recognition model to obtain identification information as leaf shape version information. This identification recognition model can be an OCR model. For example, the leaf shape version information could be "V1".

[0031] The above technical solution, combined with steps 102 to 104 and related content, serves as an inventive point of this disclosure, solving the technical problem of "increased frequency of blade clamping failure or blade loosening and vibration during processing." Factors leading to blade clamping failure or increased frequency of blade loosening and vibration during processing are often as follows: When the pre-etched character markings on the blade surface experience localized wear and corrosion due to long-term use, resulting in broken character strokes, or when slight oil stains cause a decrease in contrast between the character area and the background, directly processing the blade marking image will yield incorrect character recognition results. Incorrect character recognition results lead to incorrect blade shape version information, causing the positioning module to switch to a preset receiving station that does not match the actual blade shape, resulting in increased frequency of blade clamping failure or blade loosening and vibration during processing. Solving these factors can reduce the frequency of blade clamping failure or blade loosening and vibration during processing. To achieve this effect, firstly, the blade marking image is subjected to grayscale conversion and noise reduction processing to obtain a pre-processed blade marking image. Therefore, the acquired colored leaf label images can be converted into single-channel grayscale images, and high-frequency random noise in the image can be suppressed by filtering to improve the image signal-to-noise ratio, providing a clear base image for subsequent character region detection and repair. Next, distortion correction processing is performed on the preprocessed leaf label image to obtain a distortion-corrected leaf label image. This eliminates image geometric distortion caused by camera lens optical characteristics or tilted shooting angle, restoring the character labels to their standard geometric shape and ensuring the geometric accuracy of subsequent connected component detection and character recognition. Then, connected component detection processing is performed on the distortion-corrected leaf label image to obtain a sequence of character candidate regions. This allows adjacent white foreground pixels in the binarized image to be clustered into independent connected regions and sorted by spatial location to obtain candidate location regions for each character, decomposing the character localization problem across the entire image into a local processing problem for individual character regions. Finally, for each character candidate region in the character candidate region information sequence, stroke breakage detection processing is performed on the corresponding region image in the distortion-corrected leaf label image to obtain a stroke breakage detection score. Therefore, through skeleton extraction and endpoint analysis, the presence of stroke breaks caused by wear or corrosion in each character candidate region can be automatically detected, and the degree of breakage can be quantified. Next, in response to determining that the stroke breakage detection score is greater than a preset score threshold, stroke repair processing is performed on the regions corresponding to the character candidate region information in the distortion-corrected blade identifier image. Thus, for the detected stroke breakage regions, morphological dilation operations can be used to automatically connect and complete the broken character strokes, restoring the complete shape of the character and making characters that were originally difficult to recognize due to wear and breakage recognizable as complete and recognizable characters. Afterwards, the distortion-corrected blade identifier image after stroke repair processing is determined as the blade identifier image to be recognized.Therefore, using the complete character image after stroke restoration as input for subsequent recognition reduces the number of stroke breaks in the character image input to the recognition model that could affect the recognition results. Finally, the aforementioned blade identification image is input into a pre-trained identification recognition model to obtain identification information as blade shape version information. Thus, optical character recognition of the restored complete character image using an OCR model improves the accuracy of the output blade shape version information. Combined with steps 102 to 104, the stroke restoration process also reduces recognition errors caused by broken or blurred character strokes, ensuring accurate identification of blade shape version information even with minor local defects in the character identification. This avoids incorrect switching of the positioning module and subsequent clamping failures due to version misjudgment, thereby reducing the number of blade clamping failures or blade loosening and vibration during processing.

[0032] In addressing the technical problems mentioned in the background section using the above-described technical solutions, for the application scenario—an automated clamping production line for multi-version mixed machining of gas turbine blades—non-contact automatic identification of blade shape versions often presents the following technical challenges: When pre-etched or pasted character markings on the blade surface become inaccurate due to oil contamination, severe wear, rust, etching errors, or incorrect pasting, relying on these markings for model detection will yield incorrect blade shape version information. Incorrect blade shape version information may cause the positioning module to be driven to a preset receiving position that does not match the actual blade shape, preventing the contact surface of the adjustable positioning and fastening device from fitting the blade profile, and causing the support height of the adjustable support device to mismatch with the blade's bottom surface. Ultimately, this leads to clamping failure or increased blade loosening and vibration during machining. The following requirements are necessary for this application scenario: The mixed-version machining of gas turbine blades requires accurate and efficient identification of blade shape version information. This ensures that the identified blade shape version information is unaffected by the markings on the blade surface, guaranteeing that the positioning module is accurately driven to the preset receiving position that matches the actual blade shape, thus achieving reliable clamping and fixing of multiple blade versions. To address these technical challenges, we have decided to adopt the following solution: In some optional implementations of certain embodiments, the aforementioned execution entity can perform model detection processing on the gas turbine blade to be processed, which is located in the detection area at the fixture slide rail, through the following steps to obtain blade shape version information: The first step is to acquire multi-angle images of the gas turbine blade to be processed, located in the detection area of ​​the fixture slide rail, to obtain at least one image of the gas turbine blade. In practice, the aforementioned execution entity can use a ring camera array composed of multiple industrial cameras positioned around the detection area of ​​the fixture slide rail to simultaneously trigger and capture images of the gas turbine blade's shape at different angles, thereby obtaining at least one image of the gas turbine blade.

[0033] The second step involves extracting blade feature point information from each of the at least one gas turbine blade image to obtain a set of blade feature point information. In practice, the execution entity can input the gas turbine blade image into a pre-trained feature point extraction model to obtain the corresponding set of blade feature point information. This feature point extraction model can be an HS-ResNet feature extraction network. The blade feature point information in the set represents the information corresponding to the feature points on the gas turbine blade. This blade feature point information includes, but is not limited to, at least one of the following: two-dimensional coordinates and intensity information. The intensity information can represent the grayscale value of the feature point.

[0034] The third step involves inputting the obtained blade feature point information groups into a pre-trained feature point matching prediction model to obtain matching blade feature point information groups. The matching blade feature point information in these groups represents the blade feature point information from different perspectives corresponding to a single point on the gas turbine blade to be processed. The feature point matching prediction model includes convolutional layers, self-attention layers, fully connected layers, and an output layer. Specifically, firstly, the blade feature point information groups are input into the convolutional layers to obtain feature vector groups. Each feature vector represents a blade feature point. Each feature vector group corresponds one-to-one with each blade feature point in the blade feature point information group. Next, the feature vector groups are input into the self-attention layer to generate correlation weight information representing the correlation between every two feature vectors that are not in the same group. This correlation weight information represents the weight of the similarity between two feature vectors. The correlation weight information can include weight values ​​and corresponding feature vectors. The similarity can be cosine similarity. Subsequently, the correlation weights can be input into the fully connected layer to obtain the prediction information groups for each matching feature point. Each prediction information group can represent information including the blade feature points from different perspectives corresponding to a point on the gas turbine blade to be processed, and the confidence level of the predicted matching feature point group. Finally, the execution entity can input the prediction information groups for each matching feature point into the output layer to obtain the matching blade feature point information groups.

[0035] The fourth step involves reconstructing the feature point information groups of each matched blade using feature point cloudification to obtain the three-dimensional point cloud data of the blade. In practice, the aforementioned execution entity can use the SFM (Structure-from-Motion) algorithm to perform motion estimation on the feature point information groups of each matched blade to obtain camera extrinsic parameters. Then, the aforementioned execution entity can use the SFM algorithm to perform three-dimensional structure restoration and reconstruction based on the feature point information groups of each matched blade and the camera extrinsic parameters to obtain the three-dimensional point cloud data of the blade.

[0036] The fifth step involves inputting the aforementioned 3D point cloud data of the blade into the point cloud receiving layer of a pre-trained leaf shape version recognition model to obtain initial point cloud feature information. This leaf shape version recognition model includes the aforementioned point cloud receiving layer, a spatial pose normalization layer, a point-by-point low-dimensional feature dilation layer, a feature manifold alignment layer, a point-by-point high-level semantic extraction layer, a max-pooling layer, a global feature extraction layer, and a leaf shape version information mapping layer. This leaf shape version recognition model refers to a deep learning neural network model used to identify the leaf shape version of the blade based on the 3D point cloud data. The aforementioned point cloud receiving layer is the input layer of the leaf shape version recognition model that receives the input 3D point cloud data and converts it into vectors as the initial point cloud feature information to be passed to the next network layer.

[0037] Step 6: Input the initial point cloud feature information into the spatial pose normalization layer to obtain pose-normalized point cloud feature information. The spatial pose normalization layer can be configured to learn and predict a rotation matrix, which is then used to perform rotation transformation on the input point cloud, normalizing the point cloud's pose to a standard orientation (T-Net rotation network). The pose-normalized point cloud feature information refers to the point cloud feature vector obtained after pose alignment processing by the spatial pose normalization layer.

[0038] Step 7: Input the pose-normalized point cloud feature information into the point-by-point low-dimensional feature dilation layer to obtain low-dimensional dilated point cloud feature information. The point-by-point low-dimensional feature dilation layer can be a multilayer perceptron that maps each point from low-dimensional coordinates to a high-dimensional feature space using a multilayer perceptron with shared weights. The low-dimensional dilated point cloud feature information can be the point cloud feature vector obtained after dimensionality upscaling by the point-by-point low-dimensional feature dilation layer.

[0039] Step 8: Input the aforementioned low-dimensional dilated point cloud feature information into the aforementioned feature manifold alignment layer to obtain manifold-aligned point cloud feature information. The feature manifold alignment layer can be configured to learn and predict a feature space transformation matrix, which uniformly aligns the high-dimensional feature distributions of different samples using this matrix, eliminating feature distribution offsets between samples (T-Net rotation network). The aforementioned manifold-aligned point cloud feature information refers to the point cloud feature vector obtained after feature space alignment processing by the aforementioned feature manifold alignment layer.

[0040] Step nine involves inputting the manifold-aligned point cloud feature information into the aforementioned point-by-point high-level semantic extraction layer to obtain point-by-point semantic feature information. The aforementioned point-by-point high-level semantic extraction layer refers to a network layer used to progressively upscale the features of each point in the input point cloud from a low-dimensional geometric description to a high-dimensional semantic description. This layer can be a multi-layer perceptron layer that performs multi-level dimensionality upscaling for each point using a multi-layer perceptron with shared weights. The aforementioned point-by-point semantic feature information refers to the point cloud feature vector obtained after high-level semantic extraction processing by the aforementioned point-by-point high-level semantic extraction layer.

[0041] Step 10: Input the above point-by-point semantic feature information into the max pooling layer to obtain pooled point-by-point semantic feature information. The pooled point-by-point semantic feature information can be the point cloud feature vector obtained after max pooling.

[0042] Step 11: The pooled point-by-point semantic feature information is input into the global feature extraction layer to perform global feature refinement and compression on the pooled point-by-point semantic feature information, resulting in global point cloud semantic feature information. This global feature refinement and compression can be achieved by using a multilayer perceptron to progressively compress and map the pooled point-by-point semantic feature information from a high-dimensional space to a low-dimensional space. The global point cloud semantic feature information refers to the feature vector containing the overall geometric semantic description of the blade, obtained after global feature refinement and compression by the global feature extraction layer; that is, the global geometric semantic feature vector.

[0043] Step 12: Input the aforementioned global point cloud semantic feature information into the leaf-shaped version information mapping layer to obtain leaf-shaped version information. This leaf-shaped version information mapping layer includes a fully connected layer and an output layer. The fully connected layer maps the global point cloud semantic feature information representing the global geometric semantic feature vector to classification scores for each leaf-shaped version and outputs the final classification result. The output layer converts the classification scores into probability distributions for each category using a normalized exponential function, and takes the category with the highest probability as the final leaf-shaped version information.

[0044] The above technical solution, combined with steps 102 to 104 and related content, serves as an inventive point of this disclosure, solving the technical problem of "causing clamping failure or increased blade loosening and vibration during processing." Factors leading to clamping failure or increased blade loosening and vibration during processing are often as follows: When the pre-etched or pasted character markings on the blade surface cannot be accurately identified due to oil contamination, severe wear, rust, or etching or pasting errors, relying on the character markings for model detection will yield incorrect blade shape version information. Incorrect blade shape version information may cause the positioning module to be driven to a preset receiving position that does not match the actual blade shape, preventing the contact surface of the adjustable positioning fastening device from fitting the blade profile, and causing the support height of the adjustable support device to mismatch with the blade bottom surface, ultimately leading to clamping failure or increased blade loosening and vibration during processing. Solving these factors can reduce blade clamping failure or blade loosening and vibration during processing. To achieve this effect, firstly, multi-angle images of the gas turbine blade to be processed are acquired in the detection area set at the fixture slide rail, obtaining at least one gas turbine blade image. Therefore, at least one image of a gas turbine blade from different perspectives can be obtained. Next, blade feature point information is extracted from each of these at least one gas turbine blade image to obtain a set of blade feature point information. Then, each set of blade feature point information is input into a pre-trained feature point matching prediction model to obtain a set of matched blade feature point information. Afterward, feature point cloud reconstruction processing is performed on each set of matched blade feature point information to obtain 3D blade point cloud data. Thus, the principle of multi-view geometry can be used to restore the 2D matched feature points into 3D blade point cloud data describing the actual shape of the blade surface. Next, the 3D blade point cloud data is input into the point cloud receiving layer of a pre-trained blade shape version recognition model to obtain initial point cloud feature information. The blade shape version recognition model includes the point cloud receiving layer, a spatial pose normalization layer, a point-by-point low-dimensional feature dilation layer, a feature manifold alignment layer, a point-by-point high-level semantic extraction layer, a max pooling layer, a global feature extraction layer, and a blade shape version information mapping layer. Next, the initial point cloud feature information is input into the spatial pose normalization layer to obtain pose-normalized point cloud feature information. This allows for spatial pose alignment of the input point cloud by learning the rotation matrix, ensuring the model's invariance to different blade orientations within the detection region. Then, the pose-normalized point cloud feature information is input into the point-by-point low-dimensional feature dilation layer to obtain low-dimensional dilated point cloud feature information. This progressively increases the dimensionality of each point's three-dimensional coordinates to a higher-dimensional feature space, providing a richer local geometric description for each point. Finally, the low-dimensional dilated point cloud feature information is input into the feature manifold alignment layer to obtain manifold-aligned point cloud feature information.Therefore, the high-dimensional feature distributions of different leaf samples are aligned in the feature space to eliminate feature distribution offsets caused by individual or batch differences in leaf blanks. Then, the manifold-aligned point cloud feature information is input to the point-by-point high-level semantic extraction layer to obtain point-by-point semantic feature information. This allows the features of each point to be progressively upgraded from a low-dimensional geometric description to a high-dimensional description with global semantic meaning. Next, the point-by-point semantic feature information is input to the max-pooling layer to obtain pooled point-by-point semantic feature information. Then, the pooled point-by-point semantic feature information is input to the global feature extraction layer for global feature refinement and compression, resulting in global point cloud semantic feature information. Finally, the global point cloud semantic feature information is input to the leaf shape version information mapping layer to obtain leaf shape version information. Combining the content of steps 102 to 104, and because the three-dimensional point cloud data of the blade itself is processed from spatial alignment, feature dilation, manifold alignment to semantic extraction, version recognition based on the actual geometry of the blade is realized. This eliminates the dependence on the status of character markings on the blade surface, avoids misidentification caused by marking damage or errors, and ensures that the positioning fixture module is accurately driven to the preset receiving station that matches the actual blade shape. This reduces the number of blade clamping failures or blade loosening and vibration during processing.

[0045] Step 102: Drive the positioning fixture module corresponding to the blade version information, move it along the slide rail connected to the detection area at the fixture slide rail to the preset receiving station where the gas turbine blade to be processed is located, so as to receive the gas turbine blade to be processed.

[0046] In some embodiments, the execution entity can drive the positioning fixture module corresponding to the blade version information to move along a slide rail connected to the detection area at the fixture slide rail to a preset receiving station where the gas turbine blade to be processed is located, so as to receive the gas turbine blade to be processed. The positioning fixture module includes at least one adjustable positioning fastening device and at least one adjustable support device. The positioning fixture module refers to a modular positioning unit mounted on the slide rail that can move linearly along the slide rail. The positioning fixture module can be a device for positioning, clamping, and supporting the gas turbine blade to be processed (such as a processing fixture for the exhaust edge (trailing edge) of a turbine blade). The positioning fixture module includes at least one adjustable positioning fastening device and at least one adjustable support device. The slide rail refers to a precision linear guide rail laid on the fixture base. The slide rail is used to guide the positioning fixture module to move linearly in a preset direction. The slide rail can be a rolling linear guide rail or a sliding linear guide rail. The preset receiving station refers to a stop position of the positioning fixture module pre-set on the slide rail that corresponds one-to-one with the blade version information. The aforementioned adjustable positioning and fastening device refers to a device used to horizontally conform to the side surface of the gas turbine blade to be processed and apply positioning constraints to the gas turbine blade. The aforementioned adjustable positioning and fastening device may be equipped with a displacement sensor and a locking mechanism. The aforementioned displacement sensor is a sensor used to measure the distance between the contact surface of the aforementioned adjustable positioning and fastening device and the corresponding surface of the gas turbine blade to be processed. The aforementioned displacement sensor can be any one of an eddy current displacement sensor, a capacitive displacement sensor, or a laser triangulation sensor. The aforementioned locking mechanism is a mechanism used to lock the aforementioned adjustable positioning and fastening device in its current position. The aforementioned locking mechanism can be a hydraulic brake-type locking device or a wedge-shaped self-locking device. The aforementioned adjustable support device refers to a device used to vertically support the bottom surface of the gas turbine blade to be processed and perform height adaptive compensation for the gas turbine blade to be processed. The aforementioned adjustable support device may be equipped with a laser triangulation sensor and a support device locking mechanism. The aforementioned laser triangulation sensor is a sensor that uses a laser beam projected onto the surface of an object and calculates the distance from the sensor to the surface of the object based on the positional offset of the reflected light spot. The aforementioned support device locking mechanism refers to a mechanism used to lock the aforementioned adjustable support device at the current support height.

[0047] In some optional implementations of certain embodiments, the execution entity can drive the positioning fixture module corresponding to the blade version information to move along the slide rail connected to the detection area at the fixture slide rail to the preset receiving station where the gas turbine blade to be processed is located, in order to receive the gas turbine blade to be processed: The first step is to generate a positioning fixture module switching instruction based on the aforementioned blade version information. In practice, the executing entity can first read the blade version number corresponding to the aforementioned blade version information. Then, it can look up the preset switching instruction corresponding to the aforementioned blade version number in a pre-stored mapping table of version numbers and preset switching instructions as the positioning fixture module switching instruction. The aforementioned positioning fixture module switching instruction can be a pre-written control instruction that moves the aforementioned positioning fixture module from the current workstation to the aforementioned preset receiving workstation.

[0048] The second step involves, based on the aforementioned positioning fixture module switching command, driving the positioning fixture module corresponding to the aforementioned leaf-shaped version information to move along the aforementioned slide rail towards the aforementioned preset receiving station. In practice, the aforementioned execution entity can execute the switching task corresponding to the aforementioned positioning fixture module switching command, driving the positioning fixture module corresponding to the aforementioned leaf-shaped version information to move along the aforementioned slide rail towards the aforementioned preset receiving station.

[0049] The third step involves acquiring the position information of the positioning fixture module on the slide rail at each preset interval, and storing the acquired position information in a preset queue. The preset interval refers to a pre-defined time interval for sampling the position information of the positioning fixture module on the slide rail. In practice, the executing entity can use a linear or magnetic ruler mounted on the slide rail to read the coordinates of feature points on the positioning fixture module as position information at each preset interval.

[0050] Fourth, based on the preset queue, perform the following adjustment steps: The first sub-step, in response to determining that the number of position information in the preset queue is greater than a preset number and that an enqueue operation exists, determines the position information corresponding to the enqueue operation as the current position information. Here, the preset number can be 1. The current position information can be represented by a coordinate.

[0051] The second sub-step is to determine the position information preceding the current position information in the preset queue after the enqueue operation as the reference position information.

[0052] The third sub-step, in response to determining that the current position information is the same as the reference position information, stops acquiring the position information of the positioning fixture module on the slide rail, and generates position deviation information based on the current position information and the target position information corresponding to the preset receiving station. The target position information can be the preset coordinates (two-dimensional coordinates, including ordinate and abscissa) of a feature point on the positioning fixture module when it is at the preset receiving station. The position deviation information can be the vector between the feature point coordinates represented by the current position information and the feature point coordinates represented by the target position information (i.e., the feature point coordinates represented by the current position information minus the feature point coordinates represented by the target position information).

[0053] The fourth sub-step involves adjusting the positioning fixture module to a preset receiving position based on the position deviation information to receive the gas turbine blades being processed. This adjustment is performed by driving a slide rail drive mechanism mounted on the slide rail. The slide rail drive mechanism is a power actuator that drives the positioning fixture module to move linearly along the slide rail and precisely stop at the target position. This mechanism can be any of a servo motor with a ball screw pair, a linear motor, or a pneumatic slide table. A servo motor is a motor capable of precisely controlling the angle and speed according to pulse or bus commands. In practice, the actuator can first generate a position correction command based on the position deviation information, which includes compensation distance and direction of movement. Then, the position correction command is sent to the controller of the slide rail drive mechanism, which can be a servo driver or a PLC positioning fixture module. Next, the slide rail drive mechanism drives the positioning fixture module to move along the slide rail by the compensation distance according to the position correction command, so as to eliminate the position deviation represented by the position deviation information and bring the positioning fixture module to the preset receiving position. As an example, assuming the slide rail direction is along the X-axis, the current position coordinates are (100.15, 50.00), the target position coordinates are (100.00, 50.00), then the deviation vector is (0.15, 0.00), the deviation direction is along the negative direction of the slide rail (if it needs to move to the left), and the compensation distance is 0.15mm.

[0054] The fifth sub-step involves, in response to the determination that the current position information is different from the reference position information, continuing at each preset interval, acquiring the position information of the positioning fixture module on the slide rail, storing the acquired position information in a preset queue, and then executing the above adjustment steps again based on the preset queue.

[0055] It should be noted that the deviation calculation and fine-tuning are performed only after the detection positioning fixture module stops moving on the slide rail (the position information is the same twice consecutively). This eliminates position overshoot or undershoot caused by factors such as motion inertia, slide rail clearance, and transmission error, ensuring that the module accurately reaches the preset receiving position and avoiding failure to receive the blade or damage to the blade due to inaccurate positioning.

[0056] Step 103: For each of the at least one adjustable positioning fastening devices, control the adjustable positioning fastening device to perform a gap detection process on the gas turbine blade to be processed in its working area, so as to lock the blade in its working area.

[0057] In some embodiments, the executing entity can control each of the at least one adjustable positioning and fastening device to perform gap detection processing on the gas turbine blade to be processed in its working area, so as to lock the blade in its working area. Each adjustable positioning and fastening device is equipped with a displacement sensor, a locking mechanism, and a feeding mechanism. The displacement sensor is a sensing element capable of measuring mechanical displacement or distance, such as a laser displacement sensor. The feeding mechanism is an actuating element that drives the adjustable positioning and fastening device to move towards the blade, such as a servo motor-driven screw and nut pair or a pneumatic / hydraulic cylinder. The locking mechanism is an actuating element used to fix the position of the adjustable positioning and fastening device to prevent it from loosening, such as a pneumatic clamp, a hydraulic locking nut, or an electromagnetic brake.

[0058] In some optional implementations of certain embodiments, the aforementioned execution entity may control the aforementioned adjustable positioning and fastening device to perform gap detection processing on the aforementioned gas turbine blade to be processed in its working area through the following steps, so as to lock the blade in its working area: The first step involves using a displacement sensor mounted on the adjustable positioning and fastening device to collect the distance between the contact surface of the adjustable positioning and fastening device and the corresponding surface of the gas turbine blade to be processed, as the contact gap distance information. Here, the contact surface of the adjustable positioning and fastening device refers to the plane or curved surface on the device that directly contacts the gas turbine blade to be processed. The contact gap distance information refers to the spatial distance between the contact surface and the corresponding surface of the blade. The displacement sensor can be embedded inside the contact surface of the adjustable positioning and fastening device, and the front end of the displacement sensor is flush with the contact surface. The corresponding surface refers to the local surface area on the gas turbine blade to be processed that is directly opposite the contact surface; the corresponding surface can be the blade facet or the blade back facet.

[0059] The second step is to generate the feed compensation value of the feeding mechanism based on the above-mentioned contact gap distance information. The feed compensation value refers to the additional feed distance required to make the contact surface fit precisely against the blade surface. It is usually equal to the currently measured contact gap distance information minus a preset target contact gap (for example, the target contact gap is usually zero or a very small preload value, such as 0.05mm).

[0060] The third step is to adjust the positioning and fastening device based on the above feed compensation value to lock the blades in its working area.

[0061] In some optional implementations of certain embodiments, the aforementioned execution entity may adjust the aforementioned positioning and fastening device based on the aforementioned feed compensation value through the following steps to lock the blades in its area of ​​action: The first step involves driving the feed mechanism of the aforementioned positioning and fastening device to perform a feed operation based on the feed compensation value, so that the contact surface of the adjustable positioning and fastening device fits against the corresponding surface of the gas turbine blade to be processed. In practice, the feed compensation value can be sent to the servo driver of the feed mechanism (e.g., via EtherCAT bus or pulse direction signal). The driver controls the motor to rotate, driving the lead screw to move the positioning and fastening device forward (i.e., horizontally) by the distance corresponding to the compensation value. During the movement, the actuator can optionally read the feedback value from the displacement sensor in real time, forming a closed-loop control until the contact gap distance information approaches 0 (or a preset threshold).

[0062] The second step is to generate a locking command for the positioning and fastening device after the feed operation is completed. In practice, the aforementioned execution entity can generate a binary or digital command, such as "locking request = 1, indicating a high level", after confirming that the feed operation has been completed (e.g., triggered by a servo motor positioning signal or a stable displacement sensor reading).

[0063] The third step involves controlling the locking mechanism of the adjustable positioning and fastening device to perform a position locking process based on the aforementioned locking command, thereby locking the blades in its effective area. In practice, the executing entity sends the locking command to the control valve (such as a solenoid directional valve) of the locking mechanism. For example, when the locking command is a binary or digital signal indicating a high level, the solenoid valve is energized, compressed air enters the pneumatic clamp, and the clamp locks the lead screw or guide rail of the feed mechanism, preventing it from moving in the opposite direction.

[0064] Step 104: For each of the at least one adjustable support devices, control the adjustable support device to perform height support adaptive compensation processing on the gas turbine blade to be processed in its working area to support the blade in its working area.

[0065] In some embodiments, the executing entity can control each of the at least one adjustable support device to perform height-adaptive compensation processing on the gas turbine blade to be processed in its working area, thereby supporting the blade in its working area and achieving a processing fixation effect. Each adjustable support device is equipped with a laser triangulation sensor, a height adjustment mechanism, and a locking mechanism. The adjustable support device refers to a device mounted on the positioning fixture module for providing support force to the bottom surface of the gas turbine blade to be processed from a vertical direction. The working area of ​​the adjustable support device refers to the specific local area on the blade bottom surface corresponding to the support surface of the adjustable support device and for which support force is applied. The support surface refers to the bearing surface at the top of the adjustable support device that directly contacts the blade bottom surface. The height adjustment mechanism is a mechanical actuator for driving the support surface of the adjustable support device to move vertically. The height adjustment mechanism can be a differential thread mechanism. The locking mechanism is a mechanism for rigidly locking the vertical position of the adjustable support device after height adjustment, preventing displacement during processing. For example, the locking mechanism described above can be, but is not limited to, one of the following: a hydraulic brake-type locking device or a wedge-shaped self-locking device.

[0066] In some optional implementations of certain embodiments, the aforementioned execution entity can control the aforementioned adjustable support device to perform height-adaptive compensation processing on the gas turbine blade to be processed in its working area through the following steps, so as to support the blade in its working area: The first step involves using a laser triangulation sensor in the adjustable support device to detect the distance between the current height of the support surface and the gas turbine blade to be processed within its area of ​​operation. This distance serves as the height adaptive compensation value. Specifically, the height adaptive compensation value refers to the actual distance by which the support surface of the adjustable support device needs to move upwards to align with the bottom surface of the blade.

[0067] The second step, in response to the determination that the height adaptive compensation value is greater than the preset fitting threshold, is to execute the following driving steps: The first sub-step involves driving the height adjustment mechanism of the adjustable support device to perform adaptive height adjustment based on the height adaptive compensation value. In practice, the executing entity can first convert the height adaptive compensation value into driving parameters for the height adjustment mechanism. When the height adjustment mechanism is a differential screw mechanism, the driving parameters can be the number of rotations or angle required for the differential screw, calculated based on the height adaptive compensation value and the pitch difference of the differential screw mechanism. Then, the executing entity sends the driving parameters to the drive motor of the height adjustment mechanism. Next, the drive motor drives the height adjustment mechanism to perform the motion corresponding to the driving parameters, causing the support surface to move vertically by the distance corresponding to the height adaptive compensation value, so that the support surface faces upward and contacts the bottom surface of the blade.

[0068] The second sub-step involves, after the adaptive adjustment of the support height, using a laser triangulation sensor to re-collect the distance from the current support surface height of the adjustable support device to the gas turbine blade to be processed within its operating area, in order to update the adaptive height compensation value. In practice, the aforementioned executing entity can update the adaptive height compensation value with the re-collected distance from the current support surface height of the adjustable support device to the gas turbine blade to be processed within its operating area.

[0069] The third sub-step is to execute the above driving steps again in response to the determination that the updated height adaptive compensation value is greater than the preset fitting threshold.

[0070] The fourth sub-step involves controlling the locking mechanism of the adjustable support device to perform a support posture locking process on the adjustable support device in response to determining that the updated height adaptive compensation value is less than or equal to the aforementioned preset fitting threshold, thereby supporting the blades in its working area. In practice, the executing entity can rigidly lock the current position of the adjustable support device in the vertical direction after the height adaptive adjustment is completed, ensuring that the adjustable support device remains in its current position during subsequent processing. Specifically, the executing entity can first generate a support device locking command. Then, the support device locking command is sent to the controller of the locking mechanism. Next, the locking mechanism performs a locking action according to the support device locking command. When the locking mechanism is a hydraulic brake-type locking device, the locking action is achieved by hydraulic pressure driving the brake element to grip the guide rail surface of the height adjustment mechanism, preventing the adjustable support device from moving in the vertical direction. When the locking mechanism is the wedge-shaped self-locking device, the locking action is to drive the wedge-shaped block in the opposite direction to generate a self-locking friction force between it and the mating inclined surface, so that the adjustable support device cannot be reversibly displaced when subjected to processing force. After the above support posture locking process is completed, the adjustable support device switches from the adjustable state to the rigid support state to provide stable support for the blades in its working area during subsequent processing.

[0071] Optionally, the aforementioned implementing entity may also perform the following steps: The first step involves controlling the machining equipment to process the gas turbine blades on the positioning fixture module. The aforementioned machining equipment refers to a machine tool used to remove material from the gas turbine blades. This equipment can be a five-axis CNC machining center. The machining process involves using the cutting tools of the aforementioned machining equipment to cut, grind, or mill the surface of the gas turbine blades to remove excess material and achieve the target blade profile.

[0072] The second step involves real-time monitoring and processing of the vibration state of the gas turbine blades to be processed during the machining process to obtain blade vibration characteristic information. The machining process refers to the time interval from when the cutting tool of the machining equipment begins to contact the surface of the gas turbine blade to be processed until the machining is completed. The blade vibration characteristic information refers to a set of characteristic parameters describing the vibration state of the gas turbine blades to be processed due to cutting force excitation during machining. This information may include vibration amplitude and dominant vibration frequency. The vibration amplitude refers to the maximum displacement of the gas turbine blades to be processed from their equilibrium position in the vibration direction. The dominant vibration frequency refers to the frequency component with the most concentrated energy in the vibration signal of the gas turbine blades to be processed. In practice, the actuator can use a piezoelectric accelerometer installed inside the support surface of the adjustable support device to collect and process the vibration acceleration of the gas turbine blades to be processed in real time during the machining process, obtaining a time-domain vibration acceleration signal. Then, the time-domain vibration acceleration signal is processed by a Fast Fourier Transform to obtain vibration spectrum data. Next, the frequency corresponding to the maximum peak value is extracted from the vibration spectrum data as the main vibration frequency, and the amplitude corresponding to the main vibration frequency is extracted as the vibration amplitude. The main vibration frequency and the vibration amplitude are combined to form the blade vibration characteristic information. The piezoelectric accelerometer mentioned above refers to an inertial sensor that uses the positive piezoelectric effect of piezoelectric materials to convert the vibration acceleration of the measured object into an electric charge or voltage signal.

[0073] The third step involves generating vibration deviation information based on the aforementioned blade vibration characteristic information and a preset stable processing vibration threshold. In practice, the executing entity can numerically compare the vibration amplitude in the aforementioned blade vibration characteristic information with the aforementioned preset stable processing vibration threshold. If the difference between the aforementioned vibration amplitude and the aforementioned preset stable processing vibration threshold is greater than zero, this difference is used as the aforementioned vibration deviation information, indicating that the current vibration exceeds the allowable range for stable processing. If the aforementioned difference is less than or equal to zero, it is determined that the current vibration is within the preset stable processing vibration state, and the aforementioned vibration deviation information is not generated. The aforementioned preset stable processing vibration threshold refers to a pre-set upper limit value for the vibration amplitude used to determine whether the vibration of the gas turbine blade to be processed is within a stable and controllable range during processing. When the vibration amplitude of the gas turbine blade to be processed exceeds the aforementioned preset stable processing vibration threshold, it indicates that the blade vibration is too large, which may cause processing chatter and affect processing accuracy and surface quality. The aforementioned preset stable processing vibration threshold can be determined in advance through modal testing or cutting testing based on the material properties, structural stiffness characteristics, and processing parameters of the gas turbine blade to be processed. For example, the aforementioned preset stable processing vibration threshold can be 0.01 mm.

[0074] The fourth step involves generating a dynamic compensation command for the support force based on the vibration deviation information, which indicates that the current vibration of the gas turbine blade to be processed exceeds the preset stable processing vibration threshold. In practice, the executing entity can first obtain a preset vibration-to-support force mapping table, which stores the correspondence between different vibration deviation values ​​and corresponding additional feed displacements. This correspondence can be pre-established by performing modal tests and static stiffness tests on the gas turbine blade to be processed. Then, using the vibration deviation value in the vibration deviation information as an index, the corresponding additional feed displacement is searched in the vibration-to-support force mapping table. Finally, the found additional feed displacement and the corresponding execution direction are encapsulated into the dynamic compensation command for the support force. This dynamic compensation command for the support force refers to a control command that instructs the adjustable support device to adjust the support force applied to the gas turbine blade to be processed during the processing. The aforementioned dynamic compensation command for support force includes an additional feed displacement, which refers to the slight upward movement of the support surface of the adjustable support device to increase the support force based on the current support height.

[0075] Fifth, based on the aforementioned dynamic compensation command for the support force, the height adjustment mechanism of at least one adjustable support device is driven to perform dynamic tuning of the support stiffness, so that the support force of the adjustable support device on the gas turbine blade to be processed is dynamically compensated during processing, maintaining the gas turbine blade to be processed within a preset stable processing vibration state. In practice, the executing entity can first convert the additional feed displacement in the aforementioned dynamic compensation command for the support force into the driving parameters of the height adjustment mechanism. When the height adjustment mechanism is a differential screw mechanism, the driving parameters can be the additional number of turns or angle required for the differential screw to rotate, which are calculated based on the additional feed displacement and the pitch difference of the differential screw mechanism. Then, the executing entity sends the driving parameters to the drive motor of the height adjustment mechanism. Next, the drive motor drives the height adjustment mechanism to execute the small movement corresponding to the driving parameters, causing the support surface to move upward in the vertical direction by the additional feed displacement, increasing the support force of the adjustable support device on the gas turbine blade to be processed. After the support force is increased, the vibration state of the gas turbine blade to be processed continues to be monitored by a piezoelectric accelerometer installed inside the support surface. If the vibration amplitude still exceeds the preset stable processing vibration threshold, steps two through five are repeated until the vibration amplitude of the gas turbine blade to be processed recovers to a stable processing vibration state less than or equal to the preset stable processing vibration threshold. If the vibration amplitude does not decrease after multiple adjustments, a processing abnormality alarm signal is generated to indicate that the cutting tool may be worn or the processing parameters need to be adjusted. Through the dynamic tuning process described above, the removal of processed material can be compensated in real time on the basis of the original clamping and fitting, suppressing processing chatter in the thin-walled area of ​​the blade, and further improving the final processing accuracy and surface quality. At the same time, when the vibration amplitude continues not to decrease, a processing abnormality alarm signal is generated to indicate that the cutting tool may be worn or the processing parameters need to be adjusted, further improving the reliability of processing.

[0076] The above-described technical solution and its related content, as an inventive point of this disclosure, solve the technical problem that "during the processing of gas turbine blades, the local stiffness of the blade decreases in real time due to the removal of material layer by layer, and the blade generates machining chatter under the excitation of cutting force, affecting the final machining accuracy and surface quality of the blade." Factors leading to a decrease in blade machining accuracy and surface quality are often as follows: After the adjustable support device is fully fitted and locked before processing, its support state remains fixed during processing. When the tool removes material layer by layer from the gas turbine blade, the local structural stiffness of the thin-walled area of ​​the blade decreases in real time as the material thickness decreases, and the fixed support force cannot compensate for this dynamic stiffness loss. Under continuous cutting force, the thin-walled area of ​​the blade is prone to vibrations exceeding the allowable range of stable processing (i.e., machining chatter), resulting in vibration marks, dimensional deviations, and even accelerated tool wear or localized blade breakage on the machined surface, leading to a decrease in blade machining accuracy and surface quality. Solving these factors can achieve the desired improvement in blade machining accuracy and surface quality. To achieve this effect, firstly, during the machining process, the vibration state of the gas turbine blade to be processed is monitored and processed in real time to obtain blade vibration characteristic information. This allows for the real-time sensing of key vibration characteristics such as the vibration amplitude and dominant frequency of the blade under cutting force excitation throughout the entire process of material removal by the cutting tool. Next, based on the blade vibration characteristic information and a preset stable machining vibration threshold, vibration deviation information is generated. This allows for a quantitative comparison between the real-time monitored blade vibration amplitude and the upper limit of stable machining vibration determined in advance through modal tests or cutting tests, accurately quantifying the extent to which the current vibration exceeds the stable range. Then, in response to the vibration deviation information indicating that the current vibration of the gas turbine blade to be processed exceeds the preset stable machining vibration threshold, a dynamic compensation command for the support force is generated based on the vibration deviation information. This allows the vibration deviation value to be promptly converted into an additional feed displacement command for dynamic adjustment of the support device at the instant the blade vibration just exceeds the allowable range of stable machining, ensuring that the compensation amplitude of the support force matches the amplitude of the vibration deviation. Finally, based on the aforementioned dynamic compensation command for support force, the height adjustment mechanism of the at least one adjustable support device is driven to perform dynamic tuning of support stiffness, so that the support force of the adjustable support device on the gas turbine blade to be processed is dynamically compensated during the processing, and the gas turbine blade to be processed is kept within a preset stable processing vibration state.Therefore, without interrupting the machining process, the local support force on the thin-walled area of ​​the blade can be increased by adjusting the height of the support surface in real time. This compensates for the loss of local structural stiffness due to material removal and suppresses the vibration of the blade within the range allowed for stable machining in real time. It realizes real-time tracking and dynamic compensation of stiffness changes caused by material removal during blade machining, thus breaking the technical limitation of the traditional clamping scheme where the support state remains unchanged during the machining stage. It reduces blade machining chatter caused by cutting force excitation, thereby improving the final machining accuracy and surface quality of the blade.

[0077] Further reference Figure 4 , Figure 4 This is a structural schematic diagram of some embodiments of the positioning clamp module according to the present disclosure. The positioning clamp module includes at least one adjustable positioning fastening device 401 and at least one adjustable support device 402.

[0078] Further reference Figure 2 As an implementation of the methods shown in the figures, this disclosure provides some embodiments of an adaptive clamping and fixing device for gas turbine blades. These device embodiments are similar to... Figure 1 Corresponding to the method embodiments shown, the device can be specifically applied to various electronic devices.

[0079] like Figure 2 As shown, the gas turbine blade adaptive clamping and fixing device 200 in some embodiments includes: a model detection unit 201, a drive unit 202, a first control unit 203, and a second control unit 204. The model detection unit 201 is configured to perform model detection processing on the gas turbine blade to be processed, which is located in the detection area of ​​the clamp slide rail, to obtain blade shape version information. The drive unit 202 is configured to drive the positioning clamp module corresponding to the blade shape version information to move along the slide rail connected to the detection area of ​​the clamp slide rail to the preset receiving position where the gas turbine blade to be processed is located, so as to receive the gas turbine blade to be processed. The positioning clamp module includes at least one adjustable positioning fastening device and at least one adjustable support device. The first control unit 203 is configured to control each of the at least one adjustable positioning fastening device to perform gap detection processing on the gas turbine blade to be processed in its working area, so as to lock the blade in its working area. The second control unit 204 is configured to control each of the at least one adjustable support device to perform height support adaptive compensation processing on the gas turbine blade to be processed in its working area, so as to support the blade in its working area.

[0080] It is understandable that the units described in the device 200 are related to the reference. Figure 1 The steps in the method described above correspond to each other. Therefore, the operations, features, and beneficial effects described above for the method also apply to the device 200 and the units contained therein, and will not be repeated here.

[0081] The following is for reference. Figure 3 It shows a schematic diagram of the structure of an electronic device 300 suitable for implementing some embodiments of the present disclosure. Figure 3 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.

[0082] like Figure 3 As shown, the electronic device 300 may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 into a random access memory (RAM) 303. The RAM 303 also stores various programs and data required for the operation of the electronic device 300. The processing unit 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0083] Typically, the following devices can be connected to I / O interface 305: input devices 306 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 307 including, for example, positioning clamp modules, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 308 including, for example, magnetic tapes, hard disks, etc.; and communication devices 309. Communication device 309 allows electronic device 300 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 3 An electronic device 300 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 3 Each box shown can represent a device or multiple devices as needed.

[0084] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication device 309, or installed from storage device 308, or installed from ROM 302. When the computer program is executed by processing device 301, it performs the functions defined in the methods of some embodiments of this disclosure.

[0085] It should be noted that, in some embodiments of this disclosure, the computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may 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. In some embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which 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 computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0086] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0087] The computer-readable medium may be included in an electronic device or may exist independently without being assembled into the electronic device. The computer-readable medium carries one or more programs that, when executed by the electronic device, cause the electronic device to: perform model detection processing on the gas turbine blade to be processed, located in the detection area at the clamping slide rail, to obtain blade shape version information; drive the positioning clamping module corresponding to the blade shape version information to move along the slide rail connected to the detection area at the clamping slide rail to a preset receiving position where the gas turbine blade to be processed is located, to receive the gas turbine blade to be processed, wherein the positioning clamping module includes at least one adjustable positioning fastening device and at least one adjustable support device; for each of the at least one adjustable positioning fastening devices, control the adjustable positioning fastening device to perform gap detection processing on the gas turbine blade to be processed in its working area to lock the blade in its working area; for each of the at least one adjustable support device, control the adjustable support device to perform height support adaptive compensation processing on the gas turbine blade to be processed in its working area to support the blade in its working area.

[0088] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof. Programming languages ​​include object-oriented programming languages—such as Java, Smalltalk, and C++—and conventional procedural programming languages—such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0089] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0090] The units described in some embodiments of this disclosure can be implemented in software or hardware. The described units can also be housed in a processor; for example, a processor may be described as including a model detection unit, a drive unit, a first control unit, and a second control unit. The names of these units do not necessarily limit the specific unit; for example, the model detection unit may also be described as "a unit that performs model detection processing on the gas turbine blades to be processed, located in the detection area of ​​the fixture slide rail, to obtain blade shape version information."

[0091] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), and so on.

[0092] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of technical features, but should also cover other technical solutions formed by arbitrary combinations of technical features or their equivalents without departing from the inventive concept. For example, technical solutions formed by substituting features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A method for adaptive clamping and fixing of gas turbine blades, comprising: The model detection process is performed on the gas turbine blades to be processed, which are located in the detection area at the fixture slide rail, to obtain the blade shape version information; The positioning fixture module corresponding to the blade version information is driven to move along the slide rail connected to the detection area at the fixture slide rail to the preset receiving station where the gas turbine blade to be processed is located, so as to receive the gas turbine blade to be processed. The positioning fixture module includes at least one adjustable positioning fastening device and at least one adjustable support device. For each of the at least one adjustable positioning fastening device, the adjustable positioning fastening device is controlled to perform a gap detection process on the gas turbine blade to be processed in its working area, so as to lock the blade in its working area. For each of the at least one adjustable support device, the adjustable support device is controlled to perform height support adaptive compensation processing on the gas turbine blade to be processed in its working area, so as to support the blade in its working area.

2. The method according to claim 1, wherein, The process of performing model detection on the gas turbine blades to be processed, located in the detection area at the fixture slide rail, to obtain blade shape version information includes: An image is acquired of a preset area of ​​the gas turbine blade to be processed to obtain a blade identification image, wherein the preset area has pre-etched character markings or pre-pasted character markings; The leaf identification image is processed to obtain leaf shape version information.

3. The method according to claim 1, wherein, The positioning fixture module corresponding to the blade version information moves along the slide rail connected to the detection area at the fixture slide rail to the preset receiving station where the gas turbine blade to be processed is located, in order to receive the gas turbine blade to be processed, including: Based on the leaf-shaped version information, a positioning fixture module switching instruction is generated; Based on the positioning fixture module switching command, the positioning fixture module corresponding to the leaf-shaped version information is driven to move along the slide rail toward the preset receiving station; At each preset interval, the position information of the positioning fixture module on the slide rail is acquired, and the acquired position information is stored in a preset queue; Based on the preset queue, perform the following adjustment steps: In response to the determination that the number of position information in the preset queue is greater than the preset number and that there is an enqueue operation, the position information corresponding to the enqueue operation is determined as the current position information; After the enqueue operation, the position information preceding the current position information in the preset queue is determined as the reference position information; In response to the determination that the current position information is the same as the reference position information, the acquisition of the position information of the positioning fixture module on the slide rail is stopped, and the position deviation information is generated based on the target position information corresponding to the current position information and the preset receiving station. Based on the position deviation information, the positioning fixture module is adjusted to the preset receiving station to receive the gas turbine blades to be processed; In response to the determination that the current position information is different from the reference position information, the position information of the positioning fixture module on the slide rail is obtained at each preset interval, and the obtained position information is stored in a preset queue. The adjustment step is then executed again based on the preset queue.

4. The method according to claim 1, wherein, Each adjustable positioning and fastening device is equipped with a displacement sensor, a locking mechanism, and a feeding mechanism. The method of controlling the adjustable positioning and fastening device to perform gap detection processing on the gas turbine blade to be processed within its working area, in order to lock the blade within its working area, includes: The distance between the contact surface of the adjustable positioning fastening device and the corresponding surface of the gas turbine blade to be processed is collected by the displacement sensor installed on the adjustable positioning fastening device as the fitting gap distance information. Based on the fitting gap distance information, the feed compensation value of the feeding mechanism is generated; Based on the feed compensation value, the positioning and fastening device is adjusted to lock the blades in its area of ​​action.

5. The method according to claim 4, wherein, The step of adjusting the positioning and fastening device based on the feed compensation value to lock the blades in its area of ​​action includes: Based on the feed compensation value, the feed mechanism of the positioning and fastening device is driven to perform a feed operation so that the contact surface of the adjustable positioning and fastening device is in contact with the corresponding surface of the gas turbine blade to be processed. After the feed operation is completed, a locking command for the positioning and fastening device is generated; Based on the locking command of the positioning and fastening device, the locking mechanism of the adjustable positioning and fastening device is controlled to perform position locking processing on the adjustable positioning and fastening device to lock the blades in its working area.

6. The method according to claim 1, wherein, Each adjustable support device is equipped with a laser triangulation sensor, a height adjustment mechanism, and a locking mechanism. The system controls the adjustable support device to provide height-adaptive compensation for the gas turbine blades to be processed within its operating area, thereby supporting the blades in that area. The laser triangulation sensor in the adjustable support device detects the distance between the current support surface height of the adjustable support device and the gas turbine blade to be processed in its area of ​​action, and uses this distance as the height adaptive compensation value. In response to the determination that the height adaptive compensation value is greater than the preset fitting threshold, the following driving steps are executed: Based on the height adaptive compensation value, the height adjustment mechanism of the adjustable support device is driven to perform adaptive adjustment of the support height. After the height adaptive adjustment process, the distance from the current support surface height of the adjustable support device to the gas turbine blade to be processed in its area of ​​action is collected again by the laser triangulation sensor in order to update the height adaptive compensation value. In response to the determination that the updated height adaptive compensation value is greater than the preset fitting threshold, the driving step is executed again; In response to determining that the updated height adaptive compensation value is less than or equal to the preset fitting threshold, the locking mechanism of the adjustable support device is controlled to perform a support posture locking process on the adjustable support device to support the blades in its working area.

7. A self-adaptive clamping and fixing device for gas turbine blades, comprising: The model detection unit is configured to perform model detection processing on the gas turbine blades to be processed, which are set in the detection area at the fixture slide rail, to obtain blade version information; The drive unit is configured to drive the positioning fixture module corresponding to the blade version information to move along the slide rail connected to the detection area at the fixture slide rail to the preset receiving station where the gas turbine blade to be processed is located, so as to receive the gas turbine blade to be processed. The positioning fixture module includes at least one adjustable positioning fastening device and at least one adjustable support device. The first control unit is configured to control each of the at least one adjustable positioning fastening device to perform a gap detection process on the gas turbine blade to be processed in its area of ​​action, so as to lock the blade in its area of ​​action. The second control unit is configured to control each of the at least one adjustable support device to perform height-adaptive compensation processing on the gas turbine blade to be processed in its working area, so as to support the blade in its working area.

8. An electronic device, comprising: Positioning clamp module; One or more processors; A storage device on which one or more programs are stored; When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1 to 6.

9. A computer-readable medium having a computer program stored thereon, wherein, When the program is executed by the processor, it implements the method as described in any one of claims 1 to 6.

Images