A substation inspection robot

Abstract

The utility model discloses a transformer substation is with inspection robot, including inspection robot subassembly, including line control chassis, the car shell of fixed in line control chassis top, the six -axis mechanical arm of fixed in car shell top rear end, the detection subassembly and three -color pilot lamp of fixed in car shell top front end and the crash beam of fixed in line control chassis front and back, avoid obstacle subassembly, including fixed in car shell top's 3D laser radar, fixed in car shell front's binocular camera. The utility model avoid obstacle subassembly through adopt a variety of different types' sensor, and sensor distribution is at the different position of robot, can perceive the surrounding environment from multiple angles, multiple dimensions, and adopt multi -sensor data fusion algorithm, and the data of different sensors are fused and handled, thereby comprehensively, accurately perceive the surrounding environment, greatly reduce the probability of collision accident's occurrence, ensure that robot is safe in the process of inspection and avoids the barrier.

Description

Technical Field

[0001] This utility model relates to the field of substation inspection robot technology, specifically a substation inspection robot. Background Technology

[0002] Substation inspection robots are intelligent devices that integrate sensors, navigation technology, and AI algorithms. They are mainly used for automated inspection, fault detection, and data collection of substation equipment, which can significantly improve inspection efficiency and quality and reduce operation and maintenance costs.

[0003] Inspection robots, equipped with lidar and ultrasonic sensors, can autonomously navigate within substations and avoid obstacles. They utilize infrared thermal imaging sensors to monitor equipment temperature, high-definition cameras to inspect appearance, and temperature, humidity, and gas sensors to monitor the environment. By analyzing the collected data, they assess equipment status based on preset models and issue warnings, allowing maintenance personnel to remotely monitor the inspection process and adjust the robot's tasks. However, in actual use, the lack of a reasonable task allocation mechanism leads to unreasonable task allocation for robot inspections, resulting in low efficiency. Furthermore, current inspection robots only use a single sensor for obstacle avoidance, failing to comprehensively and accurately perceive the surrounding environment, thus making them prone to collision accidents. Utility Model Content

[0004] The purpose of this utility model is to provide a substation inspection robot that has the advantages of improving inspection efficiency and obstacle avoidance. It solves the problems in actual use, such as the lack of a reasonable task allocation mechanism, which leads to unreasonable task allocation and low efficiency. Moreover, the current inspection robots only use a single sensor for obstacle avoidance, which cannot fully and accurately perceive the surrounding environment and is prone to collision accidents.

[0005] To achieve the above objective, the present utility model provides the following technical solution: An inspection robot for substation, comprising an inspection robot assembly, which includes a drive-by-wire chassis, a vehicle shell fixed on the top of the drive-by-wire chassis, a six-axis robotic arm fixed at the rear end of the top of the vehicle shell, a detection assembly and a three-color indicator light fixed at the front end of the top of the vehicle shell, and anti-collision beams fixed on the front and back sides of the drive-by-wire chassis; an obstacle avoidance assembly, which includes a 3D LiDAR fixed on the top of the vehicle shell, a binocular camera fixed on the front side of the vehicle shell, millimeter-wave radars fixed on both sides of the front and back of the vehicle shell, ultrasonic radars fixed on both sides of the vehicle shell, an RGB-D camera fixed at the end of the six-axis robotic arm, a multi-dimensional force sensor fixed at the connection between the six-axis robotic arm and the actuator, and an IMU sensor fixed on the top of the drive-by-wire chassis; a control mainboard, which integrates a 128-core GPU and a deep learning accelerator, has a built-in task collaborative allocation algorithm and anomaly detection mechanism, and is fixed on the top of the drive-by-wire chassis; a communication assembly, which includes a communication module fixed on the top of the drive-by-wire chassis, and a transmission antenna fixed at the rear end of the vehicle shell, and the communication assembly adopts a dual-queue priority mechanism; a photovoltaic power generation assembly, which includes a solar panel fixed at the rear end of the vehicle shell, a storage battery, an MPPT controller and an inverter fixed on the top of the drive-by-wire chassis, and the storage battery, the MPPT controller and the inverter are all fixed on the top of the drive-by-wire chassis, the photovoltaic power generation assembly is used to supply electric energy.

[0006] Preferably, the task collaborative allocation algorithm of the control mainboard is based on the task allocation model of the improved Hungarian algorithm, takes equipment priority, robot endurance and path cost into consideration, and introduces game theory to achieve load balancing, with its formula as follows: .

[0007] Preferably, the dual-queue priority mechanism of the communication assembly includes a real-time queue and a non-real-time queue, the real-time queue is used for control instructions and emergency stop signals, and the non-real-time queue is used for collected images and environmental data. The communication module includes a millimeter-wave transmission module and a 5G private network communication module, which adopts the Raft consensus algorithm to realize seamless transfer of control right, with a fault recovery time of less than 500 ms. The millimeter-wave module adopts the IEEE 802.11ay standard, and realizes directional transmission through beamforming.

[0008] Preferably, the obstacle avoidance assembly adopts a multi-sensor data fusion algorithm, with its formula as follows:

[0009]

[0010] Herein, is the weight of visual data, is the weight of temperature data, is the weight of ultrasonic data, .

[0011] Preferably, the anomaly detection mechanism of the control motherboard adopts a federated learning framework and establishes a device multimodal feature database, and the main control chip of the control motherboard adopts NVIDIA Jetson AGX Xavier.

[0012] Preferably, the detection assembly includes a servo cylinder fixed to the top of the vehicle body, a two-axis rotating gimbal fixed to the output end of the servo cylinder, a visible light camera and an infrared thermal imager fixed to both sides of the two-axis rotating gimbal, and an SF6 sensor fixed to one side of the vehicle body.

[0013] Preferably, the output terminals of the obstacle avoidance component, the visible light camera, the infrared thermal imager, and the SF6 sensor are all connected to the input terminals of the control motherboard, and the output terminals of the control motherboard are respectively connected to the input terminals of the wire-controlled chassis, the six-axis robotic arm, the servo cylinder, the two-axis rotating gimbal, and the three-color indicator light.

[0014] Preferably, the output and input terminals of the communication module are interconnected with the output and input terminals of the control motherboard, and the output and input terminals of the communication module are also interconnected with the output and input terminals of the transmission antenna.

[0015] Compared with the prior art, the beneficial effects of this utility model are as follows:

[0016] 1. This utility model uses a task collaborative allocation algorithm based on an improved Hungarian algorithm built into the control motherboard. It comprehensively considers equipment priority, robot endurance and path cost, and introduces game theory to achieve load balancing. It can reasonably allocate inspection tasks and improve inspection efficiency. The control motherboard also uses a federated learning framework to establish a multimodal feature database of equipment, which can detect abnormal conditions of equipment in a timely and accurate manner, making it convenient for maintenance personnel to handle them in a timely manner and ensuring the safe and stable operation of the substation.

[0017] 2. This utility model obstacle avoidance component employs various types of sensors, including 3D LiDAR, binocular cameras, millimeter-wave radar, ultrasonic radar, RGB-D cameras, multi-dimensional force sensors, and IMU sensors. These sensors are distributed at different locations on the robot, enabling it to perceive the surrounding environment from multiple angles and dimensions. Furthermore, it uses a multi-sensor data fusion algorithm to fuse data from different sensors, fully leveraging the advantages of each sensor to comprehensively and accurately perceive the surrounding environment, greatly reducing the probability of collision accidents and ensuring the robot's safe obstacle avoidance during inspections. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the main structure of this utility model;

[0019] Figure 2 This is a schematic diagram of the rear view structure of this utility model;

[0020] Figure 3 This is a schematic diagram showing the connection structure between the wire-controlled chassis and the control motherboard of this utility model;

[0021] Figure 4 This is a schematic diagram of the system flow of this utility model;

[0022] Figure 5 This is a flowchart illustrating the obstacle avoidance component of this utility model.

[0023] Figure 6 This is a schematic diagram of the process of the photovoltaic power generation module of this utility model.

[0024] In the diagram: 100, Inspection robot component; 110, Wired chassis; 120, Vehicle body; 130, Six-axis robotic arm; 140, Detection component; 141, Servo cylinder; 142, Two-axis rotating gimbal; 143, Visible light camera; 144, Infrared thermal imager; 145, SF6 sensor; 150, Anti-collision beam; 160, Three-color indicator light; 200, Obstacle avoidance component; 210, 3D LiDAR; 220, Binocular camera; 230, Millimeter-wave radar; 240, Ultrasonic radar; 250, RGB-D camera; 260, Multi-dimensional force sensor; 270, IMU sensor; 300, Control motherboard; 400, Communication component; 410, Communication module; 420, Transmission antenna; 500, Photovoltaic power generation component; 510, Solar panel; 520, Battery; 530, MPPT controller; 540, Inverter. Detailed Implementation

[0025] The technical solutions of the present utility model will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present utility model, and not all embodiments. Based on the embodiments of the present utility model, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of the present utility model.

[0026] Example 1

[0027] like Figure 1-6The image shows the first embodiment of this utility model, which provides a substation inspection robot, including an inspection robot assembly 100, comprising a wire-controlled chassis 110, a vehicle body 120 fixed to the top of the wire-controlled chassis 110, a six-axis robotic arm 130 fixed to the rear end of the top of the vehicle body 120, a detection assembly 140 and a three-color indicator light 160 fixed to the front end of the top of the vehicle body 120, and anti-collision beams 150 fixed to the front and back of the wire-controlled chassis 110, and an obstacle avoidance assembly 200, including a 3D LiDAR 210 fixed to the top of the vehicle body 120, a binocular camera 220 fixed to the front of the vehicle body 120, millimeter-wave radars 230 fixed to the front and back sides of the vehicle body 120, ultrasonic radars 240 fixed to the sides of the vehicle body 120, an RGB-D camera 250 fixed to the end of the six-axis robotic arm 130, and a multi-dimensional force sensor fixed to the connection between the six-axis robotic arm 130 and the actuator. 260, and an IMU sensor 270 fixed to the top of the drive-by-wire chassis 110; a control motherboard 300, which integrates a 128-core GPU and a deep learning accelerator, and has a built-in task collaborative allocation algorithm and anomaly detection mechanism, and the control motherboard 300 is fixed to the top of the drive-by-wire chassis 110; a communication component 400, including a communication module 410 fixed to the top of the drive-by-wire chassis 110 and a transmission antenna 420 fixed to the rear end of the vehicle body 120, and the communication component 400 adopts a dual-queue priority mechanism; a photovoltaic power generation component 500, including a solar panel 510 fixed to the rear end of the vehicle body 120, a battery 520 fixed to the top of the drive-by-wire chassis 110, an MPPT controller 530 and an inverter 540, and the battery 520, MPPT controller 530 and inverter 540 are all fixed to the top of the drive-by-wire chassis 110, and the photovoltaic power generation component 500 is used to provide electrical energy.

[0028] like Figure 1-6As shown, the obstacle avoidance component 200 employs various types of sensors, including a 3D LiDAR 210, a binocular camera 220, a millimeter-wave radar 230, an ultrasonic radar 240, an RGB-D camera 250, a multi-dimensional force sensor 260, and an IMU sensor 270. These sensors are distributed at different locations on the robot, enabling it to perceive the surrounding environment from multiple angles and dimensions. Furthermore, a multi-sensor data fusion algorithm is used to fuse data from different sensors, fully leveraging the advantages of each sensor to comprehensively and accurately perceive the surrounding environment, significantly reducing the probability of collisions and effectively avoiding obstacles. The output of the obstacle avoidance component 200 is connected to the input of the control motherboard 300. The control motherboard 300 can adjust the robot's actions in a timely manner based on the data fed back by the sensors, ensuring that the robot safely avoids obstacles during the inspection process. The inspection robot component 100 is equipped with a six-axis robotic arm 130. The end effector of the six-axis robotic arm 130 can be equipped with corresponding actuators according to the inspection requirements, thereby realizing complex movements, expanding the detection range and operation capabilities, and facilitating the inspection and operation of equipment at different positions and angles within the substation. The control motherboard 300 integrates a 128-core GPU and a deep learning accelerator, and the main control chip uses NVIDIA. The Jetson AGX Xavier boasts powerful computing capabilities, enabling rapid processing of large amounts of complex data. The communication component 400 employs a dual-queue priority mechanism to separate real-time and non-real-time queues, ensuring timely transmission of critical instructions and guaranteeing the robot's safe and stable operation. The communication module 410 utilizes millimeter-wave transmission and a 5G private network communication module, employing the Raft consensus algorithm for seamless control transfer with a fault recovery time of <500ms, ensuring reliable and stable communication and minimizing the impact of communication failures on inspection tasks. The photovoltaic power generation component 500 provides power to the robot, while the solar panels 510 convert solar energy into electrical energy. The MPPT controller 530 performs maximum power point tracking, improving solar energy utilization efficiency. The inverter 540 converts DC power to AC power to supply various robot components. The battery 520 stores excess energy to ensure normal operation even in low-light conditions. Through a reasonable task allocation algorithm and multi-sensor fusion obstacle avoidance design, the system effectively addresses issues of unreasonable task allocation and incomplete obstacle avoidance, improving inspection efficiency and safety.

[0029] Example 2

[0030] Reference Figure 1-3 This is the second embodiment of the present invention, which is based on the previous embodiment.

[0031] In this embodiment, the task coordination allocation algorithm of the control motherboard 300 is based on the task allocation model of the improved Hungarian algorithm, taking into account device priority, robot endurance and path cost, and introducing game theory to achieve load balancing. Its formula is as follows: .

[0032] The communication component 400 has a dual-queue priority mechanism consisting of a real-time queue and a non-real-time queue. The real-time queue is for control commands and emergency stop signals, while the non-real-time queue is for acquiring images and environmental data. The communication module 410 is a millimeter-wave transmission module and a 5G private network communication module. It uses the Raft consensus algorithm to achieve seamless transfer of control, with a fault recovery time of <500ms. The millimeter-wave module adopts the IEEE 802.11ay standard and achieves directional transmission through beamforming.

[0033] The obstacle avoidance component 200 employs a multi-sensor data fusion algorithm, the formula of which is:

[0034]

[0035] in, For visual data weights, Assigning weights to temperature data, Weighting of ultrasound data, .

[0036] The anomaly detection mechanism of the control motherboard 300 adopts a federated learning framework and establishes a device multimodal feature database. The main control chip of the control motherboard 300 adopts NVIDIA Jetson AGX Xavier.

[0037] like Figure 1-3 As shown, the control motherboard 300 integrates a task collaborative allocation algorithm based on an improved Hungarian algorithm task allocation model. This algorithm considers factors such as equipment priority, robot endurance, and path cost, and also introduces game theory to achieve load balancing. This allows for the reasonable allocation of tasks to the inspection robot, avoiding inefficiency caused by unreasonable task allocation. Important equipment will be prioritized for inspection. At the same time, the robot's endurance and path planning will be combined to enable the robot to utilize resources more efficiently when completing tasks. The control motherboard 300 also adopts a federated learning framework and establishes a multimodal feature database of equipment, which can detect abnormal conditions of equipment in a timely manner. During task allocation, task priority can be adjusted according to the abnormal state of the equipment to further optimize task allocation and improve inspection efficiency.

[0038] Example 3

[0039] Reference Figure 1 , 2 4 and 5 are the third embodiment of this utility model, which is based on the first two embodiments.

[0040] In this embodiment, the detection component 140 includes a servo cylinder 141 fixed to the top of the vehicle body 120, a two-axis rotating gimbal 142 fixed to the output end of the servo cylinder 141, a visible light camera 143 and an infrared thermal imager 144 fixed to both sides of the two-axis rotating gimbal 142, and an SF6 sensor 145 fixed to one side of the vehicle body 120.

[0041] The outputs of the obstacle avoidance component 200, the visible light camera 143, the infrared thermal imager 144, and the SF6 sensor 145 are all connected to the inputs of the control motherboard 300. The outputs of the control motherboard 300 are connected to the inputs of the wired chassis 110, the six-axis robotic arm 130, the servo cylinder 141, the two-axis rotating gimbal 142, and the three-color indicator light 160, respectively.

[0042] The output and input terminals of the communication module 410 are interconnected with the output and input terminals of the control motherboard 300, and the output and input terminals of the communication module 410 are also interconnected with the output and input terminals of the transmission antenna 420.

[0043] like Figure 1 , 2 As shown in Figures 4 and 5, the wire-controlled chassis 110 provides the robot with mobility, allowing it to move within the substation according to a preset route or real-time commands. The chassis 120 protects the internal equipment. The servo cylinder 141 in the detection assembly 140 can adjust the height of the two-axis rotating gimbal 142. The two-axis rotating gimbal 142 can drive the visible light camera 143 and the infrared thermal imager 144 to perform multi-angle image acquisition and temperature detection. The SF6 sensor 145 is used to detect the concentration of SF6 gas in the substation. The six-axis robotic arm 130 can perform some specific operational tasks, such as opening and closing switches, by installing corresponding end effectors. The three-color indicator light 160 is used to display the robot's working status.

[0044] When in use, the photovoltaic power generation module 500 provides power to the robot, the solar panel 510 converts solar energy into electrical energy, and the maximum power point tracking is performed by the MPPT controller 530 to improve the utilization efficiency of solar energy. The inverter 540 converts DC power into AC power to power the various components of the robot, and the battery 520 is used to store excess electrical energy to ensure that the robot can still work normally when there is insufficient light.

[0045] The wire-controlled chassis 110 provides the robot with mobility, allowing it to move within the substation according to a preset route or real-time commands. The chassis 120 protects the internal equipment. The servo cylinder 141 in the detection assembly 140 can adjust the height of the two-axis rotating gimbal 142. The two-axis rotating gimbal 142 can drive the visible light camera 143 and the infrared thermal imager 144 to perform multi-angle image acquisition and temperature detection. The SF6 sensor 145 is used to detect the concentration of SF6 gas in the substation. The six-axis robotic arm 130 can perform some specific operational tasks. The three-color indicator light 160 is used to display the robot's working status.

[0046] The communication component 400 is responsible for the information interaction between the robot and the outside world. The communication module 410 adopts a millimeter wave transmission module and a 5G private network communication module. It uses the Raft consensus algorithm to achieve seamless transfer of control and ensure the stability and reliability of communication. The dual queue priority mechanism divides the communication data into real-time queues and non-real-time queues to ensure that important information can be transmitted first. The transmission antenna 420 is used to enhance the transmission distance and strength of the communication signal.

[0047] The obstacle avoidance component 200 achieves omnidirectional obstacle detection through multiple sensors. The 3D LiDAR 210 can acquire real-time 3D information about the surrounding environment. The binocular camera 220 is used to identify objects and obstacles in front. The millimeter-wave radar 230 and ultrasonic radar 240 detect obstacles at different distances and in different scenarios. The RGB-D camera 250 is mounted at the end of the six-axis robotic arm 130 to assist the robotic arm in obstacle avoidance during operation. The multi-dimensional force sensor 260 is used to detect the force applied when the robotic arm contacts an object, preventing damage due to excessive force. The IMU sensor 270 is used to detect the robot's posture and... The motion status data from these sensors is processed through a multi-sensor data fusion algorithm to improve the accuracy and reliability of obstacle avoidance. The control motherboard 300 is the core control unit of the robot, integrating a 128-core GPU and a deep learning accelerator, which can quickly process large amounts of data. Its built-in task collaborative allocation algorithm is based on a task allocation model of an improved Hungarian algorithm, which comprehensively considers equipment priority, robot endurance and path cost, and introduces game theory to achieve load balancing, thereby rationally allocating inspection tasks. The anomaly detection mechanism adopts a federated learning framework and establishes a multimodal feature database of equipment. The main control chip NVIDIA Jetson AGX Xavier analyzes the collected data, thereby enabling timely detection of anomalies in substation equipment. Through the collaborative work of various components, autonomous inspection, obstacle avoidance, control, communication and power supply functions are realized, enabling efficient and accurate completion of substation inspection tasks.

[0048] All standard parts used in this application can be purchased from the market, and can be customized according to the description and drawings. The specific connection methods of each part adopt conventional methods such as bolts, rivets, and welding that are mature in the prior art. The machinery, parts and equipment adopt conventional models in the prior art. The control method is automatic control through a controller. The control circuit of the controller can be implemented by simple programming by those skilled in the art and is common knowledge in the field. Since this application is mainly used to protect mechanical devices, the control method and circuit connection will not be explained in detail in this application.

[0049] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0050] Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the present invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A substation inspection robot, characterized in that: include, The inspection robot assembly (100) includes a wire-controlled chassis (110), a vehicle body (120) fixed to the top of the wire-controlled chassis (110), a six-axis robotic arm (130) fixed to the rear end of the top of the vehicle body (120), a detection assembly (140) and a three-color indicator light (160) fixed to the front end of the top of the vehicle body (120), and anti-collision beams (150) fixed to the front and back of the wire-controlled chassis (110). The obstacle avoidance assembly (200) includes a 3D LiDAR (210) fixed to the top of the vehicle body (120), a binocular camera (220) fixed to the front of the vehicle body (120), a millimeter-wave radar (230) fixed to the front and back sides of the vehicle body (120), an ultrasonic radar (240) fixed to the sides of the vehicle body (120), an RGB-D camera (250) fixed to the end of the six-axis robotic arm (130), a multi-dimensional force sensor (260) fixed to the connection between the six-axis robotic arm (130) and the actuator, and an IMU sensor (270) fixed to the top of the drive-by-wire chassis (110). A control motherboard (300) is provided, which integrates a 128-core GPU and a deep learning accelerator, and is fixed to the top of the wire-controlled chassis (110). The communication component (400) includes a communication module (410) fixed to the top of the drive-by-wire chassis (110) and a transmission antenna (420) fixed to the rear end of the vehicle body (120). The photovoltaic power generation module (500) includes a solar panel (510) fixed to the rear end of the vehicle body (120), a battery (520) fixed to the top of the drive-by-wire chassis (110), an MPPT controller (530) and an inverter (540), and the battery (520), MPPT controller (530) and inverter (540) are all fixed to the top of the drive-by-wire chassis (110). The photovoltaic power generation module (500) is used to provide electrical energy.

2. The substation inspection robot according to claim 1, characterized in that: The detection assembly (140) includes a servo cylinder (141) fixed to the top of the vehicle body (120), a two-axis rotating gimbal (142) fixed to the output end of the servo cylinder (141), a visible light camera (143) and an infrared thermal imager (144) fixed to both sides of the two-axis rotating gimbal (142), and an SF6 sensor (145) fixed to one side of the vehicle body (120).

3. The substation inspection robot according to claim 2, characterized in that: The output terminals of the obstacle avoidance component (200), visible light camera (143), infrared thermal imager (144), and SF6 sensor (145) are all connected to the input terminals of the control motherboard (300). The output terminals of the control motherboard (300) are respectively connected to the input terminals of the wire-controlled chassis (110), six-axis robotic arm (130), servo cylinder (141), two-axis rotating gimbal (142), and three-color indicator light (160).

4. The substation inspection robot according to claim 1, characterized in that: The output and input terminals of the communication module (410) are interconnected with the output and input terminals of the control motherboard (300), and the output and input terminals of the communication module (410) are also interconnected with the output and input terminals of the transmission antenna (420).

5. The substation inspection robot according to claim 1, characterized in that: The communication module (410) is a millimeter-wave transmission module and a 5G private network communication module.

6. The substation inspection robot according to claim 1, characterized in that: The main control chip of the control motherboard (300) is NVIDIA Jetson AGX Xavier.

Images