Substation inspection robot and intelligent inspection system thereof

Abstract

The invention provides a transformer substation inspection robot which is used for intelligent inspection tasks in a transformer substation environment and comprises a four-wheel-drive movable chassis, a driving device and a controller. The chassis is provided with a single-line laser radar, a multi-line laser radar, a depth camera, a binocular camera, a 5G communication module, a power management module, an ultrasonic sensor, a temperature sensor and a high-precision holder used for multi-sensor fusion SLAM and automatic inspection tasks. The invention further relates to an intelligent inspection system which supports two inspection modes of automatic inspection and remote control inspection and can be switched between the two modes at will: in the process of executing the multi-target-point automatic inspection task, if certain abnormal equipment needs to be monitored, the current automatic inspection task can be paused, and the mode is switched to the remote control inspection mode; the robot is remotely controlled to quickly arrive at the position of the abnormal equipment; after the task is completed, the robot can be switched back to the automatic inspection mode, the previous automatic inspection task continues to be executed, and the automatic inspection robot has the advantages of being high in automation and intelligentization.

Description

Technical Field

[0001] The present invention relates to the technical field of robot unmanned inspection, and in particular to a substation inspection robot and an intelligent inspection system thereof. Background Art

[0002] As the core hub of the power system, substations are responsible for the transmission and distribution of electric energy. Their safe and stable operation is crucial to ensuring social and economic activities and the daily lives of residents. With the continuous growth of electricity demand, the scale and complexity of substations are also increasing, which makes regular inspections of substation equipment particularly important. Inspections can timely discover and deal with potential safety hazards, prevent accidents, and ensure the reliability of the power grid and the continuity of power supply.

[0003] However, traditional substation inspections mainly rely on manual work, which has many limitations. Manual inspections are labor-intensive and inefficient, and they are carried out in harsh environments such as high voltage and high temperature, so safety is difficult to guarantee. In addition, manual inspections are greatly affected by subjective factors, and there may be cases of missed inspections or misjudgments, and they are difficult to carry out under extreme weather conditions. These problems limit the quality and efficiency of inspections and increase the operating risks of power systems. With the continuous development of robotics technology, inspection robots have become an effective solution to improve the safety and efficiency of substation inspections. However, inspection robots currently have high requirements for sensor accuracy, data processing capabilities, and equipment safety performance, poor modeling capabilities for complex environments, and the adjustment range of the cameras used for detection is relatively limited, which may result in blind spots in the field of vision, resulting in the inability to cover all areas that need to be inspected.

[0004] Therefore, the invention of a substation inspection robot and intelligent inspection system can realize the automatic and intelligent inspection of substation equipment by integrating multiple sensor technologies, communication technologies and artificial intelligence algorithms, improve the efficiency and accuracy of inspections, and reduce the safety risks of manual inspections. It has important practical significance and broad application prospects. It can not only improve the intelligence level of substations, but also provide strong guarantees for the safe and stable operation of power systems, and has extremely high practical value and market potential. Summary of the invention

[0005] In view of the problems existing in the prior art, the purpose of the present invention is to provide a substation inspection robot and an intelligent inspection system thereof, which integrates a variety of sensor technologies, communication technologies and artificial intelligence algorithms, can realize automated and intelligent inspection of substation equipment, improve the efficiency and accuracy of inspection, and reduce the safety risks of manual inspection. To achieve the above purpose, the technical solution of the present invention is as follows:

[0006] A substation inspection robot is used for intelligent inspection tasks in a substation environment, including:

[0007] The robot body includes a four-wheel drive mobile chassis, on which are installed ultrasonic sensors, single-line laser radars, temperature sensors, binocular cameras, depth cameras, multi-line laser radars, 5G communication modules, power management modules, and high-precision gimbals for multi-sensor fusion SLAM and automatic inspection tasks; the depth cameras and multi-line laser radars are integrated on the high-precision gimbals and move synchronously with them; ultrasonic sensors are distributed at the front and rear ends of the four-wheel drive mobile chassis;

[0008] A driving device, arranged on the four-wheel drive mobile chassis, for driving the chassis to move;

[0009] The controller is installed in the four-wheel drive mobile chassis shell and communicates with the driving device, ultrasonic sensor, single-line laser radar, temperature sensor, binocular camera, depth camera, multi-line laser radar, 5G communication module and high-precision gimbal respectively.

[0010] Furthermore, the single-line laser radar is installed at the front end of the four-wheel drive mobile chassis and is used for laser SLAM, navigation positioning and path planning in the intelligent inspection system.

[0011] Furthermore, the binocular camera, depth camera, multi-line laser radar and high-precision gimbal are all electrically connected to the controller through corresponding power management modules.

[0012] Furthermore, the binocular camera and the depth camera are powered by a 48V to 5V power management module, and data is transmitted with the controller through a Type-C to USB interface; a double-layer bracket is installed on the top of the four-wheel drive mobile chassis, including a lower bracket and an upper bracket, and the binocular camera is fixedly installed at the front end of the lower bracket, which is used for continuous monitoring and detection of obstacles ahead on the robot's route; a high-precision gimbal is installed on the upper bracket, and a small bracket is installed on the top of the high-precision gimbal, and the depth camera is installed above the small bracket for intelligent detection and reading of substation instruments, and together with the multi-line laser radar and its built-in IMU, completes multi-sensor fusion SLAM.

[0013] Furthermore, the multi-line laser radar is powered by a 48V to 12V power management module, and data is transmitted with the controller through an M12 aviation plug to RJ45 interface. The multi-line laser radar is installed on the top of a small bracket and completes multi-sensor fusion SLAM together with the depth camera.

[0014] Furthermore, the high-precision gimbal is powered by a 48V to 24V power management module, and data is transmitted with the controller through an RS485 to USB interface. The high-precision gimbal, the depth camera and the multi-line lidar jointly complete multi-sensor fusion SLAM and intelligent detection and reading of substation instruments.

[0015] It also relates to an intelligent inspection system, comprising the substation inspection robot described in any of the above technical solutions and a robot control host arranged in a central control room, wherein the robot control host is communicatively connected with the controller of the substation inspection robot, and the robot control host is used to issue robot control instructions to the controller of the substation inspection robot to control the entire intelligent inspection process, further control the two inspection modes of automatic inspection and remote control inspection of the substation inspection robot, and can switch between the two modes at will; at the same time, it receives various detection data and status data sent back by the controller of the substation inspection robot, and can automatically analyze the detection data; the information contained in the detection data is the information collected by the binocular camera, depth camera, single-line laser radar, multi-line laser radar, ultrasonic sensor, temperature sensor and high-precision gimbal used for multi-sensor fusion SLAM and automatic inspection tasks; the information contained in the status data is the motion state of the inspection robot;

[0016] The intelligent inspection system further comprises a remote control receiving device in communication connection with the robot controller, the remote control receiving device being used to receive a remote control instruction sent by a remote control device and send the remote control instruction to the robot controller;

[0017] Furthermore, the remote control device mainly includes a remote control device for driving the chassis to move and a remote control device for driving the pan / tilt head to rotate, and the remote control device is a mobile device such as a remote control handle, a remote control keyboard or a mobile phone;

[0018] Furthermore, during the entire intelligent inspection process, obstacles are continuously monitored and detected by the robot's front-end binocular camera and the front and rear ultrasonic sensors for real-time obstacle avoidance.

[0019] Furthermore, the intelligent inspection process includes the following steps:

[0020] Step 1: System initialization, substation inspection robot self-inspection, high-precision PTZ self-inspection;

[0021] Step 2: By controlling the movement of the four-wheel drive mobile chassis and the rotation of the high-precision gimbal, the depth camera, multi-line laser radar and its built-in IMU are used to perform multi-sensor fusion SLAM using the R3LIVE algorithm to reconstruct the surrounding environment in three dimensions. The constructed three-dimensional dense point cloud map with RGB color information is transmitted back to the robot control host in the central control room for overall observation and control of the working environment.

[0022] Step 3: By controlling the movement of the four-wheel drive mobile chassis and the rotation of the high-precision gimbal, a single-line laser radar is used to perform laser SLAM using the Cartographer algorithm to build a two-dimensional map of the operating environment for navigation, positioning and path planning;

[0023] Step 4: The robot control host in the central control room selects multiple points in the constructed two-dimensional map and gives them to the controller of the substation inspection robot as multiple target positions for subsequent automatic inspection;

[0024] Step 5: During the automatic inspection process, the amcl function package in ROS is used to locate the robot in the two-dimensional map in real time; the proposed LH global path planning algorithm is used to plan the global path, and the optimal route from the robot to the next target position is calculated as the global route of the robot; the Dynamic Window Approaches algorithm is used to search for multiple paths for avoidance and travel, and the optimal path is selected based on various evaluation criteria, and dynamic obstacle avoidance is achieved;

[0025] Step 6. Whenever a patrol target position is reached, the depth camera follows the high-precision gimbal to perform an all-round scan according to the predetermined trajectory, and uses the target detection algorithm to perform intelligent detection on the current equipment instrument until a clear image of the area where the instrument is located is obtained. The instrument image in the image of the area where the instrument is located is extracted based on the target detection network model, and the numerical area in the instrument image is segmented. The value of the instrument is determined according to the segmented image, thereby completing the intelligent detection and reading of the instrument during the intelligent patrol process; the detected image and the read value are transmitted back to the robot control host in the central control room for analysis, and an alarm is issued if it is higher or lower than its threshold;

[0026] Step 7: During the execution of the multi-target automatic inspection task, if the status of a certain substation equipment needs to be locked and monitored, the current automatic inspection task can be suspended and switched to the remote control inspection mode. By remotely controlling the movement of the four-wheel drive mobile chassis and the rotation of the high-precision pan / tilt, the robot can be controlled to quickly reach the location of the abnormal equipment, and the equipment can be checked and the alarm information can be verified in time.

[0027] Step 8. After completing the task, the robot can switch back to the automatic inspection mode and continue to perform the previous automatic inspection task.

[0028] Furthermore, in step 5, the basic principle of the LH global path planning algorithm is to use the current position of the automatic inspection robot as the initial node, evaluate the expansion node of the node, determine the best child node in the expansion through the evaluation function, and then evaluate the expandable nodes of this child node again, and iterate in turn until there is a target node in the expandable nodes of a certain node; the evaluation function of the LH global path planning algorithm is:

[0029] f(n)=g(n)+α(d,t)·h(n)

[0030] Among them, h(n) is the heuristic function, which is the optimal path evaluation value between the current node and the target node. The calculation of h(n) is directly related to the effect of the LH global path planning algorithm; f(n) represents the cost from the starting node through the intermediate node to the target node; g(n) is the cost function, which represents the cost value between the initialization node and the current node, which is equivalent to the path length from the starting position to the current position. When the robot moves to a certain n node, it represents the cost corresponding to the optimal route from the starting node to the current position; α(d,t) is the dynamic adjustment factor, and its theoretical formula is:

[0031]

[0032] The path density weight represents the obstacle density gradient around node n, represents the maximum density gradient in the global environment; λ in the time decay factor is the decay rate; the global adjustment factor δ is the heuristic global adjustment ratio, which is used to control the strength of the overall dynamic adjustment factor to adapt to the substation environment.

[0033] Compared with the prior art, the present invention has the following beneficial technical effects:

[0034] The substation inspection robot of the present invention is equipped with a pan-tilt platform and a variety of sensor equipment, integrating a variety of sensor technologies, communication technologies and artificial intelligence algorithms. The inspection robot has a stronger ability to model complex environments, a stronger obstacle avoidance ability, a higher adaptability to road surfaces, and a smoother walking. It greatly reduces the work pressure of the staff and truly realizes all-weather, unmanned intelligent inspection of substations.

[0035] The intelligent inspection system of the present invention constructs a three-dimensional dense point cloud map with RGB color information for the robot working environment through the LiDAR-Inertial-Visual fusion SLAM algorithm, thereby enhancing the robot's modeling ability for complex environments; by using ultrasonic sensors, laser radars, binocular cameras and multiple sensors to perform dynamic obstacle avoidance during the robot's working process, the robot's obstacle avoidance ability in complex environments is improved; by programming and controlling the high-precision pan / tilt, the camera's inspection field of view covers the area to be detected in all directions, thereby making the robot's detection and reading ability of the substation instrument more accurate; by issuing control instructions to the robot controller through the remote robot control host, the substation inspection robot can realize two inspection modes: automatic inspection and remote control inspection, and switch between the two modes according to task requirements, so that the robot has a more intelligent autonomous inspection ability. In the complex environment of the substation, by utilizing the proposed LH global path planning algorithm, it can achieve rapid expansion in sparse areas, accurate planning in complex areas, and dynamically adapt to time changes, so that the substation inspection robot can complete low-time and high-efficiency path planning tasks. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 This is a three-dimensional schematic diagram of a substation inspection robot according to an embodiment of the present invention;

[0037] Figure 2 for Figure 1 The schematic diagram of the main view of the substation inspection robot shown;

[0038] Figure 3 for Figure 1 A schematic side view of a substation inspection robot is shown;

[0039] Figure 4 for Figure 1 A schematic top view of a substation inspection robot is shown;

[0040] In the figure,

[0041] 1- Four-wheel drive mobile chassis; 2- Ultrasonic sensor; 3- Single-line laser radar; 4- Temperature sensor; 5- Binocular camera; 6- Depth camera; 7- Multi-line laser radar; 8- 48V to 5V power management module, 9- 48V to 12V power management module, 10- 48V to 24V power management module; 11- High-precision gimbal; 12- Lower bracket; 13- Upper bracket; 14- Small bracket. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical solution and advantages of the present invention more clear, the substation inspection robot and intelligent inspection system of the present invention are further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention and are not used to limit the present invention.

[0043] like Figures 1 to 4 As shown, a substation inspection robot according to an embodiment of the present invention is used for intelligent inspection tasks in a substation environment. The substation inspection robot comprises a four-wheel drive mobile chassis 1, a drive device and a controller.

[0044] The robot body, i.e. the four-wheel drive mobile chassis 1, is equipped with an ultrasonic sensor 2, a single-line laser radar 3, a temperature sensor 4, a binocular camera 5, a depth camera 6, a multi-line laser radar 7, a 5G communication module, a power management module and a high-precision gimbal 11 for multi-sensor fusion SLAM and automatic inspection tasks.

[0045] The driving device is arranged in the four-wheel drive mobile chassis 1 to drive the chassis to move. The peak current of the driver can reach more than three times the rated current. It has high dynamic performance, large peak torque, smooth low-speed operation, strong anti-interference ability, and uses CAN and RS485 interfaces to control the driver.

[0046] The controller is installed in the shell of the four-wheel drive mobile chassis 1, and is respectively connected to the ultrasonic sensor 2, the single-line laser radar 3, the temperature sensor 4, the binocular camera 5, the depth camera 6, the multi-line laser radar 7, the 5G communication module, the power management module and the high-precision gimbal through a double-layer bracket and a small bracket 14 composed of aluminum profiles, stainless steel plates and copper columns.

[0047] Specifically, for the double-layer bracket above the chassis, 2020 European standard aluminum profile is used as the overall frame of the bracket, and 2028 angle code is used as the connecting part, and T-type screws and flange nuts are used for fixing; the lower bracket 12 is fixed on the aluminum profile frame with a stainless steel plate of 220mm*200mm size, and is fixed with T-type screws and flange nuts; the upper bracket 13 is fixed on the aluminum profile frame with a stainless steel plate of 240mm*260mm size, and is also fixed with T-type screws and flange nuts; and for the small bracket 14 fixed above the high-precision gimbal 11, two layers of stainless steel plates are used to fix the depth camera 6 and the multi-line laser radar 7 respectively, and the two layers of stainless steel plates are connected and fixed with M4*40 copper columns.

[0048] The single-line laser radar 3 is installed at the front end of the four-wheel drive mobile chassis 1 and is used for laser SLAM, navigation positioning and path planning in the intelligent inspection system.

[0049] The binocular camera 5, the depth camera 6, the multi-line laser radar 7 and the high-precision gimbal 11 are all electrically connected to the chassis controller through corresponding power management modules.

[0050] Specifically, the binocular camera 5 is connected to a 5V power supply hub, and the hub is powered by a DC-DC 48V to 5V power management module 8. Data is transmitted with the chassis controller through a Type-C to USB interface. The binocular camera 5 is fixedly installed at the front end of the lower bracket 12 and is used for continuous monitoring and detection of obstacles ahead on the robot's travel route.

[0051] Specifically, the depth camera 6 is also connected to the 5V power supply hub, and the hub is powered by the DC-DC 48V to 5V power management module 8. Data is transmitted with the chassis controller through the Type-C to USB interface. The depth camera 6 is installed above the small bracket 14 of the high-precision gimbal 11, and is used for intelligent detection and reading of substation instruments, and together with the multi-line laser radar 7 and its built-in IMU, it completes multi-sensor fusion SLAM.

[0052] Specifically, the multi-line laser radar 7 is powered by a DC-DC 48V to 12V power management module 9, and data is transmitted with the chassis controller through an M12 aviation plug to RJ45 interface. The multi-line laser radar 7 is installed on the top of the small bracket 14 of the high-precision gimbal 11, and completes the multi-sensor fusion SLAM together with the depth camera 6.

[0053] Specifically, the high-precision gimbal 11 is powered by a DC-DC 48V to 24V power management module 10, and data is transmitted with the chassis controller through an RS485 to USB interface. The high-precision gimbal 11 is installed above the upper bracket 13, and a small bracket 14 is used above the upper bracket to carry a depth camera 6 and a multi-line laser radar 7. The high-precision gimbal 11, the depth camera 6 and the multi-line laser radar 7 jointly complete multi-sensor fusion SLAM and intelligent detection and reading of substation instruments.

[0054] An intelligent inspection system according to an embodiment of the present invention comprises the substation inspection robot in the above-mentioned embodiment and a robot control host arranged in a central control room, wherein the robot control host is communicatively connected with the controller of the substation inspection robot, and the robot control host is used to issue robot control instructions to the controller of the substation inspection robot to control the entire intelligent inspection process, further control the two inspection modes of automatic inspection and remote control inspection of the substation inspection robot, and can switch between the two modes at will; at the same time, various detection data and status data sent back by the controller of the substation inspection robot are received, and the detection data can be automatically analyzed; the detection data contains information collected by the binocular camera 5, the depth camera 6, the single-line laser radar 3, the multi-line laser radar 7, the ultrasonic sensor 2, the temperature sensor 4 and the high-precision gimbal 11 used for multi-sensor fusion SLAM and automatic inspection tasks; the status data contains information about the motion state of the inspection robot.

[0055] The intelligent inspection system also includes a remote control receiving device that is communicatively connected to the robot controller, and the remote control receiving device is used to receive remote control instructions sent by the remote control device and send the remote control instructions to the robot controller. The remote control device mainly includes a remote control device that drives the chassis to move and a remote control device that drives the pan / tilt to rotate, and the remote control device is a mobile device such as a remote control handle, a remote control keyboard or a mobile phone.

[0056] During the entire intelligent inspection process, obstacles are continuously monitored and detected through the robot's front-end binocular camera and the front and rear ultrasonic sensors for real-time obstacle avoidance.

[0057] The intelligent inspection process of the intelligent inspection system includes the following steps:

[0058] Step 1, system initialization, substation inspection robot self-inspection, high-precision gimbal 11 self-inspection; system initialization refers to the process in which when the inspection robot is started, the robot management system of the robot control host performs the initial system settings according to the preset motion parameters and network parameters; the substation inspection robot self-inspection refers to the inspection robot detecting its own power status, system status, and whether the system setting parameters are missing, and informing the robot management system of the current situation. After the self-inspection is correct, the robot management system controls the inspection robot to enter the standby state; the high-precision gimbal 11 self-inspection is a process of comprehensive inspection of the gimbal system, which can ensure that all components of the gimbal are in normal working condition. Through self-inspection, possible faults or abnormalities of the gimbal can be discovered in time, thereby avoiding unexpected situations during operation.

[0059] Step 2: By controlling the movement of the four-wheel drive mobile chassis 1 and the rotation of the high-precision gimbal 11, the depth camera 6, the multi-line laser radar 7 and its built-in IMU are used to perform multi-sensor fusion SLAM using the R3LIVE algorithm to reconstruct the surrounding environment in three dimensions, and the constructed three-dimensional dense point cloud map with RGB color information is transmitted back to the robot control host in the central control room for overall observation and control of the working environment.

[0060] Among them, before using the depth camera 6, the multi-line laser radar 7 and its built-in IMU for multi-sensor fusion SLAM, it is necessary to complete the multi-sensor joint calibration task.

[0061] Specifically, the depth camera internal parameters need to be calibrated first. The camera internal parameter matrix K and distortion coefficients k1, k2, p1, p2, k3 obtained by calibration are used for subsequent external parameter calibration and SLAM algorithm. The camera internal parameter matrix is ​​in the following form:

[0062]

[0063] Furthermore, according to the camera intrinsic parameter matrix and distortion coefficient obtained by calibration, the external parameters of the camera and the multi-line laser radar are jointly calibrated. Through external parameter calibration, the transformation matrix T from the depth camera to the multi-line laser radar is obtained. The transformation matrix T contains the rotation matrix R and the translation matrix t, which is the rotation and translation information from the depth camera 6 to the multi-line laser radar 7. The final depth camera intrinsic parameter matrix, distortion coefficient and the transformation matrix from the depth camera 6 to the multi-line laser radar 7 will be provided to the R3LIVE algorithm for multi-sensor fusion SLAM, and a three-dimensional dense point cloud map with RGB color information will be constructed for the robot working environment. Since the IMU in this embodiment uses the built-in IMU of the multi-line laser radar, there is no need to calibrate the IMU.

[0064] Step 3: By controlling the movement of the four-wheel drive mobile chassis 1 and the rotation of the high-precision gimbal 11, the single-line laser radar 3 is used to perform laser SLAM using the Cartographer algorithm to build a two-dimensional map of the working environment for navigation, positioning and path planning.

[0065] SLAM mapping technology usually refers to the distance characterization of target points at different locations in the field by laser radar. Laser SLAM uses 2D or 3D laser radar. The emergence and popularization of laser radar makes measurement faster and more accurate, and the information is richer. The object information collected by the laser radar presents a series of scattered points with accurate angle and distance information, which is called a point cloud. The laser SLAM system calculates the distance of the relative movement and the change of posture of the laser radar by matching and comparing two point clouds at different times, thereby completing the positioning of the robot itself.

[0066] The difference between a 3D depth camera and an ordinary camera is that, in addition to being able to obtain planar images, it can also obtain depth information of the photographed object, that is, three-dimensional position and size information. The entire computing system then obtains three-dimensional data of the environment and the object. This information can be used in target tracking, three-dimensional reconstruction, human-computer interaction, SLAM and other fields.

[0067] Step 4: The robot control host in the central control room selects multiple points in the constructed two-dimensional map and gives them to the controller of the substation inspection robot as multiple target positions for subsequent automatic inspections.

[0068] Step 5. During the automatic inspection process, the amcl function package in ROS is used to perform real-time positioning of the robot in the two-dimensional map. The proposed LH global path planning algorithm (Low-time-consuming High-efficiency RoutePlanning) is used to plan the global path and calculate the optimal path for the robot to the next target position as the global path of the robot. The Dynamic Window Approaches algorithm is used to search for multiple paths for avoidance and travel, and the optimal path is selected based on various evaluation criteria, and dynamic obstacle avoidance is achieved.

[0069] After the environmental map (established in step 3) is constructed, the LH global path planning algorithm is used to plan the optimal inspection path. The basic principle is to use the current position of the inspection robot as the initial node to evaluate the expansion node of the node, determine the best child node in the expansion through the evaluation function, and then evaluate the expandable nodes of this child node again, and repeat the cycle in sequence until there is a target node in the expandable nodes of a certain node. g(n) is the cost function, which represents the cost value between the initialization node and the current node, which is equivalent to the path length from the starting position to the current position. When the robot moves to a certain n node, it represents the cost corresponding to the optimal route from the starting node to the current position; h(n) is the heuristic function, which is the optimal path evaluation value between the current node and the target node. The calculation of h(n) is directly related to the effect of the LH global path planning algorithm; f(n) represents the cost from the starting node through the intermediate node to the target node. The evaluation function of the LH algorithm is:

[0070] f(n)=g(n)+α(d,t)·h(n)

[0071] Among them, α(d,t) is a dynamic adjustment factor. The dynamic adjustment factor model combines path density perception and time decay. The purpose is to dynamically adjust the heuristic weights to make the algorithm more accurate in complex areas and faster in sparse areas, and optimize the use of computing resources over time. The dynamic adjustment factor includes path density weight, time decay factor and global adjustment factor. Its theoretical formula is:

[0072]

[0073] The path density weight represents the obstacle density gradient around node n, Represents the maximum density gradient in the global environment; it aims to reduce the heuristic weight in areas with dense obstacles to improve accuracy, and increase the heuristic weight in sparse areas to improve planning speed.

[0074] The λ in the time decay factor is the decay rate, which aims to gradually reduce the heuristic weight as the search time increases, and thus pay more attention to the actual cost g(n); the global adjustment factor δ is the heuristic global adjustment ratio, which is used to control the strength of the overall dynamic adjustment factor to adapt to the substation environment; λ and δ are both empirical parameters, which are adjusted according to actual conditions.

[0075] Step 6. Whenever a patrol target position is reached, the depth camera follows the high-precision gimbal to perform an all-round scan along a predetermined trajectory, and uses the target detection algorithm to perform intelligent detection on the current equipment instrument until a clear image of the instrument area is obtained. The instrument image in the instrument area image is extracted based on the target detection network model, and the numerical area in the instrument image is segmented. The value of the instrument is determined based on the segmented image, thereby completing the intelligent detection and reading of the instrument during the intelligent patrol process; the detected image and the read value are transmitted back to the robot control host in the central control room for analysis. If it is higher or lower than its threshold, an alarm is issued.

[0076] Step 7. During the execution of the multi-target automatic inspection task, if it is necessary to lock and monitor the status of a certain substation equipment, the current automatic inspection task can be suspended and switched to the remote control inspection mode. By remotely controlling the movement of the four-wheel drive mobile chassis 1 and the rotation of the high-precision pan-tilt head, the robot can be controlled to quickly reach the location of the abnormal equipment, and the equipment can be checked and the alarm information can be verified in time.

[0077] Step 8. After completing the task, the robot can switch back to the automatic inspection mode and continue to perform the previous automatic inspection task.

[0078] The substation inspection robots and intelligent inspection systems of the above embodiments are equipped with a pan-tilt platform and a variety of sensor equipment, integrating a variety of sensor technologies, communication technologies and artificial intelligence algorithms. The inspection robots have stronger complex environment modeling capabilities, stronger obstacle avoidance capabilities, higher road adaptability, and smoother walking; they greatly reduce the work pressure of the staff and truly realize all-weather, unmanned intelligent inspection of substations. The intelligent inspection system uses the LiDAR-Inertial-Visual fusion SLAM algorithm to construct a three-dimensional dense point cloud map with RGB color information for the robot's working environment, thereby enhancing the robot's ability to model complex environments. By using ultrasonic sensors, lidars, and binocular cameras to perform dynamic obstacle avoidance during the robot's operation, the robot's obstacle avoidance ability in complex environments is improved. By programming and controlling the high-precision gimbal, the camera's inspection field of view covers the area to be inspected in all directions, thereby making the robot's detection and reading capabilities of substation instruments more accurate. The remote robot control host issues control instructions to the robot controller, thereby realizing two inspection modes of the substation inspection robot: automatic inspection and remote control inspection. The two modes can be switched according to task requirements, giving the robot a more intelligent autonomous inspection capability.

[0079] The above is only a specific implementation of the present invention, which enables those skilled in the art to understand or implement the present invention. Although detailed descriptions are given with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments, and they should all be covered by the protection scope of the claims.

Claims

1. A substation inspection robot, used for intelligent inspection tasks in a substation environment, characterized in that: include: The robot body comprises a four-wheel drive mobile chassis (1), on which an ultrasonic sensor (2), a single-line laser radar (3), a temperature sensor (4), a binocular camera (5), a depth camera (6), a multi-line laser radar (7), a 5G communication module, a power management module and a high-precision pan-tilt platform (11) for multi-sensor fusion SLAM and automatic inspection tasks are installed; wherein the depth camera (6) and the multi-line laser radar (7) are integrated on the high-precision pan-tilt platform (11) and move synchronously with the high-precision pan-tilt platform (11); the ultrasonic sensor (2) is distributed at the front and rear ends of the four-wheel drive mobile chassis (1); A driving device, arranged on the four-wheel drive mobile chassis (1) and used for driving the chassis to move; A controller is installed in the casing of the four-wheel drive mobile chassis (1), and communicates with the driving device, the ultrasonic sensor (2), the single-line laser radar (3), the temperature sensor (4), the binocular camera (5), the depth camera (6), the multi-line laser radar (7), the 5G communication module and the high-precision gimbal (11) respectively.

2. The substation inspection robot according to claim 1, characterized in that: The single-line laser radar (3) is installed at the front end of the four-wheel drive mobile chassis (1) and is used for the intelligent inspection system to perform laser SLAM, navigation positioning, and path planning.

3. The substation inspection robot according to claim 1, characterized in that: The binocular camera (5), the depth camera (6), the multi-line laser radar (7) and the high-precision gimbal (11) are all electrically connected to the controller via corresponding power management modules.

4. The substation inspection robot according to claim 3, characterized in that: The binocular camera (5) and the depth camera (6) are powered by a 48V to 5V power management module (8), and data is transmitted with the controller via a Type-C to USB interface; a double-layer bracket is installed on the top of the four-wheel drive mobile chassis (1), including a lower bracket (12) and an upper bracket (13); the binocular camera (5) is fixedly installed at the front end of the lower bracket (12) and is used for continuous monitoring and detection of obstacles ahead on the robot's travel route; a high-precision gimbal (11) is installed on the upper bracket (13); a small bracket (14) is installed on the top of the high-precision gimbal (11); the depth camera (6) is installed above the small bracket (14) and is used for intelligent detection and reading of substation instruments, and together with the multi-line laser radar (7) and its built-in IMU, completes multi-sensor fusion SLAM.

5. The substation inspection robot according to claim 4, characterized in that: The multi-line laser radar (7) is powered by a 48V to 12V power management module (9), and data is transmitted with the controller via an M12 aviation plug to RJ45 interface. The multi-line laser radar (7) is installed on the top of a small bracket (14) and completes multi-sensor fusion SLAM together with the depth camera (6).

6. The substation inspection robot according to claim 5, characterized in that: The high-precision pan-tilt head (11) is powered by a 48V to 24V power management module (10), and data is transmitted with the controller via an RS485 to USB interface. The high-precision pan-tilt head (11), the depth camera (6) and the multi-line laser radar (7) jointly complete multi-sensor fusion SLAM and intelligent detection and reading of substation instruments.

7. An intelligent inspection system, characterized in that: The invention comprises a substation inspection robot as claimed in any one of claims 1 to 6 and a robot control host arranged in a central control room, wherein the robot control host is communicatively connected with the controller of the substation inspection robot, and the robot control host is used to issue robot control instructions to the controller of the substation inspection robot to control the entire intelligent inspection process, and control the substation inspection robot to switch between two inspection modes, namely, automatic inspection and remote control inspection; at the same time, various detection data and status data sent back by the controller of the substation inspection robot are received, and the detection data can be automatically analyzed; the detection data contains information collected by a binocular camera (5), a depth camera (6), a single-line laser radar (3), a multi-line laser radar (7), an ultrasonic sensor (2), a temperature sensor (4) and a high-precision gimbal (11) for multi-sensor fusion SLAM and automatic inspection tasks; the status data contains information about the motion state of the substation inspection robot; The intelligent inspection system also includes a remote control receiving device that is communicatively connected to the robot controller, and the remote control receiving device is used to receive remote control instructions sent by a remote control device and send the remote control instructions to the robot controller.

8. The intelligent inspection system according to claim 7, characterized in that: The remote control device includes a remote control device for driving the chassis to move and a remote control device for driving the pan / tilt head to rotate, and the remote control device is a remote control handle, a remote control keyboard, or a mobile device including a mobile phone; During the entire intelligent inspection process, obstacles are continuously monitored and detected through the robot's front-end binocular camera and the front and rear ultrasonic sensors (2) for real-time obstacle avoidance.

9. The intelligent inspection system according to claim 7, characterized in that: The intelligent inspection process includes the following steps: Step 1: System initialization, substation inspection robot self-inspection, high-precision pan / tilt (11) self-inspection; Step 2: By controlling the movement of the four-wheel drive mobile chassis (1) and the rotation of the high-precision gimbal (11), a depth camera (6), a multi-line laser radar (7) and its built-in IMU are used to perform multi-sensor fusion SLAM using the R3LIVE algorithm to perform three-dimensional reconstruction of the surrounding environment, and the constructed three-dimensional dense point cloud map with RGB color information is transmitted back to the robot control host in the central control room for overall observation and control of the working environment; Step 3: By controlling the movement of the four-wheel drive mobile chassis (1) and the rotation of the high-precision gimbal (11), a single-line laser radar (3) is used to perform laser SLAM using the Cartographer algorithm to construct a two-dimensional map of the operating environment for navigation, positioning and path planning; Step 4: The robot control host in the central control room selects multiple points in the constructed two-dimensional map and gives them to the controller of the substation inspection robot as multiple target positions for subsequent automatic inspection; Step 5: During the automatic inspection process, the amcl function package in ROS is used to locate the robot in the two-dimensional map in real time; the proposed LH global path planning algorithm is used to plan the global path, and the optimal route from the robot to the next target position is calculated as the global route of the robot; the Dynamic Window Approaches algorithm is used to search for multiple paths for avoidance and travel, and the optimal path is selected based on various evaluation criteria, and dynamic obstacle avoidance is achieved; Step 6: Whenever a patrol target position is reached, the depth camera (6) follows the high-precision pan / tilt platform (11) to perform an all-round scan according to a predetermined trajectory, and uses a target detection algorithm to perform intelligent detection on the current equipment instrument until a clear image of the area where the instrument is located is obtained. The instrument image in the image of the area where the instrument is located is extracted based on the target detection network model, and the numerical area in the instrument image is segmented. The numerical value of the instrument is determined based on the segmented image, thereby completing the intelligent detection and reading of the instrument during the intelligent patrol process; the detected image and the read numerical value are transmitted back to the robot control host in the central control room for analysis, and if it is higher or lower than its threshold, an alarm is issued; Step 7: During the execution of the multi-target point automatic inspection task, if it is necessary to lock and monitor the status of a certain abnormal substation equipment, the current automatic inspection task can be suspended and switched to the remote control inspection mode. By remotely controlling the movement of the four-wheel drive mobile chassis (1) and the rotation of the high-precision pan-tilt head (11), the robot can be controlled to quickly reach the location of a certain abnormal substation equipment, and the equipment can be checked and the alarm information can be verified in time. Step 8. After completing the task, the robot can switch back to the automatic inspection mode and continue to perform the previous automatic inspection task.

10. The intelligent inspection system according to claim 9, characterized in that: In step 5, the basic principle of the LH global path planning algorithm is to use the current position of the automatic inspection robot as the initial node, evaluate the expansion node of the node, determine the best child node in the expansion through the evaluation function, and then evaluate the expandable nodes of this child node again, and iterate in turn until there is a target node in the expandable nodes of a certain node; the evaluation function of the LH global path planning algorithm is: f(n)=g(n)+α(d,t)·h(n) Among them, h(n) is the heuristic function, which is the optimal path evaluation value between the current node and the target node. The calculation of h(n) is directly related to the effect of the LH global path planning algorithm; f(n) represents the cost from the starting node through the intermediate node to the target node; g(n) is the cost function, which represents the cost value between the initialization node and the current node, which is equivalent to the path length from the starting position to the current position. When the robot moves to a certain n node, it represents the cost corresponding to the optimal route from the starting node to the current position; α(d, t) is the dynamic adjustment factor, and its theoretical formula is: The path density weight represents the obstacle density gradient around node n, represents the maximum density gradient in the global environment; λ in the time decay factor is the decay rate; the global adjustment factor δ is the heuristic global adjustment ratio, which is used to control the strength of the overall dynamic adjustment factor to adapt to the substation environment.

Images