A power equipment quality detection method and system based on X-ray backscattering imaging technology

Abstract

The application provides a power equipment quality detection method and system based on X-ray backscattering imaging technology, which comprises a desktop batch detection device and a backscattering X-ray detection device carried by a unmanned aerial vehicle, and both can implement X-ray detection; the desktop batch detection device comprises an X-ray source module, a multistage adjustable collimator module, a semiconductor detector array, a touch screen, an equipment rotating object table, a translation table, a data acquisition and processing module; the backscattering X-ray detection device carried by the unmanned aerial vehicle comprises a translation table, an anti-shake holder, a quick-mounting and quick-dismounting assembly, a camera and a laser centering device, a backscattering X-ray source and a detector, an airborne micro data processor and a wireless data transmitter. Through equipment pretreatment, scanning parameter setting, multi-angle scanning imaging, image reconstruction and processing, defect identification and analysis, data analysis and report generation, nondestructive detection of power in-service and newly-arrived composite insulators and other equipment is realized.

Description

Technical Field

[0001] This invention relates to the field of non-destructive testing technology for power equipment, and in particular to a method and system for quality testing of power equipment based on X-ray backscatter imaging technology. Background Technology

[0002] With the increasing use of electricity, the internal components are prone to hidden defects such as insulation aging, metal cracks, and internal bubbles due to the combined effects of electrical stress, thermal stress, mechanical stress, and complex environments. If these defects are not detected in time, they can easily lead to electrical safety accidents, causing significant economic losses and adverse social impacts.

[0003] Existing power equipment testing methods are mainly divided into four categories: manual inspection, electrical performance testing, infrared thermography, and traditional X-ray transmission imaging, each with obvious shortcomings: 1. Manual inspection: Relying on personnel's eyesight and experience, it can only identify surface faults of equipment, cannot detect hidden internal defects, and is inefficient and highly subjective. 2. Electrical tests (insulation resistance, withstand voltage tests, etc.): can detect the electrical performance of equipment to a certain extent, but it is difficult to accurately obtain information such as the location, shape and size of specific defects inside the equipment; 3. Infrared thermometry: It can only detect abnormal temperature rise on the surface of the equipment and is not capable of detecting deep internal defects in the equipment. 4. Traditional X-ray transmission imaging: requires the placement of X-ray sources and detectors on both sides of the equipment, which is extremely difficult to deploy on-site for large enclosed power equipment.

[0004] X-ray backscattering imaging eliminates the need for detectors on the back of equipment and is highly sensitive to low atomic number insulating materials, making it suitable for in-situ non-destructive testing of power equipment. However, the industry currently lacks a complete backscattered X-ray inspection system and standardized inspection and analysis process specifically adapted to the operating conditions of power equipment. Dedicated image reconstruction and defect recognition algorithms are also lacking, failing to meet the high-precision, intelligent, and rapid online inspection requirements of power equipment.

[0005] Therefore, there is an urgent need for an integrated backscatter X-ray inspection system and supporting inspection methods that can be adapted to offline batch inspection and airborne inspection of in-operation equipment, in order to solve the problems of weak internal defect detection capability, low inspection efficiency, insufficient quantification of defect information, and the need for power outage inspection in existing technologies. Summary of the Invention

[0006] This invention aims to overcome the shortcomings of existing power equipment inspection technologies, such as insufficient ability to detect hidden defects, low inspection efficiency, inability to quantify defect parameters, the need for power outages to inspect in-operation equipment, and the risk of secondary damage to equipment. It provides a power equipment quality inspection method and system based on X-ray backscatter imaging technology.

[0007] The technical solution of this invention is as follows: A power equipment inspection system based on X-ray backscatter imaging technology, comprising a desktop batch inspection device and a UAV-mounted backscatter X-ray inspection device; both the desktop batch inspection device and the UAV-mounted backscatter X-ray inspection device can perform X-ray detection. The desktop batch inspection device includes an X-ray source module, a multi-level adjustable collimator module, a semiconductor detector array, a touch screen, a rotating stage, a translation stage, and a data acquisition and processing module; the UAV-mounted backscatter X-ray inspection device includes a translation stage, a gimbal stabilization unit, a quick-release assembly, a camera and laser alignment device, a backscatter X-ray source and detector, an airborne micro data processor, and a wireless data transmitter. The X-ray source module is supported and stabilized by a bracket, and is used to adjust the accelerating voltage and output current to control the X-ray dose and energy. The multi-stage adjustable collimator module is also supported and stabilized by a bracket, and includes a first-stage large-aperture collimator and a second-stage slit adjustable collimator, which are used to constrain the ray beam and optimize the backscattered photon receiving angle. The semiconductor detector array is fixed above the touch screen and is used to receive X-ray backscattered signals and convert them into electrical signals; the touch screen and the data acquisition and processing module are integrated together and placed on the mounting base; The rotating platform of the device is on a translation platform and moves with the translation platform. The translation platform is placed on the ground. The rotating platform of the device has a rotation function. The device to be tested is placed on the rotating platform of the device and completes multi-angle scanning as the rotating platform of the device rotates. The data acquisition and processing module includes a high-speed data acquisition card, an image reconstruction subunit, a defect intelligent identification subunit, a comprehensive data analysis subunit, and a report generation subunit. The anti-shake gimbal is fixed to the bottom of the drone. The quick-release assembly is fixed to the anti-shake gimbal and is used to install the housing. The housing contains a camera and laser alignment device, a backscatter X-ray source and detector, an airborne micro data processor, and a wireless data transmitter. The camera and laser alignment device are used to acquire images of high-altitude power equipment and complete positioning and alignment with the laser. The backscatter X-ray source and detector are used to emit X-rays, receive X-ray backscatter signals, and convert them into electrical signals. The airborne micro data processor packages the X-ray backscatter signal data and image data into data packets, which are transmitted to the wireless data transmitter. The wireless data transmitter then sends the data packets to the data acquisition and processing module.

[0008] The image reconstruction subunit integrates the filtering back projection algorithm and the iterative reconstruction algorithm, and combines the differential scattering and attenuation model of the metal and insulating materials of the power equipment to reconstruct two-dimensional and three-dimensional internal images. It also configures histogram equalization, median filtering and edge enhancement algorithms to complete image noise reduction and defect contour enhancement.

[0009] The intelligent defect identification subunit is equipped with a deep learning model trained on a power equipment defect sample database. It automatically identifies defects such as insulation bubbles, internal cracks, and insulation aging, outputs quantitative parameters of defect location, length, area, and depth, and classifies the defect risk level.

[0010] The integrated data analysis subunit establishes a correlation model between equipment operating parameters and defect image data, assesses the risk of defect thermal failure, and conducts trend analysis on multi-cycle detection data to predict the defect development rate and remaining equipment lifespan.

[0011] The built-in lithium battery of the UAV-mounted backscatter X-ray detection device has a continuous operating time of no less than 4 hours.

[0012] A method for quality inspection of power equipment based on X-ray backscatter imaging technology, implemented using the aforementioned inspection system, includes the following steps: S1: Equipment pretreatment: Clean the surface of the electrical equipment to be inspected of dust and impurities, and record the basic information of the equipment and real-time operating voltage, current and temperature parameters; S2: Scanning parameter settings: Configure the scanning parameters of the X-ray source, detector, collimator, and motion control mechanism according to the material, thickness, and shape of the equipment; S3: Multi-angle scanning imaging: The desktop batch inspection device scans in conjunction with the translation stage and the rotating platform; the UAV-mounted backscatter X-ray inspection device hovers in the air to collect X-ray backscatter signals from multiple angles and transmits electrical signals to the data acquisition module in real time. S4: Image Reconstruction and Processing: The X-ray source and detector move relative to each other around the power equipment, scanning the power equipment from multiple angles. Two-dimensional / three-dimensional images of the equipment's interior are generated through image reconstruction algorithms, and image enhancement algorithms are used to remove noise and highlight defects. S5: Defect identification and classification: The deep learning model compares the pre-trained deep learning sample database to automatically identify the defect type and quantify the size parameters, and classifies the risk level according to the defect location and size. S6: Comprehensive Data Analysis: Establish a data association model, associate defect data with equipment operating parameters to assess failure risk, compare historical inspection data to analyze defect development trends, and predict the remaining lifespan of the equipment; S7: Report Generation: Automatically generates standardized inspection reports, which include equipment information, defect details, risk level and graded maintenance recommendations, and emergency handling plans for major defects.

[0013] In step S3, the UAV-mounted backscatter X-ray detection device can complete non-contact, non-destructive scanning without interrupting power to the equipment, and without needing to disconnect the equipment or contact the live parts.

[0014] The image reconstruction algorithms in step S4 include the filtered back projection algorithm and the iterative reconstruction algorithm.

[0015] In step S5, the pre-trained deep learning sample database contains multiple image samples of power equipment in normal condition, bubbles, cracks, and insulation aging. The deep learning model continuously iterates and optimizes the recognition accuracy as new detection samples are added.

[0016] The dimensional parameters in step S5 include the area, length, and depth of the defect.

[0017] The beneficial effects of this invention are: 1. High sensitivity detection of minute insulation defects: X-ray backscatter imaging is highly sensitive to low atomic number insulating materials and can detect microbubbles, fine cracks, and deep aging defects in the insulation layer that cannot be identified by traditional detection methods, thus greatly improving the accuracy of defect detection.

[0018] 2. Non-contact, uninterrupted, non-destructive testing: There is no need to disassemble the equipment or contact the internal live parts, avoiding secondary damage during inspection; in-operation equipment can be inspected online using UAV-borne devices, without the need for power outages for maintenance, ensuring continuous power supply to the grid and reducing economic losses caused by power outages.

[0019] 3. Accurately locate defects from all angles and perspectives: The motion control mechanism enables 360° layered scanning, multi-dimensional reconstruction of the equipment's internal structure, precise location of the three-dimensional coordinates of defects, and quantification of defect dimensions, providing a complete and intuitive basis for maintenance.

[0020] 4. Detection efficiency is greatly improved: High-speed acquisition cards acquire signals in parallel, and dedicated reconstruction and recognition algorithms shorten image processing time. Compared with manual inspection and traditional withstand voltage tests, the detection time of a single device is significantly reduced, making it suitable for batch equipment inspection in substations.

[0021] 5. Intelligent automatic recognition eliminates human error: Deep learning models automatically identify various defects, avoiding subjective biases in human judgment; the database is continuously updated with samples, and the model's recognition accuracy continues to improve with the amount of detection, realizing intelligent detection throughout the entire process.

[0022] 6. Full lifecycle data supports operational and maintenance decisions: Multi-cycle detection data trend analysis can predict the rate of defect expansion and assess the remaining lifespan of equipment; data association models quantify the risk of defect thermal failure and insulation breakdown, intelligently assess the severity of defects and output maintenance suggestions, improve the detection accuracy and maintenance efficiency of power equipment, assist power grid companies in formulating planned maintenance and replacement schemes, and reduce overall maintenance costs. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the overall structure of the desktop batch testing device of the present invention; Figure 2 This is a schematic diagram of the overall structure of the UAV-mounted backscatter X-ray detection device of the present invention.

[0024] Reference numerals: 1-X-ray source module; 2-Multi-level adjustable collimator module; 3-Semiconductor detector array; 4-Touch screen; 5-Equipment rotating stage; 6-Translation stage; 7-Anti-shake gimbal; 8-Quick-install and quick-release assembly; 9-Camera and laser alignment device; 10-Backscattered X-ray source and detector; 11-Airborne micro data processor; 12-Wireless data transmitter. Detailed Implementation

[0025] A power equipment inspection system based on X-ray backscatter imaging technology The system includes two types of detection devices: a desktop batch detection device and a UAV-mounted backscattered X-ray detection device. Both types of devices can perform X-ray detection. The desktop batch detection device includes an X-ray source module (1), a multi-level adjustable collimator module (2), a semiconductor detector array (3), a touch screen (4), a rotating stage (5), a translation stage (6), and a data acquisition and processing module. The UAV-mounted backscattered X-ray detection device includes a translation stage (6), a gimbal (7), a quick-release assembly (8), a camera and laser alignment device (9), a backscattered X-ray source and detector (10), an airborne micro data processor (11), and a wireless data transmitter (12).

[0026] X-ray source module 1: Stable and supported by a bracket, it employs a pulsed X-ray source, allowing for precise control of X-ray emission dose and pulse time interval; the radiation energy and intensity are dynamically adjusted by regulating the accelerating voltage and output current. For electrical equipment with thick casings and complex structures, increase X-ray energy to meet the requirements of deep defect penetration imaging; For thin-walled insulating and radiation-sensitive components, reduce radiation dose and minimize radiation damage.

[0027] Multi-stage adjustable collimator module 2: Stable and supported by a bracket, with a two-stage adjustable constraint structure: Large-aperture collimator: constrains the divergence angle of the X-ray beam, ensuring that the rays uniformly cover the surface to be inspected. Secondary slit adjustable collimator: The slit width and deflection angle can be freely adjusted to optimize the backscattered photon receiving angle and improve imaging contrast and detail resolution; the parameters can be flexibly adjusted according to the shape and size of the equipment.

[0028] Semiconductor detector array 3: Fixed above the touchscreen 4, it adopts a high-sensitivity semiconductor array, which has high spatial resolution and high count rate; the size optimization design takes into account both imaging resolution and detection coverage, and reduces detection blind spots; it can collect backscattered X-ray photons in real time and convert them into standard electrical signal output.

[0029] Equipment rotating platform 5, translation platform 6: The system comprises a tabletop motion control mechanism. The rotating platform 5 of the equipment rests on the translation stage 6, which moves with the equipment / inspection head, achieving precise horizontal displacement with a movement accuracy down to the sub-millimeter level. The rotating platform 5 enables multi-angle circumferential scanning; through linkage control, it completes 360° all-round layered scanning of the equipment, collecting multi-dimensional internal structural data; and the accompanying touch screen 4 allows for manual parameter adjustment and scan start / stop control.

[0030] Data acquisition and processing module: Integrated with the touch screen 4 in a single chassis and placed on a stand, it includes a high-speed data acquisition card, an image reconstruction subunit, a defect intelligent identification subunit, a comprehensive data analysis subunit, and a report generation subunit; Includes a high-speed acquisition card, an image reconstruction subunit, an intelligent defect identification subunit, a comprehensive data analysis subunit, and a report generation subunit. High-speed data acquisition card: Simultaneously acquires electrical signals from multiple detectors and converts them into digital signals in real time for storage. It enables rapid acquisition of large amounts of data, ensuring that no important information is lost during the detection process. Image reconstruction subunit: It integrates filtered backprojection and iterative reconstruction algorithms, combined with differential scattering / attenuation models for metals and insulating materials, to reconstruct two-dimensional and three-dimensional images of the device's interior; it also includes histogram equalization, median filtering, and edge enhancement algorithms to eliminate image noise and enhance defect contours. The intelligent defect identification subunit is equipped with a deep learning model trained on a power equipment defect sample database, which contains a large number of normal / defective power equipment samples. It automatically identifies defects such as bubbles, cracks, and insulation aging, and outputs quantitative parameters of defect location, length, area, and depth to classify defect risk levels. The comprehensive data analysis subunit: correlates equipment operating voltage, current, temperature and other parameters with defect image data, establishes a correlation model to assess the risk of defect thermal failure; performs trend analysis on multi-cycle detection data to predict the defect development rate and the remaining life of the equipment; Report generation sub-unit: Automatically integrates basic equipment information, testing parameters, defect details, risk level, and maintenance and rectification suggestions to generate a standardized testing report.

[0031] Anti-shake gimbal 7: Fixed to the bottom of the drone, it eliminates flight shaking and ensures stable imaging; Quick-install and quick-release component 8: Fixed on the anti-shake gimbal 7, enabling quick installation and removal of the housing; 9. Camera and laser alignment device; 10. Backscattered X-ray source and detector; 11. Airborne micro data processor; 12. Wireless data transmitter. Installed in the housing, the built-in lithium battery supports up to 4 hours of continuous outdoor inspection work. The camera and laser alignment device 9 are used to acquire images of high-altitude power equipment and, in conjunction with the laser, locate and align the equipment to be inspected. The onboard micro data processor 11 packages the X-ray backscatter signal data and image data into data packets and transmits them back to the ground-based data acquisition and processing module.

[0032] A method for quality inspection of power equipment based on X-ray backscatter imaging technology includes the following steps: Step S1: Equipment Pre-processing Clean the surface of the electrical equipment to be inspected of dust and oil to eliminate scattering interference caused by impurities; record the basic parameters (model, location) and real-time operating parameters (voltage, current, surface temperature) of the equipment as a benchmark for subsequent defect risk assessment.

[0033] Step S2: Adaptive configuration of scan parameters Based on the equipment material, thickness, shape, and estimated defect type, set the complete set of scanning parameters: X-ray source pulse frequency, emission dose, and acceleration voltage; detector integration time and gain; collimator slit width and angle; tabletop motion control mechanism moving speed and single rotation angle.

[0034] Step S3: Multi-angle layered scanning imaging Desktop device: The rotating stage 5 and the translation stage 6 of the equipment are linked together, and the X-ray source module 1 and the semiconductor detector array 3 perform multi-angle motion scanning of the device under test. Unmanned aerial vehicle (UAV) device: The UAV carries the detection mechanism and hovers outside the equipment. The anti-shake gimbal 7 adjusts the detection angle without power outages or contact with the equipment. The backscattered X-ray source and detector 10 emits X-rays that penetrate the surface of the equipment and are scattered by the internal medium. The backscattered photons are collected by the detector and converted into electrical signals, which are then transmitted to the data acquisition and processing module in real time. Through multi-angle scanning, information from different layers and directions inside the equipment can be obtained, providing richer data for subsequent image reconstruction and defect identification.

[0035] Step S4: Image Reconstruction and Enhancement Processing Image reconstruction algorithms (such as filtering back projection algorithms and iterative reconstruction algorithms) are used to generate two-dimensional / three-dimensional internal structure images; image distortion is corrected by combining the scattering model of metal and insulating materials of power equipment; noise is removed by filtering, equalization and edge enhancement algorithms to highlight the internal structural features and possible defects of the equipment.

[0036] Step S5: Deep Learning Defect Identification and Grading Assessment The processed images are input into a pre-trained deep learning model, which compares them against a pre-trained deep learning sample database to automatically identify defect types, locate defect coordinates, and quantify dimensional parameters. Based on critical structural thresholds of the equipment, the model categorizes defect severity into minor, general, and critical / emergency defects. Depending on the severity level, subsequent handling methods are determined, including repair or replacement. The pre-trained deep learning sample database contains a large number of images of normal and defective power equipment (such as cracks, bubbles, and insulation aging) used for deep learning training.

[0037] Step S6: Multi-dimensional comprehensive data analysis By binding scanned image data, defect identification results, and equipment operating parameters to model and analyze the associated risks of abnormal temperature and electrical performance in defect areas, and comparing historical inspection data, defect development trend curves are plotted to predict equipment health information. For example, correlation analysis is performed between the internal temperature distribution of the equipment and the location of defects. If abnormal temperature increases are found near the defect area, the potential thermal failure risk caused by the defect is further assessed. Simultaneously, trend analysis methods are used to compare inspection data of the same equipment at different time points to observe defect development trends and predict the remaining lifespan of the equipment.

[0038] Step S7: Standardized test report automatically generated The report includes equipment file information (such as equipment name, model, manufacturer, installation location, etc.), inspection time, equipment operating parameters, scanning configuration, defect details (type / location / size / risk level), graded maintenance recommendations, and the signature of the inspection personnel; major defects are highlighted and emergency response plans are pushed out simultaneously.

[0039] Example 1: Benchtop batch testing device for testing newly purchased composite insulators Pre-treatment: Clean the dust off the surface of the insulator and record the insulator model and factory parameters; Parameter settings: For epoxy resin insulation materials, reduce the X-ray acceleration voltage to decrease the radiation dose and narrow the secondary collimation slit to improve the imaging resolution of the insulation layer; Scanning and imaging: The composite insulator is fixed to the rotating stage of the equipment, the translation stage is fed at a constant speed, and the rotating stage of the equipment pauses every 15° to collect images, thus completing a full circumference scan; Image reconstruction: Iterative reconstruction of the internal structure of the insulation in three dimensions, followed by median filtering to remove particulate noise; Defect identification: Deep learning models identify internal air bubbles and core rod cracks in insulation, and quantify the bubble diameter and crack length; Data analysis: Determine whether the defects exceed the factory acceptance threshold; Generate a quality inspection report, mark the substandard composite insulators, and provide suggestions for return or replacement.

[0040] Example 2: Online Inspection of High-Voltage Bushings Using an UAV-Mounted Backscattered X-ray Detection Device Pre-processing: Record the real-time operating current and surface infrared temperature of the high-voltage bushing without power outage; Parameter settings: Increase X-ray energy to match the metal sleeve shell, adjust the pan-tilt angle to align with the sleeve flange and insulation section; Aerial multi-angle scanning: The drone hovers on the outside of the casing, and the anti-shake gimbal adjusts the pitch and horizontal angle to collect X-ray backscatter signals in layers. The data is wirelessly transmitted back to the ground in real time. Image enhancement and reconstruction to extract the aging insulation area inside the high-voltage bushing; Defect classification; identifying internal insulation cracks in high-voltage bushings and classifying them as major defects; Trend analysis: Retrieving the inspection data of the same casing from the previous quarter confirms that the crack continues to expand; The report includes recommendations for emergency shutdown and maintenance, and is simultaneously sent to the maintenance team.

Claims

1. A power equipment inspection system based on X-ray backscatter imaging technology, characterized in that, The device includes a desktop batch inspection device and a UAV-mounted backscatter X-ray inspection device. Both the desktop batch inspection device and the UAV-mounted backscatter X-ray inspection device can perform X-ray detection. The desktop batch inspection device includes an X-ray source module (1), a multi-level adjustable collimator module (2), a semiconductor detector array (3), a touch screen (4), a rotating stage (5), a translation stage (6), and a data acquisition and processing module. The UAV-mounted backscatter X-ray inspection device includes a translation stage (6), a gimbal stabilizer (7), a quick-release assembly (8), a camera and laser alignment device (9), a backscatter X-ray source and detector (10), an airborne micro data processor (11), and a wireless data transmitter (12). The X-ray source module (1) is supported and stabilized by a bracket and is used to adjust the accelerating voltage and output current to control the X-ray dose and energy. The multi-stage adjustable collimator module (2) is also supported and stabilized by a bracket, and includes a first-stage large-aperture collimator and a second-stage slit adjustable collimator, which are used to constrain the ray beam and optimize the backscattered photon receiving angle. The semiconductor detector array (3) is fixed above the touch screen (4) and is used to receive X-ray backscattered signals and convert them into electrical signals; the touch screen (4) is integrated with the data acquisition and processing module and placed on the placement platform; The rotating platform (5) of the equipment is on the translation platform (6) and moves with the translation platform (6). The translation platform (6) is placed on the ground. The rotating platform (5) of the equipment has a rotation function. The equipment to be tested is placed on the rotating platform (5) of the equipment and rotates with the rotating platform (5) to complete multi-angle scanning. The data acquisition and processing module includes a high-speed data acquisition card, an image reconstruction subunit, a defect intelligent identification subunit, a comprehensive data analysis subunit, and a report generation subunit. The anti-shake gimbal (7) is fixed to the bottom of the UAV. The quick-release assembly (8) is fixed to the anti-shake gimbal (7). The quick-release assembly (8) is used to install the housing. The housing is equipped with a camera and laser alignment device (9), a backscatter X-ray source and detector (10), an airborne micro data processor (11), and a wireless data transmitter (12). The camera and laser alignment device (9) is used to acquire images of high-altitude power equipment and complete positioning and alignment with the laser. The backscatter X-ray source and detector (10) is used to emit X-rays, receive X-ray backscatter signals and convert them into electrical signals. The airborne micro data processor (11) packages the X-ray backscatter signal data and image data into data packets. The data packets are transmitted to the wireless data transmitter (12). The wireless data transmitter (12) sends the data packets to the data acquisition and processing module.

2. The power equipment inspection system based on X-ray backscatter imaging technology according to claim 1, characterized in that, The image reconstruction subunit integrates the filtering back projection algorithm and the iterative reconstruction algorithm, and combines the differential scattering and attenuation model of the metal and insulating materials of the power equipment to reconstruct two-dimensional and three-dimensional internal images. It also configures histogram equalization, median filtering and edge enhancement algorithms to complete image noise reduction and defect contour enhancement.

3. The power equipment inspection system based on X-ray backscatter imaging technology according to claim 1, characterized in that, The intelligent defect identification subunit is equipped with a deep learning model trained on a power equipment defect sample database. It automatically identifies defects such as insulation bubbles, internal cracks, and insulation aging, outputs quantitative parameters of defect location, length, area, and depth, and classifies the defect risk level.

4. The power equipment inspection system based on X-ray backscatter imaging technology according to claim 1, characterized in that, The integrated data analysis subunit establishes a correlation model between equipment operating parameters and defect image data, assesses the risk of defect thermal failure, and conducts trend analysis on multi-cycle detection data to predict the defect development rate and remaining equipment lifespan.

5. A power equipment inspection system based on X-ray backscatter imaging technology according to claim 1, characterized in that, The built-in lithium battery of the UAV-mounted backscatter X-ray detection device has a continuous operating time of no less than 4 hours.

6. A method for quality inspection of power equipment based on X-ray backscatter imaging technology, characterized in that, The detection system described in any one of claims 1-5 is implemented by comprising the following steps: S1: Equipment pretreatment: Clean the surface of the electrical equipment to be inspected of dust and impurities, and record the basic information of the equipment and real-time operating voltage, current and temperature parameters; S2: Scanning parameter settings: Configure the scanning parameters of the X-ray source, detector, collimator, and motion control mechanism according to the material, thickness, and shape of the equipment; S3: Multi-angle scanning imaging: The desktop batch inspection device scans in conjunction with the translation stage and the rotating platform; the UAV-mounted backscatter X-ray inspection device hovers in the air to collect X-ray backscatter signals from multiple angles and transmits electrical signals to the data acquisition module in real time. S4: Image Reconstruction and Processing: The X-ray source and detector move relative to each other around the power equipment, scanning the power equipment from multiple angles. Two-dimensional / three-dimensional images of the equipment's interior are generated through image reconstruction algorithms, and image enhancement algorithms are used to remove noise and highlight defects. S5: Defect identification and classification: The deep learning model compares the pre-trained deep learning sample database to automatically identify the defect type and quantify the size parameters, and classifies the risk level according to the defect location and size. S6: Comprehensive Data Analysis: Establish a data association model, associate defect data with equipment operating parameters to assess failure risk, compare historical inspection data to analyze defect development trends, and predict the remaining lifespan of the equipment; S7: Report Generation: Automatically generates standardized inspection reports, which include equipment information, defect details, risk level and graded maintenance recommendations, and emergency handling plans for major defects.

7. The method for quality inspection of power equipment based on X-ray backscatter imaging technology according to claim 6, characterized in that, In step S3, the UAV-mounted backscatter X-ray detection device can complete non-contact, non-destructive scanning without interrupting power to the equipment, and without needing to disconnect the equipment or contact the live parts.

8. The method for quality inspection of power equipment based on X-ray backscatter imaging technology according to claim 6, characterized in that, The image reconstruction algorithms in step S4 include the filtered back projection algorithm and the iterative reconstruction algorithm.

9. The method for quality inspection of power equipment based on X-ray backscatter imaging technology according to claim 6, characterized in that, In step S5, the pre-trained deep learning sample database contains multiple image samples of power equipment in normal condition, bubbles, cracks, and insulation aging. The deep learning model continuously iterates and optimizes the recognition accuracy as new detection samples are added.

10. The method for quality inspection of power equipment based on X-ray backscatter imaging technology according to claim 6, characterized in that, The dimensional parameters in step S5 include the area, length, and depth of the defect.

Images