Patrol robot system for ultrahigh-voltage converter station

Abstract

The invention provides an ultrahigh-voltage converter station inspection robot system which comprises a relay protection room / power distribution room rail-mounted robot which is installed on top guide rails of a relay protection room and a power distribution device room, carries a high-definition zoom camera, an infrared thermal imager and a mechanical arm, and scans the states of a screen cabinet meter and a switch along the guide rails; the indoor wheeled robot is provided with a laser SLAM (Simultaneous Localization and Mapping) module and an anti-collision system, performs autonomous navigation in a GIS (Geographic Information System) indoor area and is integrated with an infrared imaging analyzer; the outdoor wheeled robot adopts an all-terrain hydraulic chassis and an IP65 protection grade, is equipped with an ultrasonic obstacle avoidance and weather station, and executes inspection of converter transformers, direct current field equipment and lightning arresters; and the central control platform is used for dynamically distributing three-type robot tasks based on a multi-agent reinforcement learning algorithm, and generating an equipment health report by fusing the infrared spectrum and the mechanical vibration data. According to the technical scheme, the technical problem that in the prior art, a single robot cannot achieve full-scene coverage inspection of the ultrahigh-voltage converter station can be effectively solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric power automation inspection, in particular to a kind of superhigh voltage converter station inspection robot system. BACKGROUND

[0002] Superhigh voltage converter stations play a crucial role in modern power systems, as they not only connect different voltage levels of power grids but also undertake the important tasks of power conversion and transmission. However, due to the dense equipment, complex structure, and harsh operating environment of superhigh voltage converter stations, regular inspection of the equipment poses extremely high requirements. Currently, converter station inspection mainly relies on regular manual inspection and remote monitoring based on fixed cameras, which has the following shortcomings: 1. Limitations of single robot solutions Existing robot inspection technologies for converter stations usually focus on developing robots suitable for specific environments, such as outdoor all-terrain wheeled robots or indoor mobile wheeled robots. However, such robots often struggle to meet the needs of both indoor precision equipment inspection and outdoor complex environment inspection. Indoor track-type robots, due to their installation on fixed rails, can accurately capture and collect data from relay protection screen cabinets and power distribution devices, but they cannot move freely indoors to adapt to the inspection needs of GIS and other indoor equipment. On the other hand, outdoor wheeled robots, although equipped with all-terrain off-road capabilities, are not suitable for indoor narrow space inspection, and there is an obvious complementary relationship between the two.

[0003] 2. Data silos and limitations of anomaly detection Current robot inspection systems usually operate independently, lacking effective data interaction and anomaly review mechanisms. This means that once a robot detects an abnormal signal from equipment, such as infrared overheating, partial discharge, or mechanical vibration anomalies, it is difficult to confirm whether the anomaly is real or a sensor false alarm without the support of data from other robots. The existence of data silos increases the uncertainty of anomaly detection, which may lead to false positives or false negatives, affecting the efficiency of inspection and the accuracy of equipment safety assessment.

[0004] 3. Environmental adaptability challenges for outdoor inspection In the outdoor areas of superhigh voltage converter stations, robots not only face complex terrain and large equipment inspection, but also extreme weather conditions. Rain and fog can significantly reduce the performance of visible light and infrared sensors, while wind, sand, and ice can affect the stability of the robot and the accuracy of the sensors. Although some robots are equipped with basic protective measures, such as waterproof housings and obstacle avoidance systems, these measures are often insufficient to ensure that the robot can work normally and collect effective data in all weather conditions in such a special environment as a superhigh voltage converter station.

[0005] In summary, the existing robot inspection technology of converter station has certain limitations in covering the whole scene, realizing data collaboration and coping with bad weather, which directly affects the comprehensiveness, accuracy and continuity of the inspection work. Therefore, it is urgent to develop a new robot inspection system to solve the technical problems mentioned above. On this basis, the present application proposes a multi-robot collaborative inspection system suitable for the complex environment of an ultra-high voltage converter station, aiming to comprehensively improve the inspection quality and efficiency and ensure the safe and stable operation of the converter station. However, the specific solutions and related technical details will be described in detail in the subsequent chapters. Here, we only focus on showing the shortcomings of the existing technical solutions to lay the foundation for further technological innovation and breakthrough.

[0006] The above is only a typical case in the technical field, and is intended to provide background information for the present application, and does not constitute an exhaustive description of the prior art. Those skilled in the art should understand that the above general description and the specific embodiments given below are part of the background art. SUMMARY

[0007] One aspect of the present application provides an ultra-high voltage converter station inspection robot system, which covers track robots in relay protection rooms / distribution rooms, indoor wheeled robots, outdoor wheeled robots and a central control platform, and realizes automatic inspection of the whole scene through multi-modal data fusion and task collaboration mechanism.

[0008] Further, the track robot in the relay protection room / distribution room is equipped with a high-definition zoom camera, an infrared thermal imager and a mechanical arm, and moves on the preset guide rail to conduct detailed inspection of the facilities in the relay protection screen cabinet and the indoor distribution device room.

[0009] The robot also includes a liftable gimbal mechanism with a vertical lifting stroke of 2 meters and a horizontal rotation angle of ±180°, as well as a guide rail self-cleaning module, to ensure smooth operation of the robot and prevent it from being affected by dust accumulation.

[0010] Further, the indoor wheeled robot is equipped with a laser SLAM module and a collision avoidance system for autonomous navigation in the GIS indoor area.

[0011] It integrates an infrared imaging analyzer specifically for monitoring equipment status, and a multi-sensor fusion positioning system that combines UWB anchor points, laser radar and visual odometry information to achieve accurate positioning within 5 centimeters.

[0012] The robot also has a device meter OCR module and a voiceprint anomaly detection unit, which respectively use an Attention-UNet network to identify key values such as oil temperature gauge and SF6 pressure gauge readings, and a Mel-spectrogram convolution network to detect potential abnormal sounds during circuit breaker operation.

[0013] Further, the outdoor wheeled robot adopts an all-terrain hydraulic chassis and an IP65 protection level, suitable for harsh weather conditions, with ultrasonic obstacle avoidance function and weather station, monitoring the working state of converter transformer, DC field equipment and surge arrester.

[0014] The robot is also equipped with an insulating obstacle crossing chassis, a multi-spectral contamination detector, and an autonomous charging pile docking system, which uses visual guidance technology to achieve high-precision charging insertion, ensuring the continuity of the robot's operation.

[0015] Further, the central control platform uses multi-agent reinforcement learning algorithm to dynamically adjust and allocate tasks to various robots, and can analyze and generate device health status reports from infrared spectrum, partial discharge signal, and mechanical vibration data.

[0016] The platform also supports cross-robot collaborative review, quickly dispatching nearby robots for multi-angle data verification when abnormalities are found, improving detection accuracy and reducing false positives and false negatives.

[0017] Further, the inspection method for the ultra-high voltage converter station includes: the outdoor wheeled robot responds to the central control platform instructions, preliminarily collects converter transformer infrared spectrum data, and identifies hot spots.

[0018] If the hot spot temperature exceeds the set threshold T1, the system will trigger a review of the converter transformer oil temperature and cooling system state.

[0019] Finally, the central control platform integrates infrared data, oil temperature readings, and cooling system operation information to form a fault confidence report on the health status of the converter transformer.

[0020] The introduction of this system significantly improves the efficiency and accuracy of automatic inspection of ultra-high voltage converter stations, reducing the frequency and labor intensity of manual inspection.

[0021] Multi-modal data fusion technology enables robots to comprehensively perceive device status, and intelligent task scheduling algorithms optimize robot path planning and resource utilization.

[0022] In addition, the cross-robot collaborative review mechanism further enhances system reliability and reduces maintenance costs.

[0023] Overall, this system not only improves the safety and efficiency of device operation and maintenance, but also promotes the digital transformation of the power industry. BRIEF DESCRIPTION OF DRAWINGS

[0024] The drawings described herein are intended to provide further understanding of the present application, and form a part of the present application. The illustrative embodiments of the present application and their description serve to explain the present application without unduly limiting it. In the drawings: Fig. 1 System block diagram of an ultra-high voltage converter station inspection robot Fig. 2 Flow chart of an ultra-high voltage converter station inspection method DETAILED DESCRIPTION

[0025] In order for those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should fall within the scope of protection of the present application.

[0026] It should be noted that the terms "first", "second", and the like in the specification and claims of the present application and the above-described drawings are used to distinguish similar objects, and do not necessarily have to be used to describe a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device that includes a series of steps or units does not have to be limited to only those steps or units clearly listed, but can include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0027] Reference should be made to Figs. 1-2 An aspect of the present application provides an ultra-high voltage converter station inspection robot system, which includes a relay protection room power distribution room track robot, an indoor wheeled robot, an outdoor wheeled robot, and a central control platform. The robots realize automatic inspection of full-scene coverage through multi-modal data fusion and task coordination mechanism. In terms of technology, the present embodiment realizes comprehensive coverage of indoor and outdoor equipment of the converter station by integrating different types of robots and combining their respective advantages, solving the problem that a single robot is difficult to adapt to the complex environment of the converter station. In principle, the multi-modal data fusion technology can comprehensively analyze various sensor data such as vision, infrared, and acoustics, improving the accuracy and reliability of anomaly detection. In terms of effect, the task coordination mechanism enables dynamic allocation of inspection tasks between robots, ensuring inspection efficiency and quality. In other embodiments, an unmanned aerial vehicle inspection module can be added to further improve the inspection capability of high-altitude equipment and solve the problem of inspection of areas that ground robots cannot reach.

[0028] Further, in one aspect of the present application, the track robot in the relay protection room and power distribution room is equipped with a high-definition zoom camera, an infrared thermal imager, and a mechanical arm, which can move on the preset track and conduct detailed inspection on the equipment in the relay protection screen cabinet and power distribution device room. In terms of technology, the high-definition zoom camera and infrared thermal imager can conduct high-precision visual and temperature detection on the equipment, and the mechanical arm can operate or sample specific equipment. In terms of principle, combined with the track moving mode, the robot can stably and accurately position to each detection point for data collection. In terms of effect, the robot in the embodiment can conduct dead-angle-free inspection on the precision equipment in the relay protection room and power distribution room, and timely discover equipment abnormalities. In other embodiments, the positioning accuracy and obstacle avoidance ability of the robot in complex environments can be improved by adding laser radar or ultrasonic sensors.

[0029] Further, in one aspect of the present application, the track robot in the relay protection room and power distribution room is equipped with a high-definition zoom camera, an infrared thermal imager, and a mechanical arm, which can move on the preset track and conduct detailed inspection on the equipment in the relay protection screen cabinet and power distribution device room. In terms of technology, the high-definition zoom camera and infrared thermal imager can conduct high-precision visual and temperature detection on the equipment, and the mechanical arm can operate or sample specific equipment. In terms of principle, combined with the track moving mode, the robot can stably and accurately position to each detection point for data collection. In terms of effect, the robot in the embodiment can conduct dead-angle-free inspection on the precision equipment in the relay protection room and power distribution room, and timely discover equipment abnormalities. In other embodiments, the positioning accuracy and obstacle avoidance ability of the robot in complex environments can be improved by adding laser radar or ultrasonic sensors.

[0030] Further, in one aspect of the present application, the track robot in the relay protection room and power distribution room is equipped with a high-definition zoom camera, an infrared thermal imager, and a mechanical arm, which can move on the preset track and conduct detailed inspection on the equipment in the relay protection screen cabinet and power distribution device room. In terms of technology, the high-definition zoom camera and infrared thermal imager can conduct high-precision visual and temperature detection on the equipment, and the mechanical arm can operate or sample specific equipment. In terms of principle, combined with the track moving mode, the robot can stably and accurately position to each detection point for data collection. In terms of effect, the robot in the embodiment can conduct dead-angle-free inspection on the precision equipment in the relay protection room and power distribution room, and timely discover equipment abnormalities. In other embodiments, the positioning accuracy and obstacle avoidance ability of the robot in complex environments can be improved by adding laser radar or ultrasonic sensors.

[0031] Further, the central control platform dynamically adjusts and assigns tasks to the three types of robots using a multi-agent reinforcement learning algorithm, which can generate a device health report from infrared maps, partial discharge signals, and mechanical vibration data. Multi-agent reinforcement learning algorithms are efficient solutions for complex tasks through the interaction and cooperation or competition of multiple agents in the environment, including cooperative methods (such as QMIX, MADDPG), competitive methods (such as independent Q learning), and hybrid methods (such as HEPN). Technically, the multi-agent reinforcement learning algorithm can dynamically optimize the inspection path according to the device state and robot position, improving the inspection efficiency. In principle, by analyzing multi-modal data, the central control platform can comprehensively judge the health status of the device and generate accurate reports. In terms of effects, the central control platform in this embodiment realizes intelligent management of robot inspection tasks, improving the automation level and data processing capability of inspection. In other embodiments, deep learning models can be added to further improve the accuracy of anomaly detection and the ability of fault prediction.

[0032] Further, the relay room power distribution room track robot includes a liftable holder mechanism with a vertical lifting stroke of ≥2 meters and a horizontal rotation angle of ±180°, and a guide rail self-cleaning module for removing dust on the robot motion guide rail to ensure smooth operation of the robot. Technically, the liftable holder mechanism and horizontal rotation design enable the robot to observe the device status in all directions, and the guide rail self-cleaning module ensures the continuity and stability of the robot motion. In principle, the holder mechanism is driven by a motor to realize lifting and rotation, and the self-cleaning module uses a brush and negative pressure dust collection principle to remove dust on the guide rail. In terms of effects, the technical solution in this embodiment ensures that the robot can effectively detect high equipment while maintaining smooth operation. In other embodiments, a guide rail lubrication system can be added to further improve the efficiency and stability of the robot movement on the guide rail.

[0033] Further, the indoor wheeled robot has a multi-sensor fusion positioning system that can fuse UWB anchor points, laser radar, and vision odometer information to achieve high-precision positioning of ≤5 centimeters. It also has a device meter OCR module that uses an Attention-UNet network to identify key values such as oil temperature and SF6 pressure. It also has a voiceprint anomaly detection unit that can identify abnormal sounds during the opening and closing of circuit breakers through a Mel spectrogram convolution network. Technically, the multi-sensor fusion positioning system combines UWB, laser radar, and vision odometer to improve positioning accuracy and reliability. In principle, the Attention-UNet network can focus on key areas in images to improve the accuracy of meter value identification, and the Mel spectrogram convolution network uses acoustic features to analyze the operating state of circuit breakers. In terms of effects, the robot in this embodiment can accurately identify device status and promptly identify potential fault risks. In other embodiments, temperature sensors or humidity sensors can be added to further improve environmental condition monitoring capabilities and provide more comprehensive data support for device status evaluation.

[0034] Further, the outdoor wheeled robot is equipped with an insulating obstacle-crossing chassis that can effectively deal with special environments near power equipment. It also carries a multi-spectral contamination detector that can evaluate the salt density on the surface of insulators through visible-near infrared spectral analysis technology. In addition, the robot is equipped with an autonomous charging pile docking system that uses visual guidance technology to achieve millimeter-level charging insertion, ensuring the robot's continuous working ability. Technically, the insulating obstacle-crossing chassis allows the robot to move safely near power equipment, and the multi-spectral contamination detector can accurately assess the contamination level of insulators. In principle, the autonomous charging pile docking system uses visual sensors to position the charging socket for precise docking. In terms of effects, the robot in this embodiment can adapt to harsh environments and effectively patrol outdoor equipment while ensuring the continuity of its own energy supply. In other embodiments, wireless charging technology or solar charging panels can be added to further improve the robot's endurance and environmental adaptability.

[0035] Further, in one aspect of the present application, in addition to dynamically scheduling tasks, the central control platform can also implement cross-robot collaborative review, that is, when a robot discovers an anomaly, the central control platform can quickly dispatch nearby robots to perform multiple angle data verification, reducing the possibility of false positives and false negatives. Technically, the cross-robot collaborative review mechanism realizes data sharing and task cooperation between multiple robots through intelligent scheduling of the central control platform. In principle, the central control platform dynamically allocates review tasks according to the device location and robot state, ensuring the accuracy and reliability of abnormal data. In terms of effects, the technical solution in the embodiment effectively improves the accuracy and efficiency of the inspection, and reduces the operation and maintenance cost. In other embodiments, a remote expert system can also be added to realize remote diagnosis of complex faults, further improving the timeliness and accuracy of fault handling.

[0036] Further, in one aspect of the present application, the inspection method of the ultra-high voltage converter station inspection robot system includes the following steps: after receiving the scheduling instruction of the central control platform, the outdoor wheeled robot first collects the infrared spectrum of the converter transformer and identifies whether there is a hot spot; if the hot spot temperature exceeds the set threshold T1, the system will trigger to review the oil temperature of the converter transformer and the operation of the cooling system; finally, the central control platform comprehensively generates a fault confidence report on the health status of the converter transformer based on the infrared data, the oil temperature of the converter transformer, and the operation status of the cooling system. Technically, the infrared detection technology of the outdoor wheeled robot can monitor the temperature of the converter transformer and discover potential thermal faults in a timely manner. In principle, combined with the intelligent analysis of the central control platform, the health status of the equipment can be comprehensively judged to generate an accurate fault report. In terms of effects, the technical solution in the embodiment realizes comprehensive monitoring of the converter transformer, improving the timeliness and accuracy of fault detection. In other embodiments, online monitoring sensors such as oil chromatographic analyzers or vibration sensors can also be added to further improve the real-time monitoring capability of the converter transformer state and realize continuous tracking of the health status of the equipment.

[0037] The technical scheme of the present application relates to a working process, in combination with the content of the above-mentioned embodiment module, the working process of the present application is as follows: after the system is started, the central control platform dynamically generates a patrol task list according to the distribution and state of the converter station equipment, and distributes the track robot in the relay protection room and the power distribution room, the indoor wheeled robot and the outdoor wheeled robot. The track robot moves along the preset guide rail in the relay protection room and the power distribution room, scans the screen cabinet meter, switch state by using the high-definition zoom camera and the infrared thermal imager, and the mechanical arm can be used for operation or sampling. The indoor wheeled robot autonomously navigates in the GIS indoor area, monitors the equipment state by using the infrared imaging analyzer, identifies the key value by using the equipment meter OCR module, and identifies the abnormal sound of the circuit breaker opening and closing by using the voiceprint abnormality detection unit. The outdoor wheeled robot patrols the outdoor facilities such as the converter transformer and the DC field, monitors the environment by using the ultrasonic obstacle avoidance and the weather station, and evaluates the insulator salt density value by using the multispectral contamination detector. When a robot detects abnormal data, the central control platform will dispatch the adjacent robot to review, comprehensively analyze the infrared spectrum, the partial discharge signal and the mechanical vibration data, and generate a device health report. The whole process realizes full-scene and dead-angle-free coverage of the converter station, effectively improves the patrol efficiency and the equipment operation and maintenance level.

[0038] Obviously, those skilled in the art should understand that the above-mentioned modules or steps of the present application can be realized by using general computing devices, which can be concentrated on a single computing device or distributed on a network composed of multiple computing devices, which can be realized by program codes executable by the computing device, so that they can be stored in the storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in different order, or they can be manufactured into individual integrated circuit modules, or multiple modules or steps can be manufactured into a single integrated circuit module. Thus, the present application is not limited to any specific combination of hardware and software.

[0039] The above only describes the preferred embodiments of the present application and is not used to limit the present application. For those skilled in the art, the present application can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. within the principles of the present application shall be included in the protection scope of the present application.

Claims

1. An ultrahigh voltage converter station inspection robot system, characterized in that, The system comprises a relay protection room track robot, an indoor wheeled robot, an outdoor wheeled robot, and a central control platform, wherein the robots realize automatic inspection of full-scene coverage through multi-modal data fusion and task coordination mechanism.

2. The ultrahigh voltage converter station patrol robot system according to claim 1, characterized in that The relay protection room track robot is equipped with a high-definition zoom camera, an infrared thermal imager, and a mechanical arm, and can move on a preset guide rail and conduct detailed inspection on equipment in the relay protection screen cabinet and indoor power distribution device.

3. The ultrahigh voltage converter station patrol robot system according to claim 1, characterized in that The indoor wheeled robot is equipped with a laser SLAM module and a collision avoidance system, can autonomously navigate in the GIS indoor area, and integrates an infrared imaging analyzer for monitoring equipment status.

4. The ultrahigh voltage converter station patrol robot system according to claim 1, characterized in that The outdoor wheeled robot adopts an all-terrain hydraulic chassis and an IP65 protection level, has the ability to work in harsh weather, is equipped with an ultrasonic obstacle avoidance function and a weather station to monitor the operating conditions of the converter transformer, DC field equipment, and lightning arrester.

5. The ultrahigh voltage converter station patrol robot system according to claim 1, characterized in that The central control platform dynamically adjusts and allocates tasks to the three types of robots using multi-agent reinforcement learning algorithm, and can generate a device health status report from infrared spectrum, partial discharge signal, and mechanical vibration data.

6. The ultrahigh voltage converter station patrol robot system according to claim 2, characterized in that The relay protection room track robot further comprises a liftable gimbal mechanism with a vertical lifting stroke of ≥2 meters and a horizontal rotation angle of ±180°, and a guide rail self-cleaning module for removing dust on the robot motion guide rail to ensure smooth operation of the robot.

7. The ultrahigh voltage converter station patrol robot system according to claim 3, characterized in that, The indoor wheeled robot has a multi-sensor fusion positioning system that can fuse UWB anchor points, laser radar, and visual odometry information to achieve high-precision positioning of ≤5 cm; it also has a voiceprint anomaly detection unit that can identify abnormal sounds generated during the opening and closing process of circuit breakers through a mel-spectrum convolution network.

8. The ultrahigh voltage converter station patrol robot system according to claim 4, characterized in that The outdoor wheeled robot is equipped with an insulation obstacle crossing chassis that can effectively deal with special environments near power equipment; it also carries a multi-spectral contamination detector that can evaluate the salt density value on the surface of insulators through visible-near infrared spectral analysis technology; in addition, the robot is equipped with an autonomous charging pile docking system that uses visual guidance technology to achieve millimeter-level charging insertion, ensuring the continuous working ability of the robot.

9. The ultrahigh voltage converter station patrol robot system according to claim 5, characterized in that In addition to dynamically scheduling tasks, the central control platform can also realize cross-robot collaborative review, that is, when a robot discovers an anomaly, the central control platform can quickly dispatch nearby robots to verify the data from multiple angles, reducing the likelihood of false positives and false negatives.

10. The method of claim 1-8, wherein The method comprises the following steps: after receiving the central control platform scheduling instruction, the outdoor wheeled robot first collects the infrared spectrum of the converter transformer and identifies whether there is a hot spot; if the temperature of the hot spot exceeds the set threshold T1, the system will trigger to review the operating conditions of the oil temperature and cooling system of the converter transformer; finally, the central control platform generates a fault confidence report on the health status of the converter transformer by comprehensively analyzing the infrared data, the oil temperature of the converter, and the operating conditions of the cooling system.

Images