An automated management system for potting irrigation

Abstract

The application discloses an automatic management system for potted plant irrigation, and relates to the technical field of agricultural automation. The system comprises a fertilizing robot and a fertilizer mixing workstation. The fertilizing robot is provided with a lifting linear module, a rotary mechanical arm and a fixed-inclination spray head. A controller calculates the target distance based on visual recognition, and when the target exceeds the length of the mechanical arm, a fluid jet ballistic compensation mode is activated, the vertical height of the spray head and the output power of the pump body are cooperatively adjusted, and the fluid projection model is used to cover the remote target. The fluid system adopts a double-loop design, has a tangential self-circulation stirring function and an alternating cleaning function cooperated with the valve pump, and effectively prevents sedimentation and spray head blockage. In addition, the system adopts a segmented homing strategy combining visual servo and blind insertion, and physically docks through electrical and mechanical double signal confirmation. The application solves the problems of limited working range of the fixed mechanical arm and easy blockage of the pipeline, and significantly improves the adaptability and reliability of the fertilizing operation.

Description

Technical Field

[0001] This invention relates to the field of agricultural automation technology, specifically to an automated management system for potted plant irrigation. Background Technology

[0002] In modern greenhouse horticulture and potted plant cultivation, precise water and fertilizer management is a crucial step in ensuring crop growth quality. Traditional methods rely on manual hand-held sprinklers, which suffer from high labor intensity, inaccurate fertilizer application control, and low efficiency. With the development of automation technology, mobile robots equipped with fertilizer application mechanisms have gradually replaced manual operation, becoming an important piece of equipment for intelligent planting.

[0003] Existing mobile fertilization robots are typically equipped with robotic arms to perform target application. To accommodate potted plants with different arrangements and growth stages, traditional solutions often employ articulated robotic arms with telescopic or multi-degree-of-freedom capabilities. However, this design increases the robot's mechanical complexity and overall weight, leading to higher production and maintenance costs. In high-humidity greenhouse environments and corrosive fertilizer conditions, telescopic mechanisms and rotary joints are prone to corrosion and jamming, affecting the equipment's reliability and lifespan. If a fixed-length robotic arm is used to simplify the structure, the robot's coverage area is limited, making it difficult to simultaneously address target potted plants at varying distances on both sides of the application path, thus failing to meet the operational needs in unstructured environments.

[0004] Meanwhile, the water-fertilizer mixture has the physical characteristics of being prone to sedimentation and crystallization. Existing fertilization devices often have relatively simple fluid pipeline designs, lacking effective online stirring and automatic cleaning mechanisms. During long periods of stagnation or intermittent operation, fertilizer particles tend to settle at the bottom of the storage tank, resulting in uneven fertilizer concentration in the output solution. At the same time, fertilizer residue at the end of the pipeline and at the nozzles is prone to crystallization after water evaporation, causing blockage of precision nozzles and leading to operation interruption.

[0005] In terms of energy and material replenishment, mobile robots need to return to the workstation periodically for charging and fertilizer refilling. Unlike simple automatic charging, automatic docking of fluid interfaces requires high positional accuracy and connection sealing. Existing vision-guided docking technology has blind spots when the camera is close to the target, which can easily lead to position loss; moreover, relying solely on visual positioning makes it difficult to perceive the actual physical coupling state of the mechanical interface. If the injection pump is started when the interface is not fully locked, leakage or electrical short circuits can easily occur. Therefore, an automated potted plant irrigation management system with a simple structure, self-cleaning function, and automatic refill capability is needed. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides an automated management system for potted plant irrigation, which solves the technical problems of limited operating range of existing fixed robotic arms, easy sedimentation and blockage of fluid pipelines, and unreliable automatic replenishment docking.

[0007] To achieve the above objectives, the present invention provides the following technical solution: The first aspect of this invention provides an automated management system for potted plant irrigation, mainly comprising a fertilizing robot and a fertilizer mixing workstation. The fertilizing robot is equipped with a mobile chassis, a fertilizer storage tank, a lifting linear module, a fertilizing robotic arm steering drive motor, a fertilizer spraying head with a fixed installation angle, a fertilizer and wastewater output pump, a potted plant recognition camera, and a fertilizing robot controller.

[0008] The core innovation of this system lies in establishing a fluid dynamics-based jet trajectory compensation mechanism to address the technical problem of a fixed-length robotic arm being unable to cover distant targets. The fertilizer robot controller is configured to calculate the spatial coordinates of the target potted plant's roots using a potted plant recognition camera, and then calculate the horizontal distance between these coordinates and the robot. When this horizontal distance exceeds the physical length of the robotic arm, the controller activates the fluid jet trajectory compensation mode.

[0009] In this mode, the controller performs coordinated calculations of the initial fluid velocity and nozzle height based on a fluid projectile motion model. The specific control strategy employs hierarchical optimization logic.

[0010] First, the controller is set to operate the fertilizer and wastewater output pumps at rated power, at which point the corresponding initial fluid velocity is... The controller is based on the projectile's kinematic equations, and the inverse solution is used to calculate the coverage radial deviation. Required theoretical vertical height The calculation model is as follows: ; in, It is the acceleration due to gravity. The installation tilt angle for the fertilizer spray head.

[0011] Secondly, the controller verifies the theoretical vertical height. Is it within the physical travel range of the lifting linear module? If not, adjust the module directly to that height; if it exceeds the upper limit of the physical travel... The controller locks the lifting linear module at its maximum height. And based on this maximum height, recalculate the required initial velocity of the compensating fluid. : ; Subsequently, the controller adjusts the output power of the fertilizer and wastewater output pumps according to the pre-calibrated pump pressure-flow velocity mapping relationship to achieve the initial velocity of the compensating fluid, thereby achieving precise coverage of non-centered targets.

[0012] The second aspect of this invention relates to the topology of the fluid loop in the above-mentioned system and a collaborative control method, aiming to solve the problems of pipeline sedimentation and nozzle clogging.

[0013] The fertilizing robot's fluid system is equipped with a fertilizing tee connector. Its input end connects to the main pump, the first output end connects to the fertilizer spray head via a fertilizing solenoid valve, and the second output end connects to the wastewater discharge port via a wastewater output solenoid valve. The controller employs interlocking control logic: during fertilization, the wastewater output solenoid valve is forcibly closed and the fertilizing solenoid valve is opened; during waste discharge, the fertilizing solenoid valve is forcibly closed and the wastewater output solenoid valve is opened.

[0014] Furthermore, to remove pipeline residues, the controller is configured to execute an alternating cleaning process: after injecting clean water, the wastewater output solenoid valve is first opened to clean the main wastewater discharge pipeline, then the wastewater output solenoid valve is closed and the fertilizer solenoid valve is opened, forcing clean water to spray out at high speed through the fertilizer spray head. This control logic utilizes the pressure shock during fluid switching to effectively remove crystallization residues inside the spray head.

[0015] In addition, the system includes an independent anti-sedimentation stirring subsystem, which draws liquid from the bottom of the fertilizer storage tank through a fertilizer mixing pump and sprays it back into the tank through a tangentially installed outlet pipe. The vortex turbulence created by the tangential spray inside the tank can resuspend the sediment at the bottom, ensuring the uniformity of fertilizer concentration.

[0016] To address the blind zone issue in visual servoing at close range, the controller employs a segmented homing strategy. At long range, the pose of the workstation's visual identifier is calculated using a charging recognition camera, and a closed-loop control law guides the robot backward. When the distance is less than a preset blind zone threshold or the visual target is lost, the controller enters a blind insertion phase, locks the steering angle, and controls the robot to retreat in a constant low speed in a straight line.

[0017] To ensure the reliability of the physical docking, the controller simultaneously monitors both electrical connection signals and mechanical positioning signals. The mechanical positioning signal is determined by the status of a microswitch installed at the interface or the stall current characteristics of the drive wheel motor. Only after both electrical and mechanical signals are confirmed to be valid can the robot and workstation establish communication and initiate the heartbeat monitoring mechanism, thereby performing subsequent fluid replenishment and charging operations. This multi-confirmation mechanism effectively prevents the risk of fluid leakage or arcing caused by docking gaps.

[0018] This invention provides an automated management system for potted plant irrigation. It has the following beneficial effects: 1. This invention effectively expands the operational coverage of a fixed-length robotic arm by establishing a fluid jet trajectory compensation model. The system employs a hierarchical control strategy that prioritizes adjusting the nozzle height and secondarily adjusts the pump output power. This enables precise fertilization of potted plants with a radial distance exceeding the physical length of the robotic arm, without requiring complex hardware structures such as telescopic robotic arms. This design significantly improves the equipment's adaptability to unstructured planting environments while ensuring a compact and lightweight mechanical structure for the fertilization robot.

[0019] 2. The valve-pump coordinated fluid circuit designed in this invention significantly improves the system's operational stability and anti-clogging capability. Through an independent self-circulating stirring branch and a tangential spray structure, hydraulic shearing is utilized to prevent fertilizer sedimentation, ensuring uniform fertilizer concentration. Simultaneously, based on the alternating cleaning logic of the three-way connector flow direction switching, clean water is forced to flow through the precision nozzle, effectively removing crystal residue at the end of the pipeline and solving the technical problems of nozzle clogging and valve jamming in traditional fertilizer applicators after long-term operation.

[0020] 3. The segmented homing and multiple physical confirmation mechanisms proposed in this invention ensure high reliability and safety for the docking of the mobile robot with the supply station. The constant-speed linear control during the blind insertion stage fills the blind zone of near-range visual servoing, solving the problem of positioning loss in the last few centimeters. Combined with physical positioning detection based on motor stall current or microswitches, and a communication heartbeat mechanism, it ensures that the fluid interface and charging interface are in a tightly coupled state before transmission is initiated, eliminating the risk of waste liquid leakage and electrical arcing caused by loose connections. Attached Figure Description

[0021] Figure 1 This is a perspective view of the present invention; Figure 2 This is a schematic diagram of the charging port of the fertilizer mixing station of the present invention; Figure 3 This is a schematic diagram of the outer casing of the fertilizer mixing workstation of the present invention; Figure 4 This is a schematic diagram of the fertilizer mixing station controller of the present invention; Figure 5 This is a schematic diagram of the fertilizer storage tank of the present invention; Figure 6 This is a schematic diagram of the battery of the present invention; Figure 7 This is a schematic diagram of the robot's wastewater inlet according to the present invention; Figure 8 This is a schematic diagram of the fertilizer application robot arm of the present invention; Figure 9 This is a block diagram of the electrical architecture of the fertilizer robot control system of the present invention; Figure 10 This is the main flowchart of the automated fertilization management method of the present invention; Figure 11 This is a schematic diagram of the fluid jet trajectory compensation model of the present invention. Figure 12 This is a schematic diagram of the valve-pump coordinated control principle of the fluid system according to the present invention.

[0022] Among them, 100, fertilizer mixing station; 110, fertilizer mixing station display screen; 120, fertilizer inlet of fertilizer mixing station; 130, fertilizer mixing station charging port; 131, fertilizer mixing station charging QR code; 132, fertilizer mixing station charging identification area; 140, fertilizer inlet of fertilizer mixing station; 150, fertilizer mixing station water injection quick-connect port; 160, fertilizer mixing station control button; 170, fertilizer mixing station shell; 180, fertilizer mixing system of fertilizer mixing station; 190, fertilizer mixing station controller; 200, fertilizer application robot; 210, potted plant recognition camera; 2100, fertilizer application robot controller; 220, fertilizer application component; 221, lifting linear module; 222, fertilizer application pipe; 223, fertilizer inlet pipe; 224, first wastewater output pipe; 225, second wastewater output pipe; 226, fertilizer flow meter; 227, fertilizer application... 228. Solenoid valve; 229. Fertilizer tee connector; 2210. Fertilizer and wastewater output pump; 2211. Fertilizer and wastewater input pipe; 2212. Wastewater output solenoid valve; 2213. Mixed fertilizer pump; 2214. Mixed fertilizer inlet pipe; 2215. Fertilizer spray head; 2216. Mixed fertilizer outlet pipe; 2217. Fertilizer robotic arm; 2218. Fertilizer robotic arm steering drive motor; 2219. Fertilizer robotic arm steering drive motor mounting component; 230. Fertilizer storage tank; 240. Two-dimensional LiDAR; 250. Vehicle body; 260. Robot display screen; 261. Fertilizer robot control button; 270. Drive wheel; 280. Robot fertilizer inlet; 281. Robot charging recognition camera; 282. Robot charging port; 283. Robot wastewater inlet; 290. Battery. Detailed Implementation

[0023] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Please see the appendix Figure 1 - Appendix Figure 12 This invention provides an automated management system for potted plant irrigation. The system includes a fixed fertilizer mixing workstation 100 and a mobile fertilizer application robot 200. The fertilizer mixing workstation 100 provides the fertilizer application robot 200 with power supply, material supply, and waste liquid recycling. The fertilizer application robot 200 autonomously moves within its operating area and performs quantitative fertilizer application and irrigation tasks.

[0025] The fertilizer mixing station 100 includes a casing 170, which integrates a fertilizer mixing system 180 and a fertilizer mixing station controller 190. The controller 190 is the core processing unit of the fertilizer mixing station 100, used to control the operation of various components within the station. The casing 170 has a display screen 110 and control buttons 160 on its exterior, used to display system status parameters and receive user commands, respectively. A quick-connect water inlet 150 is located on the side of the fertilizer mixing station 100 for connecting to an external water source. The front end of the fertilizer mixing station 100 has an interface component for interaction with the fertilizer application robot 200, including a fertilizer inlet 120, a charging port 130, and a fertilizer receiving port 140. The fertilizer inlet 120 is connected to the output of the fertilizer mixing system 180 via a pipeline. The fertilizer mixing station charging port 130 is connected to the mains power or power grid via an internal circuit. The fertilizer inlet 140 is connected to a waste liquid collection container or sewage pipe. A fertilizer mixing station charging QR code 131 and a fertilizer mixing station charging identification area 132 are provided in the vicinity of the charging port 130 as passive identifiers for the visual positioning of the fertilizer application robot 200.

[0026] The fertilizing robot 200 includes a body 250, drive wheels 270, a fertilizing robot controller 2100, a battery 290, a potted plant recognition camera 210, a 2D LiDAR 240, fertilizing components 220, and a fertilizer storage tank 230. The body 250 forms the robot's mobile chassis frame. The drive wheels 270 are mounted on the bottom of the body 250 and are driven by an internal motor to achieve planar movement of the robot. The battery 290 is installed inside the body 250 and supplies power to the robot's various electrical loads. The fertilizing robot controller 2100 is installed inside the body 250 and serves as the control center for the entire machine. The 2D LiDAR 240 is diagonally positioned at the edge of the body 250 to scan the surrounding environment and construct a 2D grid map. The robot display screen 260 and fertilizing robot control buttons 261 are mounted on the rear panel of the body 250 for human-machine interaction.

[0027] The front end of the fertilizing robot 200 is equipped with a docking component that matches the fertilizer mixing station 100. This docking component includes a robot fertilizer inlet 280, a robot charging identification camera 281, a robot charging port 282, and a robot wastewater inlet 283. The robot fertilizer inlet 280 is connected to the input port of the fertilizer storage tank 230 via a fertilizer inlet pipe 223. The robot charging identification camera 281 is installed in the center of the front end and is used to collect image information from the charging QR code 131 of the fertilizer mixing station. The robot charging port 282 is connected to the battery 290. The robot wastewater inlet 283 is used to discharge cleaning wastewater or remaining pesticide solution.

[0028] The fertilization component 220, as the core assembly for performing irrigation operations, is mounted on the upper part of the vehicle body 250. The fertilization component 220 includes a lifting linear module 221, a fertilization robotic arm steering drive motor mounting component 2218, a fertilization robotic arm steering drive motor 2217, a fertilization robotic arm 2216, a fertilizer spraying head 2214, and a fluid control piping system. The lifting linear module 221 is vertically mounted on the vehicle body 250 and is used to drive the fertilization robotic arm steering drive motor mounting component 2218 to move vertically. The fertilization robotic arm steering drive motor 2217 is fixed to the fertilization robotic arm steering drive motor mounting component 2218, and its output shaft is threadedly connected to one end of the fertilization robotic arm 2216. The length of the fertilization robotic arm 2216 is fixed and does not have a telescopic function. The fertilizer spraying head 2214 is fixedly mounted to the other end (terminal end) of the fertilization robotic arm 2216.

[0029] The fluid control piping system includes a fertilizer storage tank 230, a fertilizer and wastewater output pump 229, a fertilizer mixing pump 2212, a fertilizer flow meter 226, a fertilizer solenoid valve 227, a wastewater output solenoid valve 2211, and connecting pipes. The fertilizer storage tank 230 is fixed inside the vehicle body 250 and is used to store the mixed fertilizer solution. The two ends of the fertilizer mixing pump 2212 are connected to a fertilizer inlet pipe 2213 and a fertilizer outlet pipe 2215, respectively. Both pipes are inserted inside the fertilizer storage tank 230, forming an independent internal circulation mixing circuit. The input end of the fertilizer and wastewater output pump 229 is connected to the bottom of the fertilizer storage tank 230 via a fertilizer and wastewater inlet pipe 2210. The output end of the fertilizer and wastewater output pump 229 is connected to the first port of the fertilizer tee connector 228. The second port of the fertilizer tee connector 228 is connected to the fertilizer flow meter 226. The outlet of the fertilizer flow meter 226 is connected to the fertilizer solenoid valve 227. The fertilizer solenoid valve 227 is connected to the fertilizer spray head 2214 through the fertilizer pipe 222. The third port of the fertilizer tee connector 228 is connected to the first wastewater output pipe 224. The first wastewater output pipe 224 is connected to the input of the wastewater output solenoid valve 2211. The output of the wastewater output solenoid valve 2211 is connected to the robot wastewater inlet 283 through the second wastewater output pipe 225.

[0030] During system operation, the fertilizer robot controller 2100 control system executes the following overall workflow: During the initialization phase, the fertilizer robot controller 2100 performs a self-check to confirm the battery power 290 and the liquid level in the fertilizer storage tank 230. If the status is normal, the fertilizer robot controller 2100 uses data from the two-dimensional LiDAR 240 for positioning and controls the drive wheels 270 to move the vehicle body 250 to the work area.

[0031] During the localization and recognition phase, when the robot moves near the target location, the potted plant recognition camera 210 captures environmental images. The vision recognition module 301 inside the fertilizing robot controller 2100 processes the images, identifies the potted plant features, and calculates the spatial coordinates of the target potted plant relative to the robot.

[0032] During the fertilization execution phase, the fertilization robot controller 2100, based on the spatial coordinates of the target potted plant, first controls the lifting linear module 221 to adjust the height of the fertilizer spray head 2214. Subsequently, the fertilization robot controller 2100 controls the fertilization robotic arm steering drive motor 2217 to rotate, causing the fertilization robotic arm 2216 to adjust the horizontal angle of the fertilizer spray head 2214. After alignment, the fertilization robot controller 2100 controls the fertilization solenoid valve 227 to open, while keeping the wastewater output solenoid valve 2211 closed, and starts the fertilizer and wastewater output pump 229. The fertilizer is sprayed from the fertilizer spray head 2214 through the fertilization pipe 222. During this process, the fertilization flow meter 226 monitors the flow data in real time and feeds it back to the fertilization robot controller 2100. When the pre-set fertilization amount is reached, the system stops pumping and closes the valves.

[0033] During the waste discharge and replenishment phase, when the task is completed or pipeline cleaning is required, the fertilization robot controller 2100 controls the fertilization solenoid valve 227 to close, while simultaneously opening the wastewater output solenoid valve 2211. The robot moves to the fertilizer mixing station 100 and achieves precise docking by recognizing the fertilizer mixing station's charging QR code 131 through the robot's charging recognition camera 281. After docking, the robot's wastewater inlet 283 connects to the fertilizer inlet 140 of the fertilizer mixing station, and the fertilizer and wastewater output pump 229 starts to discharge the waste liquid. Subsequently, the robot's fertilizer inlet 280 connects to the fertilizer injection port 120 of the fertilizer mixing station to receive new fertilizer, and the robot's charging port 282 connects to the fertilizer mixing station's charging port 130 for charging.

[0034] This invention provides a workflow for an automated management system for potted plant irrigation. The workflow mainly includes the following steps: System initialization and self-test logic S100 is a prerequisite for ensuring the safe operation of the fertilizer application robot 200. This step is mainly executed by the fertilizer application robot controller 2100, and specifically includes the following processing flow: S101, Controller Hardware and Communication Link Detection. After the fertilizer robot controller 2100 is powered on, it runs the initialization program and sends detection signals to each connected peripheral device sequentially via the internal bus. The controller sends position query commands to the lifting linear module 221 and the fertilizer robotic arm steering drive motor 2217, and starts a timer to monitor whether the corresponding encoder position feedback signal is received within a preset time window; if no feedback is received after the timeout, the controller writes a fault code to the error log and pauses system startup. The controller sends a status query frame to the fertilizer flow meter 226 to check whether its communication interface responds. The controller detects the data stream of the two-dimensional lidar 240 to determine whether the point cloud data is completely output. The above hardware communication protocol and underlying driver detection method can be implemented using a standard industrial bus polling mechanism by those skilled in the art, which is common knowledge in the field and will not be elaborated further here.

[0035] S102, Energy State Assessment and Threshold Determination. The fertilizer robot controller 2100 reads the terminal voltage of the battery 290 and the state of charge (SOC) fed back by the battery management system (BMS) through the analog-to-digital converter interface. The minimum operating voltage threshold of the system is set to... This threshold is set to 1.2 times the power required for the robot to return fully loaded to the fertilizer mixing workstation 100. If the current detected voltage... This indicates that the remaining power is insufficient to support this cruise operation. The controller sets the operation status to "low power return mode" and directly jumps to execute the automatic homing and resupply process in S400, and displays the low power status on the robot display screen 260.

[0036] S103, Fluid Material Inventory Detection. The fertilizer robot controller 2100 detects the liquid level sensor signal inside the fertilizer storage tank 230. This liquid level sensor is either a non-contact ultrasonic sensor or a contact float switch. The controller will collect the real-time liquid level height. With the preset minimum aspiration protection threshold Perform a numerical comparison. If The controller will set the internal pump disable flag to forcibly cut off the PWM drive control signal output of the fertilizer and wastewater output pump 229 to prevent the pump body from being damaged due to overheating from idling; at the same time, the controller will trigger the return-to-base replenishment logic to guide the robot back to the fertilizer mixing workstation 100 for fertilizer injection.

[0037] S104, Initial pose confirmation. With both hardware and material conditions normal, the fertilizer robot controller 2100 activates the 2D LiDAR 240 to scan the environment and acquire 2D point cloud data of the current environment. The controller performs feature matching between the current point cloud data and the pre-stored global probabilistic grid map of the work area in the controller's memory to calculate the initial pose of the fertilizer robot 200 in the global coordinate system. If the pose confidence calculated by the matching algorithm is lower than a preset threshold, the controller controls the drive wheel 270 to rotate in place to scan for more environmental features until the positioning converges. After positioning is completed, the system enters the S200 autonomous movement and visual positioning stage.

[0038] S200, Autonomous Movement and Visual Positioning. This step aims to enable the fertilizer application robot 200 to move from its initial position to the vicinity of the target potted plant and establish a precise local coordinate system for the target potted plant, providing spatial parameters for subsequent fertilization control. The specific processing procedure is as follows: S201, Global Path Planning and Movement. The fertilizer robot controller 2100 extracts the approximate coordinates of the current target potted plant on the global map based on a preset task list. Using an AI algorithm or Dijkstra's algorithm, the controller plans a safe travel path from the current pose to the vicinity of the target coordinates on a global grid map with known obstacles. The controller adjusts the speed difference of the drive wheels 270 through a PID motion control algorithm, driving the vehicle body 250 along the planned path. When the vehicle body 250 reaches a stopping area within a preset range from the target coordinates, the controller controls the drive wheels 270 to stop rotating and maintains a braking state to fix the vehicle body's position.

[0039] S202, Image Acquisition and Preprocessing. With the vehicle body 250 stationary, the controller activates the potted plant recognition camera 210. The potted plant recognition camera 210 acquires RGB color image frames from directly in front of it. The controller preprocesses the raw image, including Gaussian filtering for noise reduction and histogram equalization to enhance contrast, in order to eliminate the impact of ambient lighting changes on recognition accuracy.

[0040] S203, Target Feature Extraction. The controller uses a pre-trained object detection model based on a convolutional neural network (CNN) to reason about the pre-processed image and identify the bounding box of the potted plant region in the image. The controller extracts the midpoint of the bottom edge of this bounding box as the feature point of the potted plant's roots on the image plane, and records its pixel coordinates as... .

[0041] S204, Local Spatial Coordinate Calculation. This involves calculating the two-dimensional pixel coordinates. The data is converted to the 3D physical coordinates of the fertilizing robot 200, and the controller performs inverse perspective transformation based on the pinhole camera model. The optical center of the potted plant recognition camera 210 is set as the origin of the local visual coordinate system. The fixed installation height of the camera relative to the ground is known to be... The pitch angle of the camera's optical axis relative to the horizontal plane is Assuming the potted plant is placed on a flat surface, the depth value of the feature point in the local visual coordinate system is... The solution can be obtained through trigonometric geometric relations: ; Among them, is the focal length parameter of the camera, is the coordinate of the principal point of the image. The above parameters are all included in the pre-calibrated intrinsic parameter matrix .

[0042] Based on the calculated depth value , the spatial coordinates of the root of the potted plant relative to the optical center of the camera are calculated as follows: ; S205, coordinate system conversion and deviation quantification. The controller reads the hand-eye calibration matrix pre-stored in the memory. This matrix describes the rigid body transformation relationship of the optical center of the camera relative to the center of the rotation axis of the fertilization robotic arm steering drive motor 2217. The coordinates of the root of the target potted plant in the robotic arm coordinate system are: ; Based on this coordinate , the controller calculates the horizontal Euclidean distance between the center of the root of the target potted plant and the center of the rotation axis of the robotic arm, as well as the target azimuth angle

[0043] ; ; The controller temporarily stores the above calculated and in the memory as the key input parameters for the fluid jet trajectory compensation control in the subsequent step S300. At this time, the fertilization robot controller 2100 has completed the accurate real-time positioning of the operation target.

[0044] S300, fertilization execution and fluid control. This step is the core control link of this embodiment, aiming to compensate for the size limitation of the mechanical structure by adjusting the hydrodynamic parameters to achieve precise fertilization of non-centered targets. The specific control strategy is as follows: S301, determination of the robotic arm coverage range and deviation calculation. The fertilization robot controller 2100 reads the fixed length parameter of the robotic arm stored in the memory, and compares it with the target horizontal distance calculated in step S205. Define the radial deviation . If , it indicates that the target is within the turning radius of the robotic arm, and the controller enters the vertical fixed-point fertilization mode: the controller calculates the target rotation angle To align with the target location, the vehicle body 250 is controlled to fine-tune its position or maintain the current position, ensuring that the fertilizer spray head 2214 is directly above the target; simultaneously, the height is set. The preset minimum splash protection height And set the pump control signal to the low-pressure drip irrigation value. .like This indicates that the target is outside the physical reach of the robotic arm. The controller then activates the fluid jet ballistic compensation mode and executes subsequent sub-steps.

[0045] S302, Fluid ballistic dynamics modeling. In jet compensation mode, the controller calculates the required initial fluid velocity and nozzle height based on the fluid projectile motion equations. Let the installation angle of the fertilizer spray head 2214 relative to the vertical axis be... (This angle is a fixed value, and) The acceleration due to gravity is The horizontal displacement (i.e., range) of the fluid after it leaves the nozzle and reaches its point of impact. With the initial velocity of the fluid and the vertical height of the sprinkler head relative to the surface of the potted plant The mapping model is as follows: ; in, and These are the projection coefficients of the velocity vector in the horizontal and vertical directions, respectively. S303, Nonlinear mapping of the pump control system is established. Initial fluid velocity. The output power of the fertilizer and wastewater output pump 229 is determined by the controller's establishment of the pump drive PWM signal duty cycle. With the initial velocity of the fluid The transfer function of the spray head. The effective flow cross-sectional area of ​​the spray head is known to be... Within the linear operating region of the pump, the initial fluid velocity model is: ; in, This is the flow gain coefficient of the pump. This is the system dead zone deviation constant. The above parameters were obtained through calibration experiments and are fixed in the controller parameter table.

[0046] S304, Multivariable Cooperative Control Decision. The controller bases its decisions on the calculated radial deviation. Solving makes Control parameter combinations The controller executes the following hierarchical calculation logic: First-level calculation: Height-first solution. The controller is set to operate the pump at the PWM value corresponding to the standard rated pressure. At this time, the initial velocity of the fluid is Substitute it into the ballistic model to solve for the required altitude. The revised physical calculation formula is as follows: ; Second-level calculation: constraint verification and speed correction. The controller reads the physical travel range of the lifting linear module 221. .like The controller outputs control commands: target height. Pump control signal .

[0047] like This indicates that adjusting the height alone cannot cover the deviation. The controller forcibly clamps the target height to its maximum value. And based on this maximum height, the required initial fluid velocity is calculated in reverse. : ; Subsequently, the controller follows the formula Calculate the required pump control signals. If the calculated... Exceeding the pump's rated maximum duty cycle The controller will determine that the current goal is unattainable, set the "fertilization failed" error flag, abandon the goal, and log the error.

[0048] S305, dynamic compensation for execution action and fertilizer application rate. The controller drives the lifting linear module 221 to move to the calculated target height. Once in position, the controller opens the fertilization solenoid valve 227 and operates at the calculated duty cycle. Driven by pump 229 for both fertilizer and wastewater output. Changes in pump output power will affect the flow rate per unit time. Changes were made to ensure the total amount of fertilizer applied. The controller calculates the opening duration of the solenoid valve in real time to maintain a constant value. : ; The controller starts a hardware timer, which is activated when the pump running time reaches [a certain threshold]. Then, shut down the fertilizer and wastewater output pumps 229 and the fertilizer solenoid valve 227 in sequence to complete the operation.

[0049] The auxiliary fluid function control logic S400 mainly includes anti-sedimentation self-circulation stirring and waste liquid discharge cleaning. It solves pipeline blockage and liquid sedimentation problems through timing control of the solenoid valve and pump body. The specific implementation logic is as follows: S401, anti-sedimentation self-circulating stirring logic. The fertilizer robot controller 2100 is equipped with an internal periodic timer to set the stirring cycle. With stirring duration When the timer count reaches... At this time, the controller detects the current status of the robot drive wheel 270 and the fertilizer solenoid valve 227. If the drive wheel 270 is detected to be stopped and the fertilizer solenoid valve 227 is detected to be closed, the controller outputs a drive signal to start the fertilizer mixing pump 2212. The fertilizer mixing pump 2212 draws the fertilizer solution from the bottom of the fertilizer storage tank 230 into the fertilizer mixing inlet pipe 2213, and after being pressurized by the pump body, it is sprayed back into the fertilizer storage tank 230 at high speed through the fertilizer mixing outlet pipe 2215. The end outlet of the fertilizer mixing outlet pipe 2215 is installed horizontally, and the water outlet direction is parallel to the tangential direction of the inner wall of the fertilizer storage tank 230, so that the return liquid forms a vortex turbulence in the tank, thereby resuspending and mixing the sediment at the bottom. The stirring duration reaches Afterwards, the controller shuts down the fertilizer mixing pump 2212 and resets the timer.

[0050] S402, waste discharge and cleaning mode switching logic. This function relies on the interlock control of the fertilizer solenoid valve 227 and the wastewater output solenoid valve 2211. The controller defines two mutually exclusive control modes: In the fertilization operation mode, the controller continuously outputs a low-level signal to the wastewater output solenoid valve 2211 to keep it in a normally closed state without power, and only allows the fertilization solenoid valve 227 to open according to the fertilization command. At this time, the fluid flows to the fertilizer spray head 2214 through the fertilization tee connector 228.

[0051] In the waste discharge operation mode, the controller first outputs a low-level signal to close the fertilizer solenoid valve 227, and then outputs a high-level signal to open the wastewater output solenoid valve 2211. Since the wastewater output solenoid valve 2211 is connected to the lateral diversion port of the fertilizer tee connector 228, the pipeline flow direction is switched to the second wastewater output pipe 225, which leads directly to the robot's wastewater inlet 283.

[0052] S403, Post-docking waste cleaning execution. After the fertilizer application robot 200 and the fertilizer mixing workstation 100 complete physical docking, the fertilizer application robot controller 2100 reads the signal from the Hall sensor or mechanical microswitch installed at the docking interface to confirm the connection status. After confirming successful docking, the controller executes the following cleaning sequence: Step 1: Main pipeline waste discharge. Switch the controller to waste discharge operation mode, open the wastewater output solenoid valve 2211, start the fertilizer and wastewater output pump 229 to operate at maximum rated power, and discharge the residual waste liquid in the fertilizer storage tank 230 into the fertilizer mixing station 100.

[0053] Step 2: Water Injection. The controller stops the pump, and the fertilizer mixing station 100 injects a measured amount of clean water into the fertilizer storage tank 230 through the fertilizer injection port 120.

[0054] Step 3: Pipeline Cleaning. The controller restarts the fertilizer and wastewater output pumps 229. During this stage, the controller alternately opens the fertilizer solenoid valve 227 and the wastewater output solenoid valve 2211. Specifically, the wastewater output solenoid valve 2211 is opened first for a certain period of time. Flush the main wastewater discharge pipe; then close the wastewater output solenoid valve 2211 and open the fertilizer application solenoid valve 227 for a period of time. The process involves allowing clean water to flow through the fertilizer pipe 222 and spray it out from the fertilizer spray head 2214 to remove any residual crystals inside the spray head. The sprayed clean water is collected in a water collection tank (not shown) below the fertilizer mixing station 100.

[0055] After the above steps two and three are repeated a preset number of times, the controller closes all valves and pumps, ending the cleaning process.

[0056] S500, automatic homing and resupply interaction. This step involves precise docking and multi-physics coupling (fluid, electrical, data) between the mobile robot and the stationary workstation. The specific control process is as follows: S501, Global Navigation Homecoming. When the fertilizer robot controller 2100 triggers the homecoming condition (including battery voltage below a threshold)... When the fertilizer storage tank level is below the threshold or the task list is cleared, the controller reads the global coordinates of the fertilizer mixing workstation 100 from its internal memory. The controller invokes a global path planning algorithm to plan a path from the current position to a pre-dock point at a preset distance (e.g., 1.0 meter) in front of the workstation. The drive system controls the vehicle body 250 to move to the pre-dock point and rotates in place to adjust the vehicle's posture so that the normal direction of the docking panel at the robot's tail is approximately parallel to the normal direction of the docking surface of the mixing workstation 100.

[0057] S502, segmented visual servo alignment. At the pre-dock point, the controller activates the charging recognition camera 281 located at the robot's tail. The camera captures a visual identifier (part number 131) mounted on the panel of the fertilizer mixing workstation 100. The controller performs distortion correction on the acquired image, identifies the corner points of the visual identifier, and uses the PnP algorithm to calculate the pose vector of the camera coordinate system relative to the visual identifier coordinate system. .

[0058] The controller employs a piecewise approximation control strategy: Phase 1 (Visual Loop Segment): When the longitudinal distance... When the blind zone threshold (e.g., 0.2 meters) is reached, the controller performs visual closed-loop control. This is based on the lateral deviation. ,distance and yaw angle deviation Calculate the speed of the drive wheels: ; ; in, , , This is the gain for PID control.

[0059] Second stage (blind insertion segment): when the longitudinal distance... If the visual target is lost due to excessive distance, the controller locks the current steering angle and maintains it. and at a constant low speed (For example, 0.05 m / s) Control the vehicle body to move backward in a straight line, and use the guide slope of the mechanical interface to eliminate small residual errors until the physical docking is completed.

[0060] S503, Multiple Physical Connection Confirmation. During blind insertion, the fertilizer robot controller 2100 simultaneously monitors two signals to confirm the docking status: First, the electrical connection signal is used to detect whether the workstation's pilot voltage is detected at the 280 terminal of the charging electrode. Second, a mechanical positioning signal is used to detect the status of the microswitch installed at the docking interface, or to monitor the current value of the drive wheel motor. When the drive wheel motor current instantaneously exceeds the stall threshold... And the duration exceeds At that time, it was determined that the robot had physically pressed against the workstation's limiting surface.

[0061] After both signals are confirmed to be valid, the controller establishes a connection with the mixed fertilizer station controller 190 through the wireless communication module or the power line carrier communication module. Both parties start the heartbeat monitoring mechanism. If no heartbeat response is received from the other party within the specified time, the emergency stop reset logic is immediately triggered.

[0062] S404, Replenishment Operation Sequence Execution. After a successful handshake, the system enters the replenishment process.

[0063] Step 1: Perform the waste discharge and cleaning process described in S303 to remove pipeline residues.

[0064] Step Two: Fertilizer Injection. The fertilizer robot controller 2100 sends a fertilizer injection request command and the required amount. The fertilizer mixing station controller 190 starts the fertilizer injection pump, injecting liquid into the fertilizer storage tank 230 through the physical fluid interface. The fertilizer application robot controller 2100 monitors the liquid level sensor in real time, and sends a stop fertilizer injection command when the liquid level reaches the full load threshold.

[0065] Step 3: Power supply. The fertilizer mixing station controller 190 closes the charging circuit contactor to charge the battery 290 using constant current-constant voltage (CC-CV).

[0066] If a heartbeat packet is lost, the liquid level changes abruptly, or the charging temperature becomes too high during steps two and three, the controller will immediately send an emergency separation command.

[0067] Once both power and fertilizer solution have been replenished, the fertilizer applicator controller 2100 sends a separation request. The fertilizer mixing station controller 190 disconnects the circuit from the fluid pump. The fertilizer applicator robot 200 executes a forward movement command and leaves the workstation, preparing for the next round of operations.

Claims

1. An automated management system for potted plant irrigation, characterized in that, include: Fertilizer application robot (200) and fertilizer mixing workstation (100); The fertilizer application robot (200) includes: The vehicle body (250) serves as a mobile carrier platform, with drive wheels (270) installed at its bottom and a battery (290) for power supply installed inside. A two-dimensional lidar (240) is installed on the surface of the vehicle body (250). Fertilizer storage tank (230) is mounted on the vehicle body (250); The fertilization component (220) is installed on the vehicle body (250). The fertilization component (220) includes a vertically arranged lifting linear module (221), a motor fixing component (2218) connected to the moving end of the lifting linear module (221), a drive motor (2217) fixed by the motor fixing component (2218), the drive motor (2217) steering and driving the fertilization robotic arm (2216), and a fertilizer spraying head (2214) installed at the end of the fertilization robotic arm (2216). A fertilizer and wastewater output pump (229) is used to drive fluid to be sprayed from the fertilizer spray head (2214); A potted plant recognition camera (210) is used to collect images of the work area; The robot display screen (260) and fertilizer robot control buttons (261) are used for human-computer interaction; The fertilizer robot controller (2100) is connected to the aforementioned electrical components; The fertilizer robot controller (2100) is configured to: calculate the target coordinates based on the data from the potted plant recognition camera (210), and when the horizontal distance of the target is greater than the physical length of the fertilizer robot arm (2216), activate the fluid jet trajectory compensation mode, and coordinately adjust the height of the lifting linear module (221) and the output power of the fertilizer and wastewater output pump (229) so that the fluid landing point covers the target.

2. The automated management system for potted plant irrigation according to claim 1, characterized in that, The fertilizer application robot controller (2100) is configured to perform the following when calculating the spatial coordinates of the roots of the target potted plant: Control the potted plant recognition camera (210) to acquire images and identify the bounding box of the potted plant area; Extract the midpoint of the bottom edge of the bounding box as a feature point; Obtain the fixed installation height and optical axis pitch angle of the potted plant recognition camera (210) relative to the ground; Based on the pre-set assumption of flat ground, the depth value of the feature point in the local visual coordinate system is calculated using triangular geometric relationships; By combining the intrinsic parameter matrix of the camera and the pre-stored hand-eye calibration matrix, the pixel coordinates of the feature points are converted into three-dimensional physical coordinates relative to the center of the rotation axis of the steering drive motor (2217) of the fertilizer application robot arm (2216).

3. The automated management system for potted plant irrigation according to claim 1, characterized in that, The fertilizer application robot controller (2100) executes the following hierarchical control strategy in the fluid jet trajectory compensation mode: Establish a nonlinear mapping model between fluid range, initial fluid velocity, and nozzle vertical height; First-level calculation: Set the fertilizer and wastewater output pump (229) to operate at the fluid initial velocity corresponding to the rated power, and calculate the theoretical vertical height required to cover the target horizontal distance based on the nonlinear mapping model; Second-level calculation: Determine whether the theoretical vertical height exceeds the physical travel range of the lifting linear module (221); If the height is not exceeded, control the lifting linear module (221) to move to the theoretical vertical height, and control the fertilizer and wastewater output pump (229) to output at the rated power; If the target height is exceeded, the target height is locked to the maximum value of the physical travel range, and the nonlinear mapping model is substituted based on the maximum value to recalculate the initial velocity of the compensating fluid required to cover the target horizontal distance, and the output power of the fertilizer and wastewater output pump (229) is adjusted accordingly.

4. The automated management system for potted plant irrigation according to claim 3, characterized in that, The fertilizing robot (200) also includes a fertilizing flow meter (226) installed in the fluid pipeline; the fertilizing robot controller (2100) is further configured to perform dynamic compensation of the fertilization amount. After calculating the visual control signal for driving the fertilizer and wastewater output pump (229), the flow data is fed back in real time using the fertilizer flow meter (226) to calculate the start-up duration required to reach the total amount of pre-fertilized fertilizer. After the controlled operation of the fertilizer and wastewater output pump (229) reaches the specified start-up duration, the pump operation is automatically stopped.

5. The automated management system for potted plant irrigation according to claim 1, characterized in that, The fluid system piping connection structure of the fertilizer application robot (200) includes: The fertilizer and wastewater inlet pipe (2210) is connected between the fertilizer and wastewater outlet pump (229) and the inlet end of the fertilizer tee connector (228); The fertilizer pipe (222) is connected at one end to the first output end of the fertilizer tee connector (228) and at the other end to the fertilizer spray head (2214), and a fertilizer solenoid valve (227) is provided on the pipe. The wastewater output pipeline includes a first wastewater output pipe (224) and a second wastewater output pipe (225) connected in sequence. One end of the pipe is connected to the second output end of the fertilizer tee connector (228), and the other end is connected to the robot wastewater inlet (283). A wastewater output solenoid valve (2211) is provided on the pipeline. The fertilization robot controller (2100) is configured to switch between fertilization mode and waste discharge mode by controlling the opening and closing of the fertilization solenoid valve (227) and the wastewater output solenoid valve (2211).

6. The automated management system for potted plant irrigation according to claim 1, characterized in that, The arrangement of the two-dimensional lidar (240) is as follows: The two two-dimensional lidars (240) are respectively installed at the diagonal positions of the groove in the middle of the vehicle body (250) and located on top of the robot's wastewater inlet (283) for environmental perception and navigation.

7. The automated management system for potted plant irrigation according to claim 1, characterized in that, The fertilizer application robot (200) also includes an anti-sedimentation stirring subsystem and a refueling interface: A fertilizer mixing pump (2212) is connected to the bottom of the fertilizer storage tank (230) via a fertilizer mixing water inlet pipe (2213); The fertilizer outlet pipe (2215) is connected to the output end of the fertilizer pump (2212), and its outlet is located inside the fertilizer storage tank (230) and is arranged tangentially. The robot fertilizer inlet (280) is connected to the fertilizer storage tank (230) through the fertilizer inlet pipe (223); The fertilizer robot controller (2100) is configured to periodically activate the fertilizer mixing pump (2212) to create vortex turbulence.

8. The automated management system for potted plant irrigation according to claim 1, characterized in that, The fertilizing robot (200) is equipped with a robot charging identification camera (281), a robot charging port (282), and a robot wastewater inlet (283) at its tail; the fertilizer mixing station (100) is equipped with a fertilizer mixing station charging port (130), a fertilizer mixing station charging QR code (131), and a fertilizer mixing station charging identification area (132). The fertilizer robot controller (2100) is configured to perform segmented homing control: First stage: Use the robot charging recognition camera (281) to recognize the charging QR code (131) or the charging recognition area (132) of the mixed fertilizer station, calculate the pose and guide the robot to back up; Second stage: When the distance is less than the preset blind zone threshold, control the robot to retreat in a straight line at a constant low speed until the robot charging port (282) docks with the fertilizer station charging port (130).

9. The automated management system for potted plant irrigation according to claim 8, characterized in that, The fertilizer robot controller (2100) is configured to confirm successful physical docking via the following logic: Check whether the electrical connection signal at the robot charging port (282) is valid; Simultaneously detect the mechanical positioning signal, or monitor whether the motor current of the drive wheel (270) exceeds the stall threshold; Once the docking is confirmed to be successful, a communication connection will be established.

10. The automated management system for potted plant irrigation according to claim 1, characterized in that, The fertilizer mixing station (100) also includes: The fertilizer mixing station has an outer shell (170) on which a fertilizer mixing station display screen (110) and fertilizer mixing station control buttons (160) are installed. The fertilizer mixing system (180) of the fertilizer mixing station is installed inside the housing (170) of the fertilizer mixing station and includes a pump body for driving fluid; The fertilizer mixing station controller (190) is used to control the fertilizer mixing system (180) of the fertilizer mixing station. The fertilizer injection port (120) of the fertilizer mixing station is used to connect with the fertilizer receiving port (280) of the robot to add fertilizer; The fertilizer inlet (140) of the fertilizer mixing station is used to connect with the wastewater inlet (283) of the robot to recycle waste liquid; The quick-connect spout (150) for water injection at the fertilizer mixing station is used to connect to an external water source; The fertilizer mixing station controller (190) is configured to control fertilizer injection, water injection, or waste liquid recovery operations in response to commands.

Images