Automatic detection line for engine heat insulation layer defects

Abstract

The invention provides an engine heat insulation layer defect automatic detection line, and relates to the technical field of engine detection, the engine heat insulation layer defect automatic detection line comprises a marble platform, and the top of the marble platform is provided with a measuring rod fixing welding seat. According to the invention, the position and direction of the monitoring mechanism are adjusted to adapt to different detection requirements, the mechanical arm can accurately position and control the detection tool at the position needing to be detected, errors and uncertainty of manual operation are avoided, and the accuracy and reliability of detection are improved. The mechanical arm also has high flexibility and accuracy, can freely move in a three-dimensional space, and can carry out detection work with high repeatability and long duration time, so that the working efficiency and accuracy are improved, and the detection mechanism has the advantages that the detection mechanism can be used for carrying out nondestructive detection technologies such as ultrasonic wave, laser imaging, thermal imaging and the like; the thermal insulation layer surface is scanned and analyzed to detect and identify surface defects and damage.

Description

Technical Field

[0001] The invention relates to the technical field of engine detection, and in particular to an automatic detection line for defects in an engine thermal insulation layer. Background Art

[0002] The engine insulation defect automatic detection line is an automated system for detecting engine insulation defects. The engine insulation is a protective layer on the outside of the engine that protects the parts inside the engine from high temperature and high pressure. Due to the complexity of the working environment and the importance of the insulation layer, it is critical to regularly check the insulation layer to ensure that it is intact. The engine insulation defect automatic detection line uses a variety of sensors and imaging technologies to detect any defects or damage on the surface of the insulation layer. These technologies include infrared imaging, ultrasonic imaging, lidar, thermal imagers, etc. The system will compare the detected data with pre-set standards and analyze it using algorithms and machine learning technologies to determine whether the insulation layer is defective or damaged. If any problems are detected, the system will automatically issue an alarm and record the location and type of the defect for repair. The advantages of the engine insulation defect automatic detection line are that it can improve detection efficiency and accuracy, while reducing the need for manual inspections and reducing the risk of human error. In addition, the system can also monitor the condition of the insulation layer in real time, detect and repair any defects in a timely manner, thereby extending the life of the engine and improving safety performance. Summary of the invention

[0003] In order to achieve the above-mentioned purpose, the present invention adopts the following technical scheme: an automatic detection line for engine insulation layer defects, comprising a marble platform, a measuring rod fixed welding seat is arranged on the top of the marble platform, a monitoring rod is arranged on the outer surface of the measuring rod fixed welding seat, a slide rail slider is arranged on the top of the marble platform, a drag chain is arranged on the top of the marble platform, a mechanical arm is arranged on the top of the marble platform, a detection mechanism is installed on the mechanical arm, a transmission station is installed on the top of the marble platform, and a chassis is installed on the marble platform.

[0004] Furthermore, the detection mechanism uses a deep learning computer vision algorithm.

[0005] Furthermore, the deep learning computer vision algorithm comprises the following steps:

[0006] S1. Network architecture;

[0007] S2, convolution;

[0008] S3, visual scanning;

[0009] S4, three-point cloud stitching.

[0010] Furthermore, the network architecture of step S1 includes the following steps: first, the data INPUT, i.e., the input layer, the size of the input image is normalized to 32*32, where Conv represents the convolutional layer, AvgPool represents the average pooling layer, and FC represents the fully connected layer.

[0011] Furthermore, the step S2 convolution includes the following steps: after a picture of a handwritten 0 is input into the network, information of two pooling layers and one fully connected layer is obtained, and the final recognition structure is 0.

[0012] Furthermore, the step S3 of visual scanning includes the following steps:

[0013] S31, Scan the inner wall

[0014] Moving along the axis of the engine, the camera scans a section of the engine's inner wall insulation;

[0015] Turn the engine or the camera to scan the second piece of the engine inner wall insulation layer;

[0016] Repeat the rotation and scanning until all the insulation layers on the inner wall of the engine have been scanned;

[0017] S32. Use the deep learning model to perform defect detection on the scanned film of the thermal insulation layer on the inner wall of the engine;

[0018] S33, using a point cloud registration algorithm to stitch together multiple scanning point cloud slices of the engine internal warning insulation layer to restore them into an overall three-dimensional point cloud;

[0019] S34. The detected defect positions are mapped to the point cloud defect mapping technology on the overall engine three-dimensional point cloud, and the ultrasonic air coupling detection technology is used to review and compare the defect data.

[0020] Furthermore, the step S4 of stitching three point clouds includes the following steps:

[0021] Point cloud registration: Input two point clouds P S (Source) and P t (target), output a transformation T such that T(P S ) and P t The overlap is high. For the case where T is a rigid transformation, the registration can be described as:

[0022]

[0023] Where P S (Source) and P t (and) are corresponding points in the origin point cloud and the target point cloud;

[0024] The point cloud (x, y, z) is transformed into cylindrical coordinates along the axis of the engine to (θ, z, r). Assuming that the engine has the straight line x=0, y=0 as the axis, the coordinate mapping relationship is:

[0025]

[0026] The engine point cloud is converted into a depth map with (θ, Z) as the two coordinate axes and d as the index. Deep learning is applied to the two-dimensional image to inspect defects.

[0027] Compared with the prior art, the advantages and positive effects of the present invention are:

[0028] 1. In the present invention, the measuring rod fixing welding seat can fix the measuring rod at a precise position, avoiding errors and uncertainties in manual operation, and improving the accuracy and reliability of measurement. In addition, the measuring rod fixing welding seat can also reduce the shaking and vibration of the measuring rod, thereby ensuring the stability of the measurement results. The advantages of the monitoring rod: the monitoring rod can fix the monitoring mechanism at the position to be detected, so that the monitoring mechanism can be accurately positioned and controlled at the position to be detected, improving the accuracy and reliability of monitoring.

[0029] 2. In the present invention, the monitoring rod can also adjust the position and direction of the monitoring mechanism by means of telescoping and rotating to adapt to different detection requirements. The robotic arm can accurately position and control the detection tool at the position required for detection, avoiding the errors and uncertainties of manual operation and improving the accuracy and reliability of detection. The robotic arm also has a high degree of flexibility and precision, can move freely in three-dimensional space, and can perform detection work with high repeatability and long duration, improving work efficiency and accuracy. The advantages of the detection mechanism: The detection mechanism can scan and analyze the surface of the insulation layer through non-destructive testing technology, such as ultrasound, laser imaging, thermal imaging, etc., to detect and identify surface defects and damage. The detection mechanism has a high degree of precision and reliability, and can quickly and accurately detect defects and damage on the surface of the insulation layer, avoiding the subjectivity and unreliability of manual detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 A stereogram of an automatic detection line for defects in the thermal insulation layer of an engine is provided for the present invention;

[0031] Figure 2 It is a convolutional neural network diagram of the present invention;

[0032] Figure 3 This is a flow chart of the convolutional neural network of the present invention;

[0033] Figure 4 It is a convolution operation graph in deep learning of the present invention;

[0034] Figure 5 Relationship diagram between the input layer and the convolution data of the present invention

[0035] Legend:

[0036] 1. Marble platform; 2. Measuring rod fixed welding seat; 3. Monitoring rod; 4. Slide rail slider; 5. Robotic arm; 6. Detection mechanism; 7. Chassis; 8. Transfer station; 9. Drag chain. DETAILED DESCRIPTION

[0037] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0038] Example 1

[0039] like Figure 1-5 As shown, the present invention provides a technical solution: an automatic detection line for defects in an engine insulation layer, comprising a marble platform 1, a measuring rod fixing welding seat 2 is arranged on the top of the marble platform 1, a monitoring rod 3 is arranged on the outer surface of the measuring rod fixing welding seat 2, a slide rail slider 4 is arranged on the top of the marble platform 1, a drag chain 9 is arranged on the top of the marble platform 1, a mechanical arm 5 is arranged on the top of the marble platform 1, a detection mechanism 6 is arranged on the outer surface of the mechanical arm 5, a transmission station 8 is arranged on the top of the marble platform 1, and a chassis 7 is installed on the marble platform 1.

[0040] The detection mechanism 6 uses a deep learning computer vision algorithm.

[0041] The deep learning computer vision algorithm consists of the following steps:

[0042] S1. Network architecture;

[0043] Step S1 The network architecture includes the following steps: First, the data INPUT is the input layer, and the size of the input image is normalized to 32*32, where Conv represents the convolutional layer, AvgPool represents the average pooling layer, and FC represents the fully connected layer.

[0044] S2, convolution;

[0045] Step S2 convolution includes the following steps: after a picture of a handwritten 0 is input into the network, information from two pooling layers and one fully connected layer is obtained, and the final recognition structure is 0.

[0046] S3, visual scanning;

[0047] 7. Step S3 Visual scanning includes the following steps:

[0048] S31, Scan the inner wall

[0049] Moving along the axis of the engine, the camera scans a section of the engine's inner wall insulation;

[0050] Turn the engine or the camera to scan the second piece of the engine inner wall insulation layer;

[0051] Repeat the rotation and scanning until all the insulation layers on the inner wall of the engine have been scanned;

[0052] S32. Use the deep learning model to perform defect detection on the scanned film of the thermal insulation layer on the inner wall of the engine;

[0053] S33, using a point cloud registration algorithm to stitch together multiple scanning point cloud slices of the engine internal warning insulation layer to restore them into an overall three-dimensional point cloud;

[0054] S34. The detected defect positions are mapped to the point cloud defect mapping technology on the overall engine three-dimensional point cloud, and the ultrasonic air coupling detection technology is used to review and compare the defect data.

[0055] S4, three-point cloud stitching;

[0056] Step S4: stitching three point clouds includes the following steps:

[0057] Point cloud registration: Input two point clouds P S (Source) and P t (target), output a transformation T such that T(P S ) and P t The overlap is high. For the case where T is a rigid transformation, the registration can be described as:

[0058]

[0059] Where P S (Source) and P t (and) are corresponding points in the origin point cloud and the target point cloud;

[0060] The point cloud (x, y, z) is transformed into cylindrical coordinates along the axis of the engine to (θ, z, r). Assuming that the engine has the straight line x=0, y=0 as the axis, the coordinate mapping relationship is:

[0061]

[0062] The engine point cloud is converted into a depth map with (θ, Z) as the two coordinate axes and d as the index. Deep learning is applied to the two-dimensional image to inspect defects.

[0063] In this embodiment,

[0064] The measuring rod fixed welding seat 2 in the automatic detection line for defects in the engine insulation layer is a device for fixing the monitoring rod 3. Its function is to fix the monitoring rod 3 and ensure that it can accurately scan the surface of the insulation layer. The monitoring rod 3 is usually composed of an ultrasonic or laser imaging sensor, which is used to detect defects and damage on the surface of the insulation layer. The monitoring rod 3 fixed welding seat firmly fixes the monitoring rod 3 on the aircraft engine to ensure that it will not move or shake during the detection process and can maintain a constant distance from the surface of the insulation layer. The monitoring rod 3 fixed welding seat can also provide support between the monitoring rod 3 and the surface of the insulation layer to prevent the monitoring rod 3 from bending and deforming, thereby ensuring the accuracy and reliability of the measurement. At the same time, the fixed welding seat can also keep the position and direction of the monitoring rod 3 stable, reducing the risk of errors and missed detection.

[0065] The monitoring rod 3 is a device used to monitor defects in the engine's thermal insulation layer. Its function is to monitor the state of the thermal insulation layer surface so as to promptly detect and deal with any potential defects or damage. The monitoring rod 3 is usually composed of multiple sensors and imaging devices, such as infrared cameras, laser radars, ultrasonic probes, etc., which are used to detect parameters such as temperature, shape, thickness and defects on the surface of the thermal insulation layer. The monitoring rod 3 can be fixed in order to comprehensively monitor the condition of the thermal insulation layer. The function of the monitoring rod 3 is to monitor the condition of the thermal insulation layer in real time and promptly detect any defects or damage, such as timely replacement or repair of the engine to avoid potential safety risks and business losses. The monitoring rod 3 can also record and store detection data for subsequent analysis and comparison. These data can be used to optimize maintenance plans and predict the life and performance of the thermal insulation layer;

[0066] The slide rail slider 4 is usually installed on the engine to support and fix the measuring rod. It can keep the measuring rod stable during the detection process and prevent it from shaking, rotating or moving, thereby ensuring the accuracy and reliability of the measurement results. In addition, the slide rail slider 4 can also reduce the friction and wear between the measuring rod and the surface of the insulation layer, extending the service life of the measuring rod;

[0067] The drag chain 9 is a device for protecting the measuring rod, which is usually composed of multiple links and can be bent freely according to the movement trajectory of the measuring rod. The drag chain 9 can protect the measuring rod from interference from the external environment, such as vibration, collision, wear, etc., thereby reducing damage and failure of the measuring rod. In addition, the drag chain 9 can also prevent the measuring rod from being wrapped around other components, ensuring the free movement of the measuring rod.

[0068] The device uses a marble platform 1 as a carrier, wherein the advantages of the marble platform 1 are: the marble platform 1 has excellent physical properties, such as high density, high hardness and low thermal expansion coefficient, which makes it highly stable and rigid. This can ensure that the position and direction of the measuring rod and the sensor on the platform remain stable, thereby improving the accuracy and reliability of the measurement. The marble platform 1 has a very flat surface, which can provide a smooth and defect-free working platform for the measuring rod and the sensor. This can avoid the unevenness or protrusions on the platform surface from affecting the measurement process, and ensure the accuracy and stability of the measurement results. The marble platform 1 has a good vibration isolation effect and can effectively suppress external vibrations and interference. This can ensure that the measuring rod and the sensor will not be affected by external interference such as machine vibration during the working process, thereby improving the accuracy and reliability of the measurement results. The marble platform 1 has good corrosion resistance and can resist damage to the platform surface by factors such as chemicals, water and moisture. This can extend the service life of the platform, while also ensuring the accuracy and stability of the measurement results. The surface of the marble platform 1 is smooth and flat, easy to clean and maintain, which can ensure the cleanliness and sanitation of the platform surface, and can also reduce the impact of pollution or damage to the platform surface on the measurement results.

[0069] The detection mechanism 6 is composed of multiple sensors and imaging devices, such as infrared cameras, laser radars, ultrasonic probes, etc., which are used to detect parameters such as temperature, shape, thickness and defects on the surface of the thermal insulation layer, so as to comprehensively monitor the condition of the thermal insulation layer. The detection mechanism 6 also includes a data collector and a data processor, which are used to collect, store and analyze monitoring data. The data collector can convert the data collected by the sensors and imaging devices into digital signals and transmit them to the data processor for processing. The data processor can process, analyze and compare the monitoring data to find and identify defects and damage on the surface of the thermal insulation layer. At the same time, it can also record and store the monitoring data for subsequent analysis and comparison;

[0070] The robot arm 5 in the automatic detection line for defects in the engine thermal insulation layer is a device for operating and controlling the detection tool. Its function is to accurately position and control the detection tool at the position required for detection, so as to accurately detect and analyze the surface of the thermal insulation layer. The robot arm 5 is usually composed of multiple joints and actuators, which can simulate the movement and action of human hands and has high flexibility and accuracy. In the automatic detection line for defects in the engine thermal insulation layer, the robot arm 5 can be equipped with various detection tools, such as thermal imagers, laser radars, ultrasonic probes, etc., for detecting and analyzing the surface of the thermal insulation layer. The function of the robot arm 5 is to accurately control the position and direction of the detection tool, so as to accurately detect and analyze the surface of the thermal insulation layer. The robot arm 5 can move freely in three-dimensional space and can be accurately positioned and motion controlled by an electronic control system, so as to accurately move the detection tool to the position required for detection, avoiding errors and uncertainties in manual operation. In addition, the robot arm 5 can also perform detection work with high repeatability and long duration, thereby improving work efficiency and accuracy.

[0071] The above description is only a preferred embodiment of the present invention and does not limit the present invention in other forms. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. An automatic detection line for defects in engine insulation layers, comprising a marble platform (1), characterized in that: A measuring rod fixing welding seat (2) is arranged on the top of the marble platform (1), a monitoring rod (3) is arranged on the outer surface of the measuring rod fixing welding seat (2), a slide rail slider (4) is arranged on the top of the marble platform (1), a drag chain (9) is arranged on the top of the marble platform (1), a mechanical arm (5) is arranged on the top of the marble platform (1), a detection mechanism (6) is installed on the mechanical arm (5), a transmission station (8) is installed on the top of the marble platform (1), and a cabinet (7) is installed on the marble platform (1).

2. The automatic detection line for engine insulation layer defects according to claim 1 is characterized by: The detection mechanism (6) uses a deep learning computer vision algorithm.

3. The automatic detection line for engine insulation layer defects according to claim 2 is characterized by: The deep learning computer vision algorithm includes the following steps: S1. Network architecture; S2, convolution; S3, visual scanning; S4, three-point cloud stitching.

4. The automatic detection line for engine insulation layer defects according to claim 3 is characterized by: The network architecture of step S1 includes the following steps: first, the data INPUT is the input layer, the size of the input image is normalized to 32*32, where Conv represents the convolution layer, AvgPool represents the average pooling layer, and FC represents the fully connected layer.

5. The automatic detection line for engine insulation layer defects according to claim 3 is characterized by: The step S2 convolution includes the following steps: after a handwritten 0 picture is input into the network, information of two pooling layers and one fully connected layer is obtained, and the final recognition structure is 0.

6. The automatic detection line for engine insulation layer defects according to claim 3 is characterized by: The step S3 visual scanning comprises the following steps: S31, Scan the inner wall Moving along the axis of the engine, the camera scans a section of the engine's inner wall insulation; Turn the engine or the camera to scan the second piece of the engine inner wall insulation layer; Repeat the rotation and scanning until all the insulation layers on the inner wall of the engine have been scanned; S32. Use the deep learning model to perform defect detection on the scanned film of the thermal insulation layer on the inner wall of the engine; S33, using a point cloud registration algorithm to stitch together multiple scanning point cloud slices of the engine internal warning insulation layer to restore them into an overall three-dimensional point cloud; S34. The detected defect positions are mapped to the point cloud defect mapping technology on the overall engine three-dimensional point cloud, and the ultrasonic air coupling detection technology is used to review and compare the defect data.

7. The automatic detection line for engine insulation layer defects according to claim 3 is characterized by: The step S4 of three-point cloud stitching includes the following steps: Point cloud registration: Input two point clouds P S (Source) and P t (target), output a transformation T such that T(P S ) and P t The overlap is high. For the case where T is a rigid transformation, the registration can be described as: Where P S (Source) and P t (and) are corresponding points in the origin point cloud and the target point cloud; The point cloud (x, y, z) is transformed into cylindrical coordinates along the axis of the engine to (θ, z, r). Assuming that the engine has the straight line x=0, y=0 as the axis, the coordinate mapping relationship is: The engine point cloud is converted into a depth map with (θ, Z) as the two coordinate axes and d as the index. Deep learning is applied to the two-dimensional image to inspect defects.

Images