Method for detecting and rating tiny defects of turbine blade of aero-engine

Abstract

The invention discloses an aero-engine turbine blade tiny defect detection rating method, which is realized based on a multi-mechanism optimization network, and comprises the following steps: acquiring a video or an image of an aero-engine turbine blade, and marking defect features offline; constructing an aero-engine turbine blade small defect detection network, optimizing convolution modules of a feature extraction network and a feature fusion network in the network, adding an attention mechanism, and optimizing a model boundary loss function; training parameters of the micro-defect detection network of the turbine blade of the aero-engine are set, the micro-defect detection network of the turbine blade of the aero-engine is obtained through training, the video or image of the turbine blade of the aero-engine is input into the micro-defect detection network of the turbine blade of the aero-engine obtained through training, and the defect confidence coefficient and a bounding box are output; and rating the tiny defects of the turbine blade of the aero-engine through the identified defect parameters.

Description

Technical Field

[0001] The present invention relates to the field of intelligent detection technology, and in particular to a method for detecting and rating minute defects in turbine blades of an aero-engine. Background Art

[0002] Aircraft engine turbine blades are the core components of aircraft power systems, and their performance and reliability are crucial to flight safety. Turbine blades are prone to minor defects in the working environment of high temperature, high pressure, and high-speed rotation. If these defects are not detected and handled in time, they may cause blade breakage and lead to serious accidents. The aircraft engine turbine blade minor defect detection and rating method based on a multi-mechanism optimization network is of great significance for improving the accuracy and efficiency of blade detection and ensuring flight safety, and provides strong technical support for the maintenance and overhaul of aircraft engines.

[0003] Among the existing related technologies for aircraft engine turbine blade defect detection before the present invention, there are several comparative patents and documents: CN202311193610.3, CN202311069680.8, and CN202211433839.5.

[0004] The above-mentioned specific patent comparative documents are:

[0005] 1) "A miniature wireless controllable visual device for detecting turbine blades of aircraft engines", patent number CN202311193610.3. This invention discloses a miniature wireless controllable visual device for detecting turbine blades of aircraft engines, which relates to the field of industrial robot vision technology. From top to bottom, it includes: a camera lighting module, a rotating frame, an electrothermal drive module, an electronic component module, a battery, and a support frame; the camera lighting module realizes image acquisition, lighting, sensing and other functions, the electrothermal drive module realizes the rotation of the camera lighting module based on the different principles of thermal expansion coefficients of different materials, the electronic component module realizes data processing and transmission and other functions, the battery is used as a power source, and the support frame loads the visual system onto living insects and micro robots. The camera lighting module can rotate around the vertical and horizontal directions at a specified angle, thereby realizing image acquisition in different directions. The system of the present invention has the advantages of small size, light weight, wide image acquisition range, wireless controllability, etc. It can be loaded on small organisms and micro robots to enter the aircraft engine to perform turbine blade detection tasks. The detection method of the present invention is different from the above invention. It improves the detection method for tiny defects of aircraft engine turbine blades and performs quantitative rating and classification on four of the defects.

[0006] 2) "A hollow turbine blade detection method, system and equipment", patent number CN202311069680.8. This invention discloses a hollow turbine blade detection method, system and equipment, which relates to the field of X-ray detection technology. Obtain the standard CAD model data of the ceramic core, the wax mold standard CAD model data, the mold shell standard CAD model data and the blade casting standard CAD model data of the hollow turbine blade; obtain N two-dimensional images of the ceramic core, M two-dimensional images of the wax mold, Q two-dimensional images of the mold shell and W two-dimensional images of the blade casting; obtain the three-dimensional model data of the ceramic core, the three-dimensional model data of the wax mold, the three-dimensional model data of the mold shell and the three-dimensional model data of the blade casting; obtain the comparison result of the hollow turbine blade to be detected; perform difference calculation between the comparison result and the preset threshold to obtain the difference; obtain the updated difference; determine whether the updated difference is greater than the first threshold; if so, output that the full process detection of the hollow turbine blade is unqualified; otherwise, output that the full process detection of the hollow turbine blade is qualified. The present invention improves the accuracy of the full process detection of the hollow turbine blade. The detection object of the present invention is different from that of the above inventions. The defect detection algorithm and improvement of the present invention are studied on specific aircraft engine turbine blades and can detect tiny defects.

[0007] 3) "A rapid detection device for aircraft engine turbine blades", patent number CN202211433839.5. This invention discloses a rapid detection device for aircraft engine turbine blades, which relates to the field of turbine blade detection technology. The device includes a transmission disk for transmitting turbine blades, and a plurality of support rings for supporting turbine blades are distributed in an array on the upper end of the transmission disk. The upper end surface of the support ring is provided with a rolling member for facilitating fine-tuning of the turbine blades. A detection component for detecting the turbine blades is provided on the transmission path of the transmission disk. The detection component includes a base, and a transmission groove matched with the transmission disk is provided on the base. The detection method here in the present application does not require the blades to be disassembled for detection, which effectively ensures the processing efficiency, and the detection here does not require immersion operation, which eliminates the problem of slow immersion process, effectively ensures the detection effect, and the flow-type detection method here is also convenient for loading and unloading, with high safety and strong practicality. The focus of the turbine blade detection of the present invention is different from that of the above invention. The above invention focuses on rapid detection, while the present invention focuses on accurate detection of minor defects. Summary of the invention

[0008] In order to solve the above technical problems, the purpose of the present invention is to provide a method for detecting and rating tiny defects of aircraft engine turbine blades, which is implemented based on a multi-mechanism optimization network.

[0009] The purpose of the present invention is achieved through the following technical solutions:

[0010] A method for detecting and rating micro-defects of turbine blades of an aero-engine is provided. The method is implemented based on a multi-mechanism optimization network and includes:

[0011] A. Collect videos or images of aircraft engine turbine blades and annotate defect features offline;

[0012] B. Construct a network for detecting micro-defects in turbine blades of aircraft engines, optimize the convolution modules of the feature extraction network and feature fusion network in the network, add an attention mechanism, and optimize the model boundary loss function.

[0013] C. Setting training parameters of an aircraft engine turbine blade micro-defect detection network and training an aircraft engine turbine blade micro-defect detection network, inputting an aircraft engine turbine blade video or image into the trained aircraft engine turbine blade micro-defect detection network, and outputting defect confidence and bounding box;

[0014] D. Rating the minor defects of aircraft engine turbine blades through the identified defect parameters.

[0015] Compared with the prior art, one or more embodiments of the present invention may have the following advantages:

[0016] The present invention detects tiny defects of aircraft engine turbine blades based on video or images, and grades the degree of defects, which is effectively applied to the safe maintenance of aircraft engine turbine blades. The use of this method has great practical engineering application value in improving the accuracy and efficiency of blade detection and ensuring flight safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a flow chart of the method for detecting and rating micro defects in aircraft engine turbine blades based on a multi-mechanism optimization network;

[0018] Figure 2 This is the LFD-YOLO network diagram based on multiple mechanism optimization networks;

[0019] Figure 3 It is the LDConv structure diagram;

[0020] Figure 4 It is a diagram of the DAT structure;

[0021] Figure 5 This is the effect diagram of detecting tiny defects in aircraft engine turbine blades. DETAILED DESCRIPTION

[0022] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be further described in detail below in conjunction with embodiments and drawings.

[0023] like Figure 1As shown in FIG. 1 , a method flow for detecting and rating micro-defects of turbine blades of an aero-engine based on a multi-mechanism optimization network is shown, which includes the following steps:

[0024] Step 10: collecting videos or images of aircraft engine turbine blades and annotating defect features offline;

[0025] Step 20 builds an aircraft engine turbine blade micro-defect detection network, optimizes the convolution modules of the feature extraction network and the feature fusion network in the network, adds an attention mechanism, and optimizes the model boundary loss function;

[0026] Step 30 sets training parameters of an aircraft engine turbine blade micro-defect detection network and trains the aircraft engine turbine blade micro-defect detection network, inputs an aircraft engine turbine blade video or image into the trained aircraft engine turbine blade micro-defect detection network, and outputs defect confidence and a bounding box;

[0027] Step 40 ranks the minor defects of the aircraft engine turbine blades according to the identified defect parameters.

[0028] The above-mentioned step 10 specifically includes: collecting videos or images of tiny defects of aircraft engine turbine blades through an industrial endoscope; using the open source tool labelme to mark various defects such as cracks, burns, notches, material loss, and dents.

[0029] The above step 20 specifically includes: the aircraft engine turbine blade micro-defect detection network is an LFD-YOLO network, the LFD-YOLO network uses YOLOv8 as the basic network, and uses LDConv to optimize the CBS module of the initial network; uses the DAT deformable attention mechanism to optimize the feature extraction and feature fusion network; uses Focaler-IoU to optimize the loss function module of the initial network, and completes the automatic detection of micro-defects of aircraft engine turbine blades, such as Figure 2 Shown is the LFD-YOLO network diagram.

[0030] like Figure 3 The structure diagram is shown in Figure 1. LDConv replaces some CBS modules in the YOLOv8 backbone network with LDConv. LDConv uses irregular convolution kernels for convolution operations. First, the initial sampling coordinates P of the irregular convolution kernel need to be generated. n , Table 1 is P n Generate algorithm pseudocode:

[0031] Table 1

[0032]

[0033]

[0034] After defining the irregular convolution initial coordinates P n After that, at position P 0 The corresponding convolution operation at can be defined as follows:

[0035]

[0036] Where R and w are the sampling grid and convolution parameters respectively.

[0037] First, use Algorithm 1 to generate the initial coordinates P n , which generates the corresponding sampling coordinates (P n +P 0 ); secondly, use Conv2d to perform a convolution operation on the input image, obtain the corresponding convolution kernel offset through the convolution operation, and then add the offset to the original coordinates to adjust the initial sampling shape. This step realizes the dynamic adjustment of the convolution kernel shape to adapt to the image characteristics; finally, the feature map is resampled according to the adjusted sampling shape. The resampled feature map is reshaped, convolved again, and standardized, and finally the feature map is output through the activation function SiLU.

[0038] The original Yolov8 network uses CIoU as the bounding box regression loss. The CIoU calculation formula is as follows:

[0039]

[0040] Among them, ρ 2 (B,B gt ) represents the Euclidean distance between the center points of the prediction box and the ground-truth box, c represents the B box, B gt The minimum diagonal distance between the bounding boxes, w, h, w gt 、h gt Represent the width and height of the prediction box and ground-truth box respectively. Then the CIoU loss function L CIoU The calculation formula is:

[0041] L CIoU =1-CIOU (5)

[0042] This paper uses Focaler-CIoU loss to replace the CIoU loss in Yolov8, and improves the performance of the detector in different detection tasks by focusing on different regression samples. Focaler The linear interval mapping method is used to reconstruct the IoU loss, IoU Focaler The calculation formula is as follows:

[0043]

[0044] Among them, [d,u]∈[0,1], by adjusting the d and u values, IoU Focaler More suitable for different regression samples. Combined with the above CIoU calculation process, the Focaler-CIoU loss function L Focaler-CIoU The calculation formula is as follows:

[0045] L Focaler-CIoU =L CIoU +IoU-IoU Focaler (7)

[0046] Optimized network bounding box regression loss function L B_box for:

[0047] L B_box =L DFL +L Focaler-CIoU (8)

[0048] like Figure 4 The DAT structure is shown in the figure. In the feature extraction network, DAT is added to the last layer of Backbone, so that the network can adaptively adjust the attention weight according to the input image features, flexibly control the intensity and range of attention, and enhance the network's ability to express tiny features. In the feature fusion network, DAT is added after the C2f module in the downsampling stage, so that the network can maintain high-quality perception of deep discriminative features and high-level semantic information in the feature fusion stage. The DAT module process includes: the input is a feature map x of size H×W×C, from the dimension H G ×W G ×2 uniform grid to select reference points to achieve effective downsampling of the original grid.

[0049] H G =H / r, W G =W / r (9)

[0050] r is a pre-set scaling factor. The reference point value is a linearly spaced coordinate {(0,0),…,(H G -1,W G -1)}, according to the grid shape H G ×W G The reference points are normalized to [-1, +1], (-1, -1) represents the upper left corner, and (+1, +1) represents the lower right corner. To obtain the offset of each reference point, the feature map is linearly projected onto the query tokens:

[0051] q=xW q (10)

[0052] Among them, W q is the query projection weight. Then q passes through the offset generation subnetwork to generate the offset Δp:

[0053] Δp=θ offset (q) (11)

[0054] Combine the reference point and offset information to obtain the deformed reference point, and use bilinear interpolation on the deformed reference point to obtain the sampling of the feature map x:

[0055]

[0056] right Perform linear projection, W k , W v Projection weights for key and value:

[0057]

[0058] Finally, integrating the above results and applying the multi-head attention mechanism, we can get the output z of the deformable attention module (m) :

[0059]

[0060] where σ(·) represents the softmax function, d = C / M is the size of each head, and q (m) , represents the query, key, and value vector output by the mth attention head, is the bilinear interpolation of the relative position deviation.

[0061] The above step 30 specifically includes: randomly dividing the aircraft engine turbine blade micro-defect image dataset (2000 images) into a training set, a validation set, and a test set according to a ratio of 7:2:1; the network training parameter settings mainly include: the optimizer is SGD, the initial learning rate is 0.01, the batch size is 8, and the number of training rounds is 500.

[0062] like Figure 5 The figure shows the effect of detecting small defects in turbine blades of aircraft engines; the above step 40 specifically includes: according to the identified defects, there are the following rating parameter methods for cracks, burns, notches, material loss, etc.,

[0063] 1) Crack defect rating of aircraft engine turbine blades (the parameters used for determination are the maximum crack length L and maximum crack width W), as shown in Table 2:

[0064] Table 2

[0065]

[0066] 2) Burn defect rating of aircraft engine turbine blades (the parameter used for determination is the burn area ratio P), as shown in Table 3:

[0067] Table 3

[0068]

[0069] 3) Rating of notch defects in aircraft engine turbine blades (the parameter used for determination is radial depth d), as shown in Table 4:

[0070] Table 4

[0071]

[0072]

[0073] 4) Material loss defect rating of aircraft engine turbine blades (the parameter used for determination is the contour length C of the loss area), as shown in Table 5:

[0074] Table 5

[0075]

[0076] Although the embodiments disclosed in the present invention are as above, the above contents are only embodiments adopted for facilitating the understanding of the present invention and are not intended to limit the present invention. Any technician in the technical field to which the present invention belongs can make any modifications and changes in the form and details of the implementation without departing from the spirit and scope disclosed in the present invention, but the patent protection scope of the present invention shall still be subject to the scope defined in the attached claims.

Claims

1. A method for detecting and rating micro-defects of turbine blades of an aircraft engine, characterized in that: The method is based on multiple mechanisms to optimize network implementation, including: A. Collect videos or images of aircraft engine turbine blades and annotate defect features offline; B. Construct a network for detecting micro-defects in turbine blades of aircraft engines, optimize the convolution modules of the feature extraction network and feature fusion network in the network, add an attention mechanism, and optimize the model boundary loss function. C. Setting training parameters of an aircraft engine turbine blade micro-defect detection network and training an aircraft engine turbine blade micro-defect detection network, inputting an aircraft engine turbine blade video or image into the trained aircraft engine turbine blade micro-defect detection network, and outputting defect confidence and bounding box; D. Rating the minor defects of aircraft engine turbine blades through the identified defect parameters.

2. The method for detecting and rating micro-defects of aircraft engine turbine blades according to claim 1, characterized in that: The defect characteristics in A include cracks, burns, notches, material loss and indentations.

3. The method for detecting and rating micro-defects of aircraft engine turbine blades according to claim 1, characterized in that: The aircraft engine turbine blade micro-defect detection network in B is an LFD-YOLO network. The LFD-YOLO network uses Yolov8 as the basic network, and uses LDConv to optimize the CBS module of the initial network; the DAT deformable attention mechanism is used to optimize the feature extraction network and the feature fusion network; Focaler-IoU is used to optimize the loss function module of the initial network to complete the automatic detection of micro-defects in aircraft engine turbine blades.

4. The method for detecting and rating micro-defects of aircraft engine turbine blades according to claim 3, characterized in that: Replace some CBS modules in the Yolov8 backbone network with LDConv, which uses irregular convolution kernels for convolution operations, including: Define the initial sampling coordinates P of the irregular convolution kernel n , which generates the corresponding sampling coordinates (P n +P0); use Conv2d to perform a convolution operation on the input image, obtain the corresponding convolution kernel offset through the convolution operation, and then add the offset to the original coordinates to adjust the initial sampling shape; resample the feature map according to the adjusted sampling shape, and the resampled feature map is reshaped, convolved again, and normalized, and then the feature map is output through the activation function SiLU.

5. The method for detecting and rating micro-defects of aircraft engine turbine blades according to claim 4, characterized in that: The corresponding convolution operation at position P0 is defined as follows: Among them, R and w are the sampling grid and convolution parameters respectively.

6. The method for detecting and rating micro-defects of aircraft engine turbine blades according to claim 3, characterized in that: CIoU is used as the bounding box regression loss in the Yolov8 network, and the CIoU calculation formula is as follows: Among them, ρ 2 (B,B gt ) represents the Euclidean distance between the center points of the predicted box and the ground-truth box, α represents the weight parameter, v represents the aspect ratio correction factor, and c represents the B box, B gt The minimum diagonal distance between the bounding boxes, w, h, w gt 、h gt Represent the width and height of the prediction box and the ground-truth box respectively; then the CIoU loss function L CIoU The calculation formula is: L CIoU =1-CIoU (5) Use Focaler-CIoU loss to replace CIoU loss, IoU in Yolov8 Focaler Use the linear interval mapping method to reconstruct the IoU loss, IOU Focaler The calculation formula is as follows: Among them, [d,u]∈[0,1], by adjusting the d and u values, IoU Focaler Applicable to different regression samples; combined with the calculation process of CIoU, the loss function L of Focaler-CIoU Focaler-CIoU The calculation formula is as follows: L Focaler-CIoU =L CIoU +IoU-IoU Focaler (7) Combined with DFL bounding box loss L DFL , the optimized network bounding box regression loss function L B_box for: L B_box =L DFL +L Focaler-CIoU (8) 7. The method for detecting and rating micro-defects of aircraft engine turbine blades according to claim 3, characterized in that: In the feature extraction network, DAT is added to the last layer of Backbone, so that the network can adaptively adjust the attention weight according to the input image features, control the intensity and range of attention, and enhance the network's ability to express tiny features. In the feature fusion network, DAT is added to the C2f module in the downsampling stage, so that the network can maintain high-quality perception of deep discriminative features and high-level semantic information in the feature fusion stage.

8. The method for detecting and rating micro-defects of aircraft engine turbine blades according to claim 7, characterized in that: The input of the DAT deformable attention mechanism module is a feature map x of size H×W×C, which is obtained from the dimension H G ×W G Select reference points from the ×2 uniform grid to achieve effective downsampling of the original grid: H G =H / r,W G =W / r (9) r is a preset scaling factor; the reference point value is a linearly spaced coordinate {(0,0),…,(H G -1,W G -1)}, according to the grid shape H G ×W G Normalize the reference points to [-1, +1], (-1, -1) represents the upper left corner, and (+1, +1) represents the lower right corner. To obtain the offset of each reference point, linearly project the feature map onto the query tokens: q=xW q (10) Among them, W q The query projection weight is then generated by the offset generation subnetwork q to generate the offset Δp: Δp=θ offset (q) (11) Combine the reference point and offset information to obtain the deformed reference point, and use bilinear interpolation on the deformed reference point to obtain the sampling of the feature map x: Linear projection of x, W k , W v Projection weights for key and value: Finally, the multi-head attention mechanism is applied to obtain the deformable attention module output z (m) : where σ(·) represents the softmax function, d = C / M is the size of each head, and q (m) , represents the query, key, and value vector output by the mth attention head, is the bilinear interpolation of the relative position deviation.

9. The method for detecting and rating micro-defects of aircraft engine turbine blades according to claim 1, characterized in that: In C, a dataset of images of tiny defects of aircraft engine turbine blades is obtained through videos or images of aircraft engine turbine blades, and the dataset is divided into a training set, a validation set and a test set; the network training parameters include an optimizer SGD, an initial learning rate of 0.01, a batch size of 8, and a number of training rounds of 500.

10. The method for detecting and rating micro-defects of aircraft engine turbine blades according to claim 2, characterized in that: In D, the rating of identified cracks, burns, notches, and material loss defects includes: The crack defect level of aircraft engine turbine blades is determined by the length L and width W of the crack parameters: Cracked blade surface: 0mm<L≤1mm, the defect level is Grade I; 1mm<L≤3mm and 0mm<W≤2mm, the defect level is Grade II; 1mm<L≤3mm and 2mm<W≤3mm, the defect level is Grade III; L>3mm, the defect level is Grade IV; Cracked blade root: There are cracks, and the defect level is level IV; The burn defect level of aircraft engine turbine blades is determined by the parameter burn area ratio P: Burnt rotor: 0<P≤10%, the defect level is Grade I; 10%<P≤15%, the defect level is Grade II; 15%<P≤20%, the defect level is Grade III; P>20%, the defect level is Grade IV; Burned static: 0<P≤15%, the defect level is Grade I; 15%<P≤20%, the defect level is Grade II; 20%<P≤25%, the defect level is Grade III; P>25%, the defect level is Grade IV; The defect level of the notch in the turbine blade of an aero-engine is determined by the parameter radial depth d: Notched stator: 0mm<d≤2mm, the defect level is Grade I; 2mm<d≤3mm, the defect level is Grade II; 3mm<d≤4mm, the defect level is Grade III; d>3mm, the defect level is Grade IV; Notched rotor: There is a notch on the blade, the defect level is level III; there is a notch on the blade root, the defect level is level IV; The defect level of the loss of the material of the turbine blade of an aero-engine is determined by the parameter of the loss area contour length C: Material loss stator blade: 0mm<C≤2mm, defect level is Ⅰ; 2mm<C≤4mm, defect level is Ⅱ; 4mm<C≤5mm, defect level is Ⅲ; C>3mm, defect level is Ⅳ; Material loss low-pressure blade: 0mm<C≤2mm, defect level is Ⅰ; 2mm<C≤3mm, defect level is Ⅱ; 3mm<C≤4mm, defect level is Ⅲ; C>4mm, defect level is Ⅲ; Material loss high pressure blade: There is a material loss defect, and the defect level is Level IV.

Images