A surface defect detection method for aircraft engine oil pipe

Abstract

The present invention discloses a surface defect detection method for aircraft engine oil pipes, and the steps of the surface defect detection method for aircraft engine oil pipes are as follows: step one: grid division of oil pipe parts; step two: importing various components; step three: determining the workspace of oil pipe parts; step four: calculating and marking the oil pipe detection grid area under a fixed oil pipe posture; step five: covering marking of the oil pipe surface grid; step six: generating the optimal detection posture sequence of the oil pipe; step seven: acquiring the oil pipe surface image; step eight: transmitting the image taken by the camera group to the visual inspection server; step nine: visual defect detection; step ten: recording and displaying the defect detection results. The surface defect detection method for aircraft engine oil pipes of the present invention focuses on full coverage detection of the aircraft engine oil pipe surface, provides a coverage detection method for all grids on the surface of the oil pipe parts, and reduces the missed detection rate of surface defects of the oil pipe parts.

Description

Technical Field

[0001] The invention relates to a surface defect detection method for an aviation engine oil pipe, belonging to the technical field of aviation. Background Art

[0002] Fuel pipes in aircraft engines primarily transport fuel, lubricating oil, and hydraulic oil, a vital component of both engines and aircraft. A rupture in any of these pipes can cause serious accidents, even resulting in aircraft destruction and fatalities. Due to the demanding installation locations and constant fluctuations in temperature and pressure, fuel pipes are subject to complex stresses. The reliability of fuel pipes in aircraft engines is crucial to engine performance and lifespan, making surface defect inspection of these pipes essential.

[0003] Traditional methods generally rely on manual labor to complete oil pipeline defect detection, which has the following main difficulties: First, if the defects in the pipeline material are not fully inspected and maintained, the missed defects will cause abnormal engine operation and pose a serious safety hazard; second, a typical aircraft engine has as many as hundreds of pipes with different shapes and sizes. Relying on manual methods during the inspection process is time-consuming and labor-intensive, and inefficient; third, the surface defects of aircraft engine pipes mainly include cracks, notches, pits, scratches, etc. These defects are relatively irregular. If traditional visual methods are used to extract common features, they are very sensitive to light, angle, etc., and the judgment accuracy and recall rate cannot be guaranteed.

[0004] This method addresses the problems of missed detection, low detection efficiency and accuracy in traditional inspection methods for aircraft engine oil pipes. By combining three-dimensional simulation, intelligent planning of oil pipe posture and deep learning technologies, a surface defect detection method for aircraft engine oil pipes is proposed. Summary of the Invention

[0005] The purpose of the present invention is to provide a surface defect detection method for aircraft engine oil pipes to solve the problems raised in the above background technology.

[0006] To achieve the above-mentioned object, the present invention provides the following technical solution: a method for detecting surface defects of an aircraft engine oil pipe, wherein the method comprises the following steps:

[0007] Step 1: Meshing of oil pipe parts: 3D modeling of the oil pipe parts to be inspected and surface meshing of the parts;

[0008] Step 2: Import components such as the robotic arm, gripper, and camera group into the simulation environment to ensure that the poses of all components in the simulation environment are consistent with those in reality.

[0009] Step 3: Determine the workspace of the tubing part: First, based on the workspace of the manipulator, add collision constraints between the manipulator body, gripper, and tubing part in the simulation environment to obtain the accessible workspace of the tubing part.

[0010] Step 4: Calculate and mark the inspection grid area of the oil pipe under a fixed oil pipe posture. In the simulated 3D environment, an interactive method is used to calculate the inspection grid area of the oil pipe for each oil pipe part posture in the workspace.

[0011] Step 5: Mark the coverage of the tubing surface mesh. Based on the tubing part workspace in step 3 and the tubing inspection mesh area calculation method in step 4, the corresponding tubing inspectable mesh area is calculated and marked for each tubing posture in the workspace until all postures in the tubing workspace are calculated.

[0012] Step 6: Generate the optimal detection posture sequence for the oil pipe: Based on the oil pipe grid marking results in step 5, the oil pipe posture for each grid when it is detectable can be obtained. By integrating all the oil pipe postures, several optimal detection postures can be obtained to form an optimal detection posture sequence. This sequence can complete the coverage detection of the oil pipe surface in the least number of times;

[0013] Step 7: Acquiring an image of the tubing surface: Based on the optimal inspection posture sequence obtained in Step 6, the robotic arm is controlled to position the tubing at each posture value in the sequence. The robot then calculates the cameras used to inspect the grid area at the current tubing posture and controls the focal length of each camera to obtain a high-quality, clear image of the inspection grid.

[0014] Step 8: Transmit the images captured by the camera group to the visual inspection server: Using the method in step 7, several images of the oil pipeline surface are obtained. The pixel positions of the grid area to be inspected in each image are calculated based on the simulation model. The images and corresponding pixel positions are then transmitted to the visual inspection server.

[0015] Step 9: Visual defect detection;

[0016] Step 10: Record and display defect detection results.

[0017] As a preferred technical solution of the present invention, in step 1, the grid division needs to ensure that the grid scale is not greater than a certain value and maintain scale consistency. The purpose of this operation is to divide the oil pipe part from the whole into the smallest units that can be processed separately.

[0018] As a preferred technical solution of the present invention, in step 2, the robot arm base and the camera group are fixed relative to the world coordinate position, the robot arm is 6-axis, and its gripper is installed at the end of the robot arm and the position of clamping the oil pipe remains unchanged.

[0019] As a preferred technical solution of the present invention, the calculation method in step 4 is as follows: a certain density of rays are emitted within the conical field of view of each camera model. The oil pipe meshes that the rays collide with are detected and the incident angle is determined. If the incident angle vector and the normal vector of the collision surface meet a certain angular relationship, the mesh is determined to be detectable by the camera. The mesh is then marked as the camera's detection area, and the corresponding oil pipe posture is recorded.

[0020] As a preferred technical solution of the present invention, in step nine, the visual server processes each image, intercepts the area to be detected, and then uses a deep learning algorithm to classify and identify defects.

[0021] As a preferred technical solution of the present invention, the recognition results of the visual server are received and recorded, and the defect detection results of each grid on the surface of the oil pipe are displayed.

[0022] Compared with the prior art, the present invention has the following beneficial effects: the present invention provides a surface defect detection method for an aircraft engine oil pipe,

[0023] (1) The present invention focuses on full coverage detection of the surface of the oil pipe of an aircraft engine, provides a coverage detection method for all grids on the surface of the oil pipe parts, and reduces the missed detection rate of surface defects of the oil pipe parts.

[0024] (2) The present invention can realize automatic acquisition of surface images of aircraft engine oil pipes and defect identification without human intervention, thereby improving the efficiency of oil pipe defect detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 Schematic diagram of the method of the present invention. DETAILED DESCRIPTION

[0026] The technical solutions in the embodiments of the present invention are clearly and completely described. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0027] See also Figure 1 The present invention provides a surface defect detection method for an aircraft engine oil pipe.

[0028] Step 1: Model and Mesh the Oil Pipe Component: Create a 3D model of the oil pipe component to be tested. The test pipe used in this implementation is approximately 500 mm long and 20 mm in diameter. Once the model is complete, the surface is meshed. The mesh size is approximately 10 mm in straight sections of the pipe, decreasing in areas with greater curvature.

[0029] Step 2: Importing Components: Import the robot arm, gripper, camera assembly, and other component models into the simulation environment, ensuring that the poses of all components in the simulation are consistent with those in reality. In this implementation, five cameras are installed evenly above the tubing workspace, ensuring that the camera's field of view covers the tubing workspace. The robot arm base and camera assembly are fixed relative to the world coordinate system. The robot arm has six axes, and the gripper is installed at the end of the arm, with the tubing gripper located at the base of the end.

[0030] Step 3: Determine the workspace of the tubing part: First, obtain the workspace of the robot arm. Then, in the simulation environment, add collision constraints between the robot arm body, gripper, and tubing part. Then, make the robot arm change its posture in the workspace. Finally, the accessible workspace of the tubing part can be obtained.

[0031] Step 4: Calculation and marking of the oil pipe inspection grid area under fixed oil pipe posture: In the simulated 3D environment, the oil pipe inspection grid area can be calculated using an interactive method for each fixed oil pipe part posture in the workspace. The calculation method is: Figure 1 As shown in the figure, a camera is used as an example. A density of 5 rays per degree is emitted within the camera model's conical field of view. The oil pipeline meshes that the rays collide with are detected and the incident angle is determined. If the angle between the incident angle vector and the collision surface normal vector is less than 30 degrees, the mesh is considered detectable by the camera. The mesh is then marked as the camera's detection area, and the corresponding oil pipeline posture is recorded.

[0032] Step 5: Marking the Coverage of the Tubing Surface Meshes: Based on the tubing workspace in Step 3 and the tubing inspection mesh area calculation method in Step 4, calculate and mark the corresponding detectable mesh area for each tubing pose in the workspace until all poses in the tubing workspace are calculated. If all poses in the workspace are calculated and no tubing meshes are left unmarked, adjust the camera group's pose until all meshes on the tubing surface are marked.

[0033] Step 6: Generate the optimal pipeline detection pose sequence: Based on the pipeline grid marking results from Step 5, the detectable pipeline pose for each grid can be obtained. In this implementation, a total of n different pipeline poses were obtained. All pipeline poses were integrated (adjacent poses were merged so that the new poses could effectively detect the corresponding grids). M optimal detection poses were obtained to form an optimal detection pose sequence. This sequence can complete the coverage inspection of the pipeline surface in the minimum number of attempts.

[0034] Step 7: Acquiring an image of the tubing surface: Based on the m optimal tubing detection posture sequences obtained in Step 6, the robotic arm is controlled to position the tubing at each posture value in the sequence. The simulation environment then calculates the cameras and their focal lengths used to inspect the grid area at the current tubing posture. The focal length of each camera is then controlled, and images are acquired using the relevant cameras at this focal length, resulting in high-quality, clear images of the inspection grid.

[0035] Step 8: Transmit the images captured by the camera group to the visual inspection server: Using the method in step 7, several oil pipeline surface images can be obtained. The pixel positions of the grid area of the oil pipeline to be inspected in each image are calculated based on the simulation model, and the images and corresponding pixel positions are transmitted to the visual inspection server.

[0036] Step 9: Visual Defect Detection: The visual server processes each image, intercepts the area to be inspected, and then obtains an image of the grid area to be inspected. It then uses a vision-based training and inference deep learning system to classify and identify defect features.

[0037] Step 10: Record and display defect detection results: Receive and record the recognition results of the visual server, and display the defect detection results of each grid on the oil pipeline surface. The display is divided into intuitive display and display of detailed defect information.

[0038] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A surface defect detection method for an aircraft engine oil pipe, characterized in that: The surface defect detection method of the aircraft engine oil pipe has the following steps: Step 1: Meshing of oil pipe parts: 3D modeling of the oil pipe parts to be inspected and surface meshing of the parts; Step 2: Import the components, including the robot arm, gripper, and camera group component models into the simulation environment to ensure that the poses of all components in the simulation environment are consistent with those in reality. Step 3: Determine the workspace of the tubing part: First, based on the workspace of the manipulator, collision constraints between the manipulator body, gripper, and tubing part are added in the simulation environment to obtain the accessible workspace of the tubing part. Step 4: Calculate and mark the inspection grid area of the oil pipe under fixed oil pipe posture. In the simulated 3D environment, use an interactive method to calculate the inspection grid area of the oil pipe for each oil pipe part posture in the workspace. Step 5: Mark the coverage of the tubing surface mesh. Based on the tubing part workspace in step 3 and the tubing inspection mesh area calculation method in step 4, calculate the corresponding tubing detectable mesh area for each tubing posture in the workspace and mark it until all postures in the tubing workspace are calculated. Step 6: Generate the optimal detection posture sequence for the oil pipe: Based on the oil pipe grid marking results from step 5, obtain the oil pipe posture when each grid is detectable. By integrating all the oil pipe postures, several optimal detection postures are obtained to form an optimal detection posture sequence. This sequence can complete the coverage detection of the oil pipe surface in the least number of times; Step 7: Acquiring an image of the tubing surface: Based on the optimal inspection posture sequence obtained in Step 6, the robotic arm is controlled to position the tubing at each posture value in the sequence. The robot then calculates the cameras used to inspect the grid area at the current tubing posture and controls the focal length of each camera to obtain a high-quality, clear image of the inspection grid. Step 8: Transmit the images captured by the camera group to the visual inspection server: Using the method in step 7, several images of the oil pipeline surface are obtained. The pixel positions of the grid area to be inspected in each image are calculated based on the simulation model. The images and corresponding pixel positions are then transmitted to the visual inspection server. Step 9: Visual defect detection; Step 10: Record and display defect detection results; In step 2, the robot arm base and the camera assembly are fixed relative to the world coordinates, and the gripper is mounted at the end of the robot arm and the position of the oil pipe remains unchanged; The calculation method in step 4 is as follows: a certain density of rays are emitted within the conical field of view of each camera model, and the oil pipe meshes hit by the rays are detected and the incident angle is determined. If the incident angle vector and the normal vector of the collision surface meet a certain angular relationship, the mesh is determined to be detectable by the camera, and the mesh is then marked as the detection area of the camera, and the corresponding oil pipe posture is recorded; In step nine, the visual server processes each image, captures the area to be inspected, and then uses a deep learning algorithm to classify and identify defects.

2. The surface defect detection method for an aircraft engine oil pipe according to claim 1, characterized in that: The recognition results of the visual server are received and recorded, and the defect detection results of each grid on the oil pipeline surface are displayed.

Images