A calf nursing robot with autonomous navigation control system and method of use

Abstract

This invention provides a calf feeding robot with an autonomous navigation control system and its usage method, relating to the interdisciplinary field of robotics and animal husbandry. The calf feeding robot with an autonomous navigation control system includes a movable chassis, a control screen, a main control compartment, a first camera, a second camera, a lidar base, a lidar, a milk bucket, a solenoid valve, a nipple, a CAN communication interface module, and a control board. The control board incorporates the autonomous navigation control system, including an environmental perception subsystem, a navigation control subsystem, a feeding subsystem, and a chassis control subsystem, achieving a highly integrated control architecture. This integrated control system improves the robot's operational efficiency and response speed, enabling the robot to quickly adapt to changes in the pasture environment and provide timely and accurate feeding services to calves.

Description

Technical Field

[0001] This invention belongs to the interdisciplinary field of robotics and animal husbandry, and in particular relates to a calf feeding robot with an autonomous navigation and control system and its usage method. Background Technology

[0002] Animal husbandry is an important branch of agriculture, mainly involving the raising and management of various livestock. It not only provides people with food and raw materials such as meat, dairy products, and leather, but also plays a vital role in maintaining ecological balance and promoting rural economic development.

[0003] Calf feeding is a crucial aspect of animal husbandry, particularly in dairy farming and beef cattle fattening. Proper feeding methods and management significantly impact calf health, growth, productivity, and long-term economic benefits. Calf feeding not only affects the individual animal's growth and health but also influences the efficiency and economic returns of the entire production system. Adopting scientific and rational feeding methods is of great importance for improving the efficiency of animal husbandry.

[0004] In modern animal husbandry, calf feeding techniques are constantly evolving, employing various modern methods to improve feeding efficiency and management. Manual feeding remains the primary method on many farms, with farm workers feeding the calves at set times based on their age and needs. However, traditional manual feeding is costly, inefficient, and risky. Some modern farms have introduced automated feeding equipment that can automatically provide milk according to set times and the calves' needs. However, automated and intelligent feeding systems have high equipment and maintenance costs, often unaffordable for small farms. Furthermore, current automated feeding equipment lacks autonomous navigation capabilities, making it difficult to cope with complex pasture environments and posing certain safety hazards. Therefore, developing an autonomous navigation feeding robot for pastures is an important way to improve pasture management efficiency and reduce costs. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention proposes a calf feeding robot with an autonomous navigation and control system and its usage method. This robot has autonomous navigation and intelligent feeding functions, providing livestock with precise and quantitative feed while reducing waste and improving feeding efficiency.

[0006] A calf feeding robot with an autonomous navigation and control system includes a mobile chassis, a control screen, a main control compartment, a first camera, a second camera, a lidar base, a lidar, a milk bucket, a solenoid valve, a nipple, a CAN communication interface module, and a control board.

[0007] The movable vehicle chassis serves as a support structure, carrying the control screen, main control compartment, first camera, second camera, lidar base, lidar, milk bucket, solenoid valve, flow valve, nipple, CAN communication interface module, and control board. It provides mobility through several drive motors mounted on the movable vehicle chassis. Each drive motor is connected to the CAN communication interface module and communicates with the control board using the CAN communication protocol. A power supply for powering the system is also provided on the movable vehicle chassis.

[0008] The control screen is placed outside the main control compartment and connected to the control board via an HDMI cable. It is used to display the image information captured by the first and second cameras through a graphical interface, and at the same time provide human-computer interaction functions for users.

[0009] The main control compartment is fixed above the movable vehicle chassis, with a control board fixedly installed inside, and a first camera, a second camera, a lidar base, and a lidar placed on top.

[0010] Both the first and second cameras provide infrared and LED illumination. The first and second cameras are used to capture image information from the left front and right front, respectively, and send it to the control board.

[0011] The lidar base is installed above the main control compartment to fix the lidar;

[0012] The lidar is fixed by a lidar bracket and is used to scan the surrounding environment, obtain the distance to surrounding objects and the scanning angle of the laser, and send them to the control board. At the same time, the built-in IMU module is used to obtain the current coordinates and yaw angle and send them to the control board.

[0013] The milk bucket and the main control compartment are placed side by side on the mobile vehicle chassis for storing milk;

[0014] The nipple is connected to the milk bucket via a solenoid valve and a flow valve and is used to feed calves;

[0015] The solenoid valve is used to control whether the milk container and the nipple are connected.

[0016] The flow rate valve is used to control the flow rate of the milk;

[0017] The CAN communication interface module is used to enable communication between the movable vehicle chassis and the control board via the CAN communication protocol.

[0018] The control board is used to achieve autonomous navigation based on user settings, image information collected by the first and second cameras, distance to surrounding objects collected by the lidar, and the scanning angle of the laser. It controls the movement of the movable vehicle chassis, the opening and closing of the solenoid valve, and the flow rate of the flow valve to complete the feeding of calves. At the same time, it displays the current coordinates and yaw angle on the remote control terminal.

[0019] The user settings include the user-set travel speed and the user-selected planned path;

[0020] The control board is equipped with an autonomous navigation control system, including an environmental perception subsystem, a navigation control subsystem, a vehicle chassis control subsystem, and a feeding subsystem.

[0021] The environmental perception subsystem is based on the Linux operating system and integrates the robot operating system. It generates point cloud data based on the distance to surrounding objects measured by the lidar and the scanning angle of the lidar, processes the image information and point cloud data, and sends the image information and processed point cloud data to the navigation control subsystem.

[0022] The processing includes data format conversion, coordinate system transformation, filtering, downsampling, and feature extraction;

[0023] The navigation control subsystem, based on Linux and the robot operating system, is used to receive image information and processed point cloud data. During initial operation and when the environment changes, a map building program is used to process the image information into a two-dimensional grid map using the gmapping algorithm. A positioning program is used to convert the processed point cloud data into positioning information using the Robot Operating System open-source framework. Based on the calf's position in the two-dimensional grid map and the positioning information, the calf's position is determined. Based on the two-dimensional grid map and the calf's position, A* and DWA algorithms are used to perform global and local path planning, respectively, to obtain several complete planned paths and store them to ensure safety and reliability in complex environments.

[0024] In cases other than changes in the operating and usage environment, the DWA algorithm is used to adjust the local planned path and achieve obstacle avoidance based on real-time acquired image information and point cloud data, on the user-selected planned path, and then send the adjusted local planned path to the vehicle chassis control subsystem.

[0025] The chassis control subsystem is used to generate drive signals according to the planned path and user settings, and send the drive signals to the drive motor of the movable chassis through the CAN communication interface module, so that the movable chassis moves according to the planned path.

[0026] The milking subsystem is used to control the solenoid valve to be energized after reaching the calf's location, to feed the calf milk from the milk bucket using a nipple, and to control the solenoid valve to be de-energized after feeding is completed.

[0027] The graphical interface includes a chassis control area, an environmental perception area, and a navigation control area;

[0028] The chassis control area includes a path selection box, a speed input box, a chassis start button, and a movement control area;

[0029] The path selection box is used by the user to select the planned path for this feeding task from several planned paths. After confirming the planned path, clicking the OK button will allow the vehicle chassis to move according to the selected planned path to achieve the predetermined planned path movement.

[0030] The speed input box is used by the user to input the travel speed of the movable vehicle chassis within a set range. After inputting the speed, the user can click the OK button to control the movable vehicle chassis to move at the set travel speed.

[0031] The chassis start button, when pressed before path selection, speed input, and movement control, is used to start the drive motor of the movable chassis.

[0032] The mobile control area is used to control the movement of the movable vehicle chassis in four directions: forward, backward, left, and right.

[0033] The environmental perception area includes two camera activation buttons, an image display area for the first camera, and an image display area for the second camera.

[0034] The two camera activation buttons are used to activate the first and second cameras.

[0035] The image display area of ​​the first camera and the image display area of ​​the second camera are respectively used to display image information captured in real time by the first camera and the second camera.

[0036] The navigation control area includes a navigation start button, a save map button, and a map creation button.

[0037] The map creation button is used to start the map creation program and create a two-dimensional raster map;

[0038] The Save Map button is used to save the completed 2D raster map;

[0039] The navigation start button is used to activate navigation and begin traveling along the planned route.

[0040] A method for using a calf feeding robot with an autonomous navigation and control system includes the following steps:

[0041] Step 1: Power on the movable vehicle chassis. On the graphical interface of the remote control terminal or control screen, click the two camera start buttons to start the first and second cameras. The control board connects to the drive motor on the movable vehicle chassis through the CAN communication interface module. Click the chassis start button, and the control board transmits the generated drive signal to the drive motor through the CAN communication protocol.

[0042] Step 2: Click the map creation button on the graphical interface of the remote control terminal or control screen to start the map creation program in the navigation control subsystem. Enter the set speed in the speed input box on the graphical interface of the remote control terminal or control screen, and use the motion control area on the graphical interface of the remote control terminal to control the drive motor in the movable vehicle chassis to move at the set speed and direction until a complete driving process is completed in the environment. During the driving process, the first and second cameras are used to collect image information of the surrounding environment, and the laser radar is used to measure the distance to surrounding objects and the scanning angle of the laser, and store it in the environmental perception subsystem of the control board.

[0043] Step 3: The environmental perception subsystem generates point cloud data based on the distance to surrounding objects measured by the lidar and the scanning angle of the lidar, processes the point cloud data, and sends the image information and the processed point cloud data to the navigation control subsystem.

[0044] Step 4: The map creation program processes the image information into a two-dimensional raster map using the gmapping algorithm. Click the "Save Map" button on the graphical interface of the remote control terminal or control screen to save the generated two-dimensional raster map.

[0045] Step 5: The stored two-dimensional grid map is displayed on the graphical interface of the control screen and the remote control terminal. The position of the calf in the two-dimensional grid map is set as the target point. The positioning program in the navigation control subsystem uses the Robot Operating System open source framework to convert the processed point cloud data into positioning information, thereby obtaining the positioning information of the target point, that is, the position of the calf.

[0046] Step 6: Based on the two-dimensional grid map and the target point's positioning information, the navigation control subsystem uses the A* algorithm and the DWA algorithm respectively to perform global path planning and local path planning, obtaining several complete planned paths and storing them;

[0047] Step 7: Select the planned path for feeding the calves in the path selection box on the graphical interface of the remote control terminal or control screen, and click the navigation start button. The chassis control subsystem generates a drive signal according to the selected planned path and user settings, and sends the drive signal to the drive motor of the movable chassis through the CAN communication interface module, so that the movable chassis starts to move according to the planned path.

[0048] Step 8: During the journey, the IMU module built into the lidar acquires the current coordinates and yaw angle of a calf feeding robot with an autonomous navigation control system and displays them on the remote control terminal. At the same time, the navigation control subsystem adjusts the local planned path and achieves obstacle avoidance based on the real-time acquired image information and point cloud data using the DWA algorithm. The adjusted local planned path is then sent to the vehicle chassis control subsystem, and the drive signal is sent to the drive motor of the movable vehicle chassis through the CAN communication interface module to adjust and control the travel path of the movable vehicle chassis.

[0049] Step 9: When the movable vehicle chassis stops moving at the target point, the feeding subsystem controls the solenoid valve to be energized, the nipple and the milk bucket are connected, the amount of milk to be fed that day is predicted, and the flow rate valve is used to control the milk flow rate according to the predicted amount of milk to be fed. The calf stretches out its neck and drinks the milk independently using the nipple. After the set time, the solenoid valve is de-energized.

[0050] The method for controlling the milk flow rate using a flow rate valve is as follows:

[0051] S1: Retrieve daily milk feeding records over a historical period;

[0052] S2: Store the daily milk feeding records in an array or list, and create an array of days corresponding to the milk feeding records, where each element represents the number of days corresponding to the milk feeding record;

[0053] S3: Initialize a linear regression model, input the number of days as the input feature into the linear regression model, use the daily milk feeding record as the target value, train the linear regression model, and obtain the trained linear regression model;

[0054] S4: Determine the feeding date and the corresponding number of days in the day array. Input the number of days into the linear regression model to obtain the predicted amount of milk to be fed.

[0055] S5: Calculate and set the milk flow rate based on the predicted milk volume and the set feeding time to ensure that an appropriate amount of milk is provided within the predetermined time.

[0056] The predicted milk volume and the set milk flow rate are displayed in a graphical interface;

[0057] S6: Compare the actual amount of milk fed with the predicted amount of milk fed. If the difference between the two is greater than a set threshold, adjust the parameters of the linear regression model according to the actual amount of milk fed to improve future predictions.

[0058] Step 10: The movable chassis moves to the next target point to continue feeding until the feeding task is completed. The calf feeding robot with an autonomous navigation control system returns to the starting point to wait for the next feeding task or to recharge.

[0059] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0060] This invention proposes a calf feeding robot with an autonomous navigation and control system. The autonomous navigation and control system mounted on the control board includes an environmental perception subsystem, a navigation control subsystem, a feeding subsystem, and a chassis control subsystem, realizing a highly integrated control architecture. This integrated control system improves the robot's operating efficiency and response speed, enabling the robot to quickly adapt to changes in the pasture environment and provide timely and accurate feeding services to calves.

[0061] The environmental perception subsystem achieves comprehensive perception of the pasture environment through multi-sensor fusion technology, including obstacle detection and terrain recognition. This improves the robot's perception accuracy, reduces false and false judgments, and enhances the robot's autonomous navigation capability in complex pasture environments, ensuring the safety of both the robot and the calves.

[0062] The navigation and control subsystem uses A* and DWA algorithms for global and local path planning. Based on the data provided by the environmental perception subsystem, it can calculate the optimal path in real time, enabling the robot to choose the shortest or safest path when performing tasks. This improves the efficiency of feeding tasks, reduces energy consumption, and extends the robot's endurance.

[0063] The milk feeding system can automatically adjust the amount and timing of milk feeding according to the calf's habits and needs, achieving personalized feeding. Personalized feeding not only meets the nutritional needs of different calves, but also improves the healthy growth rate of calves, while reducing milk waste and improving the economic benefits of the farm.

[0064] The chassis control subsystem achieves smooth movement and precise positioning of the robot through precise drive control. This precise drive control ensures the robot's stability during movement, reduces disturbance to the calves, and improves its ability to navigate complex terrain, thus ensuring the smooth execution of the feeding task. Attached Figure Description

[0065] Figure 1 This is a schematic diagram of the structure of a calf feeding robot with an autonomous navigation and control system according to an embodiment of the present invention;

[0066] Figure 2 This is a partial structural schematic diagram of a calf feeding robot with an autonomous navigation and control system according to an embodiment of the present invention;

[0067] Figure 3 This is a block diagram of the autonomous navigation control system of a calf feeding robot with an autonomous navigation control system according to an embodiment of the present invention.

[0068] Figure 4 This is a graphical interface diagram of a calf feeding robot with an autonomous navigation and control system according to an embodiment of the present invention.

[0069] Among them, 1-mobile vehicle chassis, 2-control screen, 3-main control compartment, 4-first camera, 5-second camera, 6-LiDAR base, 7-LiDAR, 8-milk bucket, 9-nipple, 10-CAN communication interface module, 11-control board, 12-path selection box, 13-speed input box, 14-chassis start button, 15-camera turn-on button, 16-navigation start button, 17-save map button, 18-map creation button, 19-mobile control area. Detailed Implementation

[0070] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0071] like Figure 1 and Figure 2 As shown, a calf feeding robot with an autonomous navigation and control system includes a mobile chassis 1, a control screen 2, a main control compartment 3, a first camera 4, a second camera 5, a lidar base 6, a lidar 7, a milk bucket 8, a solenoid valve, a nipple 9, a CAN communication interface module 10, and a control board 11.

[0072] The movable vehicle chassis 1 serves as a support structure, supporting the control screen 2, main control compartment 3, first camera 4, second camera 5, lidar base 6, lidar 7, milk bucket 8, solenoid valve, flow valve, nipple 9, CAN communication interface module 10, and control board 11. It provides mobility through several drive motors mounted on the movable vehicle chassis 1. The chassis is a front-to-rear-drive Ackermann structure with an integral bridge suspension design, ensuring stability in complex terrain. All drive motors are connected to the CAN communication interface module 10 and communicate with the control board 11 using the CAN communication protocol. A power supply for powering the system is also provided on the movable vehicle chassis 1.

[0073] In this embodiment, the maximum load of the movable vehicle chassis 1 is 100kg;

[0074] The control screen 2 is placed outside the main control compartment 3 and connected to the control board 11 by an HDMI cable. It is used to display the image information captured by the first camera 4 and the second camera 5 through a graphical interface, and at the same time provide users with human-computer interaction functions to facilitate operation and monitoring.

[0075] In this embodiment, the control screen 2 is a 1270*720 touch screen that can perform 5-point capacitive touch.

[0076] The main control compartment 3 is fixed above the movable vehicle chassis 1, and the control board 11 is fixedly installed inside. The first camera 4, the second camera 5, the lidar base 6 and the lidar 7 are placed on the top.

[0077] The first camera 4 and the second camera 5 provide infrared and LED illumination to ensure visual perception in low light and night conditions. The first camera 4 and the second camera 5 are used to capture image information from the left front and right front respectively and send it to the control panel 11 to provide the operator with intuitive visual feedback.

[0078] In this embodiment, the first camera 4 and the second camera 5 are arranged on both sides of the front of the vehicle to form a 45° field of view. The first camera 4 and the second camera 5 can perform 120-degree ultra-wide-angle photography.

[0079] The lidar base 6 is installed above the main control compartment 3 to fix the lidar 7;

[0080] The lidar 7 is fixed by the lidar support 6 and is used to scan the surrounding environment, obtain the distance to surrounding objects and the scanning angle of the laser, and send them to the control board 11 to assist the robot in precise positioning and navigation. Considering that the robot may encounter uneven road surfaces during the journey, which may cause the vehicle to veer off course and fail to reach the designated point, in order to avoid this situation, the built-in IMU module is used to obtain the current coordinates and yaw angle and send them to the control board 11.

[0081] The built-in IMU module in this embodiment includes a 3-axis accelerometer and a 3-axis gyroscope;

[0082] In this embodiment, the lidar 7 can detect objects with a minimum distance of 0.1 meters and a maximum distance of 100 meters, with a point cloud frame rate of 10Hz.

[0083] The milk bucket 8 and the main control compartment 3 are placed side by side on the movable vehicle chassis 1 for storing milk;

[0084] The nipple 9 is connected to the milk bucket via a solenoid valve and a flow valve, and is used to feed calves;

[0085] In this embodiment, the nipple 9 is a common livestock drinking device on the market. It will not flow out liquid on its own when no external pressure is applied. The nipple 9 can be a sucking type, a direct discharge type, or other structural types to meet the user's requirements in different scenarios.

[0086] The solenoid valve is used to control whether the milk container 8 and the nipple 9 are connected.

[0087] The flow rate valve is used to control the flow rate of the milk;

[0088] The CAN communication interface module 10 is used to enable communication between the movable vehicle chassis 1 and the control board 11 via the CAN communication protocol.

[0089] The control board 11 is used to achieve autonomous navigation based on user settings, image information collected by the first camera 4 and the second camera 5, distance to surrounding objects collected by the lidar 7, and the scanning angle of the laser. It controls the movement of the movable vehicle chassis 1, the opening and closing of the solenoid valve, and the flow rate of the flow valve to complete the calf feeding. At the same time, it displays the current coordinates and yaw angle on the remote control terminal. The user settings include the user-set travel speed and the user-selected planned path.

[0090] The control board 11 is equipped with an autonomous navigation and control system, such as Figure 3 As shown, it includes an environmental perception subsystem, a navigation control subsystem, a vehicle chassis control subsystem, and a feeding subsystem;

[0091] The environmental perception subsystem is based on the Linux operating system and integrates the Robot Operating System (ROS). It generates point cloud data based on the distance to surrounding objects measured by the LiDAR 7 and the scanning angle of the laser. It processes the image information and point cloud data and sends the image information and processed point cloud data to the navigation control subsystem to provide basic data for subsequent path planning. The processing includes data format conversion, coordinate system transformation, filtering, downsampling, and feature extraction.

[0092] In this embodiment, the raw point cloud data generated by the LiDAR 7 may need to be converted into a format suitable for processing. For example, the point cloud data may need to be converted from the specific format of the LiDAR 7 to a general point cloud format, such as PCD (Point Cloud Data) or PLY. The point cloud data is usually generated in the coordinate system of the LiDAR 7, but for path planning, this data needs to be transformed to the global coordinate system or the robot coordinate system, which involves rotation and translation using transformation matrices. The raw point cloud data may contain noise and unnecessary points, so it may be necessary to remove outliers through filters and reduce the number of points through downsampling to facilitate processing and analysis. Useful features such as edges, corners, and planes are extracted from the point cloud data, which are crucial for subsequent path planning and obstacle detection.

[0093] The navigation control subsystem, based on Linux and the robot operating system, is used to receive image information and processed point cloud data. During initial operation and when the environment changes, a map building program is used to process the image information into a two-dimensional grid map using the gmapping algorithm. A positioning program is used to convert the processed point cloud data into positioning information using the Robot Operating System open-source framework. Based on the calf's position in the two-dimensional grid map and the positioning information, the calf's position is determined. Based on the two-dimensional grid map and the calf's position, A* and DWA algorithms are used to perform global and local path planning, respectively, to obtain several complete planned paths and store them to ensure safety and reliability in complex environments.

[0094] In cases other than changes in the operating and usage environment, the DWA algorithm is used to adjust the local planned path and achieve obstacle avoidance based on real-time acquired image information and point cloud data, on the user-selected planned path, and then send the adjusted local planned path to the vehicle chassis control subsystem.

[0095] The chassis control subsystem is used to generate drive signals according to the planned path and user settings, and send the drive signals to the drive motor of the movable chassis 1 through the CAN communication interface module 10, so that the movable chassis 1 moves according to the planned path.

[0096] The feeding subsystem is used to control the solenoid valve to be energized after reaching the calf's location, and to feed the calf with milk from the milk bucket 8 using the nipple 9. After feeding is completed, the solenoid valve is de-energized.

[0097] like Figure 4 As shown, the graphical interface includes a chassis control area, an environmental perception area, and a navigation control area;

[0098] The chassis control area includes a path selection box 12, a speed input box 13, a chassis start button 14, and a movement control area 19;

[0099] The path selection box 12 is used by the user to select the planned path for this feeding task from several planned paths. After confirming the planned path, the user clicks the OK button, and the movable vehicle chassis 1 can move according to the selected planned path to achieve the predetermined planned path movement.

[0100] The speed input box 13 is used by the user to input the travel speed of the movable vehicle chassis 1 within a set range. After inputting the speed, the user clicks the OK button to control the movable vehicle chassis 1 to move according to the set travel speed.

[0101] In this embodiment, the speed setting range is 0m / s-2m / s;

[0102] The chassis start button 14 is pressed before path selection, speed input and movement control to start the drive motor of the movable vehicle chassis 1.

[0103] The movement control area 19 is used to control the movable vehicle chassis 1 to move in four directions: front, back, left, and right.

[0104] The environmental perception area includes two camera activation buttons 15, the image display area of ​​the first camera 4, and the image display area of ​​the second camera 5.

[0105] The two camera activation buttons 15 are used to activate the first camera 4 and the second camera 5.

[0106] The image display area of ​​the first camera 4 and the image display area of ​​the second camera 5 are respectively used to display the image information captured in real time by the first camera 4 and the second camera 5.

[0107] The navigation control area includes a navigation start button 16, a save map button 17, and a map creation button 18.

[0108] The map creation button 18 is used to start the map creation program and create a two-dimensional raster map.

[0109] The Save Map button 17 is used to save the completed two-dimensional raster map;

[0110] Navigation start button 16 is used to start navigation and begin traveling along the planned route;

[0111] A method for using a calf feeding robot with an autonomous navigation and control system includes the following steps:

[0112] Step 1: Power on the movable chassis 1. On the graphical interface of the remote control terminal or control screen 2, click the two camera start buttons 15 to start the first camera 4 and the second camera 5. The control board 11 connects to the drive motor on the movable chassis 1 through the CAN communication interface module 10. Click the chassis start button 14. The control board 11 transmits the generated drive signal to the drive motor through the CAN communication protocol.

[0113] Step 2: Click the map creation button 18 on the graphical interface of the remote control terminal or control screen 2 to start the map creation program in the navigation control subsystem. Enter the set speed in the speed input box 13 on the graphical interface of the remote control terminal or control screen 2, and use the movement control area 19 on the graphical interface of the remote control terminal to control the drive motor in the movable vehicle chassis 1 to move at the set speed and direction until a complete driving process is completed in the environment. During the driving process, the first camera 4 and the second camera 5 are used to collect image information of the surrounding environment, and the laser radar 7 is used to measure the distance to surrounding objects and the scanning angle of the laser, and store them in the environmental perception subsystem of the control board 11.

[0114] Step 3: The environmental perception subsystem generates point cloud data based on the distance to surrounding objects measured by the lidar 7 and the scanning angle of the laser, processes the point cloud data, and sends the image information and the processed point cloud data to the navigation control subsystem.

[0115] Step 4: The map creation program processes the image information into a two-dimensional raster map using the gmapping algorithm. Click the save map button 17 on the graphical interface of the remote control terminal or control screen 2 to save the generated two-dimensional raster map.

[0116] Step 5: The stored two-dimensional grid map is displayed on the graphical interface of the control screen 2 and the remote control terminal. The position of the calf in the two-dimensional grid map is set as the target point. The positioning program in the navigation control subsystem uses the RobotOperating System open source framework to convert the processed point cloud data into positioning information, thereby obtaining the positioning information of the target point, that is, the position of the calf.

[0117] Step 6: Based on the two-dimensional grid map and the target point's positioning information, the navigation control subsystem uses the A* algorithm and the DWA algorithm respectively to perform global path planning and local path planning, obtaining several complete planned paths and storing them;

[0118] Step 7: Select the planned path for feeding the calf in the path selection box 12 on the graphical interface of the remote control terminal or control screen 2, and click the navigation start button 16. The chassis control subsystem generates a drive signal according to the selected planned path and user settings, and sends the drive signal to the drive motor of the movable chassis 1 through the CAN communication interface module 10, so that the movable chassis 1 starts to move according to the planned path.

[0119] Step 8: During the journey, the IMU module built into the lidar 7 acquires the current coordinates and yaw angle of a calf feeding robot with an autonomous navigation control system and displays them on the remote control terminal. At the same time, the navigation control subsystem adjusts the local planned path and achieves obstacle avoidance based on the real-time acquired image information and point cloud data using the DWA algorithm, and sends the adjusted local planned path to the vehicle chassis control subsystem. The drive signal is also sent to the drive motor of the movable vehicle chassis 1 through the CAN communication interface module 10 to adjust and control the travel path of the movable vehicle chassis 1.

[0120] Step 9: When the movable chassis 1 stops moving after reaching the target point, the feeding subsystem controls the solenoid valve to be energized, the nipple 9 is connected to the milk container 8, the amount of milk to be fed that day is predicted, and the flow rate valve is used to control the milk flow rate according to the predicted amount of milk. The calf stretches out its neck and drinks the milk independently using the nipple 9. After the set time, the solenoid valve is de-energized.

[0121] The method for controlling the milk flow rate using a flow rate valve is as follows:

[0122] S1: Retrieve daily milk feeding records over a historical period;

[0123] In this embodiment, the daily milk feeding records for the most recent week are obtained, for example: [2.5, 2.6, 2.7, 2.6, 2.5, 2.4, 2.5] liters;

[0124] S2: Store the daily milk volume records in an array or list, such as milk_volumes[], and create an array of days corresponding to the milk volume records, such as days[], where each element represents the number of days corresponding to the milk volume record;

[0125] S3: Initialize a linear regression model, input the number of days as the input feature into the linear regression model, use the daily milk feeding record as the target value, train the linear regression model, and obtain the trained linear regression model;

[0126] S4: Determine the feeding date and the corresponding number of days in the day array. Input the number of days into the linear regression model to obtain the predicted amount of milk to be fed.

[0127] S5: Calculate and set the milk flow rate based on the predicted milk volume and the set feeding time to ensure that an appropriate amount of milk is provided within the predetermined time.

[0128] The predicted milk volume and the set milk flow rate are displayed in a graphical interface;

[0129] S6: Compare the actual amount of milk fed with the predicted amount of milk fed. If the difference between the two is greater than a set threshold, adjust the parameters of the linear regression model according to the actual amount of milk fed to improve future predictions.

[0130] Step 10: The movable chassis 1 moves to the next target point to continue feeding until the feeding task is completed. The calf feeding robot with an autonomous navigation control system returns to the starting point to wait for the next feeding task or to recharge.

Claims

1. A calf feeding robot with an autonomous navigation and control system, characterized in that, It includes a mobile vehicle chassis, control screen, main control compartment, first camera, second camera, lidar base, lidar, milk bucket, solenoid valve, nipple, CAN communication interface module and control board; The movable vehicle chassis serves as a support structure, carrying the control screen, main control compartment, first camera, second camera, lidar base, lidar, milk bucket, solenoid valve, flow valve, nipple, CAN communication interface module, and control board. It provides mobility through several drive motors mounted on the movable vehicle chassis. Each drive motor is connected to the CAN communication interface module and communicates with the control board using the CAN communication protocol. A power supply for powering the system is also provided on the movable vehicle chassis. The control screen is placed outside the main control compartment and connected to the control board via an HDMI cable. It is used to display the image information captured by the first and second cameras through a graphical interface, and at the same time provide human-computer interaction functions for users. The main control compartment is fixed above the vehicle chassis, with a control board fixedly installed inside, and a first camera, a second camera, a lidar base, and a lidar placed on top. Both the first and second cameras provide infrared and LED illumination. The first and second cameras are used to capture image information from the left front and right front, respectively, and send it to the control board. The lidar base is installed above the main control compartment to fix the lidar; The lidar is fixed by a lidar bracket and is used to scan the surrounding environment, obtain the distance to surrounding objects and the scanning angle of the laser, and send them to the control board. At the same time, the built-in IMU module is used to obtain the current coordinates and yaw angle and send them to the control board. The milk bucket is used to store milk; The nipple is connected to the milk bucket via a solenoid valve and a flow valve and is used to feed calves; The solenoid valve is used to control whether the milk container and the nipple are connected. The flow rate valve is used to control the flow rate of the milk; The CAN communication interface module is used to enable communication between the vehicle chassis and the control board via the CAN communication protocol. The control board is used to achieve autonomous navigation based on user settings, image information collected by the first and second cameras, distance to surrounding objects collected by the lidar, and the scanning angle of the laser. It controls the movement of the vehicle chassis, the opening and closing of the solenoid valve, and the flow rate of the flow valve to complete the feeding of the calf. At the same time, it displays the current coordinates and yaw angle on the remote control terminal. The control board is equipped with an autonomous navigation control system, including an environmental perception subsystem, a navigation control subsystem, a vehicle chassis control subsystem, and a feeding subsystem. The environmental perception subsystem is based on the Linux operating system and integrates the robot operating system. It generates point cloud data based on the distance to surrounding objects measured by the lidar and the scanning angle of the lidar, processes the image information and point cloud data, and sends the image information and processed point cloud data to the navigation control subsystem. The processing includes data format conversion, coordinate system transformation, filtering, downsampling, and feature extraction; The navigation control subsystem, based on Linux and the robot operating system, is used to receive image information and processed point cloud data. During initial operation and when the environment changes, a map building program is used to process the image information into a two-dimensional grid map using the gmapping algorithm. A positioning program is used to convert the processed point cloud data into positioning information using the Robot Operating System open-source framework. Based on the calf's position in the two-dimensional grid map and the positioning information, the calf's position is determined. Based on the two-dimensional grid map and the calf's position, A* and DWA algorithms are used to perform global and local path planning, respectively, to obtain several complete planned paths and store them to ensure safety and reliability in complex environments. In cases other than changes in the operating and usage environment, the DWA algorithm is used to adjust the local planned path and achieve obstacle avoidance based on real-time acquired image information and point cloud data, on the user-selected planned path, and then send the adjusted local planned path to the vehicle chassis control subsystem. The chassis control subsystem is used to generate drive signals according to the planned path and user settings, and send the drive signals to the drive motor of the chassis through the CAN communication interface module, so that the chassis moves according to the planned path. The milking subsystem is used to control the solenoid valve to be energized after reaching the calf's location, to feed the calf milk from the milk bucket using a nipple, and to control the solenoid valve to be de-energized after feeding is completed.

2. A calf feeding robot with an autonomous navigation and control system according to claim 1, characterized in that, The user settings include the user-set travel speed and the user-selected planned route.

3. The calf nursing robot having an autonomous navigation control system according to claim 1, characterized in that, The graphical interface includes a chassis control area, an environmental perception area, and a navigation control area; The chassis control area includes a path selection box, a speed input box, a chassis start button, and a movement control area; The path selection box is used by the user to select the planned path for this feeding task from several planned paths. After confirming the planned path, the user clicks the OK button, and the vehicle chassis runs according to the selected planned path to achieve the predetermined planned path movement. The speed input box is used by the user to input the vehicle chassis speed within a set range. After inputting the speed, the user can click the OK button to control the vehicle chassis to move at the set speed. The chassis start button, when pressed before path selection, speed input, and movement control, is used to start the drive motor of the chassis. The movement control area is used to control the vehicle chassis to move in four directions: forward, backward, left, and right. The environmental perception area includes two camera activation buttons, an image display area for the first camera, and an image display area for the second camera. The two camera activation buttons are used to activate the first and second cameras. The image display area of ​​the first camera and the image display area of ​​the second camera are respectively used to display image information captured in real time by the first camera and the second camera. The navigation control area includes a navigation start button, a save map button, and a map creation button. The map creation button is used to start the map creation program and create a two-dimensional raster map; The Save Map button is used to save the completed 2D raster map; The navigation start button is used to initiate navigation and begin traveling along the planned route.

4. The method of using a calf nursing robot having an autonomous navigation control system of claim 1, wherein, Includes the following steps: Step 1: Power on the movable vehicle chassis. On the graphical interface of the remote control terminal or control screen, click the two camera start buttons to start the first and second cameras. The control board connects to the drive motor on the movable vehicle chassis through the CAN communication interface module. Click the chassis start button, and the control board transmits the generated drive signal to the drive motor through the CAN communication protocol. Step 2: Click the map creation button on the graphical interface of the remote control terminal or control screen to start the map creation program in the navigation control subsystem. Enter the set speed in the speed input box on the graphical interface of the remote control terminal or control screen, and use the motion control area on the graphical interface of the remote control terminal to control the drive motor in the chassis of the vehicle to move at the set speed and direction until a complete driving process is completed in the environment. During the driving process, the first and second cameras are used to collect image information of the surrounding environment, and the laser radar is used to measure the distance to surrounding objects and the scanning angle of the laser, and store them in the environmental perception subsystem of the control board. Step 3: The environmental perception subsystem generates point cloud data based on the distance to surrounding objects measured by the lidar and the scanning angle of the lidar, processes the point cloud data, and sends the image information and the processed point cloud data to the navigation control subsystem. Step 4: The map creation program processes the image information into a two-dimensional raster map using the gmapping algorithm. Click the "Save Map" button on the graphical interface of the remote control terminal or control screen to save the generated two-dimensional raster map. Step 5: The stored two-dimensional grid map is displayed on the graphical interface of the control screen and the remote control terminal. The position of the calf in the two-dimensional grid map is set as the target point. The positioning program in the navigation control subsystem uses the Robot Operating System open source framework to convert the processed point cloud data into positioning information, thereby obtaining the positioning information of the target point, that is, the position of the calf. Step 6: Based on the two-dimensional grid map and the target point's positioning information, the navigation control subsystem uses the A* algorithm and the DWA algorithm respectively to perform global path planning and local path planning, obtaining several complete planned paths and storing them; Step 7: Select the planned path for feeding the calves in the path selection box on the graphical interface of the remote control terminal or control screen, and click the navigation start button. The chassis control subsystem generates a drive signal according to the selected planned path and user settings, and sends the drive signal to the drive motor of the chassis through the CAN communication interface module, so that the chassis starts to move according to the planned path. Step 8: During the journey, the IMU module built into the lidar acquires the current coordinates and yaw angle of a calf feeding robot with an autonomous navigation control system and displays them on the remote control terminal. At the same time, the navigation control subsystem adjusts the local planned path and implements obstacle avoidance based on the real-time acquired image information and point cloud data using the DWA algorithm. The adjusted local planned path is then sent to the chassis control subsystem, and the drive signal is sent to the drive motor of the chassis through the CAN communication interface module to adjust and control the movement path of the chassis. Step 9: When the movable vehicle chassis stops moving at the target point, the feeding subsystem controls the solenoid valve to be energized, the nipple and the milk bucket are connected, the amount of milk to be fed that day is predicted, and the flow rate valve is used to control the milk flow rate according to the predicted amount of milk to be fed. The calf stretches out its neck and drinks the milk independently using the nipple. After the set time, the solenoid valve is de-energized. Step 10: The movable chassis moves to the next target point to continue feeding until the feeding task is completed. The calf feeding robot with an autonomous navigation control system returns to the starting point to wait for the next feeding task or to recharge.

5. The method of using a calf nursing robot having an autonomous navigation control system of claim 4, wherein, The method for controlling the milk flow rate using a flow rate valve in step 9 is as follows: S1: Retrieve daily milk feeding records over a historical period; S2: Store the daily milk feeding records in an array or list, and create an array of days corresponding to the milk feeding records, where each element represents the number of days corresponding to the milk feeding record; S3: Initialize a linear regression model, input the number of days as the input feature into the linear regression model, use the daily milk feeding record as the target value, train the linear regression model, and obtain the trained linear regression model; S4: Determine the feeding date and the corresponding number of days in the day array. Input the number of days into the linear regression model to obtain the predicted amount of milk to be fed. S5: Calculate and set the milk flow rate based on the predicted milk volume and the set feeding time to ensure that an appropriate amount of milk is provided within the predetermined time. The predicted milk volume and the set milk flow rate are displayed in a graphical interface; S6: Compare the actual amount of milk fed with the predicted amount of milk fed. If the difference between the two is greater than a set threshold, adjust the parameters of the linear regression model according to the actual amount of milk fed to improve future predictions.

Images