A novel mobile robot system for precise feeding of indoor fish ponds

By combining tracked AGVs with laser + vision SLAM navigation, pneumatic conveying devices, splash guards, and guide feeding interfaces, the problems of passability and accuracy of feeding equipment on grating surfaces in humid environments have been solved, enabling multi-device collaborative control and precise feeding.

CN123100543APending Publication Date: 2026-11-17QINGDAO UNIV OF SCI & TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611490549.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-23
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing technologies are unreliable in wet and humid environments, magnetic strip navigation and QR code navigation are not reliable, wheeled trolleys have poor maneuverability on grid surfaces, high-level hoppers are inconvenient to replenish, and pneumatic feeding is prone to scattering and mis-dispensing, making it difficult to achieve coordinated control of multiple devices.

Method used

The system employs a tracked AGV combined with laser + vision SLAM navigation, a pneumatic conveying device and splash guard, a directional feeding interface and LoRa communication to achieve precise feeding and feeding.

Benefits of technology

It improves the passability and station reliability on multi-grid surfaces, reduces feed scattering and errors, enables multi-device collaborative control, and improves feeding accuracy and on-site redundancy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure FT_1
    Figure FT_1
  • Figure FT_2
    Figure FT_2
  • Figure FT_3
    Figure FT_3
Patent Text Reader

Abstract

This invention discloses a novel indoor fishpond precision feeding mobile robot system, comprising an original automatic feeding machine assembly, a tracked AGV vehicle, a robot control system, an automatic feeding station, and a charging pile. The original automatic feeding machine assembly is integrated into a continuous mounting platform of the tracked AGV vehicle. The AGV vehicle is equipped with a tracked chassis, a waterproof electrical compartment, an onboard LoRa terminal, and a laser + vision SLAM navigation module, suitable for walking and positioning on multi-grid surfaces. The air conveying device has an air outlet / feeding inlet facing a single target fishpond and a splash-proof guide cover at its end. The automatic feeding station automatically feeds the hopper through a guided feeding interface, arrival detection, and LoRa feeding handshake. The manual control panel serves as an onboard redundant control unit.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of aquaculture equipment technology, and in particular to a precision feeding mobile robot system suitable for indoor fish ponds with high water content, humidity, and grating floor environments. Background Technology

[0002] Indoor recirculating aquaculture systems typically consist of multiple parallel fish ponds with densely packed feeding areas and narrow passageways. The floors are often constantly damp and wet, and some passageways use grating. Traditional fixed feeding systems require extensive air delivery piping or manual feed handling, resulting in significant setup and maintenance workload. Mobile wheeled vehicles traveling on grating floors are prone to tire wear, slippage, jamming, and positioning errors.

[0003] Existing AGVs typically use magnetic strip navigation or QR code navigation. Magnetic strips are prone to detachment, corrosion, or being covered by water films on wet surfaces; the reliability of QR codes also decreases in humid, reflective, and contaminated conditions. Simply stacking general SLAM navigation, wireless communication, and automatic feeding modules is insufficient to simultaneously solve the coordination issues between slippery grid passages, single-fishpond feeding end alignment, high-level hopper feeding, and on-site manual intervention.

[0004] The original automatic feeder has a hopper, wheel feeding and pneumatic feeding structure, which can complete basic quantitative feeding. However, it lacks a mobile chassis adapted to indoor fish ponds, stable navigation, automatic feeding and multi-device collaborative control structure. When the hopper is positioned high, it will also increase the workload of manual feeding. Summary of the Invention

[0005] The purpose of this invention is to provide a novel mobile robot system for precise feeding in indoor fish ponds, which solves the problems of unreliable magnetic strip navigation and QR code navigation in humid environments, poor maneuverability of conventional wheeled vehicles on gridded surfaces, inconvenience of replenishing materials in high-level hoppers, and easy scattering, splashing, and misfeeding at the end of the pneumatic feeding system.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: The system includes an original automatic feeding machine assembly, a tracked AGV vehicle 5, a robot control system 6, an automatic replenishment station 7, and a charging pile 8. The original automatic feeding machine assembly includes a hopper 1, a feeding device 2, a pneumatic conveying device 3, and a manual operating platform 4, and forms an integrated vehicle structure with the tracked AGV vehicle 5 through a continuous installation platform 57.

[0007] The tracked AGV 5 includes a tracked chassis 51, an on-board controller 52, an on-board LoRa terminal 53, a laser + vision SLAM navigation module 54, a battery 55, a waterproof electrical compartment 56, and a guided material replenishment interface 59. The tracked chassis 51 is suitable for traveling on ground with multiple water grates 10; the continuous mounting platform 57 is located above the waterproof electrical compartment 56, so that the hopper 1, feeding device 2, pneumatic conveying device 3, manual operating platform 4, and guided material replenishment interface 59 are centrally mounted in the same on-board frame.

[0008] The pneumatic conveying device 3 includes a feeding fan 31, a conveying pipe 32, an air outlet / feeding port 33, and a splash guard 34. The air outlet / feeding port 33 is positioned facing a single target fishpond 9; the splash guard 34 is located below and on both sides of the air outlet / feeding port 33, and is used to narrow the guide opening, limit the range of feed dispersion, and reduce water vapor splashing back into the fishpond.

[0009] The laser + vision SLAM navigation module 54 includes a lidar 541, a vision camera 542, and an inertial / odometer unit 543. The vehicle-mounted controller 52 integrates the edge of the fishpond, walls, equipment outline, visual positioning features, and motion information to create a map including the target fishpond 9 and the docking position of the automatic feeding station 7. It can achieve mapping, positioning, obstacle avoidance, docking, and feeding port attitude correction without relying on magnetic strips and QR codes.

[0010] The robot control system 6 includes a host computer or scheduling controller 61, a LoRa gateway 62, and a communication interface 63. The host computer or scheduling controller 61 is connected to the LoRa gateway 62 via the communication interface 63; the LoRa gateway 62 communicates with the vehicle-mounted LoRa terminal 53 and the station-end LoRa terminal 75. The task data includes at least the device address, task number, target fishpond number, feeding amount, replenishment amount, positioning status, remaining material status, fault code, check bit, and acknowledgment bit.

[0011] The automatic feeding station 7 includes a high-level hopper 71, a feeding mechanism 72, a station-end controller 73, a feeding outlet valve 74, a station-end LoRa terminal 75, and a position detection sensor 76. The guide-type feeding interface 59 includes a flared guide port and a flexible sealing sleeve; after the position detection sensor 76 confirms that it is aligned with the feeding outlet, the station-end controller 73 controls the feeding mechanism 72 and the feeding outlet valve 74 to feed material into the hopper 1.

[0012] Compared with the prior art, the present invention limits the wheel feeder and pneumatic feeding structure to a tracked AGV 5 suitable for indoor multi-grid fish ponds. It reduces scattering and back splashing by using a single fish pond air outlet / feeding port 33 and a splash guard 34; it reduces feeding errors by using a guide feeding interface 59, a position detection sensor 76 and LoRa feeding handshake; and it improves passability, positioning reliability and on-site redundancy by using a tracked chassis 51, a laser + vision SLAM navigation module 54 and a manual operating platform 4. Attached Figure Description

[0013] Figure 1 This is a schematic diagram of the overall assembly structure of the integrated vehicle-mounted material feeding mobile robot of the present invention.

[0014] Figure 2 This is a three-dimensional structural diagram of the wheel feeder of the present invention.

[0015] Figure 3 This is a schematic diagram illustrating the communication relationship between robot control, automatic material replenishment, and vehicle-mounted material feeding in this invention.

[0016] Figure 4 This is a schematic diagram of the tracked chassis and the grid ground structure of the present invention.

[0017] Figure 5 This is a schematic diagram of the installation structure of the manual control console and SLAM navigation module of the present invention.

[0018] Explanation of reference numerals in the attached drawings: 1. Hopper; 11. First weighing sensor; 12. Second weighing sensor; 13. Manual regulating valve; 14. Storage bin; 2. Feeding device; 21. Wheel motor; 22. Wheel assembly; 3. Pneumatic conveying device; 31. Feeding fan; 32. Pneumatic conveying pipe; 33. Air outlet / feeding port; 34. Splash guard; 4. Manual control panel; 41. Local controller; 42. Touch screen and physical buttons; 43. Emergency stop switch; 5. Tracked AGV vehicle; 51. Tracked chassis; 511. Track; 512. Drive wheel; 513. Driven wheel; 514. Tensioning mechanism; 515. Sealed drive motor; 52. On-board controller; 53. 54. Vehicle-mounted LoRa terminal; 54. Laser + Vision SLAM navigation module; 541. LiDAR; 542. Vision camera; 543. Inertial / odometer unit; 55. Battery; 56. Waterproof electrical compartment; 57. Continuous installation platform; 59. Guided feeding interface; 6. Robot control system; 61. Host computer or scheduling controller; 62. LoRa gateway; 63. Communication interface; 7. Automatic feeding station; 71. High-level silo; 72. Feeding mechanism; 73. Station controller; 74. Feeding outlet valve; 75. Station LoRa terminal; 76. Position detection sensor; 8. Charging pile; 9. Single target fish pond; 10. Grille ground. Detailed Implementation

[0019] The present invention will be further described below with reference to the accompanying drawings. It should be understood that the following embodiments are only for explaining the present invention and are not intended to limit the scope of protection of the present invention.

[0020] like Figure 1 As shown, the hopper 1 is mounted above the continuous installation platform 57 via a support structure. The feeding device 2 is located below the hopper 1. The feeding fan 31, air delivery pipe 32, air outlet / feeding port 33, and splash guard 34 are arranged sequentially on the discharge side of the feeding device 2. The vehicle controller 52, vehicle LoRa terminal 53, battery 55, and waterproof electrical compartment 56 are arranged below or near the continuous installation platform 57, thereby shortening the connection distance between the actuator and the control electrical components.

[0021] The hopper 1 is equipped with a first weighing sensor 11 and a second weighing sensor 12 to obtain the weight of the feed inside the hopper 1. The feed bin 14 below the hopper 1 is used to collect the feed to be fed, and the manual adjustment valve 13 is used to adjust the feed drop section for maintenance, calibration, or manual mode. The on-board controller 52 identifies the remaining feed status, the weight difference before and after a single feeding, and the replenishment needs based on the weighing data.

[0022] like Figure 2 As shown, the feeding device 2 uses a wheel motor 21 to drive the wheel device 22 to rotate. The feeding blades or feed grids on the wheel device 22 separate the feed falling from the feed bin 14 and send it into the pneumatic conveying device 3. The on-board controller 52 can control the speed and running time of the wheel motor 21 according to the set feeding amount to realize quantitative feeding of different target fish ponds 9.

[0023] The air inlet / feeding inlet 33 adopts a single-outlet structure. After the tracked AGV 5 arrives at the corresponding feeding position, the on-board controller 52, combined with the navigation and positioning results, performs parking and attitude correction so that the air inlet / feeding inlet 33 faces the single target fish pond 9. The splash guard 34 covers the lower and sides of the air inlet / feeding inlet 33, allowing airflow and feed to enter the target fish pond along the narrowed guide opening, reducing lateral dispersion and lowering the possibility of water vapor backflowing along the air delivery pipe 32.

[0024] like Figure 4 As shown, the tracked chassis 51 includes left and right track tracks 511, drive wheels 512, driven wheels 513, a tensioning mechanism 514, and a sealed drive motor 515. The outer surface of the track tracks 511 is provided with transverse anti-slip ribs and drainage gaps, and the effective contact length between the track tracks 511 and the grid ground 10 is greater than the spacing between adjacent grid holes to reduce slippage, jamming, and localized wear. The tensioning mechanism 514 is used to maintain the track tracks 511 in a predetermined tension state, and the sealed drive motor 515 is used to adapt to wet working conditions.

[0025] like Figure 5As shown, the manual control panel 4 includes a local controller 41, a touch screen, physical buttons 42, and an emergency stop switch 43. The local controller 41 is connected to the vehicle controller 52 via RS485, CAN, or a digital I / O interface. When the robot control system 6 is under maintenance, experiencing LoRa communication malfunctions, automatic tasks are paused, or during on-site debugging, the operator can use the manual control panel 4 to perform start / stop, manual feeding, manual replenishment, reset, and emergency stop control.

[0026] The laser + vision SLAM navigation module 54 includes a lidar 541, a vision camera 542, and an inertial / odometer unit 543. The lidar 541 acquires contour information of the fishpond, walls, equipment, and passageway boundaries; the vision camera 542 identifies the edges of the fishpond, feeding stations, and replenishment stations; and the inertial / odometer unit 543 provides short-term motion estimation. These three types of information are fused by the onboard controller 52 for path tracking, obstacle avoidance, on-time stopping, and feeding port attitude correction.

[0027] like Figure 3 As shown, the host computer or scheduling controller 61 in the robot control system 6 is connected to the LoRa gateway 62 via communication interface 63. The vehicle-mounted LoRa terminal 53 is connected to the vehicle-mounted controller 52 via UART, SPI, or RS485 interface; the station-end LoRa terminal 75 is connected to the station-end controller 73 via UART or RS485 interface. The host computer or scheduling controller 61 sends the target fishpond number, feeding quantity, path point, and feeding strategy according to the device address, and the vehicle-mounted controller 52 sends back the positioning status, remaining material status, feeding completion status, and fault codes.

[0028] LoRa communication uses data frames with a parity bit and employs acknowledgment, heartbeat feedback, and timeout retransmission mechanisms. The onboard controller 52 returns an acknowledgment upon receiving a feeding task; during task execution, it periodically sends heartbeat and status frames. If the robot control system 6 does not receive an acknowledgment or heartbeat frame within a preset time, it retransmits the task or generates an error message to support the coordinated operation of multiple tracked AGVs 5 according to a task queue.

[0029] The high-level hopper 71 in the automatic feeding station 7 stores feed to be replenished, and the feeding mechanism 72 delivers the feed to the feeding outlet valve 74. After the tracked AGV 5 enters the feeding position, the guide feeding interface 59 first completes mechanical guidance using a horn-shaped guide port, and then a closed or semi-closed feeding channel is formed by a flexible sealing sleeve; after the arrival detection sensor 76 confirms the docking status, the station controller 73 controls the feeding mechanism 72 and the feeding outlet valve 74 to work.

[0030] The feeding process employs a LoRa handshake mechanism: the vehicle controller 52 sends the vehicle address, feeding request, and required feeding amount via the vehicle LoRa terminal 53; the station controller 73 returns an acknowledgment frame via the station LoRa terminal 75; after the arrival detection sensor 76 confirms the docking, the station controller 73 executes the feeding; the vehicle controller 52 determines the feeding is complete based on the weighing data and sends back the feeding status. This process reduces the risk of misaligned feeding and feed spillage.

[0031] The on-board controller 52 calculates the amount of feed already added by combining the speed or angle of the first weighing sensor 11, the second weighing sensor 12, the speed or angle of the wheel motor 21, and the operating status of the feeding fan 31. After the wheel motor 21 stops, the on-board controller 52 can control the feeding fan 31 to run for a delay to remove residual feed in the air delivery pipe 32, and verify the actual amount of feed added to a single target fish pond 9 based on the weight difference before and after feeding.

[0032] The waterproof electrical compartment 56 is equipped with sealed joints, splash-proof structures, and drainage structures, and the battery 55 is housed inside the waterproof electrical compartment 56. The charging pile 8 can be located at the end of the workshop aisle or near the automatic replenishment station 7. The tracked AGV 5 can travel to the charging pile 8 for automatic or manual assisted charging when idle or with low battery.

[0033] The workflow of this invention is as follows: the robot control system 6 generates a feeding task and sends it to the designated tracked AGV vehicle 5 via the LoRa gateway 62; the tracked AGV vehicle 5 reaches a single target fish pond 9 based on the laser + vision SLAM navigation module 54; the on-board controller 52 controls the operation of the wheel motor 21 and the feeding fan 31, and the feed is put into the target fish pond through the air outlet / feeding port 33 and the splash guard 34; after the task is completed, the vehicle returns the status and selects to continue to execute the next task, go to the automatic feeding station 7 for feeding, or go to the charging pile 8 for charging according to the remaining material and power status.

[0034] During assembly, the hopper 1, silo 14, feeding device 2, and pneumatic conveying device 3 are sequentially arranged on the continuous mounting platform 57 along the material conveying direction. The upper opening of the hopper 1 can be equipped with a reinforced frame and inspection cover, and its lower part is connected to the platform 57 via a weighing support structure. The silo 14 is located below the hopper 1 and communicates with the feeding device 2. The feeding fan 31, pneumatic conveying pipe 32, and air outlet / feeding port 33 can be fixed to the platform 57 via mounting brackets, pipe clamps, and flange connections, ensuring the airflow channel remains relatively stable under vehicle vibration conditions.

[0035] The first weighing sensor 11 and the second weighing sensor 12 are respectively arranged at different support positions of the hopper 1. The detected values ​​of the two sensors can be summed, compared, or filtered by the on-board controller 52. When the difference between the two exceeds a predetermined deviation range, the on-board controller 52 can send this state back to the robot control system 6 as weighing anomaly information and prohibit the continued execution of quantitative replenishment or quantitative feeding based on the abnormal weighing result. Thus, the weighing structure is used both for counting residual material and for closed-loop verification of feeding.

[0036] In the feeding device 2, the wheel motor 21 drives the wheel device 22 to rotate via a reduction transmission assembly. The feed grid, feeding blades, or equivalent separating components on the wheel device 22 sequentially carry the feed in the hopper 14 to the discharge position circumferentially. The manual regulating valve 13 is installed on the material drop path from the hopper 1 to the hopper 14 or from the hopper 14 to the feeding device 2, and is used to cut off or regulate the material flow during maintenance, cleaning, calibration, or manual operation. The start / stop, speed, and cumulative rotation angle of the wheel motor 21 can be recorded by the on-board controller 52, so as to be used together with the weighing change for feed quantity verification.

[0037] In the pneumatic conveying device 3, a protective grille can be installed on the air inlet side of the feeding blower 31, and the air outlet side is sealed and connected to the pneumatic conveying pipe 32. The pneumatic conveying pipe 32 can be a combination of straight pipe, curved pipe, or flexible transition pipe, and is fixed to the platform 57 by pipe clamps. A discharge end is provided at the air outlet / feeding port 33, which is unidirectionally oriented towards a single target fish pond 9. The splash guard 34 may include an upper limiting plate, two side guide plates, and a lower opening, thereby limiting the spread of feed without blocking the airflow channel. After the blower 31 stops, it can maintain short-term purging to reduce the probability of residual material remaining in the pneumatic conveying pipe 32.

[0038] The left and right tracks 511 of the tracked chassis 51 are respectively wound around the outer periphery of the drive wheel 512 and the driven wheel 513. The tensioning mechanism 514 is used to compensate for the slack of the tracks 511 caused by wear, thermal expansion and contraction, or long-term load. The output of the sealed drive motor 515 is transmitted to the drive wheel 512 through the transmission mechanism, driving the tracks 511 to move. A drainage gap is formed between the track plates of the tracks 511, and the outer surface has a lateral grounding part to improve the grounding stability when crossing the grid holes and wet areas on the grid ground 10. The driven wheel 513, the support roller, and the guide rail components together limit the deviation of the tracks 511.

[0039] During navigation, the vehicle controller 52 first uses the lidar 541 to collect the surrounding contours, then uses the vision camera 542 to acquire visual features such as the edge of the fishpond, the edge of the passage, the shape of the equipment, or the parking area, and combines this with the attitude and mileage information output by the inertial / odometer unit 543 to perform pose estimation. The map records at least the feeding station location of each target fishpond 9, the docking station location of the automatic replenishment station 7, the location of the charging pile 8, and the passable area of ​​the passage. When the vehicle approaches the feeding station location, it can reduce its driving speed, and use the orientation of the air vent / feeding port 33 relative to the fishpond 9 as the basis for end attitude correction.

[0040] When the quality of laser or visual observation deteriorates, the vehicle detects an obstacle ahead, the positioning reliability is insufficient, or there is an abnormal passage area on the grid ground 10, the on-board controller 52 can suspend automatic driving and feeding actions, and report the abnormal status through the on-board LoRa terminal 53. The manual control panel 4 serves as a redundant control unit in the field and can be used for low-speed inching, feeding tests, equipment reset, and fault diagnosis under authorized conditions; when the emergency stop switch 43 is activated, the on-board controller 52 stops the wheel motor 21, the feeding fan 31, and the vehicle's travel output.

[0041] The communication interface 63 may include one or more of a serial communication interface, a bus interface, and a digital input / output interface. The vehicle-mounted controller 52 can connect to the vehicle-mounted LoRa terminal 53 via the serial communication interface, exchange status information with the management circuits of the wheel motor 21, the sealed drive motor 515, or the battery 55 via the bus interface, and read the status of the weighing sensors 11 and 12, the position detection sensor 76, and the emergency stop switch 43 via the digital input / output interface. The station-end controller 73 is connected to the feeding mechanism 72, the feeding outlet valve 74, and the station-end LoRa terminal 75 in a corresponding manner.

[0042] LoRa data frames may include a frame header, device address, task number, command type, parameter area, status area, checksum area, and acknowledgment field. For feeding tasks, the parameter area must include at least the target fishpond number and the target feeding amount; for replenishment tasks, the parameter area must include at least the vehicle address, replenishment request, and required replenishment amount. After receiving the data frame, the vehicle-mounted LoRa terminal 53 and the station-mounted LoRa terminal 75 hand it over to the corresponding controller for address, task number, and checksum verification. After the controller verifies the data frame, it generates an acknowledgment response or execution result frame.

[0043] During task execution, the on-board controller 52 can generate status frames according to a preset cycle, including vehicle position, navigation status, hopper remaining amount, battery status, current task, and fault codes; the robot control system 6 updates the vehicle task queue accordingly. When multiple tracked AGVs 5 are set up in the same workshop, the host computer or scheduling controller 61 can allocate tasks according to the location of the target fishpond 9, vehicle remaining material, battery 55 remaining amount, and the occupancy status of the replenishment station 7, to avoid multiple vehicles occupying the same feeding channel or replenishment docking position at the same time.

[0044] During material replenishment, the tracked AGV 5 first drives into the vicinity of the automatic replenishment station 7 based on the replenishment station location on the map, and then makes low-speed fine adjustments based on the positioning characteristics of the replenishment station or the position detection sensor 76. The flared guide opening of the guide-type replenishment interface 59 allows the replenishment outlet to enter the predetermined position even if the vehicle has a small stopping deviation. The flexible sealing sleeve fits around the replenishment outlet valve 74 after docking. Only after the position detection sensor 76 outputs a valid signal and communication is successfully confirmed will the station controller 73 allow the replenishment feeding mechanism 72 and the replenishment outlet valve 74 to operate.

[0045] During the replenishment process, the station controller 73 can control the running time or number of revolutions of the replenishment feeding mechanism 72 according to the required replenishment amount sent by the vehicle controller 52, and receive the weight status of the hopper 1 during the replenishment process. When the required replenishment amount is reached, the upper limit of the hopper 1 is reached, communication timeout occurs, the arrival detection signal is lost, or an emergency stop occurs, the station controller 73 closes the replenishment outlet valve 74 and stops the replenishment feeding mechanism 72; the vehicle controller 52 then forms a replenishment completion record based on the changes in the first weighing sensor 11 and the second weighing sensor 12.

[0046] During the feeding task, after confirming that the vehicle has stopped at the corresponding position in the target fishpond 9, the on-board controller 52 first starts the feeding fan 31 to establish an airflow, and then starts the wheel motor 21 to feed the material. After the target feeding amount is reached, the on-board controller 52 stops the wheel motor 21 and keeps the feeding fan 31 running until the residual material in the air delivery pipe 32 is basically discharged; then the feeding fan 31 is turned off and the weight difference before and after feeding, execution time, and completion status of the target fishpond 9 are saved. If the expected weight change is not achieved or the posture of the air outlet / feeding port 33 is abnormal, a task pending review can be generated instead of being directly marked as normal completion.

[0047] The waterproof electrical compartment 56 can employ a door with a sealing ring, cable sealing joints, an internal isolation plate, and a bottom drainage channel to isolate the battery 55, on-board controller 52, and on-board LoRa terminal 53 from the external humid environment. The continuous mounting platform 57 can be connected to the tracked chassis 51 via bolts, reinforcing ribs, or mounting brackets, and provides a common mounting reference for the hopper 1, manual operating platform 4, air conveying pipe 32, and navigation module 54. During charging, the vehicle can place the feeding and replenishing actuators in a stopped state before establishing a charging connection with the charging pile 8.

[0048] The above description is only a preferred embodiment of the present invention. Any equivalent substitutions or modifications made by those skilled in the art to the structural position, interface type, communication parameters and control flow without departing from the spirit and substance of the present invention shall fall within the protection scope of the present invention.

Claims

1. A novel mobile robot system for precise feeding of indoor fish ponds, characterized in that: The system includes an original automatic feeder assembly, a tracked AGV vehicle 5, a robot control system 6, an automatic feeding station 7, and a charging pile 8. The original automatic feeder assembly includes a hopper 1, a feeding device 2, a pneumatic conveying device 3, and a manual operating platform 4, which are integrated with the tracked AGV vehicle 5 via a continuous installation platform 57 to form an integrated vehicle structure. The tracked AGV vehicle 5 includes a tracked chassis 51 suitable for traveling on multi-water grating ground 10, an on-board controller 52, an on-board LoRa terminal 53, a laser + vision SLAM navigation module 54, a battery 55, a waterproof electrical compartment 56, and a guide-type feeding interface 59. The pneumatic conveying device 3... The system includes an air inlet / feeding port 33 facing a single target fishpond 9 and a splash guard 34 set outside the air inlet / feeding port 33; the robot control system 6 sends task data containing the target fishpond number and feeding amount to the tracked AGV vehicle 5 via LoRa communication, and communicates with the automatic feeding station 7 for feeding; the on-board controller 52 controls the tracked AGV vehicle 5 to stop according to the positioning result of the laser + vision SLAM navigation module 54, so that the air inlet / feeding port 33 and the splash guard 34 are aligned with the single target fishpond 9 to complete the fixed-point feeding, and the manual operation table 4 is reserved as an on-board manual redundant operation unit.

2. The novel indoor fishpond precision feeding mobile robot system according to claim 1, characterized in that... : The original automatic feeding machine assembly also includes a first weighing sensor 11 and a second weighing sensor 12 set at different support positions of the hopper 1, a hopper 14 connected to the lower end of the hopper 1, and a manual regulating valve 13 set on the material drop path; the feeding device 2 includes a wheel motor 21 and a wheel device 22, the wheel device 22 has circumferentially arranged feeding blades or feed grids, and the wheel motor 21 drives the wheel device 22 to quantitatively feed the feed into the pneumatic conveying device 3.

3. The novel indoor fishpond precision feeding mobile robot system according to claim 1, characterized in that... : The pneumatic conveying device 3 includes a feeding blower 31, a pneumatic conveying pipe 32 connected to the feeding blower 31, an air outlet / feeding port 33 located at the end of the pneumatic conveying pipe 32, and a splash-proof guide cover 34 covering the lower and both sides of the air outlet / feeding port 33; the splash-proof guide cover 34 forms a narrowed guide opening toward a single target fish pond 9.

4. The novel indoor fishpond precision feeding mobile robot system according to claim 1, characterized in that... : The tracked chassis 51 includes left and right tracked tracks 511, drive wheels 512, driven wheels 513, tensioning mechanism 514 and sealed drive motor 515; the outer surface of the tracked track 511 is provided with transverse anti-slip ribs and drainage gaps, and the effective contact length between the tracked track 511 and the grid ground 10 is greater than the adjacent grid hole spacing.

5. The novel indoor fishpond precision feeding mobile robot system according to claim 1, characterized in that... : The laser + vision SLAM navigation module 54 includes a lidar 541, a vision camera 542, and an inertial / odometer unit 543. The vehicle controller 52 fuses the fishpond edge, wall and equipment outline information acquired by the lidar 541, the station position features acquired by the vision camera 542, and the motion information output by the inertial / odometer unit 543 to build a map containing the target fishpond 9 and the automatic feeding station 7. It completes mapping, positioning, obstacle avoidance and attitude correction of the air outlet / feeding port 33 without relying on magnetic strip navigation and QR code navigation.

6. The novel indoor fishpond precision feeding mobile robot system according to claim 1, characterized in that... : The robot control system 6 includes a host computer or scheduling controller 61, a LoRa gateway 62, and a communication interface 63. The LoRa gateway 62 is connected to the host computer or scheduling controller 61 via Ethernet, USB, or RS485 communication interface 63. The vehicle-mounted LoRa terminal 53 is connected to the vehicle-mounted controller 52 via UART, SPI, or RS485 interface. The station-end LoRa terminal 75 of the automatic replenishment station 7 is connected to the station-end controller 73 via UART or RS485 interface. The LoRa data frame includes at least the device address, task number, target fishpond number, feeding amount, replenishment amount, positioning status, remaining material status, fault code, check bit, and acknowledgment bit.

7. The novel indoor fishpond precision feeding mobile robot system according to claim 6, characterized in that: The LoRa communication adopts an acknowledgment response, heartbeat feedback, and timeout retransmission mechanism; the host computer or scheduling controller 61 issues feeding tasks according to the device address of the tracked AGV vehicle 5, and after the vehicle controller 52 executes the task, it transmits the positioning status, remaining material status, feeding completion status, and fault status back through the vehicle LoRa terminal 53. The LoRa gateway 62 summarizes the data transmitted back from multiple tracked AGV vehicles 5 and updates the task queue.

8. The novel indoor fishpond precision feeding mobile robot system according to claim 1, characterized in that: The automatic replenishment station 7 includes a high-level hopper 71, a replenishment feeding mechanism 72, a station-end controller 73, a replenishment outlet valve 74, a station-end LoRa terminal 75, and a position detection sensor 76. The guide-type replenishment interface 59 includes a flared guide opening and a flexible sealing sleeve. After the tracked AGV 5 arrives at the automatic replenishment station 7, the station-end controller 73 controls the replenishment feeding mechanism 72 and the replenishment outlet valve 74 to replenish the hopper 1 after the position detection sensor 76 confirms that the guide-type replenishment interface 59 is aligned with the replenishment outlet and receives the replenishment communication confirmation.

9. The novel indoor fishpond precision feeding mobile robot system according to claim 1, characterized in that: The vehicle controller 52 calculates the amount of feed added based on the weighing data of the first weighing sensor 11 and the second weighing sensor 12, the speed or angle of the wheel motor 21, and the operating status of the feeding fan 31. After the wheel motor 21 stops, the vehicle controller 52 controls the feeding fan 31 to run for a delay to remove residual feed in the air delivery pipe 32, and verifies the actual amount of feed added to a single target fish pond 9 based on the weight difference before and after feeding.

10. The novel indoor fishpond precision feeding mobile robot system according to claim 1, characterized in that: The manual control panel 4 is fixed to the frame of the tracked AGV vehicle 5 or the continuous installation platform 57, and includes a local controller 41, a touch screen, physical buttons 42, and an emergency stop switch 43. The local controller 41 is connected to the on-board controller 52 via RS485, CAN, or digital I / O interface, and is used to perform local start / stop, manual feeding, manual replenishment, reset, and emergency stop control when the robot control system 6 or LoRa communication is abnormal. The waterproof electrical compartment 56 is equipped with a sealed joint, a splash guard, and a drainage structure. The charging pile 8 is connected to the battery 55 to automatically or manually charge the tracked AGV vehicle 5.