Device and method for detecting coil integrity of high-voltage transformer winding machine

Abstract

The invention discloses a high-voltage transformer winding machine coil integrity detection device and method, and the device comprises an image collection module which is used for obtaining a surface image of a wound coil; the depth detection module is used for acquiring depth information of the wound coil and generating three-dimensional point cloud data; the image processing module is used for preprocessing the surface image and the three-dimensional point cloud data; the size and depth analysis module is used for judging whether the size and depth of the wound coil meet preset standards or not according to the preprocessed surface image and the three-dimensional point cloud data; the integrity detection module is used for judging whether the wound coil has defects or not according to the preprocessed surface image and the three-dimensional point cloud data; the report generation module is used for generating a detection report according to judgment results of the size and depth analysis module and the integrity detection module; and the control and display module is used for inputting control information of the device and displaying a detection report. According to the invention, high-efficiency and accurate detection of the coil is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of high-voltage transformer manufacturing, and in particular to a device and method for detecting the integrity of a coil of a high-voltage transformer winding machine. Background Art

[0002] During the manufacturing process of high-voltage transformers, the quality of the coils produced by winding machines directly impacts the transformer's performance and safety. Traditional coil inspection methods rely primarily on manual visual inspection or simple dimensional measurement tools. These methods suffer from low efficiency, poor accuracy, and high subjectivity, making them difficult to meet the demands of modern industry for high-precision, automated inspection. Furthermore, coil integrity testing requires precise measurement of the size and depth of each section, making traditional methods difficult to achieve comprehensive and efficient inspection. Summary of the Invention

[0003] The object of the present invention is to provide a device and method for detecting the integrity of a coil of a high-voltage transformer winding machine, so as to achieve comprehensive and efficient detection of the coil of a high-voltage transformer winding machine.

[0004] In order to solve the above technical problems, the present invention provides a high-voltage transformer winding machine coil integrity detection device, comprising: An image acquisition module, used for acquiring a surface image of the wound coil; Depth detection module, used to obtain the depth information of the wound coil and generate three-dimensional point cloud data; Image processing module, used for preprocessing surface images and 3D point cloud data; The size and depth analysis module is used to determine whether the size and depth of the wound coil meet the preset standards based on the pre-processed surface image and 3D point cloud data; The integrity detection module is used to determine whether the wound coil has defects based on the pre-processed surface image and 3D point cloud data; A report generation module is used to generate a test report based on the judgment results of the size and depth analysis module and the integrity detection module; The control and display module is used to input control information of the device and display the detection report.

[0005] According to the above scheme, the image acquisition module includes a high-resolution industrial camera and a ring-shaped LED light source.

[0006] According to the above solution, the depth detection module adopts structured light or laser scanning technology.

[0007] According to the above scheme, the preprocessing of the image processing module includes denoising, enhancement, and registration.

[0008] According to the above scheme, the integrity detection module adopts image feature extraction and defect recognition algorithms.

[0009] According to the above scheme, the inspection report generated by the report generation module includes size, depth, defect location and type.

[0010] The present invention also provides a method for detecting the integrity of a high-voltage transformer winding machine coil, comprising: S1, obtaining a surface image of the wound coil; S2, obtaining depth information of the wound coil and generating three-dimensional point cloud data; S3, preprocessing the surface image and 3D point cloud data; S4. Judging whether the size and depth of the wound coil meet preset standards based on the pre-processed surface image and three-dimensional point cloud data; S5. Judging whether the wound coil has defects based on the pre-processed surface image and three-dimensional point cloud data; S6. Generate a test report based on the judgment results of the size and depth analysis module and the integrity detection module, and display the test report.

[0011] According to the above solution, structured light or laser scanning technology is used in step S2.

[0012] According to the above scheme, the preprocessing in step S3 includes denoising, enhancement, and registration.

[0013] According to the above solution, image feature extraction and defect recognition algorithms are used in step S5.

[0014] Beneficial effects The present invention collects surface images and depth information of wound coils, and performs size and depth analysis on the surface images and depth information, thereby being able to achieve accurate measurement of the size and depth of each section of the wound coils; the device realizes full automation from image acquisition to test report generation, reduces manual intervention, significantly improves detection efficiency, and realizes real-time display of test reports, which is beneficial to quality monitoring during the production process; the device also realizes detection of the integrity of the wound coils by analyzing the surface images and depth information of the wound coils, thereby ensuring the quality of the wound coils; the device is also suitable for detecting wound coils of different types and specifications, and has a certain degree of versatility. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a schematic structural diagram of a coil integrity detection device for a high-voltage transformer winding machine according to a first embodiment of the present invention; Figure 2 This is a flow chart of a method for detecting coil integrity of a high-voltage transformer winding machine according to a second embodiment of the present invention.

[0016] In the figure: 1-winding coil, 2-image acquisition module, 3-depth detection module, 4-image processing module, 5-size and depth analysis module, 6-integrity detection module, 7-report generation module, 8-control and display module, 9-industrial camera, 10-ring LED light source, 11-laser ranging sensor. DETAILED DESCRIPTION

[0017] To make the purpose, technical solutions, and advantages of the embodiments of the present disclosure more clear, the technical solutions of the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present disclosure. Obviously, the described embodiments are part of the embodiments of the present disclosure, not all of the embodiments. Based on the described embodiments of the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present disclosure.

[0018] Traditional manual coil inspection has problems such as low efficiency, insufficient accuracy, subjective judgment bias, and contact contamination risk. It is particularly difficult to identify hidden defects such as coil breakage and loose casting. During the manual inspection process, foreign matter adhesion can easily cause coil short circuit hazards. In addition, there is a lack of standardized data recording and defect location capabilities, which seriously affects the quality control and safe operation of high-voltage transformers. Based on the above situation, see Figure 1 This embodiment discloses a high-voltage transformer winding machine coil integrity detection device, comprising: An image acquisition module 2 is used to acquire a surface image of the wound coil 1; Depth detection module 3, used to obtain depth information of the wound coil 1 and generate three-dimensional point cloud data; Image processing module 4, used for preprocessing surface images and three-dimensional point cloud data; The size and depth analysis module 5 is used to determine whether the size and depth of the wound coil 1 meet the preset standards based on the pre-processed surface image and three-dimensional point cloud data; The integrity detection module 6 is used to determine whether the wound coil 1 has defects (defects include cracks, broken wires, loose wires, etc.) based on the pre-processed surface image and three-dimensional point cloud data; A report generation module 7 is used to generate a test report based on the judgment results of the size and depth analysis module 5 and the integrity detection module 6; The control and display module 8 is used to input control information of the device and display the detection report.

[0019] Furthermore, the image acquisition module 2 includes a high-resolution industrial camera 9 (a 12-megapixel industrial camera 9 in this embodiment) and a ring-shaped LED light source 10; the image acquisition module 2 of this embodiment can achieve ±0.05mm sub-pixel surface image capture, eliminating blind spots in manual visual inspection.

[0020] Furthermore, the depth detection module 3 adopts structured light or laser scanning technology; in this embodiment, the depth detection module 3 includes a laser ranging sensor 11.

[0021] Furthermore, the preprocessing of the image processing module 4 includes denoising, enhancement, and registration.

[0022] Furthermore, the size and depth analysis module 5 adopts an improved YOLOv5 defect detection algorithm, combined with a temporal convolutional network to achieve synchronous measurement of segment width and segment spacing, accurately measures the size and depth of each segment of the wound coil 1, and determines whether it meets the preset standards; in other embodiments of the present invention, a channel-spatial attention module is introduced into the YOLOv5 backbone network to improve the detection sensitivity of small-size defects (such as cracks and broken wires) by dynamically adjusting the feature channels and spatial weights; by introducing the channel-spatial attention mechanism (CBAM) and the hybrid feature pyramid network (Hybrid-FPN), the detection sensitivity of small-size defects (such as cracks and broken wires) is greatly enhanced.

[0023] Furthermore, the integrity detection module 6 adopts image feature extraction and defect recognition algorithms.

[0024] Furthermore, the inspection report generated by the report generation module 7 includes the size, depth, defect location, and type. In this embodiment, based on the inspection results, the report generation module 7 inputs parameters such as defect type (crack, discharge, etc.), location (three-dimensional coordinate error ≤ 0.1mm), and size (depth, length) into the fault model library, automatically matching the defect level (mild / moderate / severe) and repair suggestions. The report is then generated in PDF / Excel format, including a defect heat map and historical data comparison curves. Cloud storage and encrypted transmission are supported to meet data security requirements.

[0025] Furthermore, in this embodiment, the control and display module 8 realizes high-precision motion control (XYZ axis positioning accuracy ±0.02mm), synchronously triggers sensor acquisition (such as industrial camera 9), supports multi-station parallel detection, and adapts to the rhythm of industrial production lines; it is equipped with an industrial touch screen to display detection parameters and defect three-dimensional positioning maps in real time.

[0026] In other embodiments of the present invention, the device can also add integrated infrared thermal imaging sensors and acoustic vibration analysis modules to construct a multi-dimensional detection system, including infrared detection: based on the GIS internal overheating defect assessment method, real-time monitoring of abnormal coil temperature rise areas (temperature difference threshold > 5°C) and positioning of local breakdown points of the insulation layer; vibration analysis: using FFT spectrum analysis technology to identify abnormal vibration signals in the 20-200Hz frequency band and detect coil looseness or mechanical deformation; by adding integrated infrared thermal imaging and acoustic vibration analysis modules, additional detection of abnormal coil temperature rise and mechanical looseness problems is added.

[0027] Example 2: The principles of this embodiment are basically the same as those of the first embodiment. Based on the first embodiment, this embodiment discloses a method for detecting the integrity of a coil of a high-voltage transformer winding machine. Figure 2 ,include: S1, obtaining a surface image of the wound coil 1; S2, obtaining depth information of the wound coil 1 and generating three-dimensional point cloud data; S3, preprocessing the surface image and 3D point cloud data; S4. Determine whether the size and depth of the wound coil 1 meet preset standards based on the pre-processed surface image and three-dimensional point cloud data; S5. Determine whether the wound coil 1 has defects based on the preprocessed surface image and the three-dimensional point cloud data; S6. Generate a test report based on the judgment results of the size and depth analysis module 5 and the integrity detection module 6, and display the test report.

[0028] Furthermore, structured light or laser scanning technology is used in step S2.

[0029] Furthermore, the preprocessing in step S3 includes denoising, enhancement, and registration.

[0030] Furthermore, in step S5, an image feature extraction and defect recognition algorithm is used.

[0031] For ease of understanding, the specific process and principle of the method described in this embodiment are described. The method described in this embodiment includes: 1. Device Installation and Calibration: The detection device is integrated into the output of the high-voltage transformer winding machine. A high-precision XYZ three-axis slide adjusts the coverage of the image acquisition module 2 and depth detection module 3, ensuring coordinated focus between the industrial camera 9 (12-megapixel resolution) and the structured light scanner (±0.05mm accuracy). Dual-optical path self-calibration technology completes initial parameter calibration, eliminating mechanical installation errors. The device is compatible with coil diameters ranging from 50-500mm, supporting rapid deployment on transformer winding machines of varying specifications. The device is installed at the output of the high-voltage transformer winding machine, ensuring that the image acquisition module 2 and depth detection module 3 cover the entire coil.

[0032] 2. Simultaneous Multimodal Data Acquisition: After the device is powered on, the image acquisition module 2 captures the coil surface at a frame rate of 50 fps (the brightness of the ring-shaped LED light source 10 is adjustable to suppress metal reflections). Simultaneously, the depth detection module 3 generates 3D point cloud data using line laser scanning (sampling density 200 points / mm²). (An integrated acoustic vibration sensor (frequency response 20-200Hz) and an infrared thermal imager (thermal sensitivity 0.05°C) can be added.) These modules collect the mechanical vibration spectrum and temperature rise distribution during the coil winding process in real time to construct multimodal detection data.

[0033] 3. Intelligent Data Preprocessing: Image Processing Module 4 uses algorithms to eliminate image noise caused by electromagnetic interference and combines histogram equalization with multi-scale Retinex enhancement technology to enhance the contrast of surface defects. Depth data is aligned with the 3D model using the ICP point cloud registration algorithm to eliminate motion blur and deformation errors.

[0034] 4. Collaborative Size and Defect Analysis: The Size and Depth Analysis Module 5 uses an improved YOLOv5 algorithm, combined with a temporal convolutional network (TCN), to analyze the continuity characteristics of the coil winding process. By simultaneously measuring segment width and spacing, it identifies hidden defects such as broken wires and loose casting. It accurately measures the size and depth of each coil segment and determines whether it meets preset standards.

[0035] 5. Multimodal defect fusion determination: Integrity detection module 6 uses a decision-level fusion strategy, including visual defects: using an improved YOLOv5 to identify surface cracks and burrs (confidence threshold 0.85); depth anomalies: detecting coil looseness through point cloud curvature analysis (curvature threshold > 0.3mm⁻¹); temperature rise anomalies: using an infrared thermal imager to locate local overheating areas (temperature difference threshold > 5°C); vibration anomalies: identifying mechanical looseness through FFT spectrum analysis to determine whether the detection coil has defects.

[0036] 6. Intelligent report generation and traceability. The report generation module 7 statistically analyzes the process, automatically correlating defect type (cracks / breaks / looseness), location (three-dimensional coordinate error ≤ 0.1mm), and process parameters (winding tension, temperature). The generated PDF report includes: a defect heat map that indicates defect distribution density and level (red / yellow / green warning colors); a historical comparison curve that summarizes defect trends for coils within the same batch; and a repair suggestion library that matches standard recommended treatment solutions (such as partial repainting and rewinding). Data is transmitted using AES-256 encryption, supports blockchain-based evidence storage and traceability, and is output through the display module.

[0037] 7. Human-machine collaborative control and visualization The control and display module 8 utilizes an industrial-grade touchscreen, providing real-time displays of: a 3D location map with overlaid defect coordinates and winding paths (supporting AR-assisted maintenance perspective switching); and a real-time parameter panel for monitoring sensor status (such as camera temperature and light source brightness). Operators can zoom in and out of defect areas using gestures, and combined with audio and visual alarms (buzzer + LED indicator), quickly locate defective coils to ensure product quality.

[0038] The application of the present invention has at least the following beneficial effects: 1) High-precision detection: Through 3D image depth detection methods, micron-level measurement of coil segment width, segment spacing, and insulation layer thickness is achieved, and defects such as cracks and broken wires are accurately identified to ensure the accuracy of the test results.

[0039] 2) Full-process automation: Integrating multi-axis collaborative motion control and intelligent defect classification algorithms, the entire process from image acquisition and defect analysis to report generation requires no human intervention, shortening the time of a single inspection, improving efficiency, and avoiding safety hazards such as electric shock.

[0040] 3) Comprehensiveness: It not only supports dimensional parameter measurement, but also can identify potential defects such as inter-turn short circuits and abnormal temperature rise in coils, ensuring coil quality.

[0041] 4) Real-time dynamic monitoring: Real-time processing and visual display of detection data are achieved, supporting closed-loop control of production line quality and improving fault warning response speed.

[0042] 5) High scalability and compatibility: Through modular mechanical interfaces and a configurable algorithm library, it is adaptable to high-voltage transformer coil detection of different types and specifications, and has strong versatility.

[0043] It should be pointed out that, according to the needs of implementation, the various steps / components described in this application can be split into more steps / components, or two or more steps / components or partial operations of steps / components can be combined into new steps / components to achieve the purpose of the present invention.

[0044] It will be easily understood by those skilled in the art that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A high-voltage transformer winding machine coil integrity detection device, characterized in that: include: An image acquisition module, used for acquiring a surface image of the wound coil; Depth detection module, used to obtain the depth information of the wound coil and generate three-dimensional point cloud data; Image processing module, used for preprocessing surface images and 3D point cloud data; The size and depth analysis module is used to determine whether the size and depth of the wound coil meet the preset standards based on the pre-processed surface image and 3D point cloud data; The integrity detection module is used to determine whether the wound coil has defects based on the pre-processed surface image and 3D point cloud data; A report generation module is used to generate a test report based on the judgment results of the size and depth analysis module and the integrity detection module; The control and display module is used to input control information of the device and display the detection report.

2. The high-voltage transformer winding machine coil integrity detection device according to claim 1, characterized in that: The image acquisition module includes a high-resolution industrial camera and a ring LED light source.

3. The high-voltage transformer winding machine coil integrity detection device according to claim 1, characterized in that: The depth detection module uses structured light or laser scanning technology.

4. The high-voltage transformer winding machine coil integrity detection device according to claim 1, characterized in that: The preprocessing of the image processing module includes denoising, enhancement, and registration.

5. The high-voltage transformer winding machine coil integrity detection device according to claim 1, characterized in that: The integrity detection module uses image feature extraction and defect recognition algorithms.

6. The high-voltage transformer winding machine coil integrity detection device according to claim 1, characterized in that: The inspection report generated by the report generation module includes size, depth, defect location and type.

7. A method for detecting the integrity of a high-voltage transformer winding machine coil, characterized in that: include: S1, obtaining a surface image of the wound coil; S2, obtaining depth information of the wound coil and generating three-dimensional point cloud data; S3, preprocessing the surface image and 3D point cloud data; S4. Judging whether the size and depth of the wound coil meet preset standards based on the pre-processed surface image and three-dimensional point cloud data; S5. Judging whether the wound coil has defects based on the pre-processed surface image and three-dimensional point cloud data; S6. Generate a test report based on the judgment results of the size and depth analysis module and the integrity detection module, and display the test report.

8. The method for detecting the integrity of a high-voltage transformer winding machine coil according to claim 1, wherein: In step S2, structured light or laser scanning technology is used.

9. The method for detecting the integrity of a high-voltage transformer winding machine coil according to claim 1, wherein: The preprocessing in step S3 includes denoising, enhancement, and registration.

10. The method for detecting the integrity of a high-voltage transformer winding machine coil according to claim 1, wherein: In step S5, an image feature extraction and defect recognition algorithm is used.

Images