Automatic flexible docking oil extraction system and method for transformer

Abstract

An automatic flexible docking oil extraction system and method for a transformer. The system comprises an oil extraction tank (1) and an oil extraction robot, wherein the oil extraction robot comprises an oil extraction robot chassis (3), and an automatic docking device (7) and an oil sample extraction device (5) which are mounted on the oil extraction robot chassis (3); the oil extraction tank (1) comprises a flexible docking oil extraction port (105), a fixed rod (102) and a movable rod assembly which are fixed on an oil extraction tank body (107); and the automatic docking device (7) comprises a guide docking mechanism (704) and a flexible docking oil extraction mechanism (701), wherein the guide docking mechanism (704) cooperates with the fixed rod (102) and the movable rod assembly for guide docking, and the flexible docking oil extraction mechanism (701) is flexibly docked with the flexible docking oil extraction port (105).

Description

Automatic Flexible Interlocking Oil Extraction System and Method for Transformers

[0001] This application claims priority to Chinese Patent Application No. 202411699050.3, filed with the Chinese Patent Office on November 26, 2024, the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of power robot technology, for example to an automatic flexible docking oil extraction system and method for transformers. Background Technology

[0003] Transformers are one of the key pieces of equipment in power systems, and their operating status directly affects the safety and stability of the power grid. Regularly collecting and analyzing transformer oil samples can detect the internal oil quality, promptly identify and diagnose abnormal faults such as arc discharge and overheating, thus providing early warning and preventing major accidents. However, manual transformer oil sampling is complex, labor-intensive, inefficient, susceptible to environmental interference, and poses serious safety risks. Therefore, an automated, flexible, docking-based transformer oil sampling system has been proposed.

[0004] However, the automatic flexible docking oil sampling system for transformers in the relevant technologies has the following problems: (1) Most of them rely on the robotic arm to drive the oil sampling connector to dock with the oil sampling port, which requires high precision robotic arm control, which increases the computing cost, and the control of the robotic arm is easily affected by the high temperature and electromagnetic influence of the power scene; (2) The oil sampling docking method mostly relies on the precise control of the oil sampling connector. Even if there is a certain floating design, the high precision of the robotic arm is required to achieve docking, which makes it difficult to dock the oil sampling connector with the oil sampling port; (3) The current automated oil sampling system lacks the self-checking function before work. The automated oil sampling system has to work near the transformer. The automatic oil sampling system's own fault (such as line fault or battery fault) will bring a large safety risk to the transformer. Summary of the Invention

[0005] This application provides an automatic flexible docking oil sampling system and method for transformers. By replacing manual operation with the automated cooperation of an oil sampling robot and an oil sampling tank, and by applying an automatic multiple flexible docking device, the operation process is optimized, the operation and maintenance efficiency and safety of the power system are improved, the risks of manual operation are reduced, and the high reliability of power supply is ensured. It has important application value.

[0006] The technical solution adopted in this application is as follows:

[0007] In a first aspect, this application provides an automatic flexible docking oil sampling system for transformers, comprising: an oil sampling tank and an oil sampling robot, wherein the oil sampling robot includes an oil sampling robot chassis and an automatic docking device and an oil sample extraction device installed on the oil sampling robot chassis;

[0008] The oil tank includes a flexible docking oil inlet fixed to the oil tank body, a fixed rod and a movable rod assembly. The automatic docking device includes a guiding docking mechanism and a flexible docking oil inlet mechanism. The guiding docking mechanism cooperates with the fixed rod and movable rod assembly to perform guiding docking, and the flexible docking oil inlet mechanism performs flexible docking with the flexible docking oil inlet.

[0009] Secondly, this application provides an active adaptive flexible docking method, utilizing the transformer automatic flexible docking oil sampling system described in the first aspect of this application, including:

[0010] The oil-collecting robot navigates quickly to the oil-collecting tank, moves slowly after reaching the initial position in front of the tank, and stops moving forward after reaching the set position.

[0011] The automatic docking device moves horizontally along the X direction. After the fixed rod touches the guide docking mechanism, it continues to move. After moving a set distance and confirming that the fixed rod is within the docking range of the automatic docking device, the automatic docking device continues to move forward. At this time, the movable rod assembly begins to be compressed. After being compressed a set distance, the oil tank lifting door opens, and the automatic docking device stops moving forward.

[0012] After the oil tank lifting door is opened, the automatic docking device continues to move horizontally along the X direction. After moving a set distance, it stops moving forward and detects the deflection angle data. The automatic docking device rotates and adjusts according to the deflection angle data to make the automatic docking device perpendicular to the oil tank cabinet surface, which facilitates automatic docking and oil extraction.

[0013] The pole-holding mechanism firmly holds the fixed pole. During the holding process, the automatic docking device moves and adjusts along the Y direction to make the flexible docking oil-taking mechanism and the flexible docking oil-taking port align. The flexible docking oil-taking mechanism moves forward until the flexible docking oil-taking port and the flexible docking oil-taking mechanism complete docking. Here, the X direction is the forward movement of the oil-taking robot, and the Y direction is the direction in the horizontal plane perpendicular to the X direction.

[0014] Thirdly, this application provides an active oil sample collection and washing method, utilizing the transformer automatic flexible docking oil sampling system described in the first aspect of this application, including:

[0015] The pipeline is flushed with an oil sample, and the waste oil from the flushing enters the waste oil collection device. Then, the first syringe and pipeline are flushed N times. Then, the first syringe is used to collect oil, where N is greater than or equal to 2.

[0016] The process of flushing and collecting oil from multiple syringes and tubing is repeated sequentially until all syringes have collected oil and the oil sample is collected. After this, the oil collection robot begins to withdraw.

[0017] Fourthly, this application provides a key landmark-assisted precise navigation and positioning method, utilizing the transformer automatic flexible docking oil sampling system described in the first aspect of this application, including:

[0018] In the substation scenario, key landmarks are selected;

[0019] Images of key landmarks are acquired from multiple angles using a depth camera. The acquired images are then geometrically corrected and denoised, and feature points are extracted to generate a feature image library of key landmarks. The feature images of key landmarks are then bound to GPS coordinates and azimuth information and stored as a searchable database.

[0020] The relative position of the oil-retrieving robot to the oil tank is initially determined using GPS and IMU, and it navigates according to a predetermined path. The movement path of the oil-retrieving robot is monitored in real time, the movement trajectory of the depth camera is analyzed, and the displacement and attitude of the oil-retrieving robot relative to the current environment are estimated.

[0021] Based on the target location and key landmarks, the route is planned. When the GPS signal is normal, the global location information is provided by GPS. If the GPS signal weakens, the navigation is switched to IMU-assisted navigation. The movement trajectory of the oil-retrieving robot is adjusted in real time based on the navigation data.

[0022] When the distance between the oil-collecting robot and the target location is less than a set threshold, relevant landmark images are retrieved from the searchable database, feature point matching is performed, the image that best matches the current environment with the searchable database is retrieved, and image mapping is performed based on the image matching results to determine the position and posture deviation of the oil-collecting robot in the current environment.

[0023] Calculate the relative pose of the depth camera with respect to key landmarks, and based on the position and pose deviation of the oil-retrieving robot in the current environment, combined with the known landmark position information, correct the displacement and pose of the oil-retrieving robot relative to the key landmarks.

[0024] Real-time images of key landmarks are continuously acquired at fixed acquisition intervals and image matching is performed to correct the robot's pose.

[0025] Based on multiple image matching results, it was confirmed that the oil-collecting robot had accurately reached the target location.

[0026] Fifthly, this application provides a fault self-diagnosis and remote assisted repair method for use in the transformer automatic flexible docking oil sampling system described in the first aspect of this application;

[0027] The transformer automatic flexible docking oil extraction system further includes: an autonomous operation master controller, a host computer control system, a navigation and positioning system, a mobile platform control system, a power management system, and an oil sample collection and detection control system arranged on the robot chassis. The power management system is connected to the power supply battery of the oil extraction robot, and the oil sample collection and detection control system is connected to the telescopic mechanism and the extraction mechanism respectively to perform telescopic and extraction control.

[0028] The navigation and positioning system includes a navigation controller. The autonomous operation master controller is communicatively connected to the host computer control system and the mobile platform control system. The mobile platform control system includes a motion controller, a motor driver, a drive motor, a steering motor driver, and a steering motor. The motion controller is communicatively connected to the navigation controller, the motion controller is connected to the motor driver, the motor driver is connected to the drive motor, the motion controller is connected to the steering motor driver, and the steering motor driver is connected to the steering motor.

[0029] The method includes:

[0030] The system performs self-tests on the power management system, autonomous operation main controller, host computer control system, navigation and positioning system, mobile platform control system, and oil sample collection and detection control system. The host computer control system sends all self-test results to the cloud server, which then generates auxiliary repair strategies based on these results and sends them to the host computer control system. The host computer control system then performs automatic repairs based on the received auxiliary repair strategies or sends the strategies to maintenance personnel.

[0031] Sixthly, this application provides a multimodal deep fusion-based embodied perception task monitoring method for use in the transformer automatic flexible docking oil sampling system described in the first aspect of this application, comprising:

[0032] The large model receives multimodal information from multiple sensors, performs comprehensive analysis on the real-time acquired multimodal information, judges anomalies or task deviations, and generates early warnings or handling solutions.

[0033] When the large model identifies potential anomalies or task deviations, it queries the built-in expert knowledge base to find expert advice and historical experience related to the anomalies or task deviations, and optimizes the oil extraction docking process based on the expert advice and historical experience. Attached Figure Description

[0034] Figure 1 is a schematic diagram of the overall structure of the transformer automatic flexible docking oil sampling system provided in this application;

[0035] Figure 2 is a partial structural schematic diagram of the transformer automatic flexible docking oil sampling system provided in this application;

[0036] Figure 3 is a schematic diagram of the overall structure of the oil tank provided in this application;

[0037] Figure 4 is a partial structural diagram of the oil tank provided in this application;

[0038] Figure 5 is a schematic diagram of the overall structure of the flexible docking oil intake provided in this application;

[0039] Figure 6 is a partial structural diagram of the oil tank provided in this application;

[0040] Figure 7 is a schematic diagram of the overall structure of the oil-collecting robot shell provided in this application;

[0041] Figure 8 is a schematic diagram of the overall structure of the oil sampling device provided in this application;

[0042] Figure 9 is a schematic diagram of the overall structure of the pipeline switching system provided in this application;

[0043] Figure 10 is a schematic diagram of the overall structure of the automatic docking device provided in this application;

[0044] Figure 11 is a partial structural schematic diagram of the automatic docking device provided in this application;

[0045] Figure 12 is a schematic diagram of the flexible docking oil sampling mechanism provided in this application;

[0046] Figure 13 is a schematic diagram of the pole-holding mechanism provided in this application;

[0047] Figure 14 is a schematic diagram of the deflection angle detection method provided in this application;

[0048] Figure 15 is a schematic diagram of the multiple control systems provided in this application;

[0049] Figure 16 is a schematic diagram of the navigation and positioning system provided in this application;

[0050] Figure 17 is a schematic diagram of the mobile platform control system provided in this application;

[0051] Figure 18 is a structural diagram of the connecting pipeline provided in this application;

[0052] Figure 19 is a flowchart of the active adaptive flexible docking method provided in this application;

[0053] Figure 20 is a flowchart of the active rinsing method for oil sample collection provided in this application;

[0054] Figure 21 is a flowchart of the key landmark-assisted precise navigation and positioning method provided in this application;

[0055] Figure 22 is a flowchart of the fault self-diagnosis and remote assisted repair method provided in this application;

[0056] Figure 23 is a flowchart of the multimodal deep fusion embodied perception task monitoring method provided in this application;

[0057] in:

[0058] 1. Oil tank; 101. Characteristic marker; 102. Fixed rod; 103. Movable rod; 104. Base; 105. Flexible docking oil inlet; 10501. Conical docking interface; 10502. Flat self-locking male connector; 10503. Adapter; 10504. First joint bearing; 10505. First bearing retaining ring; 10506. First adjusting connector; 10507. First docking interface adjusting spring; 10508. First baffle; 10509. Docking interface base; 105010. Docking interface bracket; 106. Oil tank lifting door; 107. Oil tank body; 108. Oil tank door; 109. Flat 110. Face lock; 111. Grounding busbar; 112. Transformer upper branch; 113. Transformer middle branch; 114. Transformer lower branch; 115. Circulating pump; 116. Oil return branch; 117. Ball valve; 118. First solenoid valve; 119. Oil intake branch; 120. Movable rod limit sensor; 121. Movable rod limit stop; 122. Linear bearing assembly; 123. Movable rod fixing seat; 124. Control unit; 125. Limit switch; 126. Through-type lead screw motor; 127. Through-type lead screw; 128. Lifting door guide rail; 129. Oil intake tank door frame; 130. Return spring; 130. Oil intake pipeline assembly;

[0059] 2. Oil-receiving robot shell; 201. LiDAR; 202. LiDAR mount; 203. Depth camera; 204. Front door; 205. Communication antenna; 206. Left door; 207. Magnetic door lock; 208. Emergency stop switch; 209. Anti-collision strip; 210. Rubber strip; 211. Reflector; 212. Docking sealing cover; 213. Right door; 214. Rear door; 215. Ultrasonic sensor;

[0060] 3. Oil extraction robot chassis; 301. Wheel and leg assembly; 302. Charging port; 303. Debugging interface assembly;

[0061] 4. Motion control module for the oil extraction robot;

[0062] 5. Oil sampling device; 501. Syringe nylon seat; 502. Syringe baffle; 503. Syringe; 504. Two-position two-way valve; 505. Polyurethane (PU) tubing; 506. Valve assembly bracket; 507. Rubber cap column; 508. Pipeline switching system; 50801. Secondary solenoid valve; 50802. Secondary oil outlet of the system; 50803. Pressure sensor; 50804. Valve assembly mounting base; 508041. Connecting pipeline; 50805. Booster pump; 50806. Pump mounting base; 50807. Cut-off... Replace the secondary oil inlet of the system; 50808, primary oil outlet of the system; 50809, primary oil inlet of the system; 50810, temperature sensor; 50811, reserved oil outlet; 50812, waste oil outlet; 509, rubber cap; 510, first oil receiving box; 511, elastic pin; 512, online detection sensor; 513, syringe tray; 514, electrical control module of oil collection system; 515, linear module limit sensor; 516, syringe push module; 517, module fixing base; 518, second oil receiving box; 519, oil sample detection camera;

[0063] 6. Waste oil collection device; 601. Weighing sensor; 602. Waste oil tank; 603. Fixing plate; 604. Straps;

[0064] 7. Automatic docking device; 701. Flexible docking oil extraction mechanism; 70101. Flat self-locking female connector; 70102. Pagoda connector; 70103. Fixed base adapter plate; 70104. Docking joint fixed base; 70105. Flange connector; 70106. Female connector sleeve; 70107. Second spherical bearing; 70108. Second bearing retaining ring; 70109. Second adjusting connector; 70110. Second docking interface adjusting spring; 70111. Second baffle;

[0065] 702. Telescopic mechanism; 703. Vertical frame; 704. Guide docking mechanism; 70401. Pole-holding mechanism connector; 70402. Rear sensor of fixed pole; 70403. Trigger plate bracket; 70404. Trigger plate of movable pole; 70405. Guide plate bracket; 70406. Front sensor of fixed pole; 70407. Guide plate; 70408. Pole-holding mechanism; 704081. Pole-holding seat; 704082. Claw; 704083. Rack; 704084. Guide rail pair; 704085. Drive motor housing; 704086. Pole-holding drive motor; 704087. Motor output shaft; 704088. Gear; 704089. Bearing; 70408 10. Opposite side gripper; 7040811. Pole clamping mechanism fixing plate; 70409. Docking detection camera; 704010. Guide docking mechanism fixing plate; 704011. Pole clamping mechanism adjusting plate; 705. Rotation mechanism; 70501. Large gear; 70502. Rotation drive motor; 70503. Small gear; 706. Y-axis moving slide; 70601. Cable chain; 70602. Y-axis compression spring; 70603. Zero position scale; 70604. Pointer; 70605. Linear motion component; 70606. Y-axis guide rail pair; 707. X-axis moving slide; 70701. X-axis drive motor; 70702. X-axis guide rail pair; 70703. X-axis lead screw and nut pair;

[0066] 8. Charging station; 801. Charging connector. Detailed Implementation

[0067] The present application will now be described in conjunction with the accompanying drawings and embodiments. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0068] In this implementation, an automatic flexible docking oil sampling system for transformers is proposed, as shown in Figures 1 and 2. It includes an oil sampling tank 1, an oil sampling robot, and a charging pile 8. The oil sampling robot includes an oil sampling robot chassis 3, an oil sampling robot shell 2, an automatic docking device 7, an oil sample extraction device 5, a waste oil collection device 6, and a motion control module.

[0069] As shown in Figure 3, the oil tank 1 mainly includes an oil tank body 107, a feature marker 101, a flexible docking oil outlet 105, an oil tank lifting door 106, an oil tank door 108, a fixed rod 102, a movable rod assembly, an oil pipeline assembly 130, a grounding bar 110, and a control unit 123, etc.

[0070] The oil tank 1 is installed on the base 104 (the base 104 is fixed to the ground). The inside of the oil tank door 108 is equipped with a silicone waterproof strip. A flat lock 109 is installed on the oil tank door 108. The oil tank door 108 and the oil tank body 107 are structurally pressed together to achieve waterproofing. A grounding bar 110 is provided on the outside of the oil tank body 107 to achieve grounding. A control unit 123 is installed inside the oil tank body 107 to control and detect the operating status of the oil tank 1 and communicate with the host computer.

[0071] The feature marker 101 is fixed above the oil tank body 107 for the oil-retrieving robot's lidar 201 to perform feature recognition, so as to locate and navigate the oil tank 1 and realize the docking of the oil-retrieving robot with the oil tank 1 when retrieving oil.

[0072] As shown in Figure 5, the flexible docking oil inlet 105 includes a tapered docking interface 10501, a planar self-locking male connector 10502, an adapter 10503, a first joint bearing 10504, a first bearing retaining ring 10505, a first adjusting connector 10506, a first docking interface adjusting spring 10507, a first baffle 10508, a docking interface base 10509, and a docking interface bracket 105010. The docking interface bracket 105010 is fixed on the oil tank body 107. A fixing rod 102 is fixed below the docking interface bracket 105010. The front end of the fixing rod 102 is designed as a semi-circular spherical surface and is made of sturdy stainless steel. When the oil-taking robot docks with the oil tank 1, the gripping mechanism 70408 grips the fixing rod 102 to realize the positioning adjustment and correction of the automatic docking device 7 in the up-down and left-right directions.

[0073] The interface base 10509 is fixed above the interface bracket 105010. The first bearing retaining ring 10505 and the first baffle 10508 fix the outer ring of the first spherical bearing 10504 to the interface base 10509. One end of the first adjusting connector 10506 passes through the inner ring of the first spherical bearing 10504 and connects to one end of the adapter 10503. The other end of the first adjusting connector 10506 leads out to the oil inlet. A first interface adjusting spring 10507 is installed between the first adjusting connector 10506 and the first baffle 10508. The tension of the first interface adjusting spring 10507 can be adjusted by turning the first adjusting connector 10506.

[0074] The other end of the adapter 10503 is fixed to one end of the planar self-locking male connector 10502. The planar self-locking male connector 10502 has functions such as self-locking, leakage prevention, dust prevention and impact resistance. The tapered mating interface 10501 is fixed to the other end of the planar self-locking male connector 10502, which can guide the planar self-locking female connector 70101 to achieve smooth docking and locking of the planar self-locking female connector 70101 and the planar self-locking male connector 10502. Under the combined action of the inner ring and outer ring of the first joint bearing 10504 and the first mating interface adjusting spring 10507, the docking oil port can achieve flexible rotation and swing at multiple angles.

[0075] As shown in Figure 6, the oil tank lifting door 106 is fixed on the door frame of the oil tank body 107. It mainly includes a lifting door, a limit switch 124, a through screw motor 125, a through screw 126, and two sets of lifting door guide rails 127. The lifting door guide rails 127 and the limit switch 124 are arranged on the oil tank door frame 128. The through screw motor 125 is fixed on the oil tank body 107. The through screw 126 is installed on the lifting door. The rotor of the through screw motor 125 rotates and drives the through screw 126 to move up and down, thereby realizing the lifting of the oil tank lifting door 106.

[0076] The movable rod assembly is fixed on the oil tank body 107 and located below the fixed rod 102. It includes two sets of linear bearing assemblies 121, movable rod fixing seat 122, two sets of movable rod limit sensors 119, movable rod limit baffles 120 of different lengths, and a return spring 129. The movable rod 103 is fixed on the movable rod fixing seat 122 and slides along the linear bearing assembly 121. The movable rod limit baffles 120 are triggered according to the different compression of the linear bearing assembly 121 to realize the opening and closing of the lifting door and the opening and closing of the first solenoid valve 117 on the oil pipeline assembly. The linear bearing assembly 121 is reset by the return spring.

[0077] The oil intake pipeline assembly is installed inside the oil intake tank 1, with its upper part connected to the oil inlet of the flexible docking oil intake port 105, as shown in Figure 4. The lower part of the oil intake branch 118 consists of three branches that connect to the upper branch 111, the middle branch 112, and the lower branch 113 of the transformer, and return the oil to the transformer via the circulating pump 114 and the return oil branch 115. The upper branch 111, the middle branch 112, and the lower branch 113 are each equipped with a first solenoid valve 117 and a ball valve 116, allowing for manual or automatic switching of the oil circuit. A solenoid valve 117 is a normally closed valve. A circulation pump 114 is installed at the front end of the return oil branch 115. The return oil branch 115 is also equipped with a first solenoid valve 117 and a ball valve 116. It is mainly used to realize the circulation of transformer upper branch 111, transformer middle branch 112 and transformer lower branch 113 with the transformer oil in order to obtain qualified oil samples. The oil sampling pipeline assembly can realize the connection between transformer upper branch 111, transformer middle branch 112 and transformer lower branch 113 and oil sampling branch 118 to complete the collection of three oil samples.

[0078] The oil-collecting robot chassis 3 includes the oil-collecting robot body and wheel-leg assembly 301, etc. The front end of the oil-collecting robot body is equipped with a charging port 302 and a debugging interface assembly 303, which can be charged and debugged autonomously. The rear end is equipped with a control panel with interfaces for debugging, oil-collecting robot switch, manual charging, etc. The wheel-leg assembly 301 adopts a four-wheel drive form, which is flexible, stable and reliable.

[0079] An automatic docking device 7, an oil sample extraction device 5, a waste oil collection device 6, and a motion control module are installed on the oil extraction robot chassis 3. The oil extraction robot shell 2 is installed on the oil extraction robot chassis 3, as shown in Figure 7. The oil extraction robot shell 2 and the oil extraction robot chassis 3 are tightly connected by a rubber strip 210. The oil extraction robot shell 2 includes a lidar 201, a lidar support 202, a depth camera 203, an ultrasonic sensor 215, a front door 204, a communication antenna 205, a left door 206, an emergency stop switch 208, a crash bar 209, a rubber strip 210, a docking sealing cover 212, a right door 213, and a rear door 214. A Global Positioning System (GPS) and an Inertial Measurement Unit (IMU) are arranged on the oil extraction robot chassis 3.

[0080] A collision protection strip 209 is installed at the front end of the oil-collecting robot shell 2, and ultrasonic sensors 215 are distributed on both sides of the bottom. A touch switch is installed inside the collision protection strip 209, which has the function of mitigating impact and alarming when a collision occurs. When the oil-collecting robot stops at the oil tank 1 with a deflection angle, the ultrasonic sensors 215 on both sides measure the distance, calculate the deflection angle data, and transmit it to the motion control module. The automatic docking device 7 adjusts the deflection angle.

[0081] The lidar support 202 is set above the front end of the oil-receiving robot shell 2. The lidar 201 is installed above the lidar support 202 to realize the positioning and navigation of the oil-receiving robot. The depth camera 203 is arranged below the lidar support 202 to monitor the working environment in real time.

[0082] A docking sealing cover 212 is installed inside the front end of the oil-collecting robot shell 2 to prevent dust from entering the automatic docking device 7. The oil-collecting robot shell 2 is equipped with a front door 204, a rear door 214, a left door 206, and a right door 213 on its four sides. The shell 2 is made of acrylonitrile butadiene styrene (ABS) plastic. The rear door 214, left door 206, and right door 213 are all equipped with magnetic door locks 207 for convenient on-site operation and maintenance. The front door 204 is a lifting door that can automatically rise and fall during automatic docking. A reflector 211 is installed on the inner side of the left door 206 to allow the oil sample detection camera 519 to reflect the oil sample extraction details from the syringe 503.

[0083] The method for detecting the deflection angle of the ultrasonic sensor 215 is shown in Figure 14. The distance a between the two ultrasonic sensors 215 is known. By detecting the distances l1 and l2 from the two ultrasonic sensors to the surface of the oil tank 1, the deflection angle θ of the oil-retrieving robot relative to the oil tank 1 when it stops can be calculated using the formula θ=arctan((l2-l1) / a). If the value of θ is positive, the automatic docking device 7 rotates clockwise; if the value of θ is negative, the automatic docking device 7 rotates counterclockwise to ensure that the automatic docking device 7 is perpendicular to the surface of the oil tank 1, which facilitates automatic docking and oil extraction.

[0084] As shown in Figures 10, 11, 12, and 13, the automatic docking device 7 includes a guiding docking mechanism 704, a flexible docking oil-taking mechanism 701, an X-axis moving slide 707, a Y-axis moving slide 706, a telescopic mechanism 702, a vertical frame 703, and a rotating mechanism 705. The guiding docking mechanism 704 is fixed to the vertical frame 703 via a guiding docking mechanism fixing plate 704010. The guiding docking mechanism 704 includes a pole-holding mechanism adjusting plate 704011, a pole-holding mechanism connector 70401, a pole-holding mechanism 70408, a guide plate 70407, a guide plate bracket 70405, a movable rod trigger plate 70404, a trigger plate bracket 70403, a fixed rod front sensor 70406, a fixed rod rear sensor 70402, a pole-holding mechanism fixing plate 7040811, and a guiding docking mechanism fixing plate 704010.

[0085] The pole-holding mechanism 70408 is fixed on the guide docking mechanism fixing plate 704010 via the pole-holding mechanism adjusting plate 704011 and the pole-holding mechanism connecting piece 70401. The pole-holding mechanism 70408 includes a pole-holding seat 704081, a gripper 704082, a opposite gripper 7040810, a gear 704088, a rack 704083, a guide rail pair 704084, a pole-holding drive motor 704086, a drive motor housing 704085, a motor output shaft 704087, a bearing 704089, and a pole-holding mechanism fixing plate 7040811. By adjusting the position of the pole-holding mechanism fixing plate 7040811 on the pole-holding mechanism adjusting plate 704011, the position of the pole-holding mechanism 70408 can be adjusted to better grip the fixed pole 102.

[0086] The mast drive motor 704086 is fixed on the mast seat 704081. Gear 704088 and bearing 704089 are mounted on the motor output shaft 704087. Bearing 704089 provides support. Rack 704083 is arranged on both sides of gear 704088 and meshes tightly with it. Guide rail pair 704084 is arranged on rack 704083 for guiding and load-bearing functions. Claw 704082 and opposite claw 7040810 are respectively mounted on… Mounted on both sides of the rack 704083, the gripper 704082 and the opposite gripper 7040810 are designed as single-claw and double-claw structures respectively. When the gear 704088 rotates, it drives the rack 704083 and the gripper 704082 to move relative to each other. The gripper 704082 and the opposite gripper 7040810 overlap together, which can firmly hold the fixing rod 102 without slipping. The drive motor housing 704085 is installed on the outside of the rod drive motor 704086 for protection.

[0087] The guide plate 70407 and two sets of guide plate brackets 70405 are connected and fixed on the guide docking mechanism fixing plate 704010. The guide plate 70407 is designed with a smooth slope structure and is made of hard and wear-resistant material. During operation, the fixing rod 102 of the oil tank 1 remains stationary, and the front end of the guide plate 70407 contacts the spherical surface of the fixing rod 102 and slides passively along the slope to lift the guide docking mechanism 704.

[0088] The movable rod trigger plate 70404 is connected to the guide plate bracket 70405 through two sets of trigger plate brackets 70403. The movable rod trigger plate 70404 is made of hard and wear-resistant material. During operation, the movable rod trigger plate 70404 moves closer to the oil tank 1. The front of the movable rod trigger plate 70404 contacts the movable rod 103, pushing the movable rod 103 to move, thereby triggering different switches to realize the lifting and lowering of the oil tank lifting door 106 and the on and off functions of the first solenoid valve 117.

[0089] The front sensor 70406 and the rear sensor 70402 of the fixed rod are respectively installed at the front and rear ends of the guide plate bracket 70405, and are triggered by the fixed rod 102 during operation.

[0090] The docking detection camera 70409 is installed at the front of the guide docking mechanism 704, which can observe the status of the oil-taking robot docking with the oil-taking tank 1 in real time and report any problems in a timely manner.

[0091] The flexible docking oil extraction mechanism 701 is mounted on the telescopic mechanism 702 and moves with the telescopic mechanism 702. The flexible docking oil extraction mechanism 701 mainly includes a planar self-locking female connector 70101, a pagoda connector 70102, a fixed seat adapter plate 70103, a docking connector fixed seat 70104, a flange connector 70105, a female connector sleeve 70106, a second joint bearing 70107, a second bearing retaining ring 70108, a second adjusting connector 70109, a second docking interface adjusting spring 70110, a second baffle 70111, etc. The fixed seat adapter plate 70103 is mounted on the telescopic mechanism 702, and the docking connector fixed seat 70104 is mounted on the fixed seat adapter plate 70103.

[0092] The second bearing retaining ring 70108 and the second baffle 70111 fix the outer ring of the second spherical plain bearing 70107 to the mating joint fixing seat 70104. One end of the second adjusting joint 70109 passes through the inner ring of the second spherical plain bearing 70107 and connects to one end of the flange joint 70105. The other end of the second adjusting joint 70109 connects to the pagoda joint 70102 and leads out an oil outlet. The pagoda joint 70102 delivers the oil sample to the oil sample extraction device 5 through the PU tube 505. A second pair of interface adjusting springs 70110 is installed between the second adjusting joint 70109 and the second baffle 70111. The tension of the spring can be adjusted by turning the second adjusting joint 70109. The other end of the flange joint 70105 is fixed with a flat self-locking female joint 70101. The flat self-locking female joint 70101 has functions such as self-locking, leakage prevention, dust prevention, and impact resistance. A female joint sleeve 70106 is installed on the outside of the flat self-locking female joint 70101. Under the guidance of the tapered mating interface 10501, the flat self-locking female joint 70101 achieves smooth docking and locking with the flat self-locking male joint 10502. Under the combined action of the inner and outer rings of the second joint bearing 70107 and the adjusting spring, the mating oil outlet can achieve flexible rotation and swing at multiple angles.

[0093] The X-axis moving slide 707 includes two sets of guide rail pairs, one set of lead screw and nut pairs, a drive motor, and a structural body. The X-axis moving slide 707 mainly realizes the contact between the automatic docking device 7 and the oil sampling pile guide rod and the movable rod 103. The X-axis moving slide 707 includes an X-axis drive motor 70701, an X-axis guide rail pair 70702, and an X-axis lead screw and nut pair 70703.

[0094] The Y-axis moving slide 706 adopts a follower structure, including a Y-axis guide rail pair 70606, a linear motion assembly 70605 (linear bearing + guide rod assembly), a Y-axis compression spring 70602, a drag chain 70601, a zero-position scale 70603, and a pointer 70604. Y-axis compression springs 70602 are respectively installed on the guide rods on both sides of the linear bearing. When no external force is applied, the Y-axis moving slide 706 is at the zero position. When subjected to external force from the system, the Y-axis moving slide 706 follows the movement within the allowable range.

[0095] The telescopic mechanism 702 is rigidly connected to the automatic docking device 7 via a connecting plate. The telescopic mechanism 702 adopts a screw + guide rail configuration, and the drive unit is a servo motor, which enables the telescopic mechanism 702 to accurately and stably perform the telescopic function.

[0096] The vertical frame 703 adopts a multi-ribbed plate structure, which is lightweight and sturdy and can withstand the large thrust impact when the automatic docking device 7 docks horizontally.

[0097] The rotating mechanism 705 includes a rotary drive motor 70502, a large gear 70501, and a small gear 70503. The large gear 70501 is fixed on the upper part of the Y-axis movable slide 706. The rotary drive motor 70502 is installed at the bottom of the vertical frame 703. The small gear 70503 is installed on the output shaft of the rotary drive motor 70502. The small gear 70503 meshes with the large gear 70501. When the rotary drive motor 70502 rotates, it drives the small gear 70503 to rotate along the large gear 70501, thereby realizing the rotation of the vertical frame 703.

[0098] As shown in Figures 8 and 9, the oil sampling device 5 includes an oil sample pushing device, a pipeline switching system 508, a first oil receiving box 510, a second oil receiving box 518, and an oil sampling system electrical control module 514.

[0099] The oil sample delivery device includes a syringe nylon seat 501, a syringe baffle 502, a syringe 503, a two-position two-way valve 504, a flexible tube (e.g., a PU tube 505), an elastic pin 511, an online detection sensor 512, a syringe tray 513, a linear module limit sensor 515, a syringe delivery module 516, and a module fixing seat 517. The syringe nylon seat 501 has six D-shaped opening slots for syringe barrels. A syringe baffle 502 is installed at each opening slot to secure the syringe 503 and prevent it from falling out. An online detection sensor 512 is installed inside the D-shaped opening slots to detect whether the syringe 503 is properly installed. The syringe nylon seat 501 has a handle on each side. The syringe nylon seat 501 is mounted on the module fixing seat 517 via elastic pins 511 at both ends. By removing the elastic pins 511, the handles can be grasped to remove the syringe nylon seat 501 along with the six sets of extracted oil samples from the syringes 503, and deliver them to the laboratory for testing.

[0100] The syringe tray 513 is set on the syringe push module 516 to support the piston handle of the syringe 503. The syringe 503 is made of glass to facilitate observation of the internal liquid. The syringe outlet is equipped with a two-position two-way valve 504 and is connected to the six outlets of the pipeline switching system 508 through a PU tube 505. When the oil sample extraction is completed, the two-position two-way valve 504 is closed. The PU tube 505 is a high-temperature resistant, corrosion-resistant, and explosion-proof oil tube.

[0101] The syringe pushing module 516 is fixed on the module fixing base 517. The syringe pushing module 516 mainly consists of a motor, lead screw, guide rail, structural body, and linear module limit sensor 515. The syringe pushing module 516 pushes the oil sample from the syringe 503 through the PU tube 505 and the pipeline switching system 508 to the waste oil collection device 6. The oil sample is pushed smoothly and accurately, and is limited by the linear module limit sensor 515.

[0102] The valve group bracket 506 is equipped with a pipeline switching system 508, an oil sample detection camera 519, a rubber cap column 507, and a rubber cap 509. The rubber cap 509 is fitted onto the rubber cap column 507 to block the oil outlet of the syringe 503. The oil sample detection camera 519 is automatically illuminated, and the oil sample collection details are observed by looking at the image of the syringe 503 in the reflector 211 on the back of the left door 206. The pipeline switching system 508 includes a second solenoid valve 50801, a valve assembly mounting base 50804, a pressure sensor 50803, a temperature sensor 50810, a booster pump 50805, and a pump mounting base 50806. The second solenoid valve 50801 realizes pipeline on / off switching. The booster pump 50805 is fixed on the pump mounting base 50806. The valve assembly mounting base 50804 is designed with a connecting pipeline 508041 inside, as shown in Figure 18. The primary oil inlet 50809 of the system is connected to the primary oil outlet 50808 of the system. The secondary oil inlet 50807 of the switching system is connected to the secondary oil outlet 50802, waste oil outlet 50812, and reserved oil outlet 50811 of the six-way system. The booster pump 50805 is installed between the primary oil outlet 50808 of the system and the secondary oil inlet 50807 of the switching system. The booster pump 50805 is used to flush the pipeline and extract oil samples. A temperature sensor 50810 and a pressure sensor 50803 are installed between the primary oil inlet 50809 and the primary oil outlet 50808 of the system. The temperature sensor 50810 is used to detect the temperature of the transformer oil, and the pressure sensor 50803 is used to detect the pressure of the transformer oil. Based on the detection results, an alarm is triggered or the operation is stopped. The secondary oil outlet 50802 of the six-way system is connected to the syringe 503 through the PU tube 505 to inject the oil sample into the syringe 503. A pressure sensor 50803 is installed at the front end of the outlet to detect the pressure injected into the syringe 503 and prevent the syringe 503 from bursting due to excessive pressure.

[0103] The electrical control module 514 of the oil sampling system is located on the back side of the module fixing base 517. The oil sample leaked by the pipeline switching system 508 is collected through the first oil collection box 510 and transported to the first oil collection box 510 through the PU tube 505. The oil sample leaked by the oil sample pushing device is collected by the second oil collection box 518. The second oil collection box 518 can be manually removed when it is full.

[0104] The waste oil collection device 6 includes a weighing sensor 601, a waste oil tank 602, a fixing plate 603, and a strap 604. The weighing sensor is used to measure the amount of waste oil discharged, the waste oil tank is used to collect waste oil used for cleaning pipelines, and the fixing plate and strap are used to fix the waste oil tank above the weighing sensor.

[0105] The motion control module includes an oil-collecting robot motion control module 4 and an oil-collecting system electrical control module 514. The oil-collecting robot motion control module 4 has functions such as navigation, mobile oil-collecting robot trajectory planning, real-time monitoring of the on-site working environment, real-time monitoring of the on-site scene, and data transmission. The oil-collecting system electrical control module 514 can perform tasks such as automatic docking of the automatic docking device 7 with the oil tank 1, opening and closing of the first solenoid valve 117 in the oil tank 1, raising and lowering of the lifting door, and flushing and extracting the oil sample extraction device 5, ensuring that the operation tasks are completed accurately and in real time.

[0106] The charging pile 8 is placed in a suitable location in the substation. The charging connector 801 adopts a telescopic structure. The oil-retrieving robot stops at a suitable location on the charging pile 8 through positioning and navigation. The charging connector 801 automatically extends and connects to the charging port 302 of the oil-retrieving robot to achieve autonomous charging.

[0107] The working method of the above-mentioned automatic flexible docking oil sampling system for transformers includes:

[0108] Based on the positioning and navigation of the lidar, the oil-collecting robot moves quickly towards the oil-collecting tank, then moves slowly after reaching the set initial position in front of the oil-collecting pile, and stops moving forward after reaching the set position of the oil-collecting pile.

[0109] At this time, the automatic docking device moves horizontally along the X direction. After the fixed rod touches the lower edge of the guide plate, it continues to move, triggering the sensor in front of the fixed rod. After confirming that the status is normal, the automatic docking device continues to move forward. At this time, the trigger plate of the movable rod touches the movable rod, and the movable rod begins to be compressed. After being compressed by a set distance, the limit switch is triggered, the oil tank lifting door is raised, and the automatic docking device stops moving forward.

[0110] After the oil tank lifting door is raised, the automatic docking device continues to move horizontally in the X direction. After the sensor behind the fixed rod is triggered, it stops moving forward. The deflection angle data is detected by the ultrasonic sensor. The automatic docking device rotates and adjusts according to the deflection angle data to make the automatic docking device perpendicular to the oil tank cabinet surface, which facilitates automatic docking and oil extraction.

[0111] The pole-holding mechanism starts to move and firmly holds the fixed pole. During the holding process, the automatic docking device moves and adjusts along the Y direction so that the planar self-locking female joint of the flexible docking oil-taking mechanism is within the conical interface range of the flexible docking oil-taking port.

[0112] The telescopic mechanism moves, driving the flexible docking oil extraction mechanism forward. The planar self-locking female connector of the flexible docking oil extraction mechanism enters the conical interface of the flexible docking oil extraction port until the planar self-locking male connector of the flexible docking oil extraction port and the planar self-locking female connector of the flexible docking oil extraction mechanism are docked.

[0113] Once a suitable pipeline is connected in the oil tank, the oil sampling device begins to collect oil samples. It is understood that the suitable pipeline can be the upper branch, middle branch, or lower branch of the transformer. This application can select to take oil from the upper branch, middle branch, or lower branch of the transformer by controlling the solenoid valve.

[0114] Temperature and pressure sensors provide real-time feedback of oil circuit information, and the booster pump increases pressure in real time.

[0115] First, the oil sample pipeline is flushed, and the waste oil enters the waste oil collection device. Then, the No. 1 syringe and pipeline are flushed three times. Next, the No. 1 syringe is used to collect oil.

[0116] Syringes No. 2 through No. 6 follow the same pattern; different syringes can be used to collect oil samples from different transformers.

[0117] After the oil sample collection is completed, the oil extraction robot begins to exit, and the exit process is the reverse of the docking process.

[0118] Optionally, in other implementations, the transformer automatic flexible docking oil extraction system, as shown in Figure 15, may also include a host computer control system, an autonomous operation master controller, a power management system, a navigation and positioning system, a mobile platform control system, and an oil sample collection and detection control system. The power management system is connected to the power supply battery of the oil extraction robot, and the oil sample collection and detection control system is connected to the telescopic mechanism and the extraction mechanism respectively for telescopic and extraction control.

[0119] The host computer control system includes a remote controller and a back-end system. Users can choose to issue operation commands through the back-end system and have the oil-collecting robot perform tasks autonomously, or they can choose to use the remote controller so that operators can control the oil-collecting robot to perform oil sample collection tasks.

[0120] The autonomous operation master controller interacts with other systems through communication modules such as network, serial port, and Controller Area Network (CAN) bus. After receiving the task from the host computer control system, the autonomous operation master controller decomposes the task into sub-tasks and issues them to the corresponding systems according to the operation logic. It waits for the system to report the execution status of the sub-tasks. After receiving the feedback of successful execution, it executes the next sub-task until the overall task is completed.

[0121] The power management system includes a voltage conversion module and a charging management module. The voltage conversion module converts the 48VDC power supply of the mobile platform control system into 12VDC and 24VDC to power the corresponding devices. The charging management module monitors the current power of the mobile platform control system in real time. When the power is below 20%, an alarm message is issued. When it is below 10%, the current task is stopped. The mobile platform control system autonomously goes to the charging pile location to charge. After the battery is fully charged, it disconnects from the charging pile and resumes task execution.

[0122] Optionally, in other implementations, as shown in Figure 16, the navigation and positioning system includes a navigation controller, a lidar, an inertial measurement unit (IMU), a depth camera, and a global positioning system (GPS) (i.e., the positioning unit). The lidar and depth camera collect information about the external environment, the IMU acquires the attitude and acceleration information of the mobile platform control system and, together with GPS, provides accurate positioning information, and the navigation controller receives the above information to construct a map of the external environment, thereby realizing the navigation and positioning function of the oil-collecting robot.

[0123] Optionally, in some other implementations, as shown in Figure 17, the mobile platform control system includes a motion controller, motor drivers (including the first motor driver, the second motor driver, the third motor driver, and the fourth motor driver), drive motors (including the first drive motor, the second drive motor, the third drive motor, and the fourth drive motor), steering motor drivers (including the first steering motor driver, the second steering motor driver, the third steering motor driver, and the fourth steering motor driver), and steering motors (including the first steering motor, the second steering motor, the third steering motor, and the fourth steering motor). The motion controller receives movement commands from the navigation controller, controls the oil-collecting robot to reach the oil sample collection location, and sends control commands to the motor drivers via the CAN bus. The motor drivers generate corresponding pulse width modulation (PWM) signals according to the commands to control the speed and direction of the drive motors.

[0124] Understandably, in some other implementations, the oil sample collection and detection control system may include an oil sample collection and detection controller, an oil sample collection module, an oil-gas separation module, a gas detection module, and a data analysis module. After the oil circuit is opened, the oil sample collection and detection controller sends commands to the oil sample collection module, the oil-gas separation module, the gas detection module, and the data analysis module according to the instructions of the autonomous operation controller of the oil sampling robot. Based on the feedback information from all modules, the system autonomously completes the entire process from oil sample collection to data analysis.

[0125] Optionally, this implementation also provides a self-test method for the automatic flexible docking oil sampling system of a transformer, including the following process:

[0126] (1) After the oil-retrieving robot is powered on, the power indicator light will light up, indicating that the power has been connected. At this time, the power management system inside the oil-retrieving robot will begin to check the status of the battery pack, including whether the battery power, voltage and temperature are within the normal range.

[0127] Battery power detection: The built-in power sensor accurately measures the current remaining battery power and compares it with a preset safety threshold. If the power level is lower than the preset safety threshold, a low battery warning is issued and the user is prompted to charge.

[0128] Subsystem power management check: By reading the voltage and current data of the power management unit of each subsystem, it is determined whether the power supply part of each subsystem is normal; the subsystems include the host computer control system, navigation and positioning system, mobile platform control system, and oil sample collection and detection control system.

[0129] (2) Self-test of autonomous operation controller. After the autonomous operation controller is powered on, the bootloader will run automatically. The bootloader will first check whether the serial port, network port, Universal Serial Bus (USB) port and other hardware of the autonomous operation controller are working properly. If any component is abnormal, an error alarm will be issued. If the check is normal, it will start to check whether there is a new version of code in the download area that has not been updated. If there is, the new version of code will be copied from the download area to the application (APP) area, and then the application in the APP area will be started to boot. If there is no new version of code, the application in the APP area will be started directly.

[0130] (3) The host computer control system performs a self-test. After the host computer control system is powered on, it first reads the version information of the host computer software and compares it with the list of compatible versions stored internally. If there is a new version of the software, it will update the application. Then, it starts to check the network connection between the host computer control system and the autonomous operation controller. If the network connection fails, an error alarm will be issued. After the connection is normal, it starts to read the relevant configuration parameters of the autonomous operation controller and displays the relevant configuration parameters on the screen. It also verifies whether the relevant configuration parameters meet the current operation requirements. If they do not meet the requirements, an error alarm will be issued to prompt the operator to check and reset the relevant configuration parameters.

[0131] (4) Navigation and positioning system self-test: After the navigation and positioning system is powered on, it first checks whether the hardware of the GPS receiver, inertial measurement unit (IMU) and lidar is normal. If there is a hardware abnormality, an error alarm will be issued. At the same time, it confirms whether each sensor has been correctly initialized and is in working condition. Then, it starts power-on repositioning and positioning accuracy calibration. The accuracy of the navigation and positioning system is evaluated by comparing the difference between the actual position and the expected position. For the IMU, zero bias calibration and gyroscope drift test are performed to determine whether the IMU meets the required accuracy. If it does not meet the requirements, an alarm message will be issued.

[0132] (5) Self-test of mobile platform control system: After the mobile platform control system is powered on, the motors and drivers on the mobile platform control system are checked one by one to confirm whether the communication is normal and whether there is a fault alarm. If there is an error, an alarm is issued.

[0133] (6) Self-test of oil sample collection and detection control system: After the oil sample collection and detection control system is powered on, it first resets the six sets of syringes. During the reset, it detects the photoelectric detection switch of each syringe drive module. When a photoelectric detection switch outputs a signal, the reset stops and reads the encoder value of each syringe drive motor as the zero point of each syringe movement. During the reset process, the output torque of the six syringe drive motors will be continuously detected. When the output torque reaches the alarm threshold, the drive motor movement will be stopped immediately and an alarm message will be issued to prompt the user that the syringe reset is abnormal.

[0134] After all six components have completed their self-tests, the oil extraction robot enters a ready state, awaiting operation control commands.

[0135] Optionally, a process for a robot to navigate to an oil collection tank is also provided, including:

[0136] Before the first operation, a map of the working environment needs to be built. Click the map building button in the background system and the oil-collecting robot will check the connection status of sensors such as lidar, IMU, depth camera, and GPS. If no sensor data is received, the system will issue a corresponding prompt to remind the staff to check the corresponding sensor hardware connection. If all sensors are connected and in good condition, the system will enter the map building mode.

[0137] After entering the mapping mode, the operator uses a remote control to move the oil-collecting robot in the outdoor environment, using multiple sensors to collect environmental information. During this process, the oil-collecting robot will continuously update its position and posture, while recording the feature points of the surrounding environment.

[0138] After traversing the surrounding environment, click the "Build Map" button in the backend system. The backend system will generate a map of the current environment based on the collected environmental information. After the map is successfully generated, it can be manually edited. Virtual walls can be generated in areas of the substation where passage is prohibited, and the work location can be marked as the target point. This allows the navigation and positioning system to automatically block prohibited areas during the oil extraction robot's journey to the target point, preventing safety accidents. After the map is edited, click the "Save Map" button to name the created map. The map will be saved for use in executing tasks.

[0139] When performing an oil extraction task, select the corresponding map, check the target point, and click "Execute Task." The oil extraction robot will then begin navigating to the work site. During navigation, the oil extraction robot first confirms its position relative to the map. It performs self-localization by rotating around its center position to scan the surrounding environmental features and matching them with map features. To simplify the matching process and shorten the operation time, the program sets a default preset point as the oil extraction robot's starting position. It then matches the current environmental features around the oil extraction robot with the features of that starting position on the map. If the oil extraction robot starts at this preset point, the matching process will be completed instantly.

[0140] After the oil-collecting robot completes its self-localization, it combines the current location information, target location information, and obstacle information between the two locations to generate an optimal global path plan. The oil-collecting robot will move along this path. Based on the real-time path information, a local path plan is generated to adjust the global path plan, so that the oil-collecting robot avoids obstacles and moves along the real-time relatively optimal path.

[0141] The process of generating global and local path planning incorporates information about cable trench covers on the map. During mapping, the locations of cable trench covers are visually identified and recorded. Weights are added to the global path planning to ensure that paths with high cable trench cover coverage are optimal. Similarly, paths along cable trench covers are selected as locally optimal, guiding the oil-receiving robot to navigate along these covers rather than on outdoor lawns, thus ensuring a smooth and safe journey to the target location. Upon arrival at the target location, the position of the characteristic marker 101 on top of the oil-receiving pile is identified, allowing for fine-tuning of the robot's position and precise docking accuracy.

[0142] Optionally, this implementation also proposes an embodied perception task monitoring method. Through a large model and expert knowledge base, the safety, accuracy and success rate of the automated oil extraction process can be further improved. Based on the real-time acquisition and fusion of multimodal information, intelligent judgment and response can be made by perceiving the task status of the oil extraction robot, environmental changes and potential anomalies.

[0143] This implementation method collects and fuses multimodal information from different sensors (such as vision, lidar, ultrasound, IMU, force sensor, etc.) to monitor the task status of the oil sampling robot in real time during the process of collecting transformer oil samples. It can judge the progress of the operation, capture abnormal behaviors (such as docking failure, oil sample abnormality, etc.), and reduce the risk of dangerous operations through preset safety control strategies.

[0144] In this implementation, the large model receives data from various sensors, including visual cameras, lidar, ultrasonic sensors, IMUs, force sensors, etc. (all of which can be mounted on the oil-collecting robot).

[0145] The large model comprehensively analyzes the multimodal information collected in real time, judges the current docking status of the oil extraction robot through visual information, and uses force sensor data to detect abnormal forces that may exist during the docking process. During the operation, if abnormal sensor data is detected, the large model can infer the potential cause of the failure by semantic understanding, comparing historical data and failure modes, and generating early warnings or handling solutions.

[0146] When the large model identifies potential anomalies or task deviations, it queries the built-in expert knowledge base to find expert advice and historical experience related to the current anomaly or task deviation. For example, if an abnormal oil pressure is detected during oil sample collection, the system can recommend shutdown or recalibration based on the fault type matching process in the knowledge base.

[0147] The expert knowledge base includes best practices for transformer maintenance, troubleshooting strategies, and oil sample testing standards. For example, when the system determines that the oil sampling docking has failed, the expert knowledge base will provide specific troubleshooting procedures (such as checking the docking angle and cleaning the docking interface) and suggest adjusting the posture of the oil sampling robot and its auxiliary components.

[0148] Optionally, this implementation also achieves collaborative optimization between the large model and the expert knowledge base, including:

[0149] (1) Dynamic decision-making and feedback learning: The large model can adjust the task monitoring strategy in real time and generate optimized operation procedures based on the results of multimodal perception. By combining with the expert knowledge base, the system can feed back the actual operation of the task to the expert knowledge base and continuously optimize and expand the expert knowledge base. For example, in the process of oil sample collection, the system gradually optimizes the pressure control algorithm for oil extraction through multiple operation feedback learning.

[0150] (2) Anomaly detection and emergency handling: The big model automatically generates an in-depth analysis report on operational anomalies by fusing visual and sensor data. For example, if a docking posture deviation is detected during the operation, the big model will generate posture adjustment suggestions and automatically correct the in-depth analysis report in combination with the processing scheme of the expert knowledge base.

[0151] As shown in Figure 19, this application also provides an active adaptive flexible docking method, including:

[0152] S101, the oil-collecting robot navigates quickly to the oil-collecting tank, moves slowly after reaching the set initial position in front of the oil-collecting tank, and stops moving forward after reaching the set position of the oil-collecting tank;

[0153] S102. The automatic docking device moves horizontally along the X direction. After the fixed rod touches the guide docking mechanism, it continues to move. After moving a set distance and confirming that the fixed rod is within the docking range of the automatic docking device, the automatic docking device continues to move forward. At this time, the movable rod assembly begins to be compressed. After being compressed a set distance, the oil tank lifting door opens, and the automatic docking device stops moving forward.

[0154] S103. After the oil tank lifting door is opened, the automatic docking device continues to move horizontally in the X direction. After moving a set distance, it stops moving forward and detects the deflection angle data. The automatic docking device rotates and adjusts according to the deflection angle data to make the automatic docking device perpendicular to the oil tank cabinet surface, which facilitates automatic docking and oil extraction.

[0155] S104. The clamping mechanism firmly clamps the fixed rod. During the clamping process, the automatic docking device moves and adjusts along the Y direction to make the flexible docking oil extraction mechanism and the flexible docking oil extraction port relative to each other. The flexible docking oil extraction mechanism moves forward until the flexible docking oil extraction port and the flexible docking oil extraction mechanism complete docking. Here, the X direction is the forward movement of the oil extraction robot, and the Y direction is the direction in the horizontal plane perpendicular to the X direction.

[0156] As shown in Figure 20, this application also provides an active rinsing method for oil sample collection, comprising:

[0157] S201. Use oil sample to flush the pipeline. The flushed waste oil enters the waste oil collection device. Then flush the first syringe and pipeline N times. Then use the first syringe to collect oil. Where N is greater than or equal to 2.

[0158] S202. The process of flushing and collecting oil from multiple syringes and tubing is repeated in sequence until all syringes have collected oil and the oil sample is collected. Then the oil collection robot begins to withdraw.

[0159] As shown in Figure 21, this application also provides a key landmark-assisted precise navigation and positioning method, including:

[0160] S301. In the substation scenario, select key landmarks;

[0161] S302. Use a depth camera to acquire images of key landmarks from multiple angles, perform geometric correction and noise reduction on the acquired images, extract feature points, generate a feature image library of key landmarks, bind the feature images of key landmarks with GPS coordinates and azimuth information, and store them as a searchable database.

[0162] S303. Using GPS and IMU, the relative position of the oil-retrieving robot with respect to the oil tank is initially determined, and navigation is performed according to the predetermined path. The movement path of the oil-retrieving robot is monitored in real time, the movement trajectory of the depth camera is analyzed, and the displacement and attitude of the oil-retrieving robot relative to the current environment are estimated.

[0163] S304. Based on the target location and key landmarks, plan the route. If the GPS signal is normal, rely on GPS to provide global location information. If the GPS signal weakens, switch to IMU-assisted navigation and adjust the movement trajectory of the oil-retrieving robot in real time according to the navigation data.

[0164] S305. When the distance between the oil-collecting robot and the target location is less than a set threshold, relevant landmark images are retrieved from the searchable database, feature point matching is performed, the image that best matches the current environment and the searchable database is retrieved, and image mapping is performed based on the image matching results to determine the position and posture deviation of the oil-collecting robot in the current environment.

[0165] S306. Calculate the relative pose of the depth camera with respect to key landmarks. Based on the position and attitude deviation of the oil-receiving robot in the current environment, and combined with the known landmark position information, correct the displacement and attitude of the oil-receiving robot with respect to key landmarks.

[0166] S307. Continuously acquire real-time images of key landmarks at fixed acquisition intervals and perform image matching to correct the robot's pose.

[0167] S308. Based on the pose convergence results after multiple image matching, it is confirmed that the oil-retrieving robot has accurately reached the target position.

[0168] This application also provides a fault self-diagnosis and remote assisted repair method. The transformer automatic flexible docking oil extraction system further includes: an autonomous operation master controller, a host computer control system, a navigation and positioning system, a mobile platform control system, a power management system, and an oil sample collection and detection control system arranged on the robot chassis. The power management system is connected to the power supply battery of the oil extraction robot, and the oil sample collection and detection control system is connected to the telescopic mechanism and the extraction mechanism respectively to perform telescopic and extraction control.

[0169] The navigation and positioning system includes a navigation controller. The autonomous operation master controller is communicatively connected to the host computer control system and the mobile platform control system. The mobile platform control system includes a motion controller, a motor driver, a drive motor, a steering motor driver, and a steering motor. The motion controller is communicatively connected to the navigation controller, the motion controller is connected to the motor driver, the motor driver is connected to the drive motor, the motion controller is connected to the steering motor driver, and the steering motor driver is connected to the steering motor.

[0170] As shown in Figure 22, the method includes:

[0171] S401 performs self-tests on the power management system, autonomous operation main controller, host computer control system, navigation and positioning system, mobile platform control system, and oil sample collection and detection control system, respectively. The host computer control system sends all self-test results to the cloud server, so that the cloud server can generate auxiliary repair strategies based on all self-test results and send them to the host computer control system. The host computer control system can then perform automatic repairs based on the received auxiliary repair strategies or send the auxiliary repair strategies to maintenance personnel.

[0172] As shown in Figure 23, this application also provides a multimodal deep fusion-based embodied perception task monitoring method, including:

[0173] S501: The large model receives multimodal information from multiple sensors, performs comprehensive analysis on the real-time acquired multimodal information, judges anomalies or task deviations, and generates early warnings or handling plans.

[0174] S502. When the large model determines that there may be anomalies or task deviations, it queries the built-in expert knowledge base to find expert suggestions and historical experience related to the anomalies or task deviations, and optimizes the oil sampling docking process based on the expert suggestions and historical experience.

[0175] In summary, the technical effects of this application are:

[0176] 1. The active adaptive flexible docking method for oil sampling robots provided in this application develops an adaptive flexible docking mechanism. Through the strategy of "fixed rod clamping and positioning + active detection and rotation adjustment of deflection angle", the docking mechanism adopts a dual flexible adjustment method, which has automatic docking buffer and universal flexible adjustment functions. This enables the automatic docking device to reliably dock with the oil sampling tank docking mechanism through active flexible adjustment, solving the problems of unsuccessful docking, joint slippage and oil sample leakage. It realizes universal flexible adjustment, fast and accurate docking, joint self-locking and automatic separation functions during docking, reducing the safety hazards caused by frequent docking and improving the accuracy and success rate of automatic docking. The gripper of the clamping rod mechanism adopts a fork-overlapping structure, which can firmly clamp the fixed rod without slippage, effectively ensuring the automatic docking.

[0177] 2. The key landmark-assisted precise navigation and positioning method provided in this application is based on image mapping for key landmark fusion positioning. It compares the image information in the actual environment with the pre-stored key landmark images, and corrects and fine-tunes the current pose of the oil-retrieving robot through the image mapping algorithm. This solves the problem of inaccurate positioning caused by limited or failed GPS signals and accumulated odometer errors in complex environments, realizes the high-precision positioning and docking function of the oil-retrieving robot, reduces navigation deviation caused by accumulated GPS and IMU errors, and improves the robot's operating efficiency.

[0178] 3. The active rinsing method for oil sample collection provided in this application has developed an automatic oil sample rinsing and pipeline switching device. It adopts a solenoid valve and valve block structure, which is compact and small in size. It can switch between multiple oil samples through the solenoid valve for rinsing and collection. The oil sample is extracted and rinsed in advance by the oil sample extraction device, and circulated through the circulation pump and multiple branches to rinse the inner wall of the pipeline and remove waste oil multiple times. It solves the problem of inaccurate test results caused by oil sample residue and impurities contaminating the oil sample in the oil sampling pipeline. It realizes the automatic rinsing function of the inner wall of the pipeline and syringe, and improves the cleanliness of the oil sample and the accuracy of the automatic test results.

[0179] 4. The fault self-diagnosis and remote assisted repair method provided in this application designs a system fault self-diagnosis and pre-repair strategy technology. Before oil sample collection, the autonomous operation master controller, the host computer control system, the power management system, the navigation and positioning system, the mobile platform control system and the oil sample collection and detection control system perform functional self-diagnosis and upload fault data to the cloud server to assist in the pre-repair of faults. This solves the problem that the robot may be working with safety hazards, realizes the functional self-diagnosis and active fault repair function of the transformer automatic flexible docking oil sampling system, and improves the safety of robot operation.

[0180] 5. This application provides a multimodal deep fusion-based embodied perception task monitoring method. It constructs a multimodal information real-time acquisition and deep fusion model. By monitoring the task status during oil sample collection in real time, it perceives the robot's task status, environmental changes, and potential anomalies, captures abnormal behaviors, and reduces the risk of hazardous operations through preset safety control strategies and collaborative optimization using an expert knowledge base. This solves the problems of unattended operation and susceptibility to interference from factors such as bubbles and impurities during oil sample collection. It achieves intelligent judgment and response during oil sample collection and automatic optimization of the oil extraction task, improving the safety, accuracy, and success rate of the automated oil extraction process. Through a large model and expert knowledge base, it enhances the automation of the oil extraction process. Safety, accuracy, and operational success rate are ensured through real-time acquisition and fusion of multimodal information. The system monitors the task status of the oil sampling robot during transformer oil sample collection, assessing operational progress, sensing robot status, environmental changes, and potential anomalies, and capturing abnormal behaviors (such as docking failures or oil sample abnormalities). Pre-set safety control strategies reduce the risk of hazardous operations, achieving intelligent judgment and response. When the large model identifies potential anomalies or task deviations, it queries the built-in expert knowledge base to find expert suggestions and historical experience related to the current anomaly or task deviation, enabling automatic optimization of the oil sampling task. This collaborative optimization between the large model and the expert knowledge base further improves the monitoring accuracy of the oil sampling task.

[0181] 6. The oil sampling robot of this application has a reflector installed on the back of the left door, which facilitates the oil sample detection camera to detect the details of the oil sample extraction by the syringe through the reflection of the reflector, and the observation is realized within a small viewing distance; the nylon seat of the syringe of the oil sample pushing device is installed on the module fixing seat by elastic pins at both ends. By pulling out the elastic pins, the handle can be grabbed to remove the nylon seat of the syringe along with the multiple sets of extracted syringe oil samples and send them to the laboratory for testing, which is convenient for disassembly and assembly; the waste oil collection device uses a weighing sensor to measure the amount of waste oil discharged, which is convenient and easy to use.

[0182] The above description is merely an optional embodiment of this application and is not intended to limit this application. For those skilled in the art, this application can have various modifications and variations. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. An automatic flexible docking oil sampling system for transformers, comprising: An oil collection tank (1) and an oil collection robot, the oil collection robot including an oil collection robot chassis (3) and an automatic docking device (7) and an oil sample extraction device (5) installed on the oil collection robot chassis (3); The oil tank (1) includes a flexible docking oil inlet (105), a fixed rod (102), and a movable rod assembly fixed on the oil tank body (107). The automatic docking device (7) includes a guiding docking mechanism (704) and a flexible docking oil inlet (701). The guiding docking mechanism (704) cooperates with the fixed rod (102) and the movable rod assembly to perform guiding docking. The flexible docking oil inlet (701) performs flexible docking with the flexible docking oil inlet (105).

2. The transformer automatic flexible docking oil sampling system as described in claim 1, wherein, The oil tank (1) further includes: an oil tank door (108), the inner surface of which is equipped with a silicone waterproof strip, and a flat lock (109) is installed on the oil tank door (108). The oil tank door (108) is connected to the oil tank body (107). A grounding bar (110) is provided on the outer side of the oil tank body (107). A control unit (123) is installed inside the oil tank body (107). The control unit (123) is configured to control and detect the operating status of the oil tank (1) and communicate with the host computer.

3. The transformer automatic flexible docking oil sampling system as described in claim 2, wherein, The top of the oil tank body (107) is connected to a feature marker (101). The laser radar (201) of the oil-taking robot identifies the feature marker (101) and positions and navigates the oil tank (1) so that the oil-taking robot docks with the oil tank (1) when taking oil.

4. The transformer automatic flexible docking oil sampling system as described in any one of claims 1-3, wherein, The flexible docking oil inlet (105) includes: a tapered docking interface (10501), a flat self-locking male connector (10502), an adapter (10503), a first spherical bearing (10504), a first bearing retaining ring (10505), a first adjusting connector (10506), a first docking interface adjusting spring (10507), a first baffle (10508), a docking interface base (10509), and a docking interface bracket (105010); The interface bracket (105010) is fixed on the oil tank body (107). A fixing rod (102) is fixed below the interface bracket (105010). The front end of the fixing rod (102) is a semi-circular spherical surface. The interface base (10509) is fixed above the interface bracket (105010). The first bearing retaining ring (10505) and the first baffle (10508) fix the outer ring of the first spherical bearing (10504) on the interface base (10509). One end of the first adjusting connector (10506) passes through the inner ring of the first spherical bearing (10504) and connects to one end of the adapter (10503). The other end of the first adjusting connector (10506) leads out to the oil inlet. A first pair of interface adjusting springs (10507) is installed between the first adjusting connector (10506) and the first baffle (10508). The tightness of the first pair of interface adjusting springs (10507) is adjusted by turning the first adjusting connector (10506). The other end of the adapter (10503) is fixed with a flat self-locking male connector (10502), and the tapered mating interface (10501) is fixed at the other end of the flat self-locking male connector (10502). Under the combined action of the inner and outer rings of the first joint bearing (10504) and the first mating interface adjusting spring (10507), the flexible mating oil outlet (105) can rotate and swing flexibly at multiple angles.

5. The transformer automatic flexible docking oil sampling system as described in any one of claims 1-3, wherein, The oil tank (1) also includes an oil tank lifting door (106), which is connected to the door frame of the oil tank body (107). The oil tank lifting door (106) includes a lifting door, a limit switch (124), a through screw motor (125), a through screw (126), and a lifting door guide rail (127). The lifting door guide rail (127) and limit switch (124) are arranged on the oil tank door frame (128). The lifting door is slidably connected to the lifting door guide rail (127). The through screw motor (125) is fixed on the oil tank body (107). The output shaft of the through screw motor (125) is connected to the through screw (126). The through screw (126) is installed on the lifting door. The output shaft of the through screw motor (125) rotates, driving the through screw (126) to move up and down, so as to drive the lifting door to rise and fall along the lifting door guide rail (127).

6. The transformer automatic flexible docking oil sampling system as described in any one of claims 1-3, wherein, The movable rod assembly is fixed on the oil tank body (107). The movable rod assembly is located on the lower side of the fixed rod (102). The movable rod assembly includes two sets of linear bearing assemblies (121), a movable rod fixing seat (122), two sets of movable rod limit sensors (119), movable rod limit baffles (120) of different lengths, and a return spring. The movable rod (103) is fixed on the movable rod fixing seat (122). The movable rod fixing seat (122) is connected to two sets of linear bearing assemblies (121) and slides along the linear bearing assembly (121). According to the different compression of the linear bearing assembly (121), the movable rod limiting baffle (120) of different lengths is triggered. The linear bearing assembly (121) is reset by the reset spring, which is arranged on the linear bearing assembly (121).

7. The transformer automatic flexible docking oil sampling system as described in any one of claims 1-3, wherein, The oil tank (1) further includes an oil extraction pipeline assembly (130); the oil extraction pipeline assembly (130) is installed in the oil tank body (107), the upper part of the oil extraction pipeline assembly (130) is connected to the oil inlet of the flexible docking oil extraction port (105), and the lower part of the oil extraction pipeline assembly (130) is connected to the upper branch (111), the middle branch (112) and the lower branch (113) of the transformer, and returns the oil to the transformer through the circulating pump (114) and the return oil branch (115); The upper branch (111), middle branch (112), and lower branch (113) of the transformer are all equipped with a first solenoid valve (117) and a ball valve (116). The first solenoid valve (117) is a normally closed valve. The circulation pump (114) is installed at the front end of the return oil branch (115) to realize the circulation of the upper branch (111), middle branch (112), and lower branch (113) of the transformer with the oil inside the transformer.

8. The transformer automatic flexible docking oil sampling system as described in any one of claims 1-3, wherein, The oil-collecting robot chassis (3) is connected to a wheel leg assembly (301), and the front end of the oil-collecting robot chassis (3) is equipped with a charging port (302) that can charge autonomously.

9. The transformer automatic flexible docking oil sampling system as described in any one of claims 1-3, wherein, The oil-collecting robot further includes: an oil-collecting robot shell (2); the oil-collecting robot shell (2) is mounted on the oil-collecting robot chassis (3); The oil-collecting robot shell (2) is equipped with a lidar (201), a lidar support (202), a depth camera (203), an ultrasonic sensor (215), a front door (204), a communication antenna (205), a left door (206), an emergency stop switch (208), a crash bar (209), a rubber strip (210), a docking sealing cover (212), a right door (213), and a rear door (214). The oil-collecting robot is also equipped with a global positioning system (GPS) and an inertial measurement unit (IMU). The outer shell (2) of the oil-collecting robot is connected to the chassis (3) of the oil-collecting robot via a rubber strip (210). A collision strip (209) is installed at the front end of the outer shell (2) of the oil-collecting robot, and ultrasonic sensors (215) are distributed on both sides of the bottom. A touch switch is installed inside the collision strip (209). When the oil-collecting robot stops and has a deflection angle with the oil tank (1), the ultrasonic sensors (215) on both sides measure the distance and calculate the deflection angle data so that the automatic docking device (7) can adjust the deflection angle. A lidar support (202) is set above the front end of the oil-taking robot shell (2). A lidar (201) is installed above the lidar support (202). A depth camera (203) is arranged below the lidar support (202). A docking sealing cover (212) is set inside the front end of the oil-taking robot shell (2). An oil sample detection camera (519) is set inside the oil-taking robot shell (2). The outer shell (2) of the oil-collecting robot is provided with a front door (204), a rear door (214), a left door (206), and a right door (213) on its four sides. The rear door (214), the left door (206), and the right door (213) are all equipped with magnetic door locks (207). The front door (204) is a lifting door that can automatically lift and lower during automatic docking. The inner side of the left door (206) is equipped with a reflector (211) so that the oil sample detection camera (519) can reflect the oil sample extraction details of the syringe (503) through the reflector (211).

10. The transformer automatic flexible docking oil sampling system as described in claim 9, wherein, The distance between the two ultrasonic sensors (215) is a, and the distances from the two ultrasonic sensors (215) to the surface of the oil tank (1) are l1 and l2, respectively. The formula for calculating the deflection angle θ between the oil-retrieving robot and the oil tank (1) when it stops is as follows: θ = arctan((l2-l1) / a); where, if the value of θ is positive, the automatic docking device (7) rotates clockwise, and if the value of θ is negative, the automatic docking device (7) rotates counterclockwise.

11. The transformer automatic flexible docking oil sampling system as described in any one of claims 1-3, wherein, The automatic docking device (7) includes: a guiding docking mechanism (704), a flexible docking oil extraction mechanism (701), an X-axis moving slide (707), a Y-axis moving slide (706), a telescopic mechanism (702), a vertical frame (703), and a rotating mechanism (705); the flexible docking oil extraction mechanism (701) is mounted on the telescopic mechanism (702) and moves with the telescopic mechanism (702); the telescopic mechanism (702) is rigidly connected to the automatic docking device (7) through a connecting plate. The guide docking mechanism (704) is fixed on the vertical frame (703) by the guide docking mechanism fixing plate (704010). The guide docking mechanism (704) includes a pole holding mechanism adjusting plate (704011), a pole holding mechanism connecting piece (70401), a pole holding mechanism (70408), a guide plate (70407), a guide plate bracket (70405), a movable rod trigger plate (70404), a trigger plate bracket (70403), a fixed rod front sensor (70406), a fixed rod rear sensor (70402), and a pole holding mechanism fixing plate (7040811). The pole-holding mechanism (70408) is fixed on the guide docking mechanism fixing plate (704010) through the pole-holding mechanism adjusting plate (704011) and the pole-holding mechanism connecting piece (70401). The guide plate (70407) is connected and fixed on the guide docking mechanism fixing plate (704010) with two sets of guide plate brackets (70405). The guide plate (70407) has a sloping structure. During operation, the fixing rod (102) of the oil tank (1) remains stationary, and the front end of the guide plate (70407) contacts the spherical surface of the fixing rod (102) and slides passively along the slope to lift the guide docking mechanism (704). The movable rod trigger plate (70404) is connected to the guide plate bracket (70405) through two sets of trigger plate brackets (70403). During operation, the movable rod trigger plate (70404) moves closer to the oil tank (1). The front of the movable rod trigger plate (70404) contacts the movable rod (103) in the movable rod assembly, pushing the movable rod (103) to move, thereby triggering the movable rod limit baffles (120) of different lengths, realizing the lifting control of the oil tank (1) lifting door and the on / off control of the first solenoid valve (117). The fixed rod front sensor (70406) and fixed rod rear sensor (70402) are respectively installed at the front and rear ends of the guide plate bracket (70405) and are triggered by the fixed rod (102) during operation.

12. The transformer automatic flexible docking oil sampling system as described in claim 11, wherein, The pole-holding mechanism (70408) is fixed on the guide docking mechanism fixing plate (704010) via the pole-holding mechanism adjusting plate (704011) and the pole-holding mechanism connecting piece (70401). The pole-holding mechanism (70408) includes a pole-holding seat (704081), a gripper (704082), a opposite gripper (7040810), a gear (704088), a rack (704083), a guide rail pair (704084), a pole-holding drive motor (704086), and a motor output shaft (704087). The pole drive motor (704086) is fixed on the pole support (704081). The motor output shaft (704087) is connected to the output end of the pole drive motor (704086). The gear (704088) is connected to the motor output shaft (704087). The rack (704083) is arranged on both sides of the gear (704088) and meshes with the gear (704088). The rack (704083) is provided with a guide rail pair (704084). The gripper (704082) and the opposite gripper (7040810) are respectively installed on both sides of the rack (704083). The gripper (704082) and the opposite gripper (7040810) are designed as single-claw and double-claw structures respectively. When the gear (704088) rotates, it drives the rack (704083) and the gripper (704082) to move relative to each other. The gripper (704082) and the opposite gripper (7040810) overlap together to hold the fixing rod (102) tightly.

13. The transformer automatic flexible docking oil sampling system as described in claim 11, wherein, The flexible docking oil extraction mechanism (701) includes: a planar self-locking female connector (70101), a pagoda connector (70102), a fixed base adapter plate (70103), a docking connector fixed base (70104), a flange connector (70105), a female connector sleeve (70106), a second spherical bearing (70107), a second bearing retaining ring (70108), a second adjusting connector (70109), a second docking interface adjusting spring (70110), and a second baffle (70111). The fixed seat adapter plate (70103) is set on the telescopic mechanism (702), and the butt joint fixed seat (70104) is set on the fixed seat adapter plate (70103). The second bearing retaining ring (70108) and the second baffle (70111) fix the outer ring of the second spherical bearing (70107) on the butt joint fixed seat (70104). One end of the second adjusting joint (70109) passes through the inner ring of the second spherical bearing (70107) and connects to one end of the flange joint (70105). The other end of the second adjusting joint (70109) is connected to the pagoda joint (70102) and leads out the oil outlet. The pagoda joint (70102) is connected to the oil sample extraction device (5) through the pipeline. A second pair of interface adjusting springs (70110) is installed between the second adjusting joint (70109) and the second baffle (70111). The other end of the flange joint (70105) is fixed with a flat self-locking female joint (70101). A female joint sleeve (70106) is installed on the outside of the flat self-locking female joint (70101). Under the guidance of the tapered interface (10501), the flat self-locking female joint (70101) smoothly docks and locks with the flat self-locking male joint (10502). Under the combined action of the inner ring and outer ring of the second joint bearing (70107) and the second pair of interface adjusting springs (70110), the flexible docking oil taking mechanism (701) performs flexible rotation and swing control at multiple angles.

14. The transformer automatic flexible docking oil sampling system as described in claim 11, wherein, The X-axis moving slide (707) is fixed on the oil-collecting robot chassis (3), the Y-axis moving slide (706) is arranged on the upper part of the X-axis moving slide (707), and the vertical frame (703) is fixed on the upper part of the rotating mechanism (705). The rotating mechanism (705) includes a rotary drive motor (70502), a large gear (70501) and a small gear (70503). The large gear (70501) is fixed on the upper part of the Y-axis moving slide (706). The rotary drive motor (70502) is installed at the bottom of the vertical frame (703). The small gear (70503) is installed on the output shaft of the rotary drive motor (70502). The large gear (70501) meshes with the small gear (70503). When the rotary drive motor (70502) rotates, it drives the small gear (70503) to rotate along the large gear (70501) to realize the rotation of the vertical frame (703).

15. The transformer automatic flexible docking oil sampling system as described in any one of claims 1-3, wherein, The oil sampling robot also includes an oil sample extraction device (5) installed on the chassis (3) of the oil sampling robot. The oil sample extraction device (5) includes an oil sample pushing device and a pipeline switching system (508). The oil sample pushing device includes: a syringe nylon seat (501), a syringe baffle (502), a syringe (503), a two-position two-way valve (504), a hose, an elastic pin (511), an online detection sensor (512), a syringe tray (513), a linear module limit sensor (515), a syringe pushing module (516), and a module fixing seat (517). The syringe pushing module (516) is fixed on the module fixing seat (517). The syringe pushing module (516) pushes the syringe (503) oil sample through the hose and pipeline switching system (508) to the waste oil collection device (6), and limits it through the linear module limit sensor (515). The syringe nylon seat (501) is provided with multiple syringe barrel D-shaped opening slots. A syringe baffle (502) is provided at the opening of the syringe barrel D-shaped opening slot to fix the syringe (503). An online detection sensor (512) is provided on the inner surface of the syringe barrel D-shaped opening slot to detect whether the syringe (503) is installed in place. A handle is provided on both sides of the syringe nylon seat (501). The syringe nylon seat (501) is installed on the module fixing seat (517) through the elastic pins (511) at both ends. The syringe tray (513) is set on the syringe push module (516) to support the piston handle of the syringe (503). The oil outlet of the syringe (503) is equipped with a two-position two-way valve (504) and is connected to the multi-way oil outlet of the pipeline switching system (508) through a hose.

16. The transformer automatic flexible docking oil sampling system as described in claim 15, wherein, The pipeline switching system (508) is fixed on the valve group bracket (506). The valve group bracket (506) is also connected to a rubber cap column (507) and a rubber cap (509). The rubber cap (509) is fitted on the rubber cap column (507) and is set to block the oil outlet of the syringe (503). The pipeline switching system (508) includes a second solenoid valve (50801), a valve assembly mounting base (50804), a pressure sensor (50803), a temperature sensor (50810), a booster pump (50805), and a pump mounting base (50806). The second solenoid valve (50801) is configured to realize pipeline opening and closing and switching. The booster pump (50805) is fixed on the pump mounting base (50806). The valve assembly mounting base (50804) is designed with a connecting pipeline (508041) inside. The primary oil inlet (50809) of the pipeline switching system (508) is connected to the primary oil outlet (50808) of the pipeline switching system (508). The secondary oil inlet (50807) of the pipeline switching system (508) is connected to the secondary oil outlet (50802), waste oil outlet (50812) and reserved oil outlet (50811) of the multi-line pipeline switching system (508). A booster pump (50805) is installed between the primary oil outlet (50808) and the secondary oil inlet (50807) of the system, which is set up to flush the pipeline and extract oil samples. A temperature sensor (50810) and a pressure sensor (50803) are installed between the primary oil inlet (50809) and the primary oil outlet (50808) of the system, respectively, to detect the temperature and pressure of the transformer oil. The secondary oil outlet of the multi-line switching system (508) is connected to the syringe (503) through a hose to inject the oil sample into the syringe (503). A pressure sensor (50803) is installed at the front end of the secondary oil outlet (50802) of the multi-line switching system (508) to detect the pressure injected into the syringe (503).

17. The transformer automatic flexible docking oil sampling system as described in claim 16, wherein, The valve group bracket (506) is connected to a first oil collection box (510) and a second oil collection box (518). The oil sample leaked by the pipeline switching system (508) is collected through the first oil collection box (510) and transported to the second oil collection box (518) through a hose. The oil sample leaked by the oil sample pushing device is collected by the second oil collection box (518). The second oil collection box (518) is manually removed when it is full of oil.

18. The transformer automatic flexible docking oil sampling system as described in claim 15, wherein, The oil-collecting robot also includes a waste oil collection device (6) installed on the chassis (3) of the oil-collecting robot. The waste oil collection device (6) includes a weighing sensor (601), a waste oil tank (602), a fixing plate (603), and a strap (604). The weighing sensor (601) is used to measure the amount of waste oil discharged. The waste oil tank (602) is used to collect waste oil from the cleaning pipeline. The fixing plate (603) and the strap (604) are used to fix the waste oil tank (602) above the weighing sensor (601).

19. An active adaptive flexible docking method, utilizing the transformer automatic flexible docking oil sampling system according to any one of claims 1-18, comprising: The oil-collecting robot navigates quickly to the oil-collecting tank, moves slowly after reaching the initial position in front of the tank, and stops moving forward after reaching the set position. The automatic docking device moves horizontally along the X direction. After the fixed rod touches the guide docking mechanism, it continues to move. After moving a set distance and confirming that the fixed rod is within the docking range of the automatic docking device, the automatic docking device continues to move forward. At this time, the movable rod assembly begins to be compressed. After being compressed a set distance, the oil tank lifting door opens, and the automatic docking device stops moving forward. After the oil tank lifting door is opened, the automatic docking device continues to move horizontally along the X direction. After moving a set distance, it stops moving forward and detects the deflection angle data. The automatic docking device rotates and adjusts according to the deflection angle data to make the automatic docking device perpendicular to the oil tank cabinet surface, which facilitates automatic docking and oil extraction. The pole-holding mechanism firmly holds the fixed pole. During the holding process, the automatic docking device moves and adjusts along the Y direction to make the flexible docking oil-taking mechanism and the flexible docking oil-taking port align. The flexible docking oil-taking mechanism moves forward until the flexible docking oil-taking port and the flexible docking oil-taking mechanism complete docking. Here, the X direction is the forward movement of the oil-taking robot, and the Y direction is the direction in the horizontal plane perpendicular to the X direction.

20. An active rinsing method for oil sample collection, utilizing the transformer automatic flexible docking oil sampling system as described in claim 15 or 16, comprising: The pipeline is flushed with an oil sample, and the waste oil from the flushing enters the waste oil collection device. Then, the first syringe and pipeline are flushed N times. Then, the first syringe is used to collect oil, where N is greater than or equal to 2. The process of flushing and collecting oil from multiple syringes and tubing is repeated sequentially until all syringes have collected oil and the oil sample is collected. After this, the oil collection robot begins to withdraw.

21. A key landmark-assisted precise navigation and positioning method, utilizing the transformer automatic flexible docking oil sampling system as described in claim 9, comprising: In the substation scenario, key landmarks are selected; Images of key landmarks are acquired from multiple angles using a depth camera. The acquired images are then geometrically corrected and denoised, and feature points are extracted to generate a feature image library of key landmarks. The feature images of key landmarks are then bound to GPS coordinates and azimuth information and stored as a searchable database. The relative position of the oil-retrieving robot to the oil tank is initially determined using GPS and an inertial measurement unit (IMU), and navigation is performed according to a predetermined path. The movement path of the oil-retrieving robot is monitored in real time, the movement trajectory of the depth camera is analyzed, and the displacement and attitude of the oil-retrieving robot relative to the current environment are estimated. Based on the target location and key landmarks, the route is planned. When the GPS signal is normal, the global location information is provided by GPS. If the GPS signal weakens, the system switches to IMU-assisted navigation and adjusts the movement trajectory of the oil-retrieving robot in real time based on the navigation data. When the distance between the oil-collecting robot and the target location is less than a set threshold, relevant landmark images are retrieved from the searchable database, feature point matching is performed, the image that best matches the current environment with the searchable database is retrieved, and image mapping is performed based on the image matching results to determine the position and posture deviation of the oil-collecting robot in the current environment. Calculate the relative pose of the depth camera with respect to key landmarks, and based on the position and pose deviation of the oil-retrieving robot in the current environment, combined with the known landmark position information, correct the displacement and pose of the oil-retrieving robot relative to the key landmarks. Real-time images of key landmarks are continuously acquired at fixed acquisition intervals and image matching is performed to correct the robot's pose. Based on the pose convergence results after multiple image matchings, it was confirmed that the oil-collecting robot had accurately reached the target location.

22. A fault self-diagnosis and remote assisted repair method, used in the transformer automatic flexible docking oil sampling system as described in claim 9; The transformer automatic flexible docking oil sampling system also includes: The autonomous operation master controller, host computer control system, navigation and positioning system, mobile platform control system, power management system and oil sample collection and detection control system are arranged on the robot chassis. The power management system is connected to the power supply battery of the oil-collecting robot. The oil sample collection and detection control system is connected to the telescopic mechanism and the extraction mechanism respectively to perform telescopic and extraction control. The navigation and positioning system includes a navigation controller. The autonomous operation master controller is communicatively connected to the host computer control system and the mobile platform control system. The mobile platform control system includes a motion controller, a motor driver, a drive motor, a steering motor driver, and a steering motor. The motion controller is communicatively connected to the navigation controller, the motion controller is connected to the motor driver, the motor driver is connected to the drive motor, the motion controller is connected to the steering motor driver, and the steering motor driver is connected to the steering motor. The method includes: The system performs self-tests on the power management system, autonomous operation main controller, host computer control system, navigation and positioning system, mobile platform control system, and oil sample collection and detection control system. The host computer control system sends all self-test results to the cloud server, which then generates auxiliary repair strategies based on these results and sends them to the host computer control system. The host computer control system then performs automatic repairs based on the received auxiliary repair strategies or sends the strategies to maintenance personnel.

23. The fault self-diagnosis and remote assisted repair method as described in claim 22, wherein, The power management system self-test includes: after the oil-retrieving robot is powered on, the power indicator light turns on, indicating that the power is connected, and the power management system inside the oil-retrieving robot checks the status of the battery pack, including whether the battery power, voltage and temperature are within the normal range. The self-test of the autonomous operation controller includes: after the autonomous operation controller is powered on, it automatically runs a bootloader program. The bootloader program checks whether the serial port, network port and universal serial bus USB port of the autonomous operation controller are working properly. In response to the abnormal check result, it issues an error alarm. In response to the normal check result, it checks whether there is a new version of code that has not been updated in the download area. In response to the check result that there is a new version of code that has not been updated, it copies the new version of code from the download area to the application APP area and boots the application in the APP area. In response to the check result that there is no new version of code, it boots the application in the APP area. The host computer control system performs a self-test, including: after powering on, reading the version information of the host computer software and comparing it with the internally stored list of compatible versions; updating the application if a new version of the software is available; checking the network connection between the host computer control system and the autonomous operation controller; issuing an error alarm if the network connection fails; and reading the relevant configuration parameters of the autonomous operation controller and displaying them on the screen if the network connection is normal. It also verifies whether the configuration parameters meet the current operational requirements; if the configuration parameters do not meet the current operational requirements, it issues an error alarm, prompting the operator to check and reset the relevant configuration parameters. The navigation and positioning system self-test includes: after powering on, checking whether the hardware of the positioning unit, inertial measurement unit, and lidar is normal; in response to the check result of hardware abnormality, issuing an error alarm, and confirming whether the positioning unit, inertial measurement unit, and lidar have been correctly initialized and are in working condition; performing power-on repositioning and positioning accuracy calibration, evaluating the accuracy of the positioning unit by comparing the difference between the actual position and the expected position; for the inertial measurement unit, performing zero-bias calibration and gyroscope drift test to determine whether the inertial measurement unit meets the usage accuracy requirements, and issuing an alarm message in response to the judgment result that the usage accuracy is not met. The mobile platform control system self-test includes: after the mobile platform control system is powered on, checking the motor driver, drive motor, steering motor driver and steering motor one by one to confirm whether the communication is normal and there is no fault alarm. In response to the confirmation result of communication abnormality, an alarm is issued. The oil sample collection and detection control system performs a self-test, including: after powering on, the system resets all syringes on the oil sampling robot. During the reset, it detects the photoelectric detection switch of each syringe drive module. When a photoelectric detection switch outputs a signal, the reset stops, and the encoder value of each syringe drive motor is read as the zero point of each syringe's movement. During the reset process, the output torque of each syringe drive motor is continuously detected. In response to the detection result that the output torque reaches the alarm threshold, the movement of the syringe drive motor is immediately stopped, and an alarm message is issued, indicating that the syringe reset is abnormal.

24. A multimodal deep fusion-based embodied perception task monitoring method, used in the transformer automatic flexible docking oil sampling system of claim 9, comprising: The large model receives multimodal information from multiple sensors, performs comprehensive analysis on the real-time acquired multimodal information, judges anomalies or task deviations, and generates early warnings or handling solutions. When the large model identifies potential anomalies or task deviations, it queries the built-in expert knowledge base to find expert advice and historical experience related to the anomalies or task deviations, and optimizes the oil extraction docking process based on the expert advice and historical experience.

25. The multimodal deep fusion-based embodied perception task monitoring method as described in claim 24, further comprising: The large model adjusts the task monitoring strategy in real time and generates an optimized oil extraction task operation process based on the multimodal perception results. By combining with the expert knowledge base, the actual operation of the oil extraction task is fed back to the expert knowledge base to continuously optimize and expand the expert knowledge base.

26. The multimodal deep fusion-based embodied perception task monitoring method as described in claim 24 or 25, further comprising: The large model automatically generates in-depth analysis reports on operational anomalies by fusing visual and sensor data, and automatically corrects these reports using processing solutions from an expert knowledge base.

Images