Self-adaptive electric power inspection robot and inspection method

By integrating autonomous navigation, fault detection and dynamic decision-making units in power inspection robots, using neural networks and machine learning algorithms for intelligent decision-making and path optimization, the problem of insufficient navigation and diagnosis capabilities in the existing technology is solved and the power is improved System security and efficiency.

CN119036485APending Publication Date: 2024-11-29GUANGXI UNIV

Patent Information

Application Number
CN202411453624.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-17
Publication Date
2024-11-29

AI Technical Summary

Technical Problem

The existing power inspection robots have limited autonomous navigation capabilities in complex and dynamic environments, insufficient fault detection and diagnosis capabilities, lack adaptability and learning capabilities, and cannot make intelligent decisions and adjustments based on environmental changes and new task requirements.

Method used

An adaptive power inspection robot was designed, integrating a control processing module, an action module and a perception module, including an autonomous navigation algorithm unit, a fault detection and diagnosis unit and a dynamic decision-making unit, and a neural network and machine learning algorithm for real-time data Analyze and optimize path planning to achieve intelligent decision-making and adaptive adjustment.

Benefits of technology

Improve the navigation efficiency of robots in complex environments and the accuracy of fault detection, enhance the response ability to environmental changes, realize intelligent decision-making support, improve the reliability and safety of the power system, and reduce manual inspections rely.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119036485A_ABST
    Figure CN119036485A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of automatic inspection, in particular to a self-adaptive electric inspection robot and an inspection method. The robot comprises a control processing module, an action module and a sensing module which are arranged on a robot body frame. The control processing module is used for receiving the task information, processing the environment data transmitted by the sensing module and generating a control instruction; the action module is used for receiving a control instruction of the control processing module to drive the robot to move for inspection; the sensing module is used for environment sensing and detecting the state of to-be-detected equipment, and transmitting the obtained environment data and the state data of the to-be-detected equipment to the control processing module; and the control processing module is also used for processing the state data of the to-be-inspected equipment transmitted by the sensing module to obtain an inspection result of the to-be-inspected equipment. According to the method, environment changes are responded in real time, self-adaptive adjustment is carried out, intelligent decision-making is carried out, and the state of the power equipment is automatically analyzed and diagnosed, so that intelligent decision-making support based on data is realized, and the accuracy and efficiency of decision-making are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of automated inspection, and in particular to an adaptive electric power inspection robot and an inspection method. Background Art

[0002] The safe and stable operation of the power system is an indispensable guarantee for modern society. In order to ensure the reliability of the power system, regular inspection of power equipment is a vital task. Traditional power inspection mainly relies on manual work, which is not only labor-intensive and inefficient, but also has certain safety risks, especially in dangerous environments such as high altitude and high voltage.

[0003] With the continuous advancement of science and technology, automated power inspection robots have gradually been introduced into the field of power inspection, in order to replace or assist manual inspection tasks. These robots use preset inspection paths and sensor technology to improve the efficiency and safety of inspections to a certain extent. However, power inspection robots in the existing technology still have many limitations.

[0004] First, the autonomous navigation capabilities of existing robots are limited, especially in complex and changeable natural and electromagnetic environments, it is difficult to achieve effective path planning and obstacle avoidance. For example, although the technology mentioned in CN112230660A is capable of environmental perception, its adaptability in dynamic environments still needs to be improved. Secondly, the fault detection and diagnosis capabilities of existing robots are relatively rudimentary, and they can often only perform simple status detection, but cannot deeply analyze and accurately diagnose potential problems of power equipment. In addition, existing robots generally lack adaptability and learning capabilities, and cannot be intelligently adjusted and optimized according to environmental changes and new task requirements. Although CN112123338A proposes some improvements, it still has limitations in intelligent decision-making and real-time response.

[0005] In addition, existing power inspection robots are not capable of intelligent decision-making and cannot make autonomous decisions and responses based on real-time data. These problems seriously limit the efficiency and reliability of robots in practical applications and affect the overall quality and safety of power inspection work. In order to solve these problems, this application proposes an adaptive power inspection robot and inspection method. Summary of the invention

[0006] The purpose of the present invention is to provide an adaptive electric power inspection robot and an inspection method to solve the problem that the existing electric power inspection robots are unable to perform adaptive adjustments and intelligent decisions according to environmental changes and new task requirements.

[0007] To achieve the above purpose, the following technical solution is adopted.

[0008] An adaptive power inspection robot comprises: a control processing module, an action module and a perception module arranged on a main frame of the robot;

[0009] The control processing module is used to receive task instructions, process the environmental data transmitted by the perception module and the status data of the device to be inspected according to the task instructions, generate control instructions according to the processing results, and send the control instructions to the action module;

[0010] The action module is used to receive the control instructions of the control processing module and drive the robot to move and inspect according to the control instructions;

[0011] The sensing module is used to sense the environmental data obtained through environmental perception, detect the status of the device to be inspected, obtain the status data of the device to be inspected, and transmit the obtained environmental data and the status data of the device to be inspected to the control processing module;

[0012] The control processing module is also used to process the status data of the device to be inspected transmitted by the sensing module to obtain the inspection result of the device to be inspected.

[0013] Optionally, the control processing module integrates a task unit, an autonomous navigation algorithm unit, a fault detection and diagnosis unit, and a dynamic decision-making unit;

[0014] The task unit is used to receive and analyze task instructions from an operator or a remote control system, obtain task information, and send the task information to the autonomous navigation algorithm unit and the fault detection and diagnosis unit.

[0015] The autonomous navigation algorithm unit is used to perform path planning and obstacle avoidance according to the task information provided by the task receiving unit, in combination with the environmental data collected by the perception module and the pre-stored map data, generate an initial inspection strategy, and send the initial inspection strategy to the dynamic decision-making unit;

[0016] A fault detection and diagnosis unit, which is used to use a neural network algorithm to analyze the received inspection strategy and the status data of the equipment to be inspected collected by the perception module, automatically identify and diagnose the existing faults and potential faults of the equipment to be inspected, obtain the diagnosis result, record the diagnosis result and send the diagnosis result to the dynamic decision-making unit;

[0017] The dynamic decision-making unit is used to receive the initial inspection strategy of the autonomous navigation algorithm unit and the diagnosis result of the fault detection and diagnosis unit in real time, comprehensively consider environmental changes, fault urgency and task priority, use the optimization algorithm to dynamically adjust the inspection strategy, generate the optimized inspection strategy, and send the optimized inspection strategy to the autonomous navigation algorithm unit and the fault detection and diagnosis unit, and at the same time generate control instructions according to the optimized inspection strategy and send them to the action module.

[0018] Optionally, the control processing module is further integrated with an adaptive learning unit, and the adaptive learning unit is provided with an online learning mechanism for receiving in real time the path planning data generated by the autonomous navigation algorithm unit, the diagnosis result data provided by the fault detection and diagnosis unit, and the optimized inspection strategy data formulated by the dynamic decision-making unit, and analyzing the accuracy of the path planning based on the path planning data to obtain a planning analysis result;

[0019] Based on the diagnosis result data, analyzing the accuracy of the diagnosis result to obtain a diagnosis analysis result;

[0020] Based on the optimized inspection strategy data, analyze the effectiveness of inspection strategy optimization and obtain strategy analysis results;

[0021] Based on the obtained planning analysis results, diagnosis analysis results and strategy analysis results, an optimization model is generated using a machine learning algorithm to optimize the performance of the autonomous navigation algorithm unit, the fault detection and diagnosis unit and the dynamic decision-making unit respectively.

[0022] Optionally, the adaptive learning unit is provided with an offline learning mechanism, which is used to collect and store the optimization model and historical inspection data generated by the online learning mechanism during the non-inspection task of the robot; the performance of the autonomous navigation algorithm unit, the fault detection and diagnosis unit and the dynamic decision-making unit are optimized respectively by conducting an in-depth comprehensive analysis of the collected long-term historical data through the optimization model;

[0023] Optionally, it further includes a communication module and a remote control system, wherein the communication module is used to enable the control processing module to communicate with the remote control system.

[0024] Optionally, the perception module includes:

[0025] High-definition camera, used to capture visible light images of the equipment to be inspected and identify appearance defects and anomalies of the equipment to be inspected;

[0026] Infrared thermal imager, used to detect the heat distribution on the surface of the equipment to be inspected and identify hot spots and temperature anomalies;

[0027] LiDAR sensors are used for 3D modeling of the environment and obstacle detection, assisting robots in precise spatial positioning and navigation;

[0028] Sound sensor, used to collect sound signals generated when the equipment to be inspected is running, identify the running status of the equipment to be inspected and identify potential faults;

[0029] Environmental parameter sensors are used to detect temperature, humidity and air pressure in the inspection environment;

[0030] The data fusion unit is used to integrate the data collected by the high-definition camera, the infrared thermal imager, the lidar sensor, the sound sensor and the environmental parameter sensor, and provide the control processing module with comprehensive environmental data and status data of the equipment to be inspected.

[0031] Optionally, the action module includes:

[0032] at least one set of driving devices for executing movement and positioning of the robot;

[0033] At least one transmission mechanism connected to the driving device for transmitting power and converting the form of motion;

[0034] At least one set of posture adjustment devices, used to adjust the posture and direction of the robot body or specific parts to meet different inspection requirements;

[0035] At least one set of stabilizing devices, used to keep the robot stable when it moves or performs tasks;

[0036] at least one set of end effectors for performing inspection operations;

[0037] A motion control unit is used to coordinate and control the actions of the drive device, the transmission mechanism, the posture adjustment device, the stabilization device and the end effector.

[0038] Optionally, an interaction module is also included, which is arranged on the main frame and is used to provide an interface for the operator to interact with the robot.

[0039] An adaptive power inspection method comprises the following steps:

[0040] The control processing module receives the task instruction, processes the environmental data and the status data of the device to be inspected transmitted by the perception module according to the task instruction, generates a control instruction according to the processing result, and sends the control instruction to the action module;

[0041] The action module receives the control instruction from the control processing module, and drives the robot to move and inspect according to the control instruction;

[0042] The sensing module obtains environmental data through environmental sensing and detects the status of the device to be inspected to obtain status data of the device to be inspected, and transmits the obtained environmental data and status data of the device to be inspected to the control processing module;

[0043] The control processing module processes the status data of the device to be inspected transmitted by the sensing module to obtain the inspection result of the device to be inspected.

[0044] Optionally, the control processing module receives the task instruction, processes the environmental data transmitted by the perception module and the status data of the device to be inspected according to the task instruction, generates a control instruction according to the processing result, and sends the control instruction to the action module, specifically including:

[0045] The autonomous navigation algorithm unit in the control processing module receives the task information provided by the task unit, and simultaneously obtains the environmental data and pre-stored map data from the perception module;

[0046] The autonomous navigation algorithm unit performs path planning and obstacle avoidance, generates an initial inspection strategy, and sends the initial inspection strategy to the dynamic decision-making unit;

[0047] A fault detection and diagnosis unit, which adopts a neural network algorithm to analyze the received inspection strategy and the status data of the equipment to be inspected collected by the sensing module, automatically identifies and diagnoses the existing faults and potential faults of the equipment to be inspected, obtains the diagnosis result, records the diagnosis result and sends the diagnosis result to the dynamic decision-making unit;

[0048] The dynamic decision-making unit receives the initial inspection strategy of the autonomous navigation algorithm unit and the diagnosis results of the fault detection and diagnosis unit in real time, comprehensively considers environmental changes, fault urgency and task priority, uses the optimization algorithm to dynamically adjust the inspection strategy, generates an optimized inspection strategy, and sends the optimized inspection strategy to the autonomous navigation algorithm unit and the fault detection and diagnosis unit. At the same time, it generates control instructions based on the optimized inspection strategy and sends them to the action module.

[0049] Compared with the prior art, the present invention has the following beneficial effects:

[0050] The adaptive power inspection robot of the present invention is equipped with an autonomous navigation algorithm unit and a dynamic decision-making unit, which can respond to environmental changes in real time and make adaptive adjustments, thus solving the shortcomings of existing robots in terms of environmental adaptability. By integrating a fault detection and diagnosis unit and an adaptive learning unit, the robot can make intelligent decisions, automatically analyze and diagnose the status of power equipment, thereby realizing data-based intelligent decision support and improving the accuracy and efficiency of decision-making.

[0051] The perception module of the present invention is combined with a data fusion unit to collect and process environmental and equipment status data in real time, ensuring the robot's rapid response and accurate monitoring of environmental changes, thereby improving the real-time and accuracy of inspections.

[0052] The autonomous navigation algorithm unit of the present invention dynamically optimizes path planning according to real-time environmental data and task requirements, effectively avoids obstacles, and improves the navigation efficiency of the robot in complex environments.

[0053] The present invention uses deep learning and neural network algorithms, and the fault detection and diagnosis unit can accurately identify and warn potential equipment failures, enhancing the depth and breadth of fault detection. The online and offline learning mechanisms of the adaptive learning unit enable the robot to continuously learn from inspection experience and continuously improve the performance of autonomous navigation, fault detection and decision-making.

[0054] The dynamic decision-making unit of the present invention can flexibly adjust the priority and sequence of task execution according to real-time inspection data and diagnosis results, thereby ensuring timely response to key tasks.

[0055] Through intelligent decision-making and adaptive adjustment, the robot can more effectively identify and respond to potential faults and risks, thereby improving the reliability and safety of the entire power system. The automation and intelligence of the robot reduces the reliance on manual inspections and reduces the risks of manual operation, especially in dangerous or hard-to-reach environments. The robot's adaptive adjustment and intelligent decision-making capabilities make power inspections more efficient and reduce operation and maintenance interruptions caused by environmental changes and new task requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 This is a schematic diagram of the structural modules of an adaptive power inspection robot embodiment of the present invention.

[0057] Figure 2 The figure is a control logic diagram of an adaptive power inspection robot embodiment of the present invention.

[0058] Figure 3 The figure is a flow chart of an embodiment of an adaptive power inspection method of the present invention.

[0059] Among them: 1-control processing module, 11-task unit, 12-fault detection and diagnosis unit, 13-autonomous navigation algorithm unit, 14-dynamic decision making unit, 2-action module, 3-perception module. DETAILED DESCRIPTION

[0060] The present invention will be described in detail below with reference to the accompanying drawings and in combination with embodiments. It should be noted that the embodiments and features in the embodiments of the present application can be combined with each other without conflict.

[0061] The following detailed description is an exemplary description, which is intended to provide further detailed description of the present invention. Unless otherwise specified, all technical terms used in the present invention have the same meaning as those generally understood by those skilled in the art to which the present application belongs. The terms used in the present invention are only for describing specific embodiments, and are not intended to limit exemplary embodiments according to the present invention.

[0062] like Figure 1As shown, an adaptive power inspection robot includes: a control processing module 1, an action module 2 and a perception module 3 arranged on the robot body frame; the control processing module 1 is used to receive task instructions, process the environmental data and the status data of the equipment to be inspected transmitted by the perception module 3 according to the task instructions, generate control instructions according to the processing results, and send the control instructions to the action module 2; the action module 2 is used to receive the control instructions of the control processing module 1, and drive the robot to move for inspection according to the control instructions; the perception module 3 is used to obtain environmental data through environmental perception, detect the status of the equipment to be inspected, obtain the status data of the equipment to be inspected, and transmit the obtained environmental data and the status data of the equipment to be inspected to the control processing module 1; the control processing module 1 is also used to process the status data of the equipment to be inspected transmitted by the perception module 3, and obtain the inspection result of the equipment to be inspected.

[0063] Specifically, the main frame is made of high-strength lightweight alloy material to provide sufficient stability and durability while reducing the overall weight of the robot. The power module may include a high-capacity lithium battery and an energy management system, which are installed inside the main frame to provide continuous and stable power to all modules of the robot.

[0064] The motion module 2 may include a plurality of driving devices, such as electric motors and hydraulic systems, and matching transmission mechanisms, such as gears and belts, installed at the bottom or side of the main frame to achieve movement and positioning of the robot.

[0065] The perception module 3 can integrate a variety of sensors, including high-definition cameras, infrared thermal imagers, lidar sensors, sound sensors and environmental parameter sensors, and be installed at appropriate positions on the main frame to fully perceive the environment and detect the status of the equipment to be inspected.

[0066] The control processing module 1 may include a task unit 11, an autonomous navigation algorithm unit 13, a fault detection and diagnosis unit 12 and a dynamic decision-making unit 14; it is electrically connected to the power module, the action module 2 and the perception module 3, and is used to process task information, environmental data and equipment status data, generate control instructions, and output inspection results.

[0067] Through the above specific implementation methods, the adaptive power inspection robot of the present invention can achieve adaptive adjustment and intelligent decision-making in complex environments, and effectively improve the efficiency and safety of power inspection.

[0068] like Figure 2 As shown as a preferred example, the control processing module 1 integrates a task unit 11, an autonomous navigation algorithm unit 13, a fault detection and diagnosis unit 12 and a dynamic decision-making unit 14;

[0069] The task unit 11 is used to receive and analyze task instructions from an operator or a remote control system, obtain task information, and send the task information to the autonomous navigation algorithm unit 13 and the fault detection and diagnosis unit 12.

[0070] The autonomous navigation algorithm unit 13 is used to perform path planning and obstacle avoidance according to the task information provided by the task receiving unit, combined with the environmental data collected by the perception module 3 and the pre-stored map data, generate an initial inspection strategy, and send the initial inspection strategy to the dynamic decision-making unit 14;

[0071] The fault detection and diagnosis unit 12 is used to analyze the received inspection strategy and the status data of the equipment to be inspected collected by the perception module 3 using a neural network algorithm, automatically identify and diagnose the existing faults and potential faults of the equipment to be inspected, obtain the diagnosis result, record the diagnosis result and send the diagnosis result to the dynamic decision making unit 14;

[0072] The dynamic decision-making unit 14 is used to receive the initial inspection strategy of the autonomous navigation algorithm unit 13 and the diagnosis results of the fault detection and diagnosis unit 12 in real time, comprehensively consider environmental changes, fault urgency and task priority, use the optimization algorithm to dynamically adjust the inspection strategy, generate the optimized inspection strategy, and send the optimized inspection strategy to the autonomous navigation algorithm unit 13 and the fault detection and diagnosis unit 12. At the same time, it generates control instructions based on the optimized inspection strategy and sends them to the action module 2 to adjust and control the robot in real time.

[0073] Specifically, the task unit 11 may include an instruction receiver and a parser, which are used to receive task instructions from an operator or a remote control system, and parse the instructions into task information that the robot can understand and execute.

[0074] The autonomous navigation algorithm unit 13 may include a path planner and an obstacle avoider, and utilizes the environmental data collected by the perception module 3 and the pre-stored map data to generate an initial inspection path through advanced algorithms such as A* or Dijkstra algorithm.

[0075] The perception module 3 collects environmental data in real time and transmits it to the autonomous navigation algorithm unit 13 through the control processing module 1 to achieve dynamic adjustment and optimization of the path.

[0076] The fault detection and diagnosis unit 12 uses deep learning algorithms, including convolutional neural networks (CNNs) and recurrent neural networks (RNNs), to analyze the device status data collected by the perception module 3 to achieve automatic identification and diagnosis of faults.

[0077] The dynamic decision-making unit 14 dynamically adjusts the inspection strategy by using optimization algorithms such as genetic algorithms or Monte Carlo tree search in combination with real-time environmental changes, fault urgency, and task priorities. Initially, the diagnosis result transmitted by the fault detection and diagnosis unit is a null value. At this time, only the initial inspection strategy is used, and the inspection strategy is dynamically adjusted by using the optimization algorithm to comprehensively consider environmental changes and task priorities, thereby generating an optimized inspection strategy.

[0078] The initial inspection strategy generated by the autonomous navigation algorithm unit 13 is sent to the dynamic decision making unit 14, which optimizes it to adapt to possible environmental changes and fault conditions.

[0079] The optimized inspection strategy is sent back to the autonomous navigation algorithm unit 13 and the fault detection and diagnosis unit 12 to guide the robot's navigation and fault detection activities.

[0080] According to the optimized inspection strategy, the dynamic decision-making unit 14 generates a control instruction and sends it to the action module 2 through the control processing module 1 to drive the robot to perform the corresponding action.

[0081] The action module 2 receives the control instructions and executes the movement, positioning and posture adjustment of the robot to ensure that the robot operates according to the optimized inspection strategy.

[0082] As a preferred example, the control processing module 1 is also integrated with an adaptive learning unit, which is provided with an online learning mechanism for receiving in real time the path planning data generated by the autonomous navigation algorithm unit 13, the diagnostic result data provided by the fault detection and diagnosis unit 12, and the optimized inspection strategy data formulated by the dynamic decision-making unit 14, and analyzing the accuracy of the path planning based on the path planning data to obtain the planning analysis result; analyzing the accuracy of the diagnostic result based on the diagnostic result data to obtain the diagnostic analysis result; analyzing the effectiveness of the inspection strategy optimization based on the optimized inspection strategy data to obtain the strategy analysis result; and generating an optimization model based on the obtained planning analysis result, diagnostic analysis result and strategy analysis result using a machine learning algorithm to optimize the performance of the autonomous navigation algorithm unit 13, the fault detection and diagnosis unit 12 and the dynamic decision-making unit 14, respectively.

[0083] Specifically, the control processing module 1 further integrates an adaptive learning unit, which is designed to continuously improve the performance and intelligence level of the robot. The adaptive learning unit may include an online learning mechanism that can receive and process data from the autonomous navigation algorithm unit 13, the fault detection and diagnosis unit 12, and the dynamic decision-making unit 14 in real time.

[0084] The online learning mechanism receives the path planning data generated by the autonomous navigation algorithm unit 13 , the diagnosis result data provided by the fault detection and diagnosis unit 12 , and the optimized inspection strategy data formulated by the dynamic decision-making unit 14 .

[0085] The adaptive learning unit conducts in-depth analysis of the received data to evaluate the accuracy of path planning, the accuracy of diagnostic results, and the effectiveness of inspection strategy optimization.

[0086] Based on the results of data analysis and evaluation, the adaptive learning unit generates an optimization model using machine learning algorithms, such as supervised learning, reinforcement learning, or unsupervised learning algorithms.

[0087] The optimization model is used to improve the path planning accuracy of the autonomous navigation algorithm unit 13 , improve the diagnostic accuracy of the fault detection and diagnosis unit 12 , and enhance the strategy optimization capability of the dynamic decision-making unit 14 .

[0088] The adaptive learning unit optimizes the model through historical data training and verifies the effectiveness of the model through actual inspection data to ensure the accuracy and robustness of the model.

[0089] The adaptive learning unit can update the optimization model in real time to adapt to environmental changes and new task requirements, and realize continuous iteration and optimization of the algorithm.

[0090] The optimized models and parameters are passed to the corresponding units through the feedback mechanism, achieving performance improvements in autonomous navigation, fault detection and diagnosis, and dynamic decision making.

[0091] Through continuous learning and optimization of the adaptive learning unit, the robot's intelligence level has been significantly improved, enabling it to complete power inspection tasks more efficiently and accurately.

[0092] As a preferred example, the adaptive learning unit is provided with an offline learning mechanism for collecting and storing the optimization model and historical inspection data generated by the online learning mechanism during the non-inspection tasks of the robot; the optimization model is used to conduct an in-depth and comprehensive analysis of the collected long-term historical data to optimize the performance of the autonomous navigation algorithm unit 13, the fault detection and diagnosis unit 12 and the dynamic decision-making unit 14 respectively.

[0093] Specifically, the adaptive learning unit includes an offline learning mechanism designed to work during the period when the robot is in a non-inspection task.

[0094] The offline learning mechanism is responsible for collecting the optimization model and historical inspection data generated by the online learning mechanism, and storing these data in the robot's local storage or cloud storage system.

[0095] Using stored historical inspection data, the offline learning mechanism conducts in-depth comprehensive analysis to identify patterns, trends and potential areas for improvement in inspection tasks.

[0096] The offline learning mechanism uses the optimization model to analyze historical data to verify the accuracy and efficiency of the model, and further adjusts and improves the model based on the analysis results.

[0097] Based on the analysis results of the offline learning mechanism, the performance of the autonomous navigation algorithm unit 13, the fault detection and diagnosis unit 12 and the dynamic decision making unit 14 are optimized.

[0098] Autonomous Navigation Algorithms Unit 13: Optimize path planning and obstacle avoidance algorithms to improve the efficiency and accuracy of robot navigation in complex environments.

[0099] Fault Detection and Diagnosis Unit 12: Improve neural network algorithms to enhance the sensitivity and accuracy of fault detection and reduce false positives and negatives.

[0100] Dynamic decision making unit 14: adjusts the optimization algorithm to enhance the response speed and decision quality to environmental changes and fault conditions.

[0101] Update of models and parameters: Update the models and parameters optimized by the offline learning mechanism to the corresponding units so that the robot can apply the improved algorithm in the next inspection task.

[0102] Ensure that the results of the online learning mechanism and offline learning mechanism of the adaptive learning unit can be seamlessly integrated to achieve a closed loop of robot learning.

[0103] Continuous improvement of system performance: Through the combination of online learning and offline learning, the robot system performance can be continuously improved, and the robot's ability to adapt to new environments and tasks can be improved.

[0104] Whenever possible, the offline learning mechanism is adjusted based on operator feedback to ensure that the learning outcomes meet the needs of actual inspections.

[0105] Through the above-mentioned specific implementation methods, the adaptive power inspection robot of the present invention can realize continuous optimization and upgrading of its own performance during non-inspection periods, ensuring that the robot can be more efficient and intelligent when performing inspection tasks.

[0106] As a preferred example, it further includes a communication module and a remote control system, and the communication module is used to enable the control processing module 1 to communicate with the remote control system.

[0107] Specifically, a communication module is integrated in the control processing module 1, and the module is responsible for data transmission and communication with the remote control system.

[0108] The communication module can include wireless communication interfaces such as Wi-Fi, 4G / 5G, Bluetooth or dedicated wireless protocols to ensure that the robot can communicate stably in different environments.

[0109] A remote control system is established, which can send task instructions to the robot and receive inspection data and status information sent back by the robot.

[0110] The communication module uses data encryption technology, such as TLS / SSL protocol, to ensure the security and privacy of communication data between the robot and the remote control system.

[0111] The communication module implements real-time communication functions, allowing the remote control system to monitor the robot status in real time and perform remote control and intervention when necessary.

[0112] The remote control system sends task instructions to the robot through the communication module, including inspection routes, key inspection areas or other specific tasks.

[0113] Feedback of inspection data: The robot transmits the collected inspection information such as environmental data, equipment status data, and fault diagnosis results to the remote control system in real time through the communication module.

[0114] Remote fault diagnosis and analysis: Inspection data received by the remote control system can be used for more in-depth analysis and fault diagnosis to facilitate operator decision making.

[0115] System compatibility and scalability: The communication module design takes into account compatibility with different remote control systems and has scalability to adapt to future technology upgrades.

[0116] Monitoring of communication status: The control processing module 1 monitors the working status of the communication module to ensure the stability of the communication link and take corresponding measures when the communication is interrupted.

[0117] Redundant design of communication modules: To improve the reliability of communication, communication modules can adopt redundant design, such as multi-band communication capability, to ensure that communication can be maintained when certain frequency bands are blocked.

[0118] Through the above specific implementation methods, the adaptive power inspection robot of the present invention can establish a stable and secure communication connection with the remote control system, realize remote task distribution, status monitoring and data analysis, and improve the flexibility and response speed of the inspection work.

[0119] As a preferred example, the perception module 3 includes:

[0120] High-definition camera, used to capture visible light images of the equipment to be inspected and identify appearance defects and anomalies of the equipment to be inspected;

[0121] Infrared thermal imager, used to detect the heat distribution on the surface of the equipment to be inspected and identify hot spots and temperature anomalies;

[0122] LiDAR sensors are used for 3D modeling of the environment and obstacle detection, assisting robots in precise spatial positioning and navigation;

[0123] Sound sensor, used to collect sound signals generated when the equipment to be inspected is running, identify the running status of the equipment to be inspected and identify potential faults;

[0124] Environmental parameter sensors are used to detect temperature, humidity and air pressure in the inspection environment;

[0125] The data fusion unit is used to integrate the data collected by the high-definition camera, infrared thermal imager, lidar sensor, sound sensor and environmental parameter sensor, and provide the control processing module 1 with comprehensive environmental data and status data of the equipment to be inspected.

[0126] Specifically, a high-definition camera is integrated in the perception module 3 to capture visible light images of the device to be inspected, and to identify appearance defects and anomalies of the device through image processing technology.

[0127] Install infrared thermal imagers to detect the thermal distribution map of the surface of the equipment to be inspected, identify hot spots and temperature anomalies through temperature analysis, and predict potential failures.

[0128] Equipped with lidar sensors for three-dimensional modeling of the environment, obstacle detection, and assisting the robot in spatial positioning and navigation.

[0129] The integrated sound sensor collects the sound signals generated when the equipment to be inspected is running, and identifies the equipment's operating status and potential faults through sound analysis technology.

[0130] Environmental parameter sensors are installed to detect parameters such as temperature, humidity and air pressure in the inspection environment in real time, and provide environmental adaptability data for the robot.

[0131] Development of a data fusion unit that integrates data collected by high-definition cameras, thermal imaging cameras, lidar sensors, sound sensors, and environmental parameter sensors.

[0132] The data fusion unit preprocesses the collected raw data, including filtering, denoising and normalization, to improve the data quality.

[0133] Extract key features from the preprocessed data, such as defect features in images, hotspot features in heat maps, abnormal frequency features in sound signals, etc.

[0134] The data fusion unit comprehensively analyzes the extracted features and generates a comprehensive status report of the equipment and environment.

[0135] The comprehensive analysis results are output to the control processing module 1 to provide data support for the robot's autonomous navigation, fault detection and diagnosis, and dynamic decision-making.

[0136] The control processing module 1 adjusts the robot's inspection strategy and path planning in real time according to the information provided by the data fusion unit.

[0137] Data fusion algorithms and feature extraction technologies are regularly optimized and upgraded to adapt to new inspection tasks and environmental changes.

[0138] Through the above specific implementation methods, the adaptive power inspection robot of the present invention can achieve comprehensive perception of power equipment and the environment, and improve the accuracy and efficiency of inspection.

[0139] As a preferred example, the action module 2 includes:

[0140] at least one set of driving devices for executing movement and positioning of the robot;

[0141] At least one transmission mechanism connected to the driving device for transmitting power and converting the form of motion;

[0142] At least one set of posture adjustment devices, used to adjust the posture and direction of the robot body or specific parts to meet different inspection requirements;

[0143] At least one set of stabilizing devices, used to keep the robot stable when it moves or performs tasks;

[0144] at least one set of end effectors for performing inspection operations;

[0145] Motion control unit, used to coordinate and control the actions of the drive device, transmission mechanism, attitude adjustment device, stabilizer and end effector.

[0146] Specifically, the motion module 2 includes at least one set of driving devices, which may be electric motors, hydraulic motors or pneumatic motors, for providing the power required for the movement and positioning of the robot.

[0147] At least one transmission mechanism connected to the driving device, including gears, belts, chains or gear boxes, is used to transmit power and convert motion forms, such as converting rotational motion into linear motion.

[0148] At least one set of posture adjustment devices, such as a servo motor with a connecting rod mechanism or a rotary joint, is used to adjust the posture and direction of the robot body or a specific part to adapt to different inspection perspectives and operation requirements.

[0149] At least one set of stabilizing devices, such as a balancing system, shock absorbers, or adjustable support feet, is used to keep the robot stable and reduce vibration and displacement when the robot moves or performs tasks.

[0150] At least one set of end effectors, which can be robotic arms, grippers, sensor probes or other specialized tools, are used to perform specific detection and manipulation tasks.

[0151] The motion control unit uses advanced control algorithms and software to accurately coordinate and control the actions of the drive device, transmission mechanism, attitude adjustment device, stabilizer and end effector.

[0152] The motion control unit adjusts the motion parameters of each component, such as speed, acceleration, position and direction, in real time according to the instructions of the control processing module 1 to ensure the accuracy and stability of the robot's movements.

[0153] The motion module 2 is designed as a multi-degree-of-freedom motion system, which enables the robot to flexibly inspect and operate in complex environments.

[0154] The motion module 2 adopts a modular design, which facilitates the rapid replacement or upgrading of the end effector and other components according to the specific needs of the inspection task.

[0155] Safety features such as emergency stop button, overload protection and fault detection are integrated in the motion module 2 to ensure the safety of robot operation.

[0156] The operation of the action module 2 can be manually controlled through the user interface of the interaction module to achieve human-machine collaborative work.

[0157] Through the above specific implementation methods, the motion module 2 of the adaptive power inspection robot of the present invention can achieve precise, flexible and stable movement and operation, and meet the requirements of various power inspection tasks.

[0158] As a preferred example, an interaction module is also included. The interaction module is arranged on the main frame and is used to provide an interface for the operator to interact with the robot.

[0159] Specifically, an interaction module is integrated on the main frame, and the module is a key interface for the operator to interact with the robot.

[0160] The interaction module includes a user interface, which can be a touch screen, physical buttons, joystick or voice control interface, for receiving instructions from the operator and displaying the robot's status information.

[0161] The user interface can display the robot's inspection status, environmental data, fault diagnosis results and other key information in real time.

[0162] Operators can input control instructions through the interactive module, such as starting an inspection, pausing an inspection, changing the inspection route, or performing a detailed inspection of specific equipment.

[0163] The interaction module provides feedback mechanisms, including visual, auditory, or tactile feedback, to confirm that the operator's instructions have been received and executed.

[0164] The interactive module supports remote monitoring and control functions, allowing operators to operate the robot away from the inspection site.

[0165] When the robot encounters a malfunction or abnormal situation, the interactive module provides an emergency intervention interface, allowing the operator to take quick action.

[0166] like Figure 3 As shown, an adaptive power inspection method includes the following steps:

[0167] The control processing module 1 receives the task instruction, processes the environmental data and the status data of the device to be inspected transmitted by the perception module 3 according to the task instruction, generates a control instruction according to the processing result, and sends the control instruction to the action module 2;

[0168] The action module 2 receives the control instruction from the control processing module 1, and drives the robot to move and inspect according to the control instruction;

[0169] The sensing module 3 obtains environmental data through environmental sensing and detects the status of the device to be inspected to obtain status data of the device to be inspected, and transmits the obtained environmental data and status data of the device to be inspected to the control processing module 1;

[0170] The control processing module 1 processes the status data of the device to be inspected transmitted by the sensing module 3 to obtain the inspection result of the device to be inspected.

[0171] Specifically, the power module, action module 2, perception module 3 and control processing module 1 of the adaptive power inspection robot are started to ensure that each module works normally and enters a standby state.

[0172] The control processing module 1 receives the task instructions from the remote control system through the integrated communication module, analyzes the task details, and prepares to execute the scheduled inspection task.

[0173] The high-definition camera, infrared thermal imager, lidar sensor, sound sensor and environmental parameter sensor in the perception module 3 are used to collect the environmental and equipment status data of the area to be inspected.

[0174] The control processing module 1 processes the received environmental data in real time, uses data fusion technology to integrate multi-source data, and provides accurate environmental information for decision-making.

[0175] The autonomous navigation algorithm unit 13 in the control processing module 1 plans the inspection path according to the task information and environmental data, and instructs the action module 2 to drive the robot to move to the area to be inspected.

[0176] In the area to be inspected, the perception module 3 performs detailed status detection on the power equipment and collects data such as images, thermal images, sounds, and environmental parameters of the equipment.

[0177] The control processing module 1 uses the deep learning algorithm of the fault detection and diagnosis unit 12 to analyze the equipment status data and diagnose the operating status and potential faults of the equipment.

[0178] According to the fault diagnosis results, the dynamic decision-making unit 14 adjusts the inspection strategy in real time, optimizes the inspection path, and ensures that key issues are inspected in a timely manner.

[0179] The online learning mechanism of the adaptive learning unit receives navigation and diagnosis data in real time, evaluates and optimizes the navigation path and diagnosis accuracy, while the offline learning mechanism uses historical data to further train and optimize the algorithm.

[0180] After the inspection task is completed, the control processing module 1 organizes the fault diagnosis results, generates a detailed inspection report, and sends it to the remote control system through the communication module.

[0181] The control processing module 1 receives feedback information from the remote control system, confirms the inspection results, and adjusts the robot state to standby or maintenance mode as needed.

[0182] Through the above specific implementation methods, the adaptive power inspection method of the present invention realizes the automated and intelligent inspection of robots, improves the inspection efficiency and safety of power facilities, and continuously optimizes the inspection performance through an adaptive learning mechanism.

[0183] As a preferred example, the control processing module 1 receives the task instruction, processes the environmental data and the status data of the device to be inspected transmitted by the perception module 3 according to the task instruction, generates a control instruction according to the processing result, and sends the control instruction to the action module 2, specifically including:

[0184] The autonomous navigation algorithm unit 13 in the control processing module 1 receives the task information provided by the task unit 11, and simultaneously obtains the environmental data and pre-stored map data from the perception module 3;

[0185] The autonomous navigation algorithm unit 13 performs path planning and obstacle avoidance, generates an initial inspection strategy, and sends the initial inspection strategy to the dynamic decision-making unit 14;

[0186] The fault detection and diagnosis unit 12 uses a neural network algorithm to analyze the received inspection strategy and the status data of the equipment to be inspected collected by the perception module 3, automatically identifies and diagnoses the existing faults and potential faults of the equipment to be inspected, obtains the diagnosis result, records the diagnosis result and sends the diagnosis result to the dynamic decision making unit 14;

[0187] The dynamic decision-making unit 14 receives the initial inspection strategy of the autonomous navigation algorithm unit 13 and the diagnosis results of the fault detection and diagnosis unit 12 in real time, comprehensively considers environmental changes, fault urgency and task priority, uses the optimization algorithm to dynamically adjust the inspection strategy, generates an optimized inspection strategy, and sends the optimized inspection strategy to the autonomous navigation algorithm unit 13 and the fault detection and diagnosis unit 12. At the same time, it generates control instructions based on the optimized inspection strategy and sends them to the action module 2 to adjust and control the robot in real time.

[0188] Specifically, the autonomous navigation algorithm unit 13 in the control processing module 1 receives the task information provided by the task unit 11, and analyzes the task information to determine the inspection objectives and requirements.

[0189] The autonomous navigation algorithm unit 13 synchronously obtains environmental data and pre-stored map data from the perception module 3, and these data include but are not limited to device location, obstacle distribution, environmental characteristics, etc.

[0190] The autonomous navigation algorithm unit 13 uses the acquired data to plan the path, such as using the A* algorithm or the Dijkstra algorithm, and simultaneously implements obstacle avoidance to generate an efficient inspection path.

[0191] Based on the path planning results, the autonomous navigation algorithm unit 13 generates an initial inspection strategy, including inspection routes, checkpoints, and estimated time.

[0192] The initial inspection strategy is sent to the dynamic decision making unit 14 for further optimization.

[0193] The fault detection and diagnosis unit 12 uses a neural network algorithm to analyze the status data of the equipment to be inspected collected by the perception module 3 to perform fault identification and diagnosis.

[0194] The fault detection and diagnosis unit 12 records the diagnosis result and transmits it to the dynamic decision-making unit 14 to provide a basis for strategy adjustment.

[0195] The dynamic decision-making unit 14 receives the initial inspection strategy and diagnosis results in real time, comprehensively considers environmental changes, fault urgency and task priority, and uses optimization algorithms such as genetic algorithms or Monte Carlo tree search to dynamically adjust the inspection strategy.

[0196] The dynamic decision-making unit 14 generates an optimized inspection strategy and generates control instructions based on the strategy. These instructions include the robot's movement path, speed, inspection sequence, etc., and are sent to the action module 2 through the control processing module 1.

[0197] The action module 2 receives control instructions and drives the robot to perform inspection tasks, including moving to the inspection point, adjusting the posture for inspection, etc.

[0198] During the inspection process, the control processing module 1 continuously monitors the task execution status and environmental changes, and adjusts the control instructions in real time to adapt to new situations.

[0199] After the inspection task is completed, the control processing module 1 organizes and outputs the inspection results, and feeds back to the remote control system to ensure the successful completion of the task.

[0200] Through the above specific implementation methods, the adaptive power inspection method of the present invention can realize efficient and intelligent inspection by robots, ensuring the safe and stable operation of power facilities.

[0201] It is known from common technical knowledge that the present invention can be implemented by other embodiments that do not deviate from its spirit or essential features. Therefore, the above disclosed embodiments are only illustrative in all respects and are not exclusive. All changes within the scope of the present invention or within the scope equivalent to the present invention are included in the present invention.

Claims

1. An adaptive power inspection robot, characterized in that: include: A control processing module (1), an action module (2) and a perception module (3) are arranged on the main frame of the robot; The control processing module (1) is used to receive a task instruction, process the environmental data and the status data of the device to be inspected transmitted by the perception module (3) according to the task instruction, generate a control instruction according to the processing result, and send the control instruction to the action module (2); The action module (2) is used to receive the control instructions of the control processing module (1) and drive the robot to move and inspect according to the control instructions; The sensing module (3) is used to sense the environmental data obtained through environmental sensing and detect the status of the device to be detected to obtain the status data of the device to be detected, and transmit the obtained environmental data and the status data of the device to be detected to the control processing module (1); The control processing module (1) is also used to process the status data of the device to be inspected transmitted by the sensing module (3) to obtain the inspection result of the device to be inspected.

2. The adaptive power inspection robot according to claim 1, characterized in that: The control processing module (1) is integrated with a task unit (11), an autonomous navigation algorithm unit (13), a fault detection and diagnosis unit (12) and a dynamic decision-making unit (14); The task unit (11) is used to receive and analyze task instructions from an operator or a remote control system, obtain task information, and send the task information to the autonomous navigation algorithm unit (13) and the fault detection and diagnosis unit (12); The autonomous navigation algorithm unit (13) is used to perform path planning and obstacle avoidance according to the task information provided by the task receiving unit, in combination with the environmental data and pre-stored map data collected by the perception module (3), generate an initial inspection strategy, and send the initial inspection strategy to the dynamic decision-making unit (14); A fault detection and diagnosis unit (12) is used to use a neural network algorithm to analyze the received inspection strategy and the status data of the device to be inspected collected by the sensing module (3), automatically identify and diagnose the existing faults and potential faults of the device to be inspected, obtain a diagnosis result, record the diagnosis result and send the diagnosis result to the dynamic decision-making unit (14); The dynamic decision-making unit (14) is used to receive in real time the initial inspection strategy of the autonomous navigation algorithm unit (13) and the diagnosis result of the fault detection and diagnosis unit (12), comprehensively consider environmental changes, fault urgency and task priority, use the optimization algorithm to dynamically adjust the inspection strategy, generate an optimized inspection strategy, and send the optimized inspection strategy to the autonomous navigation algorithm unit (13) and the fault detection and diagnosis unit (12), and at the same time generate a control instruction based on the optimized inspection strategy and send it to the action module (2).

3. The adaptive power inspection robot according to claim 2, characterized in that: The control processing module (1) is further integrated with an adaptive learning unit, wherein the adaptive learning unit is provided with an online learning mechanism for receiving in real time the path planning data generated by the autonomous navigation algorithm unit (13), the diagnosis result data provided by the fault detection and diagnosis unit (12), and the optimized inspection strategy data formulated by the dynamic decision-making unit (14); Analyze the accuracy of path planning based on path planning data and obtain planning analysis results; Based on the diagnosis result data, analyzing the accuracy of the diagnosis result to obtain a diagnosis analysis result; Based on the optimized inspection strategy data, analyze the effectiveness of inspection strategy optimization and obtain strategy analysis results; Based on the obtained planning analysis results, diagnosis analysis results and strategy analysis results, an optimization model is generated using a machine learning algorithm to optimize the performance of the autonomous navigation algorithm unit (13), the fault detection and diagnosis unit (12) and the dynamic decision-making unit (14), respectively.

4. The adaptive power inspection robot according to claim 3, characterized in that: The adaptive learning unit is provided with an offline learning mechanism for collecting and storing the optimization model and historical inspection data generated by the online learning mechanism during the non-inspection task of the robot; The collected long-term historical data is analyzed in depth and comprehensively through the optimization model, and the performance of the autonomous navigation algorithm unit (13), the fault detection and diagnosis unit (12) and the dynamic decision-making unit (14) are optimized respectively.

5. The adaptive power inspection robot according to claim 1, characterized in that: It also comprises a communication module and a remote control system, wherein the communication module is used to enable the control processing module (1) to communicate with the remote control system.

6. The adaptive power inspection robot according to claim 1, characterized in that: The perception module (3) comprises: High-definition camera, used to capture visible light images of the equipment to be inspected and identify appearance defects and anomalies of the equipment to be inspected; Infrared thermal imager, used to detect the heat distribution on the surface of the equipment to be inspected and identify hot spots and temperature anomalies; LiDAR sensors are used for 3D modeling of the environment and obstacle detection, assisting robots in precise spatial positioning and navigation; Sound sensor, used to collect sound signals generated when the equipment to be inspected is running, identify the running status of the equipment to be inspected and identify potential faults; Environmental parameter sensors are used to detect temperature, humidity and air pressure in the inspection environment; A data fusion unit is used to integrate the data collected by the high-definition camera, the infrared thermal imager, the laser radar sensor, the sound sensor and the environmental parameter sensor, and provide the control processing module (1) with comprehensive environmental data and status data of the equipment to be inspected.

7. The adaptive power inspection robot according to claim 1, characterized in that: The action module (2) comprises: at least one set of driving devices for executing movement and positioning of the robot; At least one transmission mechanism connected to the driving device for transmitting power and converting the form of motion; At least one set of posture adjustment devices, used to adjust the posture and direction of the robot body or specific parts to meet different inspection requirements; At least one set of stabilizing devices, used to keep the robot stable when it moves or performs tasks; at least one set of end effectors for performing inspection operations; A motion control unit is used to coordinate and control the actions of the drive device, the transmission mechanism, the posture adjustment device, the stabilization device and the end effector.

8. The adaptive power inspection robot according to claim 1, characterized in that: It also includes an interaction module, which is arranged on the main frame and is used to provide an interface for operators to interact with the robot.

9. An adaptive power inspection method, based on an adaptive power inspection robot according to any one of claims 1 to 8, characterized in that: The following steps are involved: The control processing module (1) receives the task instruction, processes the environmental data and the status data of the device to be inspected transmitted by the perception module (3) according to the task instruction, generates a control instruction according to the processing result, and sends the control instruction to the action module (2); The action module (2) receives the control instruction of the control processing module (1), and drives the robot to move and inspect according to the control instruction; The sensing module (3) senses the environment data obtained through environmental sensing and detects the state of the device to be detected to obtain the state data of the device to be detected, and transmits the obtained environmental data and the state data of the device to be detected to the control processing module (1); The control processing module (1) processes the status data of the device to be inspected transmitted by the sensing module (3) to obtain the inspection result of the device to be inspected.

10. The adaptive power inspection method according to claim 9, characterized in that: The control processing module (1) receives the task instruction, processes the environmental data and the status data of the device to be inspected transmitted by the perception module (3) according to the task instruction, generates a control instruction according to the processing result, and sends the control instruction to the action module (2), specifically comprising: The autonomous navigation algorithm unit (13) in the control processing module (1) receives the task information provided by the task unit (11), and simultaneously obtains environmental data and pre-stored map data from the perception module (3); The autonomous navigation algorithm unit (13) performs path planning and obstacle avoidance, generates an initial inspection strategy, and sends the initial inspection strategy to the dynamic decision-making unit (14); A fault detection and diagnosis unit (12) adopts a neural network algorithm to analyze the received inspection strategy and the status data of the device to be inspected collected by the sensing module (3), automatically identifies and diagnoses the existing faults and potential faults of the device to be inspected, obtains a diagnosis result, records the diagnosis result and sends the diagnosis result to the dynamic decision-making unit (14); The dynamic decision-making unit (14) receives in real time the initial inspection strategy of the autonomous navigation algorithm unit (13) and the diagnosis result of the fault detection and diagnosis unit (12), comprehensively considers environmental changes, fault urgency and task priority, uses an optimization algorithm to dynamically adjust the inspection strategy, generates an optimized inspection strategy, and sends the optimized inspection strategy to the autonomous navigation algorithm unit (13) and the fault detection and diagnosis unit (12), and at the same time generates a control instruction based on the optimized inspection strategy and sends it to the action module (2).

Citation Information

Patent Citations

  • Intelligent inspection robot system for transformer substation capable of supporting deep learning acceleration

    CN112123338A

  • System architecture of electric power inspection robot

    CN112230660A

  • Multi-target path optimization method and system considering fault priority

    CN117808183A

  • Distribution network machine patrol operation management system and management method

    CN118014557A

  • Substation intelligent autonomous patrol system and method based on patrol path

    CN118137355A

Cited By

  • Robot rescue emergency control system based on voice interaction

    CN119658721A

  • Electric power inspection robot based on remote control

    CN119704133A

  • A remotely controlled power inspection robot and power inspection method

    CN119704133B

  • Electric power inspection method, device, equipment, storage medium and product

    CN119766369A

  • Power equipment inspection and health scoring robot

    CN120363150A