An electrical equipment detection device based on machine vision

Abstract

This invention discloses a machine vision-based electrical equipment inspection device, specifically relating to the field of electrical equipment inspection technology. It includes a mobile platform, a lifting mechanism, a rotating gimbal, a multispectral imaging unit, an environmental adaptive lighting system, and a main control unit. The lifting mechanism is mounted above the mobile platform for adjusting the imaging height. The rotating gimbal is mounted on top of the lifting mechanism for adjusting the imaging angle. The multispectral imaging unit is mounted above the rotating gimbal for acquiring multispectral images of the equipment. This machine vision-based electrical equipment inspection device, by incorporating a mobile platform, a lifting mechanism, and a two-degree-of-freedom rotating gimbal, enables flexible multi-degree-of-freedom adjustment of the imaging unit in height, horizontal direction, and pitch, effectively expanding the inspection coverage of a single device, reducing system deployment costs, and adapting to the inspection needs of complex scenarios such as large substations and multi-bay distribution rooms.

Description

Technical Field

[0001] This invention relates to the field of electrical equipment testing technology, and in particular to an electrical equipment testing device based on machine vision. Background Technology

[0002] Electrical equipment plays a crucial role in the transmission, distribution, and control of electrical energy in power systems, and the stability of its operating status directly affects the safety and reliability of the entire power system. With the continuous expansion of power grid scale and the increasing complexity of equipment operating environments, traditional manual inspection methods are no longer sufficient to meet the needs for real-time, accurate, and efficient monitoring of the status of electrical equipment.

[0003] Currently, the inspection of electrical equipment mainly relies on the following methods: First, manual periodic inspections, where inspectors use handheld infrared thermal imagers, partial discharge detectors, and other equipment to conduct on-site checks. This method suffers from low inspection efficiency, high subjectivity, susceptibility to the influence of personnel experience on inspection results, and difficulty in achieving continuous monitoring. Second, online monitoring systems, which collect equipment operating parameters in real time by installing sensors, such as temperature sensors, current transformers, and vibration sensors, on key parts of the equipment. However, this type of method usually requires modification of the equipment or shutdown for installation, resulting in high construction difficulty, limited applicability, and high sensor deployment costs. Third, infrared thermal imaging-based detection methods, mainly used to identify abnormal temperature rises caused by poor contact, overload, etc., but this method is difficult to effectively identify visually identifiable faults such as equipment appearance defects, abnormal meter readings, and insulator damage.

[0004] In recent years, machine vision technology has been widely used in industrial inspection due to its advantages of being non-contact, high-precision, and highly automated. Some attempts have been made to apply machine vision to the inspection of electrical equipment, such as using fixed cameras to capture images of the equipment and then using image processing algorithms to identify surface defects or meter readings. However, existing solutions generally suffer from the following problems: First, existing visual inspection devices mostly adopt fixed installation methods, and the position and angle of the camera are difficult to adjust flexibly, which limits the detection range of a single device. For large substations or multi-bay power distribution rooms, a large number of cameras need to be deployed, making the system complex and costly.

[0005] Secondly, the on-site environment of electrical equipment is complex, with factors such as drastic changes in lighting, cluttered backgrounds, reflections, and dust obstruction. Existing image acquisition and processing systems lack targeted optical design and adaptive algorithms, resulting in unstable image quality and decreased detection accuracy.

[0006] Finally, most existing devices only perform a single function, such as infrared temperature measurement or appearance recognition, making it difficult to complete multiple detection tasks simultaneously and failing to meet the needs of comprehensive condition assessment of electrical equipment. Summary of the Invention

[0007] The main objective of this invention is to provide an electrical equipment inspection device based on machine vision, which can effectively solve the problems of limited detection range, complex system and high cost caused by the fixed installation method of existing vision inspection devices. It can also solve the problems of unstable image quality and reduced detection accuracy caused by drastic changes in on-site lighting, cluttered background and reflection interference. Furthermore, it can address the problem that existing devices have single functions and cannot simultaneously complete multiple tasks such as appearance recognition, infrared temperature measurement and partial discharge detection, thus failing to meet the needs of comprehensive condition assessment of electrical equipment.

[0008] To achieve the above objectives, the technical solution adopted by the present invention is as follows: An electrical equipment inspection device based on machine vision includes a mobile platform, a lifting mechanism, a rotating gimbal, a multispectral imaging unit, an environmental adaptive lighting system, and a main control unit. The lifting mechanism is mounted above the mobile platform for adjusting the imaging height. The rotating gimbal is mounted on top of the lifting mechanism for adjusting the imaging angle. The multispectral imaging unit is mounted above the rotating gimbal for acquiring multispectral images of the equipment. The environmental adaptive lighting system is mounted on one side of the rotating gimbal for automatically supplementing light according to ambient illumination. The main control unit is mounted on top of the mobile platform and located in front of the lifting mechanism, for image processing and system control. This structure enables multi-degree-of-freedom adjustment and multi-source data acquisition in the inspection device, improving inspection flexibility and information integrity.

[0009] Preferably, the lifting mechanism includes a drive column, a servo motor, a lifting threaded rod, a limiting slide rod, a lifting plate, a push plate, and a lifting platform. The bottom of the drive column has a motor slot, and the servo motor is fixedly installed inside the motor slot. The side wall of the drive column has a through-type lifting groove, and the output shaft of the servo motor passes through the inner cavity of the lifting groove. The lifting threaded rod is fixedly installed on the output shaft of the servo motor. The lifting plate is threaded onto the outer wall of the lifting threaded rod, and the limiting slide rod is fixedly installed between the upper and lower sides of the inner cavity of the lifting groove. The lifting plate and the limiting slide rod are slidably connected. The push plate is fixedly installed on both sides of the top of the lifting plate, and the top of the push plate extends through to the top of the drive column. The lifting platform is fixedly installed on the top of the push plate, and the rotating gimbal is installed on the top of the lifting platform. This structure enables precise height adjustment of the imaging unit, improving adaptability to devices of different heights.

[0010] Preferably, the rotating gimbal is a two-degree-of-freedom gimbal, including horizontal rotational degree of freedom and pitch rotational degree of freedom, driven by a servo motor. This configuration allows the imaging unit to rotate flexibly in both horizontal and pitch directions, effectively expanding the detection coverage area of ​​a single deployment.

[0011] Preferably, the multispectral imaging unit includes a visible light camera, an infrared thermal imager, and an ultraviolet imager. The visible light camera is a high-resolution industrial camera, and a polarizing filter is provided at the front end of the visible light camera. Through multispectral imaging and polarization filtering design, the appearance, temperature, and partial discharge information of the equipment can be acquired simultaneously, surface reflection interference can be suppressed, and image quality and detection accuracy can be improved.

[0012] Preferably, the environmental adaptive lighting system includes a multi-angle LED light source and a light sensor. The multi-angle LED light source is installed on the top of the rotating gimbal, and the light sensor is installed on the top of the lifting platform. This system automatically adjusts the brightness and angle of the light source according to real-time lighting conditions, effectively eliminating shadows and reflections and ensuring imaging stability.

[0013] Preferably, the main control unit includes a control box, an image processing module, and a wireless communication module. The control box has a hinged access door on its outer wall, and both the image processing module and the wireless communication module are located within the control box's interior. This structure facilitates system maintenance and remote data transmission, improving the maintainability and information technology level of the device.

[0014] Preferably, the mobile platform is also equipped with obstacle avoidance sensors and a positioning module. The positioning module is a lidar or UWB positioning module, and the main control unit is connected to the remote monitoring platform through the wireless communication module. This configuration enables the device to achieve autonomous navigation and precise positioning, improving the automation level of inspections.

[0015] Preferably, the machine vision-based electrical equipment inspection device further includes a drive wheel assembly, which is installed at the bottom of the mobile platform to enable the overall movement of the device.

[0016] Preferably, the drive wheel assembly includes a wheel assembly bracket, a drive motor, and omnidirectional wheels. The wheel assembly bracket is fixedly installed at the bottom of the mobile platform, and the omnidirectional wheels are disposed on both sides of the wheel assembly bracket. The omnidirectional wheels are drive-connected to the drive motor. Through the omnidirectional wheel structure, the device can flexibly turn and move in narrow or complex terrain, adapting to various field environments such as substations and power distribution rooms.

[0017] Preferably, the image processing module incorporates an image enhancement algorithm, a defect identification model, a meter reading identification model, and an infrared temperature analysis model. This module can perform integrated analysis of acquired multispectral images, enabling simultaneous identification of appearance defects, meter readings, and temperature anomalies, thereby improving detection efficiency and intelligence.

[0018] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention provides an electrical equipment inspection device based on machine vision. By setting up a mobile platform, a lifting mechanism and a two-degree-of-freedom rotating gimbal, the imaging unit can be flexibly adjusted in multiple degrees of freedom in the height, horizontal and pitch directions, effectively expanding the detection coverage of a single device, reducing system deployment costs, and adapting to the inspection needs of complex scenarios such as large substations and multi-bay power distribution rooms.

[0019] 2. This invention provides an electrical equipment inspection device based on machine vision. By setting up a multispectral imaging unit and an environmental adaptive supplementary lighting system, it integrates visible light, infrared and ultraviolet imaging functions, and combines a light sensor and a multi-angle LED light source to achieve automatic supplementary lighting adjustment, suppress reflection and shadow interference, and simultaneously acquire appearance, temperature and partial discharge information, thereby improving image quality and detection accuracy.

[0020] 3. This invention provides an electrical equipment inspection device based on machine vision. By integrating image enhancement algorithms, defect recognition models, meter reading recognition models, and infrared temperature analysis models into the main control unit, it performs integrated intelligent analysis of multispectral images to achieve simultaneous identification of appearance defects, meter abnormalities, and temperature faults. Combined with a wireless communication module and a remote monitoring platform, it improves detection efficiency and intelligence level. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the overall structure of the present invention; Figure 2 This is a schematic diagram of the lifting mechanism structure of the present invention; Figure 3 This is a schematic diagram of the multispectral imaging unit and the environmental adaptive supplementary lighting system of the present invention; Figure 4 This is a schematic diagram of the main control unit structure of the present invention; Figure 5 This is a schematic diagram of the mobile platform structure of the present invention; Figure 6 This is a schematic diagram of the drive wheel assembly structure of the present invention.

[0022] In the diagram: 1. Mobile platform; 11. Obstacle avoidance sensor; 12. Positioning module; 2. Lifting mechanism; 21. Drive column; 22. Motor slot; 23. Servo motor; 24. Lifting slot; 25. Lifting threaded rod; 26. Limiting slide bar; 27. Lifting plate; 28. Push plate; 29. ​​Lifting platform; 3. Rotating gimbal; 4. Multispectral imaging unit; 41. Visible light camera; 42. Infrared thermal imager; 43. Ultraviolet imager; 44. Polarizing filter; 5. Environmental adaptive lighting system; 51. Multi-angle LED light source; 52. Light sensor; 6. Main control unit; 61. Control box; 62. Inspection door; 63. Image processing module; 64. Wireless communication module; 7. Drive wheel set; 71. Wheel set bracket; 72. Drive motor; 73. Omnidirectional wheel. Detailed Implementation

[0023] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0024] like Figure 1 As shown, an electrical equipment inspection device based on machine vision includes a mobile platform 1, a lifting mechanism 2, a rotating gimbal 3, a multispectral imaging unit 4, an environmental adaptive lighting system 5, and a main control unit 6. The lifting mechanism 2 is installed above the mobile platform 1, the rotating gimbal 3 is installed at the top of the lifting mechanism 2, the multispectral imaging unit 4 is installed above the rotating gimbal 3, the environmental adaptive lighting system 5 is installed on one side of the rotating gimbal 3, the main control unit 6 is installed at the top of the mobile platform 1, the main control unit 6 is located in front of the lifting mechanism 2, and a drive wheel set 7 is installed at the bottom of the mobile platform 1.

[0025] like Figure 2 As shown, the lifting mechanism 2 includes a drive column 21, a servo motor 23, a lifting threaded rod 25, a limiting slide rod 26, a lifting plate 27, a push plate 28, and a lifting platform 29. The bottom of the drive column 21 has a motor slot 22, and the servo motor 23 is fixedly installed inside the motor slot 22. The side wall of the drive column 21 has a lifting groove 24 that runs through the front and back. The output shaft of the servo motor 23 passes through the inner cavity of the lifting groove 24. The lifting threaded rod 25 is fixedly installed on the output shaft of the servo motor 23. The lifting plate 27 is threadedly installed on the outer wall of the lifting threaded rod 25. The limiting slide rod 26 is fixedly installed between the upper and lower sides of the inner cavity of the lifting groove 24. The lifting plate 27 is slidably connected to the limiting slide rod 26. The push plate 28 is fixedly installed on the top two sides of the lifting plate 27. The top of the push plate 28 extends through to the top of the drive column 21. The lifting platform 29 is fixedly installed on the top of the push plate 28. The rotating gimbal 3 is installed on the top of the lifting platform 29.

[0026] Specifically, when the lifting mechanism 2 is working, the main control unit 6 sends a pulse control signal to the servo motor 23 according to the preset height value of the target equipment or automatically identifies the equipment height information through the image processing module 63, driving the lifting threaded rod 25 to rotate. When the lifting plate 27 touches the limit switches set at the upper and lower ends of the lifting groove 24, the servo motor 23 immediately stops running to prevent overtravel operation.

[0027] The rotating gimbal 3 is a two-degree-of-freedom gimbal, including horizontal rotational freedom and pitch rotational freedom, and is driven by a servo motor.

[0028] Specifically, the rotating gimbal 3 is controlled by the main control unit 6 via CAN bus or PWM signal to move horizontally and in pitch. The horizontal rotation angle range is 0 to 360°, and the pitch angle range is -30° to +60°. During image acquisition, the main control unit 6 can combine the real-time image feedback from the multispectral imaging unit 4 to automatically adjust the gimbal angle so that the target device is always located in the center area of ​​the image, realizing dynamic tracking and multi-angle scanning.

[0029] like Figure 3 As shown, the multispectral imaging unit 4 includes a visible light camera 41, an infrared thermal imager 42, and an ultraviolet imager 43. The visible light camera 41 is a high-resolution industrial camera, and a polarizing filter 44 is provided at the front end of the visible light camera 41. The environmental adaptive supplementary lighting system 5 includes a multi-angle LED light source 51 and a light sensor 52. The multi-angle LED light source 51 is installed on the top of the rotating gimbal 3, and the light sensor 52 is installed on the top of the lifting platform 29.

[0030] Specifically, the light sensor 52 collects the ambient illuminance value in real time, and the main control unit 6 automatically adjusts the output current of the multi-angle LED light source 51 according to the preset illuminance threshold, thereby achieving linear brightness adjustment. At the same time, the main control unit 6 dynamically adjusts the color temperature and illumination angle of the LED light source according to the grayscale histogram of the image collected by the visible light camera 41 to suppress high light reflection or shadow areas and ensure imaging consistency. The supplementary lighting system completes the adjustment within 100 ms before image acquisition to avoid affecting the acquisition rhythm.

[0031] like Figure 4 As shown, the main control unit 6 includes a control box 61, an image processing module 63, and a wireless communication module 64. The outer wall of the control box 61 is hinged with an inspection door 62. The image processing module 63 and the wireless communication module 64 are both located inside the control box 61. The image processing module 63 has built-in image enhancement algorithms, defect recognition models, meter reading recognition models, and infrared temperature analysis models.

[0032] Specifically, the image processing module 63 uses an embedded AI processor. The built-in image enhancement algorithm is based on Retinex theory and adaptive histogram equalization to improve the quality of low-light or high-contrast images. The defect recognition model is based on an improved YOLOv8 network structure. The meter reading recognition model adopts a U-Net segmentation network and OCR joint architecture, which supports the recognition of circular, pointer and digital meters. The infrared temperature analysis model automatically extracts abnormal temperature rise areas through temperature threshold segmentation and region growing algorithms, and performs pixel-level fusion with visible light images to output the fault location and temperature value.

[0033] like Figure 5 As shown, the mobile platform 1 is also equipped with an obstacle avoidance sensor 11 and a positioning module 12. The positioning module 12 is a lidar or UWB positioning module. The main control unit 6 is connected to the remote monitoring platform through the wireless communication module 64.

[0034] like Figure 6 As shown, the drive wheel set 7 includes a wheel set bracket 71, a drive motor 72, and omnidirectional wheels 73. The wheel set bracket 71 is fixedly installed on the bottom of the mobile platform 1, and the omnidirectional wheels 73 are arranged on both sides of the wheel set bracket 71. The omnidirectional wheels 73 are connected to the drive motor 72 for transmission.

[0035] The working principle of this machine vision-based electrical equipment inspection device will be explained in detail below.

[0036] like Figure 1-6 As shown, during operation, the mobile platform 1 moves to the front of the device to be inspected according to a preset path or remote command. The main control unit 6 controls the drive wheel group 7 to accurately position itself according to the positioning module 12 and the obstacle avoidance sensor 11. The lifting mechanism 2 is activated, the servo motor 23 drives the lifting threaded rod 25 to rotate, the lifting plate 27 rises and falls along the limit slide bar 26, and the lifting platform 29 rises and falls through the push plate 28, adjusting the multispectral imaging unit 4 to the set height. The rotating gimbal 3 adjusts the imaging angle so that the visible light camera 41, the infrared thermal imager 42, and the ultraviolet imager 43 are respectively aligned with the target detection area of ​​the device. The light sensor 52 detects the current ambient light conditions. The adaptive supplementary lighting system 5 automatically adjusts the brightness and color temperature of the multi-angle LED light source 51 according to the detection signal of the light sensor 52. The multispectral imaging unit 4 collects the multispectral image data of the device and transmits it to the main control unit 6. After receiving the multispectral image data, the image processing module 63 improves the image clarity through the image enhancement algorithm, and then calls the defect identification model to detect appearance abnormalities, the meter reading identification model to obtain meter reading values, and the infrared temperature analysis model to identify abnormal temperature rise areas. The detection results are uploaded to the remote monitoring platform by the main control unit 6 through the wireless communication module 64. After the single-point detection is completed, the device automatically moves to the next detection point.

[0037] During device operation, the main control unit 6 monitors the working status of each module in real time. If the lifting mechanism 2, the rotating gimbal 3, or the drive wheel assembly 7 is detected to be stuck, the system automatically stops operating and sends a fault alarm to the remote monitoring platform via the wireless communication module 64. The remote monitoring platform can issue inspection tasks, set inspection points, adjust image acquisition parameters, and receive inspection results and raw image data in real time. When the device is started for the first time each day, it automatically performs a self-test process, including gimbal zeroing, light source self-calibration, camera initialization, and communication link detection, to ensure that the system is in an available state.

[0038] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A machine vision-based electrical equipment inspection device, characterized in that: The system includes a mobile platform (1), a lifting mechanism (2), a rotating gimbal (3), a multispectral imaging unit (4), an environmental adaptive lighting system (5), and a main control unit (6). The lifting mechanism (2) is installed above the mobile platform (1), the rotating gimbal (3) is installed at the top of the lifting mechanism (2), the multispectral imaging unit (4) is installed above the rotating gimbal (3), the environmental adaptive lighting system (5) is installed on one side of the rotating gimbal (3), and the main control unit (6) is installed on the top of the mobile platform (1) and located in front of the lifting mechanism (2).

2. The electrical equipment inspection device based on machine vision according to claim 1, characterized in that: The lifting mechanism (2) includes a drive column (21), a servo motor (23), a lifting threaded rod (25), a limiting slide rod (26), a lifting plate (27), a push plate (28), and a lifting platform (29). The bottom of the drive column (21) has a motor slot (22), and the servo motor (23) is fixedly installed inside the motor slot (22). The side wall of the drive column (21) has a through-type lifting slot (24), and the output shaft of the servo motor (23) passes through the inner cavity of the lifting slot (24). The lifting threaded rod (25) is fixedly installed on the servo motor (26). The output shaft of 23), the lifting plate (27) is threadedly installed on the outer wall of the lifting threaded rod (25), the limiting slide rod (26) is fixedly installed between the upper and lower sides of the inner cavity of the lifting groove (24), the lifting plate (27) is slidably connected to the limiting slide rod (26), the push plate (28) is fixedly installed on the top two sides of the lifting plate (27), the top end of the push plate (28) extends through to the top of the drive column (21), the lifting platform (29) is fixedly installed on the top end of the push plate (28), and the rotating gimbal (3) is installed on the top of the lifting platform (29).

3. The electrical equipment inspection device based on machine vision according to claim 1, characterized in that: The rotating gimbal (3) is a two-degree-of-freedom gimbal, including horizontal rotational degree of freedom and pitch rotational degree of freedom, and is driven by a servo motor.

4. The electrical equipment inspection device based on machine vision according to claim 1, characterized in that: The multispectral imaging unit (4) includes a visible light camera (41), an infrared thermal imager (42), and an ultraviolet imager (43). The visible light camera (41) is a high-resolution industrial camera, and a polarizing filter (44) is provided at the front end of the visible light camera (41).

5. The electrical equipment inspection device based on machine vision according to claim 1, characterized in that: The environmental adaptive lighting system (5) includes a multi-angle LED light source (51) and a light sensor (52). The multi-angle LED light source (51) is installed on the top of the rotating gimbal (3), and the light sensor (52) is installed on the top of the lifting platform (29).

6. The electrical equipment inspection device based on machine vision according to claim 1, characterized in that: The main control unit (6) includes a control box (61), an image processing module (63) and a wireless communication module (64). The outer wall of the control box (61) is hinged with an inspection door (62). The image processing module (63) and the wireless communication module (64) are both located in the inner cavity of the control box (61).

7. The electrical equipment inspection device based on machine vision according to claim 6, characterized in that: The mobile platform (1) is also equipped with an obstacle avoidance sensor (11) and a positioning module (12). The positioning module (12) is a lidar or UWB positioning module. The main control unit (6) is connected to the remote monitoring platform through the wireless communication module (64).

8. The electrical equipment inspection device based on machine vision according to claim 1, characterized in that: It also includes a drive wheel assembly (7), which is mounted on the bottom of the mobile platform (1).

9. The electrical equipment inspection device based on machine vision according to claim 8, characterized in that: The drive wheel set (7) includes a wheel set bracket (71), a drive motor (72) and an omnidirectional wheel (73). The wheel set bracket (71) is fixedly installed at the bottom of the mobile platform (1). The omnidirectional wheel (73) is located on both sides of the wheel set bracket (71). The omnidirectional wheel (73) is connected to the drive motor (72) in a transmission connection.

10. The electrical equipment inspection device based on machine vision according to claim 6, characterized in that: The image processing module (63) has built-in image enhancement algorithm, defect recognition model, meter reading recognition model and infrared temperature analysis model.

Images