Robotic system for inspection and detection in high-voltage electrical installations and a method for controlling the robot

Abstract

The present invention pertains to the field of program-controlled manipulators and discloses a robotic technology for inspection and detection in high-voltage electrical installations. In high-voltage electrical installations, inspections are carried out by operators to identify faults and hazardous situations in the equipment. Modernizing said inspections by using new technologies will contribute to occupational safety and efficiency of maintenance methods, reducing faults in the electrical installations and avoiding economic losses. In light of this, the invention discloses a ground-based, mobile, autonomous robotic system with multiple instruments which carries out detection and inspection on high-voltage electrical equipment in power generation and transmission installations. The system uses an integrated navigation and inspection method, allowing precise location and independent movement in indoor and outdoor environments, replacing manual inspections carried out by operators.

Description

ROBOTIC INSPECTION AND DETECTION SYSTEM IN HIGH VOLTAGE ELECTRICAL INSTALLATIONS AND A ROBOT CONTROL METHOD FIELD OF APPLICATION

[0001] The present invention applies to the field of program-controlled manipulators. The present invention discloses a robotic technology for inspection and detection in high-voltage electrical installations. BACKGROUND OF THE INVENTION

[0002] In high-voltage power generation and transmission facilities, inspections performed by operators are essential for detecting equipment failures and hazardous situations that could compromise the availability of power supply and production quality. However, these traditional inspections are limited in their regularity and efficiency due to the high occupational risks involved. To overcome these challenges, modernizing inspections through the adoption of new technologies that promote occupational safety and efficient maintenance processes is essential. This will minimize failures in electrical installations, avoiding economic losses resulting from these failures, and improving companies' quality indicators.

[0003] The introduction of advanced inspection and detection technologies, such as autonomous robotic systems and specialized instruments, can significantly improve monitoring and troubleshooting capabilities in high-voltage electrical installations. These modern solutions enable the accurate detection of hot spots, smoke, oil leaks, and electrical integrity conditions. assets, in addition to identifying the presence of unauthorized persons and animals, thus reducing the occupational risks faced by operators. Furthermore, the implementation of these new technologies will result in an overall improvement in the quality of inspections, preventing failures, reducing interruptions in the power supply and bringing economic benefits to companies in the sector.

[0004] To solve the aforementioned problems, the present invention discloses a technology that replaces the need for operators to perform manual inspections. The invention is a ground-based and mobile robotic system capable of detecting and inspecting high-voltage electrical equipment in power generation and transmission facilities. It features an integrated navigation method, enabling precise inspections in both indoor and outdoor environments, replacing the need for manual operator intervention.

[0005] The invention provides autonomous navigation, with detection and avoidance of fixed obstacles and mobile obstacles in inhabited operational environments; automatic detection, based on artificial intelligence, of venomous animals, vehicles and people; autonomous recharging of batteries for electric autonomy; visual inspection by means of automatic detection, based on artificial intelligence, of analog oil meters and analog gas meters belonging to the industrial instrumentation of electrical equipment; visual inspection by means of capturing images from different points of view of electrical equipment; thermographic inspection with automatic capture of images of electrical equipment; hyperspectral VNIR inspection with automatic image capture and post-processed detection of mineral oil leaks in electrical equipment; fire smoke detection in indoor environments; web-based human-machine interface (HMI) for teleoperation; HMI with video transmission for monitoring and teleoperation; HMI for interactive generation of maintenance reports with indexing between data and equipment. STATE OF THE TECHNIQUE

[0006] Document CN110379036 discloses an intelligent transformer substation inspection and identification method, system, and device, as well as a storage medium. The method comprises the steps of performing data acquisition on transformer substation equipment and a field operation scene and establishing a sample database; training a sample data set using an artificial intelligence algorithm and constructing an inspection and identification model; performing inspection on a transformer substation and identifying a real-time condition of the transformer substation through the inspection and identification model to obtain an inspection judgment result; and displaying the inspection judgment result at a corresponding position in a real three-dimensional scene of the transformer substation.According to the method, the sample database is established, and the training is carried out by the artificial intelligence algorithm to generate the inspection and identification model, so that the equipment states of the transformer substation and the operating conditions of the personnel can be identified in real time, and the accuracy of the inspection and identification is. Greatly improved; and the inspection and identification results can be displayed in the real three-dimensional scene of the transformer substation in real time, so that the working personnel can check the inspection and identification results in time, and the inspection and identification results are clear and intuitive. The intelligent transformer substation inspection and identification method, system and device can be widely applied to the field of transformer substation inspection.

[0007] Unlike the present invention, the above document describes an inspection method based on an artificial intelligence algorithm for collecting, storing data, and displaying the real scene in three dimensions. However, the method is only capable of visually inspecting equipment, which is its limitation. The present invention uses artificial intelligence-based methods to operate autonomously in the inspection environment, in addition to identifying objects to be inspected and storing data for expert analysis. The present invention differs from the above document in other aspects, such as post-processed detection of mineral oil leaks in electrical equipment and fire smoke detection in indoor environments.

[0008] Document CN110266111 discloses an artificial intelligence patrol system for substations, which includes a terminal layer and a base station layer. The terminal layer includes intelligent patrol robots and a remote monitoring layer. The base station layer includes a background server, a hard disk recorder, a router, and a control and software system. Intelligent analysis. The remote monitoring layer includes an X86 architecture industrial computer, and the remote monitoring layer commands, controls, and monitors multiple intelligent patrol robots remotely and across regions. The intelligent patrol robots transmit video and image information to the background server in real time, allowing the intelligent control and analysis software system to perform control and analysis. The artificial intelligence substation patrol system has the following advantages: reduced labor intensity; reduced substation operation and maintenance costs; improved automation and intelligence levels in normal patrol operation and management; and provided innovative technical detection means and comprehensive safety assurance for intelligent substations and autonomous substations.

[0009] Unlike the present invention, the above document reports an artificial intelligence patrol system for substations. In this sense, the system proposed in the above document is not capable of autonomously carrying out inspections on equipment in high-voltage electricity generation and transmission facilities.

[0010] Document CN102082466 discloses an intelligent robotic inspection system for transformer substation equipment. The intelligent robotic inspection system for transformer substation equipment comprises a control center, a wireless communication network, and a robotic inspection system, wherein the control center performs remote control. and transmitting data with the inspection robotic system through the wireless communication network; the control center comprises an upper control host, an inspection monitoring data processing module, and a robot remote control transmission module; and the inspection robotic system comprises a robot control unit, a robot body mechanism, a robot obstacle detection and navigation device, a robot remote control receiving module, and a field signal acquisition and transmission device.

[0011] Unlike the present invention, the above document reports the development of an intelligent inspection system for substations that uses a robot to capture and transmit inspection data. However, the system proposed in the above document is only capable of identifying obstacles during autonomous navigation, which is its limitation. The present invention, during autonomous navigation, allows the identification of obstacles, as well as the presence of unauthorized persons and animals in the operating environment, thus reducing the occupational risks faced by operators.

[0012] Document CN112506205 discloses an invention that is in the technical field of artificial intelligence and provides a robot inspection task planning method, which comprises the steps of obtaining a first inspection task and environmental information of a target inspection site corresponding to the first inspection task, the environmental information comprising solar altitude and azimuth angle information of the target inspection site, and the first inspection task comprising one or more devices to inspect the target inspection site; acquire pose information of the robot at a predefined detection location, the predefined detection location being a location where the robot is located when detecting each of the one or more devices; determine a time constraint condition according to the solar altitude angle, azimuth angle information, and pose information; based on the time constraint condition and the first inspection task, determine a first inspection route according to a predefined algorithm; and allow the robot to perform the first inspection task according to the first inspection route. The method of inspecting transformer substation equipment by means of the robot replaces a traditional manual inspection mode, and the inspection efficiency of transformer substation equipment is improved.

[0013] Unlike the present invention, the above document describes a system for autonomous inspection with planned inspection execution. A method for acquiring robot pose information at a predefined detection location using environmental information is proposed, with the aim of improving data collection efficiency. However, the method is only applicable to outdoor environments, and inspection in indoor environments is not possible. The present invention has an integrated autonomous navigation method and a set of sensors, allowing precise inspections in both indoor and outdoor environments.

[0014] Document CN112199993 provides a method for recognizing a transformer substation insulator infrared image detection model in any direction Based on artificial intelligence, the method comprises the following steps: preprocessing the obtained image to form an infrared insulator image dataset; labeling the insulator on the obtained dataset and dividing the dataset into a training set and a test set; constructing an infrared image detection model from an enhanced single-stage detection network; establishing a network structure, designing a feature map, and establishing a loss function; training the model using a training set in the dataset and adjusting the parameters; and testing the trained model using a test set in the dataset to prove the model's effectiveness. Through the above steps, automatic detection and identification of the infrared image of the transformer substation insulator are achieved.By improving a single-stage sensing network and modifying the number of feature map channels of a sensing layer, the network's feature extraction capability for a target is improved, and it is ensured that the model effectively detects an infrared image of a transformer substation insulator.

[0015] Unlike the present invention, the above document describes the development of a machine learning-based software model for recognizing an infrared image of a transformer substation insulator, specifically targeting its application for autonomous inspection of high-voltage power generation and transmission facilities. The present invention features visual image recognition applied to meters. oil and pressure analogs (gauges) and recognition of high voltage current transformers (dome and head), being able to identify them and, from there, crop the respective images of interest for inspection report purposes. In a pre-programmed manner, the invention has the capacity to capture visual and thermographic images of any electrical equipment by registering their position and orientation on the map (presets), data that are later provided as targets for mission planning and then processed by the autonomous navigation system in order to complete an inspection route.

[0016] Document CN112091969 discloses an intelligent robotic inspection system monitoring system for a power line pipeline. The intelligent robotic inspection system monitoring system comprises an intelligent robotic inspection system, a mobile inspection module, an online monitoring module, and an AI decision-making module. Intelligent inspection of the power line pipeline is achieved through the intelligent robotic inspection system. The analysis, storage, extraction, monitoring, application, and presentation of related information, as well as corresponding autonomous analysis and diagnosis, are performed through the AI ​​decision-making module. The status of the power line pipeline can be effectively monitored in real time, and abnormal conditions in the power line pipeline can be detected in time. Online monitoring and robotic inspection are adopted to replace manual inspection.

[0017] Unlike the present invention, the above document reports an online monitoring method based on An intelligent robotic inspection system for a power line tunnel. This type of system is highly specific for indoor applications, such as power line tunnels, and does not offer the ability to adapt to other environments, such as outdoor environments. The present invention features an integrated autonomous navigation method and a set of sensors, enabling precise inspections in both indoor and outdoor environments.

[0018] Document CN111080634 discloses a transformer appearance defect identification method based on a robotic inspection system and a Cascade RCNN algorithm.The invention pertains to the field of image recognition, and the method comprises the following steps: S1: Image acquisition, operation and maintenance personnel, and engineers in a power plant use tools such as cameras, handheld terminals, and cameras to perform image acquisition under transformer appearance defect conditions; S2: Creating a set of image books. The method comprises the following steps: labeling an XML file based on an acquired image of a transformer; randomly dividing a training set and a test set according to a certain ratio; S3: training a Cascade RCNN model using the training set data; and S4: deploying the model on a centralized control platform of a robotic inspection system, and acquiring an appearance image of the transformer equipment using a high-definition camera to inspect the appearance status of the equipment in the transformer area.It is verified that. feasibility of identifying the appearance defects of transformer equipment based on the transformer substation robotic inspection system and deep learning technology, and promoting the landing application of artificial intelligence technology in transformer substation.

[0019] Unlike the present invention, the above document reports a method for identifying transformer appearance defects implemented in a robotic inspection control platform. However, the development does not include the functionalities of autonomous navigation, obstacle detection, recognition of people and animals, in addition to the software for operating the robotic system, such as the present invention.

[0020] Document CN102355052 discloses a route inspection system based on an intelligent robot for a transformer station and a method for monitoring the operation of the transformer station, which can meet all-weather work requirements. The method comprises the following steps: 1) remotely assuming control of an intelligent route inspection robot; 2) issuing a device detection command to the intelligent route inspection robotic system for the transformer station; 3) automatically navigating the intelligent route inspection robotic system to an operating point using an optimal path specification and sending back a real-time monitoring video; 4) waiting for an operating command after the operator is on site; 5) issuing a one-time operating instruction to the operator; 6) monitoring the operation of a person. of the scene using the video of the intelligent routing inspection robotic system; 7) remotely control the intelligent routing inspection robotic system to move to an appropriate position or adjust a bracket of the intelligent routing inspection robotic system to be at an appropriate angle; 8) after the operation of the steps is finished, wait for the next operation instruction; and 9) after all operation tasks are completed, control the intelligent routing inspection robotic system to automatically move back to a loading room for loading and complete the operation. 8) after the operation of the steps is finished, wait for the next operation instruction; and 9) after all operation tasks are completed, control the intelligent routing inspection robotic system to automatically move back to a loading room for loading and complete the operation.8) After the operation of the steps is finished, wait for the next operation instruction; and 9) After all the operation tasks are completed, control the intelligent routing inspection robotic system to automatically move back to a loading room for loading and complete the operation.

[0021] Unlike the present invention, the above document describes a routing inspection system based on an intelligent robot of a transformer station to operate in any climate. It has functionalities for autonomous navigation, route calculation, data acquisition and storage. However, the system presented in the above document is only applicable to outdoor environments, which is its limitation. The present invention has a method integrated autonomous navigation and a set of sensors, allowing precise inspections in indoor and outdoor environments, in addition to having other functionalities for system operation through local (in the robot system itself) and remote (application server) data control, capture, transmission and storage software. SUMMARY OF THE INVENTION

[0022] In high-voltage power generation and transmission facilities, inspections are performed by operators using instruments and human senses to detect signs of equipment failure and dangerous or undesirable situations, such as hot spots, smoke, oil leaks, asset integrity issues, and the presence of unauthorized personnel and animals. These situations can impact the availability of power supply and, consequently, production quality. The regularity and efficiency of these inspections are limited because they involve high occupational risks. Therefore, modernizing inspections through the introduction of new technologies that contribute to occupational safety and the efficiency of such maintenance processes will minimize failures in these electrical installations.Consequently, it will avoid economic losses caused by these failures, as well as bring improvements to the companies' quality indicators.

[0023] In view of this context, the present invention reveals an autonomous robotic, mobile and terrestrial multi-instrument system capable of performing different types of detection and inspection on high voltage electrical equipment in Electric power generation and transmission facilities. The invention also includes an integrated navigation and inspection method. The robotic system is capable of precise localization and independent movement in indoor environments (insulated cable galleries) and outdoor environments (open-air substations and industrial yards) to perform inspections that are typically performed manually by operators using instruments. BRIEF DESCRIPTION OF THE FIGURES

[0024] Figure 1 illustrates the components of the robot of the present invention.

[0025] Figure 2 illustrates an isometric view of the female electrical charging structure - docking female.

[0026] Figure 3 illustrates an isometric view of the male electrical charging structure - docking male.

[0027] Figure 4 illustrates the operation flowchart of the robotic system.

[0028] Figure 5 illustrates an exploded isometric view of the male and female electrical load structure.

[0029] Figure 6 illustrates an exploded side view of the male and female electrical load structure.

[0030] Figure 7 illustrates a 3D image of the female electrical charging structure (docking female) installed in the autonomous ground robot unit.

[0031] Figure 8 illustrates a 3D image of the assembly of the male electrical charging structure (docking male) at the charging point including ArUco type fiducial markers. DETAILED DESCRIPTION OF THE INVENTION

[0032] The invention may be better understood through the following detailed description, in accordance with the attached figures.

[0033] The present invention discloses a robotic inspection and detection system in high voltage electrical installations and a method of controlling the robot.

[0034] The robotic system for inspection and detection in high-voltage electrical installations comprises two main units: an autonomous ground robot unit, proposed from the arrangement, structures and integration of a commercial mobile robotic platform with multiple commercial sensors that, in combination, are intended for autonomous navigation and inspection; and an autonomous electrical recharging unit, of exclusive design, composed of mechanical and electrical couplings, as well as fiducial markers, as shown in figure 8.

[0035] The autonomous ground robot unit (or robot) comprises at least one mobile robotic platform (1), optionally a lifting platform (2), at least one 3D LiDAR (3), at least one inspection camera (4), at least one depth camera (5), at least one safety laser scanner (6), at least one GNSS RTK receiver (7), at least one GNSS RTK antenna (8), these two for communicating with a GNSS RTK base station of the same type and installed on the autonomous electric charging unit, optionally two LED headlights (9), optionally an audible signal (10), at least one Wi-Fi router (11), at least one support and protection structure (12), at least one on-board computer (13), at least one structure for female electric charging (14), at least one Wireless E-Stop receiver (15), at least one thermo-hygrometer (16), at least one at least one smoke sensor (17), at least one at least one Attitude and Heading Reference System (AHRS) sensor (18), at least one rear-facing monocular camera (19), and at least one antistatic tape (21).

[0036] The autonomous ground robot unit was designed using a modular architecture, meaning it can be configured with hardware elements according to the inspection task to be performed, without necessarily relying on all of its components to operate. This allows the system to save energy and operate more optimally.

[0037] The inspection camera (4) is an interchangeable component and is therefore selected according to the demand to be met, and may be, for example, a visual and thermographic pan, tilt and zoom (PTZ) type or a hyperspectral type with a rotating unit, depending on the type of inspection to be performed.

[0038] The quantity of constituents is defined according to the service demand, thus allowing the customization of the robotic system to the service.

[0039] In figure 1 it can be seen that the assembly of the autonomous ground robot unit consists of the installation on the front bumper of the mobile robotic platform (1), which is a market equipment, of two optional LED headlights (9) whose function is to assist teleoperated navigation and of a safety laser scanner (6) whose function is to detect obstacles, especially dynamic ones, in a horizontal angular opening of 275°, with a range of 40 m and a protection field of 4 m in accordance with Performance Level (PL) level d (EN ISO 13849); in the installation, on the rear bumper, of a female electrical charging structure (14); in the installation of a GNSS RTK receiver (7) that makes up the robot's precision localization system in outdoor environments and an optional audible signal (10) that allows the emission of audible alerts, in the upper front part and directly on the chassis of the mobile robotic platform (1); in the installation of an antistatic tape (21), in the lower part and directly on the chassis of the mobile robotic platform (1), in order to provide contact with the terrain and enable electrostatic discharges; in the installation of a high-performance embedded computer (13) responsible for coordinating processes and mission development, acquisition and processing of navigation, inspection and inference data for object detection by AI, and an Attitude and Heading Reference System (AHRS) sensor (18), responsible for the 3D orientation of the robot, both in the upper rear part of the chassis of the mobile robotic platform (1) and covered by the support and protection structure (12);in the installation of a support and protection structure (12) on the mobile robotic platform (1) and on which the 3D LiDAR sensor (3) is installed, responsible for acquiring data for building maps in indoor environments, and GNSS RTK antenna (8) responsible for communication with the GPS, GLONASS and BeiDou constellations, a depth camera (5) on the front, responsible for building maps in outdoor environments, a rear monocular camera (19) responsible for detecting markers for correct and precise movement in reverse and towards the autonomous electric charging unit (docking station) for charging, a Wireless E-Stop receiver (15) to receive a wireless emergency stop command, a thermo-hygrometer (16) for; measure relative humidity and ambient temperature, necessary for inspection processes, a smoke sensor (17) capable of indicating the beginning of a fire in indoor environments, a Wi-Fi router (11) for receiving, mission monitoring and downloading data to the server and an optional elevating platform (2) responsible for vertically elevating the inspection camera (4) installed on it and obtaining better framing.

[0040] The autonomous electric charging unit comprises the male electric charging structure (20) , which receives the female electric charging structure (14) in an attachable manner.

[0041] In detail, the female electrical charging structure (14) - docking female - as shown in figure 2, is fixed to the rear of the mobile robotic platform (1), in the front position equivalent to the bumper, and acts as a female connector for recharging the robotic system. It enables recharging through the male electrical charging structure (20) - docking male - shown in figure 3, making the electrical connection by mechanical coupling (fitting).

[0042] The main structure of the female electrical charging structure (14) - docking female - comprises two parts (22) which complement each other as two halves, as shown in figure 5 and are fixed to the mobile platform by the connector of the parts (23) and whose assembly can be seen in figure 7. The current passage for recharging the system is done using conductive connectors (24) made of copper or other metallic material with good conductivity.

[0043] The male electrical charging structure (20) docking male - for the mechanical docking system is also assembled by the addition of two pieces (25) that form two halves, and is compatible with the female electrical load structure input (14) - docking female. This component has a degree of rotational freedom, which is coupled onto a circular bearing (35), in addition to containing conductors (37) for the passage of electrical current to the conductors (24) of the female electrical load structure (14).

[0044] To minimize any impacts and make the fitting between the connectors (24, 37) more flexible, the male electrical charging structure (20) - docking male - has vertical linear movement, through the vertical linear guide (28), making the necessary adjustments to correct any unevenness during the coupling of the parts for recharging. A fixed telescopic slide (32) allows the horizontal linear movement of the attached mobile telescopic slide (33) and reduces possible impacts from shock absorbers (31), for example silicone.

[0045] Complementary to the two robotic system units, two other subsystems are present: a control cabin - container type - designed to accommodate the data server, GNSS RTK base station, male electrical charging structure (20) (docking male) and the robot itself in an inactive / recharging state; and a set of WiFi repeaters for wireless coverage in a high voltage electrical substation of approximately 200,000 m2 .

[0046] The robotic system comprises a non-transitory computer-readable medium comprising instructions for executing a method of controlling the robotic system.

[0047] At least 11 possible functionalities were mapped for the robotic system, which will be described below: 1. Perform Data Persistence

[0048] This feature maintains the files used in the post-processing stage of the inspection operation. Files generated during the inspection operation are stored in folders that uniquely identify each inspection step / process. For the system to store the data, it is necessary to first identify the method corresponding to the inspection process step. A check of available storage space will be performed, and the system will match the method with the path where the file corresponding to the byte set will be stored. The command will identify the system action based on the file processing, as follows: C - Create; R - Read; U - Update; D - Delete. 2. Perform Thermographic Inspection

[0049] This functionality is used for general thermographic inspections of the main equipment in the electrical substation (current transformers, gas circuit breakers, disconnect switches, and insulator strings); in the transformer yard (step-up transformers); and in the insulated cable gallery (trays). The function receives a thermal image as input, which is converted into a data matrix representing the RGB values ​​of the original image, the irradiance proportional to the radiosity of the thermographed object, corresponding to the inspected equipment, and information on the ambient temperature and humidity. 3. Perform Autonomous Navigation

[0050] This feature integrates technologies of Perception and route planning methods are used to navigate the environment, avoiding obstacles, and seeking the lowest energy cost to reach the objective. During navigation, it is possible to generate the robot's status, positioning, and collision or obstacle detection alerts. Autonomous navigation depends on external and internal system factors, including: starting location, object to be inspected, the 2D map of the substation, the pose for correct framing of the object, and battery status to determine navigation autonomy. 4. Perform Visual Inspection on Analog Oil and Gas Meters

[0051] Upon receiving an inspection command from a pressure gauge or display, the robot moves to a position that best suits the target's framing by the cameras, either in automatic or remote mode. The latter is controlled by an operator. At the end of the inspection, all data is stored by the persistence module. 5. Perform Visual Inspection on Electrical Equipment

[0052] When selecting circuit breaker and disconnect switch inspection, the robot starts from its initial position and moves to the target elements following an autonomous route. Upon reaching the destination, it checks whether the component of interest can be framed. If the position is unsatisfactory, a new position is established. Next, an image of the element is captured from different angles to maximize overall visual inspection coverage of this equipment. 6. Perform Spectral Inspection for Oil Merging

[0053] When requested to perform spectral inspection, the robot will inspect all elements containing oil. It will check the equipment one by one and capture images of all of them. It will be during post-processing that the presence or absence of oil leakage will be diagnosed. The search for these leaks is to prevent leaks, so it is possible to take the necessary measures to prevent this from happening. 7. Perform Post-Processing

[0054] This functionality is performed offline, after the mission is completed. Acquiring and storing hyperspectral images are essential for post-processing. Progress can also be monitored during the process, and the post-processing results are presented in a report at the end. 8. Perform Data Export

[0055] This functionality aims to export data in a format compatible with the electricity company's maintenance management and control software. This functionality depends on the user's action, who, after consulting the inspection results, chooses to export the data. 9. Recharging Batteries Autonomously

[0056] The autonomous battery charging system doesn't just work when the robot is discharged. In addition to continuously monitoring the battery level, it also calculates whether the remaining battery charge is sufficient for the robot to return to the charging point. To achieve this, the autonomous charging system must always have the robot's current pose. Whenever the battery charge reaches a critical level, the system calculates a route for the robot. recharge point. 10. Detect Smoke

[0057] This functionality aims to detect potential fires through smoke detectors. When smoke is detected, the prototype recalculates its route to avoid the potential fire and generates an alarm / notification to inform the operator. 11. Preflight (Check subsystems)

[0058] Upon receiving the signal to start the mission, all of the robot's subsystems and sensors will be checked, thus generating a signal to authorize the start of the mission if all systems are functioning correctly and a log with a list of possible errors detected.

[0059] The operation of the robotic system follows the flowchart shown in figure 4, the robotic system control method comprises the steps of: (a) mission selection: choose the robotic system operating mode and select the equipment and types of inspection to be performed in the same mission; (b) mission reception and subsystem verification: the robotic system receives authorization to start, performs a subsystem verification; (c) planning the trajectory and starting the journey; (d) navigation, inspection, monitoring and, if necessary, recharging; (e) data download; (f) data processing.

[0060] It is important to mention that the autonomous ground robot unit remains in the control cabin attached to the autonomous electric recharging unit (docking station) until that receives a mission from the system operator interface (HMI) used by the operator through a mobile device or computer.

[0061] In step (a), the operator can enter login and password information to access the system's features. Only users registered in the database and with a defined access profile can be authenticated by the system; access to features is restricted to authentication. Thus, the operator, aware of the maintenance plan, chooses the operating mode, which can be autonomous or teleoperator, and selects the equipment and types of inspections to be performed in a single mission.

[0062] This way, at this moment the functionalities for registering basic information for system configuration are available: • Equipment - items to be inspected by the robot; • Inspection location - hydroelectric plants, for example; • Specific location - generating unit, turbine generating units, for example; • Reason for Test - type / motivation of inspection, for example: Daily, Monthly and Preventative.

[0063] The sequential interface follows the same usability flow pattern with the user, namely: • List screen - displays a list of the basic table with information on industrial plants / plants, equipment, generating units and the reason for the selected inspection; • Registration / change screen - allows the creation or editing of a record in the basic table.

[0064] Autonomous and teleoperation modes are inspection management modes. In autonomous mode, the robotic system performs the operation independently and allows online monitoring of an inspection, i.e., viewing current information about the inspected locations and images from the robot's front camera. Teleoperation mode allows the user to guide the robot and control inspection functions remotely.

[0065] Once the mission has been defined, in step (b), the robotic system receives authorization to start and first performs a subsystem check, now called preflight. This consists of obtaining status via software commands regarding the operational or inoperative condition of the robot's sensors and actuators, which may validate or abort the continuation of the mission. For example, the minimum satellite visibility and RTK mode stabilization required for the GNSS RTK system to operate accurately must be achieved for missions in an outdoor environment; otherwise, the robot will not be able to continue until these minimum conditions are met.

[0066] Next, in step (c), the robotic system uses global and local planners. Global planning estimates a trajectory that covers the various objectives (each piece of equipment to be inspected). Global planning performs this estimate based on the previously digitized map (part of the industrial plant registration process) of the inspected area and then initiates the movement of the autonomous ground robot unit. Local planning acts on short-term movement, directly handling obstacle detection and handling, and using controllers to smooth the trajectory around the robot's movement. This way, the interaction between local and global planning can be updated throughout the route due to the possible presence of people, vehicles, or animals along the planned path. The map, therefore, also undergoes updates as the robot executes its mission and encounters new obstacles or reconfigurations of the environment.

[0067] The missions performed by the invention, with the exception of teleoperated operation, consist of a series of tasks that must be performed autonomously in a well-defined order. Some tasks may have specific conditions or even higher priority than others. To design the robot's missions, which manage the execution of all these tasks, a behavior tree (BT) structure was adopted.

[0068] Thus, navigation, step (d), using autonomous operation mode, moves the autonomous ground robot unit to the destination point, avoiding obstacles. To achieve this, some of the services implemented in the robot are used. The two main services are: dl) Planning a route to a destination position from its current position, avoiding obstacles; d.2) Generating speed commands, at 5 Hz, to the platform so that it can follow the planned route to the destination, avoiding obstacles.

[0069] In addition to these, recovery services may also be requested in the event of a navigation problem, such as: i. Clearing information about some obstacles that have been included on the map; ii. Rotating around its own axis; iii. Waiting for a specified period of 1 s; i. Performing a reverse maneuver.

[0070] Navigation will have been successful if services dl and d.2 execute and complete successfully. If either fails, a recovery action must be triggered. Initially, action i is called, and then actions dl and d.2 are called again. After six attempts, if any of them continue to fail, actions i through iv can be triggered, in that order. Not all will necessarily be executed. If any of them complete successfully, actions dl and d.2 will be called again. If any of them fail, the next one will be called.

[0071] The navigation step will be executed until any of the following three conditions below are met: • The dl and d.2 services are successful. In this case, navigation was successful and the robot reached its destination. • Services i - iv failed. In this case, none of the recovery actions were able to make the robot perform actions dl and d.2, so navigation failed and the robot was unable to reach its destination; • The maximum number, 6, of attempts to execute the dl and d.2 services is reached. In this case, navigation also failed, BT will enter recovery mode by performing the following steps: i. Clear information of some obstacles that were included in the map; ii. Rotate around its own axis; iii. Wait for a specified period of 1 s; iV. Perform a reverse maneuver.

[0072] In teleoperated navigation mode, the mission's BT is expected to only provide operator support regarding robot status, i.e., remaining time and battery charge, assisting the mission through outputs to the user interface. The outputs from the device components demonstrated in the previous sections must be available to the operator. The user interface can issue robot movement commands and capture data for inspection purposes.

[0073] The teleoperation mode also registers the objects to be inspected. This task eliminates the need for the robot to figure out which objects to inspect on its own. Instead, an operator registers these objects, showing them to the robot manually before any inspection is performed. This task only needs to be performed once for each existing object to be inspected.

[0074] Still in step (d), the robot then navigates the area, inspects, and, if necessary, recharges its batteries until it completes the entire mission. The robot alternates between navigation, inspection, and recharging until it completes its route through all the desired equipment. Recharging occurs at the unit. Initial autonomous electric charging (docking station). During navigation, the robot approaches each piece of equipment selected for the mission and, if obstacles are encountered, can detour, replanning its route to reach its objectives. During inspection, the robot positions itself appropriately and collects images of each piece of equipment. Battery charge level monitoring is performed to ensure that the robot moves away from the male electric charging structure (20) only to the minimum distance necessary to return and recharge, even if a mission has not yet been completed. The robot then heads to the electric charging structure (docking station) located in the container and maintains the mission suspended until the charging operation is complete and can resume and complete.Throughout the entire journey, the operator can remotely monitor the mission's progress, checking the route traveled, the percentage of equipment inspected, the time elapsed since the start, the battery charge level, and any alerts issued. At the end of the mission, the robot returns to the autonomous electric charging unit (docking station).

[0075] The inspection task consists of capturing data from the robot's cameras (4) (e.g., Visual Inspection Camera, Thermal Camera, Hyperspectral Camera) so that an inspection object is framed in the image. The framing process can be done by a set of actuators, starting from the location and orientation of the robot, to adjustments of the Elevating Platform (2) and the PTZ module (if the sensor is on or above it). The use of optical zoom can also assist, if necessary.

[0076] One of the types of visual inspection is intended to To enable the operator to read analog oil and pressure gauges that are included in the inspection report. This type of inspection requires the robot to perform an AI-based automatic detection process and crop the object of interest to read the captured image, eliminating other information, making framing more accurate. This applies, in this case, to analog gauges that are part of the plant's high-voltage equipment instrumentation, namely, oil level gauges on step-up transformer bushings, oil level gauges on expansion tanks of these transformers, oil level gauges on current transformers, and pressure level gauges on three-pole circuit breakers.Compared to meter inspection, the other inspections are simpler processes, which use the same standard of components, removing the need for object detection and replacing the sensor component that will perform the capture (in the case of integrity inspection, the sensor remains the same) or the need to adjust the PTZ camera.

[0077] This subsystem assumes that the mission BT has already positioned the robot appropriately for inspection using the navigation BT.

[0078] In the autonomous electric recharging unit (docking station), step (e), the robotic system positions itself and recharges the batteries, and, if there is data to be transferred from a completed mission, it downloads the data to the local server, via wireless connection (Wi-Fi network), but expandable to wired connection.

[0079] Sequentially, in step (f), the data is processed according to the mission that was carried out and the results are made available in a report for the operator to issue a report.

[0080] Inspection reports can be exported in formats such as XLSX and PDF. The following options may be included, but are not required: • Refine the search using the Test Reason and Period filters. This option allows the user to obtain the list of inspections carried out in an optimized manner; • Export listing data to XLSX or PDF; • Create the report from the data collected during inspections; • Pagination of list items, i.e., increasing or decreasing the number of items to be displayed in the list.

[0081] The present invention has been disclosed in this descriptive report in terms of its preferred embodiment. However, other modifications and variations are possible from the present description, still being within the scope of the invention disclosed herein. List of references 1 - Mobile robotic platform 2 - Lifting platform 3 - 3D LIDAR 4 - Inspection camera 5 - Depth camera 6 - Security laser scanner 7 - GNSS RTK receiver (rover unit) 8 - GNSS RTK Antenna 9 - LED headlights 10 - Audible signal - Wi-Fi Router - Support and protection structure - Onboard computer - Female electrical charging structure (female docking) - Wireless E-Stop Receiver - Thermo-hygrometer - Smoke sensor - Sensor Attitude and Heading Reference System (AHRS) - Rear monocular camera - Male electrical load structure (docking mal e) - Antistatic tape - Parts - Female Docking and - Parts connector - Female conductor connector - Parts - Docking Mal and - Aluminum profile - Vertical guide clamp - Linear vertical guide - Linear bearing - Linear bearing clamp - Shock absorber - Fixed telescopic slide - Mobile telescopic slide - Circular bearing joint - Circular bearing - Circular bearing clamp - Male conductor connector - Spring 31475 Spring 102130 T spring 31416 41 - ArUco type fiducial markers

Claims

CLAIMS 1. Modular robotic system for inspection and detection in high-voltage electrical installations. CHARACTERIZED by comprising two main units: an autonomous ground robot unit and an autonomous electrical recharging unit;wherein the autonomous ground robot unit comprises at least one mobile robotic platform (1), at least one 3D LiDAR (3), at least one inspection camera (4), at least one depth camera (5), at least one security laser scanner (6), at least one GNSS RTK receiver (7), at least one GNSS RTK antenna (8), at least one Wi-Fi router (11), at least one support and protection structure (12), at least one on-board computer (13), at least one female electrical charging structure (14), at least one Wireless E-Stop receiver (15), at least one thermo-hygrometer (16), at least one smoke sensor (17), at least one Attitude and Heading Reference System (AHRS) sensor (18), at least one rear-facing monocular camera (19), at least one antistatic tape (21); and wherein the autonomous electrical charging unit comprises a male electrical charging structure (20).; 2. Robotic system, according to claim 1, CHARACTERIZED by the fact that the assembly of the robotic system occurs in the installation on the front bumper of the mobile robotic platform (1) a safety laser scanner (6), on the rear bumper a structure for female electrical charging (14); installation, in the upper front part and directly on the chassis of the mobile robotic platform (1) of a GNSS RTK receiver (7); installation of an antistatic tape (21), in the lower part and directly on the chassis of the mobile robotic platform (1); in the upper rear part of the chassis of the mobile robotic platform (1) and covered by the support and protection structure (12) installation of an on-board computer (13) and an Attitude and Heading Reference System unit (18); installation of a support and protection structure (12) on the mobile robotic platform (1) and on which are installed the 3D LiDAR sensor (3) and GNSS RTK antenna (8); in the front part a depth camera (5), rear monocular camera (19), a Wireless E-Stop receiver (15), a thermo-hygrometer (16), a smoke sensor (17) and a Wi-Fi router (11); the female electrical charging structure unit (14) is fixed to the rear part of the mobile robotic platform (1), in the front position equivalent to the bumper.

3. Robotic system, according to claim 1 or 2, CHARACTERIZED by further comprising a lifting platform (2) and / or two LED headlights (9) and / or an audible signal (10); wherein the headlights are installed on the front bumper of the mobile robotic platform (1); installation, in the upper front part and directly on the chassis of the mobile robotic platform (1), of an audible signal (10); in the front part an optional lifting platform (2).

4. Robotic system, according to claim 1, CHARACTERIZED by the safety laser scanner (6) having a horizontal angular opening of 275°, with a range of 40 m and a protection field of 4 m.

5. Robotic system, according to claim 1, CHARACTERIZED by the inspection camera (4) being selected according to the demand to be met, which may be one of the visual and thermographic pan, tilt and zoom (PTZ) type or hyperspectral type with rotating unit depending on the type of inspection to be performed.

6. Robotic system, according to claim 1, CHARACTERIZED by the main structure of the female electrical charging structure (14) comprising two pieces (22) that complement each other as two halves, being fixed to the mobile platform (1) by the connector of the pieces (23), and in which the passage of current for recharging the system is made from conductive connectors (24) of copper or other metallic material with good conductivity.

7. Robotic system, according to claim 1, CHARACTERIZED by the male electrical load structure (20) for the mechanical fitting system being assembled by the complement of two pieces (25) that form two halves and being compatible with the input of the female electrical load structure (14); it has a degree of freedom of rotation, being coupled on a circular bearing (35), in addition to containing conductors (37) for the passage of electrical current to the conductors (24) of the female electrical load structure (14).

8. Robotic system, according to claim 1 or 7, CHARACTERIZED by the male electrical load structure (20) having vertical linear movement, through a vertical linear guide (28); a fixed telescopic slide (32) for the horizontal linear movement of the coupled mobile telescopic slide (33) and by comprising shock absorbers (31).

9. Robotic system according to any one of claims 1 to 8, characterized in that it additionally comprises two other subsystems: a control to accommodate data server, GNSS RTK base station, electrical charging structure and the autonomous ground robot in idle / recharging state; and a set of Wi-Fi repeaters for wireless coverage in a 200,000 m high-voltage electrical substation 2 .

10. The robotic system of any one of claims 1 to 9, comprising a non-transitory computer-readable medium comprising instructions for executing a method of controlling the robotic system.

11. Method of controlling the robotic system, as defined in any one of claims 1 to 10, CHARACTERIZED by comprising the steps of: (a) mission selection: choose the robotic system operating mode and select the equipment and types of inspection to be performed in the same mission; (b) mission reception and subsystem verification: the robotic system receives authorization to start, performs a subsystem verification; (c) planning the trajectory and starting the journey; (d) navigation through autonomous operation mode or teleoperator operation mode, inspection (the robot positions itself appropriately and collects images of each piece of equipment), monitoring and, if necessary, recharging; (e) data download; (f) data processing.

12. The method of claim 11, wherein the autonomous ground robot unit remains in the control cabin attached to the structure. for electrical load until it receives a mission from the system operator interface (HMI) used by the operator through a mobile device or computer.

13. Method, according to claim 11, CHARACTERIZED by, in step (a), the operator can enter the Login and Password data to access the system functionalities; only users registered in the database and with a defined access profile can be authenticated by the system, and access to the functionalities is restricted to authentications; where the operator, aware of the maintenance planning, chooses the operating mode, which can be autonomous or teleoperator, and selects the equipment and types of inspection to be carried out in the same mission.

14. Method, according to claim 13, CHARACTERIZED by the functionalities for registering basic information for system configuration being available: • Equipment - items to be inspected by the robot; • Inspection site - hydroelectric power plants, for example; • Specific location - generating unit, turbine generating units, for example; • Reason for the Test - type / motivation of the inspection, for example: Daily, Monthly and Preventive; where the sequential interface follows the same usability flow pattern with the user: List screen - displays a list of the basic table with information on industrial plants / mills, equipment, generating units and reason for the selected inspection; • Registration / change screen - allows the creation or editing of a record in the basic table.

15. Method according to any one of claims 11 to 14, characterized in that the autonomous and teleoperation modes are inspection management modes; wherein in the autonomous mode, the robotic system performs the operation independently and allows the monitoring of an inspection online; and the teleoperation mode allows the user to guide the robot and control the inspection functions remotely.

16. Method according to any one of claims 11 to 15, CHARACTERIZED by, in step (b), the verification of the subsystems occurs by obtaining status via software commands on the operating or inoperative condition of the robot's sensors and actuators, the situation of which validates or aborts the continuation of the mission.

17. Method, according to any one of claims 11 to 16, CHARACTERIZED by which, in step (c), the robotic system makes use of global and local planners, where the global planning estimates a trajectory that runs through the various objectives based on the previously digitized map of the inspected area and then starts its displacement; the local planning acts on the short-term displacement, dealing directly with the detection and treatment of obstacles and making use of controllers to smooth the trajectory being executed in the vicinity of the robot's displacement; the interaction of the local planning with the global planning can be updated along the path due to possible presence of people, vehicles or animals on the planned path; the map, therefore, also undergoes updates as the robot executes the mission and encounters new obstacles or reconfigurations of the environment.

18. Method according to any one of claims 11 to 17, CHARACTERIZED in that the execution of all these tasks is adopted the behavior tree (BT) structure.

19. Method according to any one of claims 11 to 18, CHARACTERIZED in that the navigation, step (d), by the autonomous operation mode, moves the robotic platform to the destination point avoiding obstacles, thus using two services: dl) planning a route to a destination position from its current position avoiding obstacles; d.2) generating speed commands at 5 Hz for the platform, so that it can follow the planned route to the destination avoiding obstacles.

20. Method, according to any one of claims 11 to 19, CHARACTERIZED in that the navigation, step (d), additionally comprises requesting recovery services, in the event of a problem in the navigation in d1 and d.2, such as: i) clearing information on some obstacles that were included in the map; ii) rotating around its own axis; iii) waiting for a determined period of ls; iv) performing a reverse maneuver.

21. Method according to any one of claims 11 to 20, CHARACTERIZED in that the navigation is considered executed successfully, if the dl and d.2 services are executed and completed successfully, if any of them fail, some recovery action must be triggered; initially, action i is called and then the planning and generating commands actions are executed again, and after six attempts, if any of them continue to fail, actions i - iv can be triggered, in that order.

22. Method according to any one of claims 11 to 21, CHARACTERIZED by the fact that the navigation step will be executed until any of the following three conditions below are met: • The dl and d.2 services are successful; in this case, the navigation was successful and the robot was able to reach the destination point; • services i - iv failed; in this case, none of the recovery actions were able to make the robot perform actions dl and d.2, so navigation failed and the robot was unable to reach its destination; • The maximum number of 6 attempts to execute the dl and d.2 services is reached; in this case, navigation also failed, BT will enter recovery mode by performing the following steps: i. Clear information of some obstacles that were included in the map; ii. Rotate around its own axis; iii. Wait for a specified period of 1s; i. Perform a reverse maneuver.

23. Method according to any one of claims 11 to 22, CHARACTERIZED in that during navigation, step (d), by the teleoperator operating mode, it is expected that BT of the mission only supports the operator regarding the robot's states, assisting in the mission through outputs to the user interface; the user interface can give commands for robot movement and data capture, so that the inspection can be carried out.

24. Method according to any one of claims 11 to 23, CHARACTERIZED by the fact that the teleoperation operating mode is also responsible for registering the objects to be inspected.

25. Method according to any one of claims 11 to 24, CHARACTERIZED by the fact that still in step (d), the robot navigates the area, inspects and, if necessary, recharges its batteries until it completes the entire mission; the recharge occurring in the structure for initial electrical charging; additionally, battery charge level monitoring is carried out to ensure that the robot moves away from the base only to the minimum sufficient to return and recharge, even if a mission has not yet been completed, in which the robot then heads to the base for electrical recharging and keeps the mission suspended until it completes the recharge operation, can resume it and end it.

26. Method, according to any one of claims 11 to 25, CHARACTERIZED by the fact that throughout the route, the operator, remotely, can consult the progress of the mission, checking the path traveled, percentage of equipment inspected, time elapsed since the start, battery charge level and possible alerts issued.

27. Method, according to any of the claims 11 to 26, CHARACTERIZED by the fact that at the end of the mission, the robot returns to the base - structure for electrical charging.

28. Method according to any one of claims 11 to 27, CHARACTERIZED by the fact that the inspection task is performed by capturing data from the robot's cameras (4) so that an inspection object is framed in the image, the framing process can be done by a set of actuators, starting from the location and orientation of the robot, to adjustments of the Elevating Platform (2) and the PTZ module (if the sensor is in or on it), the use of optical zoom can also help, if necessary.

29. Method according to any one of claims 11 to 28, CHARACTERIZED by the fact that in the structure for electrical charging, step (e), the robotic system positions itself and recharges the batteries, and, if there is data to be transferred from a completed mission, it downloads the data to the local server, via a wireless connection (Wi-Fi network), but expandable to a wired connection.

30. Method according to any one of claims 11 to 29, CHARACTERIZED by the fact that, in step (f), the data are processed according to the mission that was performed and the results are made available in a report for the operator to issue a report; the following options may be optionally included: • Refine the search using the Test Reason and Period filters; this option allows the user to obtain the list of inspections carried out in an optimized manner; • Export listing data to XLSX or PDF; • Create the report from the data collected during inspections; • Pagination of list items, i.e., increasing or decreasing the number of items to be displayed in the list.

Images