Variable spray control system based on vertical application structure and nozzle flow independent control method thereof

Abstract

This invention discloses a variable spray control system based on a vertical application structure and its independent nozzle flow control method. The variable spray control system has eight nozzles symmetrically installed on both sides of a vertical spray support. During operation, the system uses a two-dimensional lidar to detect the canopy characteristics of the fruit trees. Based on the vertical structure variable application model, the spraying area is segmented, and the canopy volume of each spraying area, the required output flow of each nozzle, and the duty cycle of the corresponding variable spray solenoid valve are independently calculated. The forward speed of the orchard sprayer is obtained through the BeiDou navigation system. The spraying delay is calculated by combining the two-dimensional lidar scanning cycle, the distance between the lidar and the spray structure, and the hardware response time. After delay control, the microcomputer sends variable spray commands to the main variable spray controller, achieving independent nozzle flow control for target-oriented variable spraying. This variable spray system offers high target accuracy, is environmentally friendly and reduces pesticide application, and increases efficiency.

Description

Technical Field

[0001] This invention relates to a method for independent control of nozzle flow rate, and in particular to a method for independent control of nozzle flow rate based on vertical spray structure variable spray. Background Technology

[0002] Orchard plant protection is a crucial aspect of orchard management, accounting for approximately one-quarter of the total workload. It is also key to improving fruit quality and yield. Traditional orchard spraying relies on continuous, fixed-dosage spraying, resulting in low pesticide utilization and high labor intensity. Over-application can lead to excessive pesticide residues, affecting fruit quality, polluting the environment, and even poisoning operators. Variable-rate spraying (VRF) technology uses sensors to non-contactly detect plant canopy characteristics in real time, dynamically adjusting spray parameters based on an intelligent application model to achieve on-demand application. Compared to traditional continuous spraying, VRF can save over 25% of pesticide usage, making it a highly effective orchard application technology that achieves "reduced application, increased efficiency." Conventional VRF based on canopy volume (TRV) calculation models primarily calculate the total pesticide dosage for one side of the canopy and uniformly control the variable-rate spray solenoid valve on that side, without dividing the spray area for each nozzle or individually controlling the nozzle's output flow. To address this issue, a method for independent control of nozzle flow rate in variable spraying based on a vertical application structure was designed. Summary of the Invention

[0003] To improve pesticide utilization and achieve independent control of nozzle flow rate in orchard variable displacement sprayers, a method for independent nozzle flow rate control is proposed, taking into account the characteristics of vertical application structures and combining the scanning characteristics of lidar. This method can segment the volume of the fruit tree canopy in real time, independently calculate the volume of the corresponding area for each nozzle, the output flow rate of each nozzle, and the duty cycle of the corresponding variable displacement spray solenoid valve. Combined with time-delay control, it can achieve variable displacement spraying operations with independent nozzle flow rate control that matches real-time changes in plant canopy characteristics.

[0004] The technical solution of the present invention is as follows:

[0005] The method for independent nozzle flow control of a vertical spray structure variable sprayer based on two-dimensional lidar includes the following steps:

[0006] Step 1: Establish communication between modules and set relevant parameters for variable spraying. A 2D LiDAR, based on the Robot Operating System (ROS) and installed vertically, acquires real-time plant characteristic information on both sides of the orchard rows. The BeiDou Navigation Satellite System is used to acquire the real-time travel speed of the vertical spraying structure. Step 2: Convert the polar coordinates of the 2D LiDAR to Cartesian coordinates. Based on the structure of the vertical spraying support, the spatial distribution of the nozzles, the spray angle of the nozzles, and the relative position between the LiDAR and the nozzles, the spraying area is divided. Step 3: Set a lateral scanning threshold. Based on the laser point cloud data within the threshold range, calculate the plant canopy volume of each spraying area, the required output flow rate of each nozzle, and the duty cycle of the corresponding variable spray solenoid valve. Step 4: Combine the LiDAR scanning cycle, hardware response time, algorithm processing time, and the distance between the 2D LiDAR and the spraying mechanism for delay control. The main controller receives the variable spraying command sent by the microcomputer and controls the opening and closing of the variable spray solenoid valve corresponding to each nozzle, realizing target-oriented variable spraying operation with independent nozzle flow control.

[0007] Furthermore, step 1 specifically includes:

[0008] Establish communication between the various modules of the orchard variable-rate sprayer to ensure normal communication between the microcomputer, 2D lidar, Beidou navigation system, variable-rate spray controller, and variable-rate spray actuator. Set the relevant parameters for variable-rate spraying: the distance *l* between the 2D lidar and the center positions of the fruit trees on both sides, the required spray volume per unit volume of the plant canopy *▽*, and the calibration constants *a* and *b* for the duty cycle fitting formula. Use the robot operating system ROS to acquire real-time plant canopy feature information detected by the 2D lidar and the orchard variable-rate sprayer's travel speed *v* collected by the Beidou navigation system.

[0009] Furthermore, step 2 specifically includes:

[0010] A two-dimensional lidar is installed vertically at the front end of the orchard variable sprayer. The vertical spraying structure consists of 8 ESCs, 8 fans, 8 variable spray solenoid valves, 8 nozzle assemblies, a DC motor driver, a water pump, and a pressure sensor. The 8 nozzles are symmetrically mounted on both sides of the spray bracket, with the nozzles on both sides installed at the same height. The nozzles are numbered 1 to 8, starting from the lowest nozzle on the right side and arranged counter-clockwise. The height distance between adjacent nozzles on the same side is ΔH; D is the distance between the center of the two-dimensional lidar and the ground; H is the distance between the lowest nozzle in the orchard sprayer and the ground; θ spray ΔW represents the theoretical spray angle of the nozzle; ΔW is the horizontal distance between the center of the Hokuyo lidar and the nozzle. Coordinate transformation is performed on the two-dimensional lidar:

[0011]

[0012] Among them, (ρi ,θ i (y) represents the polar coordinates of the two-dimensional lidar point cloud data; i ,z i ) represents the rectangular coordinates of the two-dimensional lidar point cloud data.

[0013] The spray patterns of adjacent nozzles on the same side will converge at a distance d from the nozzle. This is determined by the installation positions of the two-dimensional lidar and the nozzles, and the horizontal distance y between the two-dimensional lidar and the surface of the fruit tree canopy. i The height thresholds Z1, Z2, and Z3 for dividing the spraying area can be calculated. The intersection distance d of the theoretical spray width output from the nozzle:

[0014] d=ΔH×tanβ

[0015] Where β is the theoretical spray angle θ of the nozzle. spray Half of, that is:

[0016]

[0017] Combining the two equations above, the intersection distance d of the theoretical spray width of the nozzle can be obtained as follows:

[0018]

[0019] Based on the horizontal distance ΔW and d between the two-dimensional lidar and the nozzle, the lateral threshold Δy between the two-dimensional lidar and the plant canopy surface can be calculated:

[0020] Δy=ΔW+d

[0021] When the distance y from the outer layer of the fruit tree canopy i When the value is greater than Δy, the height threshold for dividing the spraying area can be calculated:

[0022] z1=-D+H

[0023]

[0024] z3=-D+H+3ΔH

[0025] After coordinate transformation, the two-dimensional lidar detection threshold range [y] is set based on the row spacing of the fruit trees. 2D_min ,y 2D_max ] Filter out laser data points outside the threshold range. Based on the horizontal distance y between the two-dimensional lidar and the fruit tree canopy surface within the threshold range. i The positive and negative values ​​are used to divide the spraying area into left and right zones. When y i When |y>0, it represents the data for the right spraying area; otherwise, it represents the data for the left spraying area. i When |>Δy, the height division thresholds Z1, Z2 and Z3 of the spraying area are calculated according to the formula.

[0026] Furthermore, step 3 specifically includes:

[0027] Step 2 yields the lateral intersection distance d of the theoretical spray width output from the nozzle of the vertical application structure, the surface threshold Δy of the 2D lidar, and the height division thresholds Z1, Z2, and Z3 of the spraying area. Using these height division thresholds, the detection area of ​​the 2D lidar is divided into four areas on each side, for a total of eight spraying areas. The area S of each spraying area is calculated separately. j The volume V of a spray unit is calculated using 4 frames of lidar data. j The output flow rate Q of the nozzles corresponding to each spraying area j Duty cycle (DUC) of the spray solenoid valve corresponding to the nozzle j .

[0028]

[0029] Among them, S j The area of ​​the j-th sprayed region is expressed in meters. 2 ; l represents the distance from the origin of the 2D lidar coordinate system to the center of the fruit tree, in meters; (y i ,z i ) represents the coordinates of the i-th laser point of the two-dimensional lidar; (y i+1 ,z i+1 ) represents the coordinates of the (i+1)th laser point of the two-dimensional lidar; the area of ​​each spraying area in a single frame can be calculated separately according to the above formula for calculating the area of ​​the spraying area.

[0030] V j =4×Δt×v×S j_max

[0031] Among them, V j The volume corresponding to the j-th spraying area within the spray unit, in meters. 3 S j_max This represents the maximum area of ​​the j-th sprayed region in 4 frames of data, in meters. 2 Δt is the scanning cycle of the two-dimensional lidar, in seconds; v is the forward speed of the orchard variable speed sprayer, in m / s; the volume of each spraying area within the spray unit can be calculated using the volume calculation formula above.

[0032] The output flow rate Q of the nozzles corresponding to each spraying area within the spray unit j :

[0033]

[0034] Among them, Q j∠ is the required output flow rate of the nozzle corresponding to the j-th spraying area within the spray unit, in L / s; ▽ is the spray volume required per unit volume of the plant canopy, in L / m². 3 The required spray volume per unit volume of plant canopy ranges from 0.05L to 0.13L per cubic meter. For sparsely foliaged or pruned plant canopies, a spray volume of 0.05L / m³ is typically used. 3 For plants with undried branches and leaves and dense canopies, a canopy density of 0.13 L / m² is typically selected. 3 .

[0035] Duty cycle (DUC) of the variable spray solenoid valve corresponding to the nozzle in each spraying area within the spray unit. j :

[0036]

[0037] In the formula, DUC j Let be the duty cycle of the variable spray solenoid valve corresponding to the nozzle in the j-th spraying area within the spray unit, expressed as a percentage; a and b are the calibration constants of the fitting formula between the output flow rate and the solenoid valve duty cycle. The duty cycle of the variable spray solenoid valve corresponding to each nozzle can be obtained from the above formula.

[0038] Furthermore, step 4 specifically includes:

[0039] Based on step 3, the area S of each sprayed zone in the plant canopy can be calculated online. j The volume V of each spraying area within the spray unit j The output flow rate Q of the nozzles corresponding to each spraying area j Duty cycle (DUC) of the spray solenoid valve corresponding to the nozzle j To achieve independent nozzle flow control for target-variable spraying operations, the main factors contributing to spray delay are analyzed, and delay control for target-variable spraying is implemented. The main factors causing spray delay are: the two-dimensional lidar scanning period Δt, and the hardware response time T. hardware_delay Algorithm processing time T soft_delay The distance L between the two-dimensional lidar and the spray mechanism c Combining the above delay factors and the forward speed v of the orchard variable sprayer, the delay T of the variable spraying software can be calculated. delay :

[0040]

[0041] Among them, T delay The time from generating the duty cycle data of the variable spray solenoid valve to sending the data to the main controller is measured in seconds (T). hardware_delayThe time it takes for the liquid medicine to travel from the tank to the nozzle, including the liquid medicine transportation time and the time required for the variable spray solenoid valve to process and respond to the control signal, is measured in seconds (s). soft_delay The time taken by the microcomputer to process the algorithm from receiving lidar data to generating variable spray solenoid valve duty cycle data is expressed in seconds. Step 3 calculates the duty cycle data of the variable spray solenoid valve corresponding to each nozzle in the spray unit online, and then... delay After a delay control, data is sent to the main controller. Each variable spray solenoid valve receives the duty cycle array and independently controls the opening and closing of the solenoid valve, realizing target variable spraying operation with independent control of nozzle flow.

[0042] The present invention provides a variable spray control system based on a vertical spraying structure, comprising a power supply system 1, a walking control system 2, a variable spraying operation mechanism 3, a microcomputer 4, a two-dimensional lidar 5, and a Beidou navigation system 6;

[0043] The power system 1 consists of a 24V 30AH lithium battery 15 and a 24V 16AH lithium battery 16, which power the airborne microcomputer 4, the two-dimensional lidar 5, the Beidou navigation system 6, the variable spray controller 7, the chassis controller 20, and the variable spray operation mechanism 3. The stable DC / DC power modules 17 and 18 convert the 24V 30AH and 24V 16AH lithium batteries to 12V and 19V voltages, respectively. The 12V power supply mainly powers the two-dimensional lidar 5, the Beidou navigation system 6, the variable spray controller 7, the chassis controller 20, and the variable spray operation mechanism 3, while the 19V power supply powers the airborne microcomputer 4.

[0044] The walking control system 2 is mainly composed of a remote controller 19, a chassis controller 20, a motor drive 21, a travel motor 22, and a steering motor 23 connected in series. The operator can directly control the movement of the orchard variable sprayer through the remote controller 19. The remote controller 19 uses the SBUS (Serial Bus) protocol to communicate with the chassis controller 20 through a 2.4G module. It controls the travel motor 22 and the steering motor 23 through remote control commands to realize the walking control of the orchard sprayer, and realize the forward, reverse, turning, and emergency braking functions of the orchard variable sprayer. The microcomputer 4 communicates with the chassis controller 20 through the CAN port.

[0045] The variable spraying mechanism 3 of the vertical spraying structure mainly consists of a DC motor driver 8, a water pump 9, a pressure sensor 10, a variable spray solenoid valve 11, a nozzle 12, an electronic speed controller 13, a fan 14, and a distributor 26; one output of the variable spray controller 7 is connected to the electronic speed controller 13 and the fan 14 in sequence; the second output of the variable spray controller 7 is connected to the DC motor driver 8, the water pump 9, and the pressure sensor 10 in sequence; the third output of the variable spray controller 7 is connected to the variable spray solenoid valve 11 and the nozzle 12 in sequence.

[0046] In the variable spraying mechanism 3 of the vertical spraying structure, before variable spraying operation, the tank switch 25 is opened in advance, and the liquid pesticide flows out from the tank 24 through the tank switch 25. When the system is working, the microcomputer 4 acquires data from the two-dimensional lidar 5 and the Beidou navigation system 6 in real time through the robot operating system ROS. Based on the vertical structure variable spraying model, it calculates the required output flow rate of each nozzle 12, obtains the duty cycle of the corresponding variable spraying solenoid valve 11 of each nozzle through the fitting formula, and sends the generated duty cycle array to the variable spraying controller 7 through the serial port. The variable spraying controller 7 receives the data. After receiving the variable spray command from the microcomputer 4, the water pump 9 is driven to pressurize the liquid and send it to the distributor 26. The pressure sensor 10 mounted on the distributor 26 detects the water pressure inside the distributor 26 in real time and feeds it back to the variable spray controller 7 to achieve constant pressure operation. The variable spray controller 7 controls the opening and closing of the variable spray solenoid valve 11 according to the duty cycle data, and at the same time adjusts the speed of the fan 14 through the electronic speed controller 13. The liquid flows out from each variable spray solenoid valve 11, is atomized into fine droplets by the nozzle 12, and then blown by the fan 14 toward the target, realizing variable spray operation with independent control of nozzle flow.

[0047] Furthermore, the 2D lidar 5 uses the Hokuyo UTM-300TX model. This lidar is based on the time-of-flight ranging principle, achieving non-contact detection of plants based on the flight speed and time of the laser beam. The laser point cloud data is returned in polar coordinates. The Hokuyo 2D lidar 5 has an IP64 protection rating, making it suitable for orchard environments. Its 25ms scan cycle, 0.1–30m detection range, 0.25° angular resolution, and 270° effective scanning angle range allow it to generate 1080 effective laser data points per frame, enabling the detection of plant canopy features. The orchard variable-speed sprayer has an extended support rod on the vertical spraying structure, and the 2D lidar is rigidly connected to this support rod. The 2D lidar is mounted vertically on the support rod, with its center 1.3m above the ground and a distance L from the spraying mechanism. c The speed is 0.6m; the Beidou Navigation System 6 can obtain the speed v of the orchard variable sprayer in real time.

[0048] The beneficial effects of this invention are as follows:

[0049] 1. An independent nozzle flow control method for orchard variable sprayers with vertical spraying structures is invented. It uses lidar to detect plant characteristics in a non-contact manner and can calculate the volume of the plant canopy in real time. It is applicable to various operating environments such as orchards and nurseries, and improves the versatility and accuracy of orchard variable sprayer detection.

[0050] 2. The canopy volume and output flow rate of each nozzle in the corresponding area were independently calculated by dividing the spraying area. The spraying area was divided by combining the spraying angle of the nozzle, the characteristics of the vertical application structure, the scanning characteristics of the lidar and the spatial distribution of the nozzles, which provided a theoretical basis for independent control of the nozzle flow rate.

[0051] 3. The variable spray control system is equipped with an independent air delivery system, with one fan for each nozzle; a two-dimensional lidar acquires plant canopy features to generate the required spray volume for the nozzle and the duty cycle of the variable spray solenoid valve, and dynamically adjusts the fan speed according to the generated duty cycle to achieve precise variable wind speed and flow rate spraying.

[0052] 4. Variable flow spraying is carried out by using independent flow control of the nozzles. The flow output of the corresponding nozzles can be independently controlled according to the characteristics of the plant canopy, which reduces the amount of pesticide used and improves the accuracy of the operation. Attached Figure Description

[0053] Figure 1 This is a hardware structure block diagram of a variable spray control system based on a vertical application structure.

[0054] Figure 2 This is a schematic diagram of a variable spray system based on a vertical application structure;

[0055] Figure 3 This is a schematic diagram of variable spraying area division based on vertical application structure;

[0056] Figure 4 This is a schematic diagram of longitudinal segmentation of plant canopy scanning based on lidar;

[0057] Figure 5 This is a schematic diagram of horizontal segmentation of plant canopy scanning based on lidar;

[0058] Figure 6 This is a schematic diagram of the working mode of a variable sprayer for orchards based on a vertical spraying structure;

[0059] Figure 7 This is a flowchart of a nozzle flow independent control method based on a vertical spray structure variable spray system;

[0060] In the diagram, 1-Power supply system, 2-Walking control system, 3-Variable spray mechanism, 4-Microcomputer, 5-Two-dimensional lidar, 6-Beidou navigation system, 7-Variable spray controller, 8-DC motor driver, 9-Water pump, 10-Pressure sensor, 11-Variable spray solenoid valve, 12-Nozzle, 13-Electronic speed controller, 14-Fan, 15-24V 30AH lithium battery, 16-24V 16AH lithium battery, 17-12V voltage regulator module, 18-19V voltage regulator module, 19-Remote controller, 20-Chassis controller, 21-Motor drive, 22-Walking motor, 23-Steering motor, 24-Medicine tank, 25-Medicine tank switch, 26-Diverter. Detailed Implementation

[0061] The present invention will be further described below with reference to the embodiments and accompanying drawings, but the embodiments of the present invention are not limited thereto.

[0062] like Figure 1 As shown, the hardware structure of the variable spray control system based on a vertical application structure used in this example includes a power supply system 1, a walking control system 2, and a variable spray operating mechanism 3. The power supply system 1 includes a 24V 30AH lithium battery 15, a 24V 16AH lithium battery 16, two 12V regulated power supply modules 17 and 19V regulated power supply module 18. The 24V 30AH lithium battery 15 supplies power to the microcomputer 4, two-dimensional lidar 5, Beidou navigation system 6, chassis controller 20, and motor drive 21 through the 12V regulated power supply module 17 and the 19V regulated power supply module 18, respectively. The 16AH lithium battery 16 supplies power to the variable spray mechanism 3 through the 12V regulated power supply module 17; the walking control system 2 includes a remote controller 19, a chassis controller 20, a motor drive 21, a travel motor 22, and a steering motor 23. The remote controller 19 sends walking control commands to the chassis controller 20 through the SBUS protocol. After receiving the commands, the chassis controller 20 sends the commands to the motor drive 21 through the CAN port. The motor drive 21 controls the travel motor 22 and the steering motor 23 through PWM to realize the walking control of the chassis; the variable spray mechanism 3 includes a DC motor driver 8, a water pump 9, a pressure sensor 10, eight variable spray solenoid valves 11, eight nozzles 12, eight electronic speed controllers 13, and eight fans 14.

[0063] In this example, the 2D lidar 5 is the Hokuyo UTM-300TX model. This lidar has an effective working angle range of 270°, an angular resolution of 0.25°, generates 1080 effective distance data points per cycle, and has a scan cycle of 25ms. The 2D lidar 5 is installed vertically at the front end of the orchard variable displacement sprayer, 1.3m above the ground and 0.6m from the nozzle. The Beidou navigation system 6 can acquire the forward speed v of the orchard variable displacement sprayer in real time. During system operation, the microcomputer 4, through the Robot Operating System (ROS), collects and processes data from both the Hokuyo 2D lidar 5 and the Beidou navigation system 6. Figure 3 Divide the spraying area and, according to Figure 4 and Figure 5 The canopy volume of each spraying area, the output flow rate of the corresponding nozzle 12, and the duty cycle of the variable spray solenoid valve 11 are calculated. The generated spray command is sent to the variable spray controller 7 via serial port. After a delay, the variable spray function with independent nozzle flow control is realized.

[0064] like Figure 2 The diagram shows a variable spray system based on a vertical application structure. The system includes an information acquisition unit, an information processing unit, and a variable spray operation mechanism 3. Before variable spraying, the medicine tank switch 25 is opened, and the liquid medicine flows from the medicine tank 24 through the switch 25. The information acquisition unit includes a two-dimensional lidar 5 and a Beidou navigation system 6; the information processing unit includes a microcomputer 4 and a variable spray controller 7. After acquiring data from the two-dimensional lidar, the microcomputer 4 calculates the required output flow rate for each nozzle 12 based on the vertical structure variable application model, obtains the duty cycle of the corresponding variable spray solenoid valve 11 for each nozzle 12 through a fitting formula, and sends the generated duty cycle array to the variable spray controller 7 via a serial port. After receiving the variable spray command from the microcomputer 4, the variable spray controller 7 drives the water pump 9 to pressurize the liquid medicine and send it to the distributor 26. The pressure sensor 10 mounted on the distributor 26 detects the internal water pressure in real time and feeds it back to the variable spray controller 7 to achieve constant pressure operation. Based on the travel speed v of the orchard variable sprayer collected by the Beidou Navigation System 6, the delay is calculated. After the delay is controlled, the variable spray solenoid valve 11 controls the opening and closing of the solenoid valve according to the received duty cycle array. The liquid flows out from each variable spray solenoid valve 11, is atomized into fine droplets through the nozzle 12, and blown towards the target by the fan 14 to realize variable spraying operation.

[0065] like Figure 3The diagram shows the variable spraying area division based on a vertical spraying structure. Eight nozzles 12 of the vertical spraying structure are symmetrically installed on both sides of the spraying bracket. The nozzles 12 on both sides are installed at the same height. The nozzles are numbered 1 to 8, starting from the lowest nozzle on the right side and arranged counter-clockwise. The height distance between adjacent nozzles 12 on the same side is ΔH; D is the distance between the center of the Hokuyo 2D lidar 5 and the ground, which is 1.3m; H is the distance between the lowest nozzle 12 in the orchard sprayer and the ground, which is 0.65m; θ spray The theoretical spray angle of nozzle 12 is 110°; ΔW is the horizontal distance between the center of Hokuyo lidar 5 and nozzle 12, which is 0.4m. Coordinate transformation is performed on the two-dimensional lidar 5:

[0066]

[0067] In the formula, (ρ i ,θ i (y) represents the polar coordinates of the 5-point cloud data from a 2D lidar system; i ,z i () represents the rectangular coordinates of the 5-point cloud data from the 2D lidar.

[0068] The spray patterns of adjacent nozzles 12 on the same side will converge at a distance d from the nozzle. This is based on the installation positions of the two-dimensional lidar 5 and the nozzles 12, and the horizontal distance y between the two-dimensional lidar 5 and the surface of the fruit tree canopy. i The height thresholds Z1, Z2, and Z3 for dividing the spraying area can be calculated. The intersection distance d of the theoretical spray width output of nozzle 12:

[0069] d=ΔH×tanβ

[0070] In the formula, β is the theoretical spray angle θ of nozzle 12. spray Half of, that is:

[0071]

[0072] Combining the two equations above, the intersection distance d of the theoretical spray width of nozzle 12 can be obtained as follows:

[0073]

[0074] Based on the horizontal distance ΔW and d between the two-dimensional lidar 5 and the nozzle 12, the lateral threshold Δy between the two-dimensional lidar 5 and the plant canopy surface can be calculated:

[0075] Δy=ΔW+d

[0076] From the above formula and parameters, Δy is 0.65m. When the distance y from the outer layer of the fruit tree canopy... i When the height is greater than 0.65m, the height threshold for dividing the spraying area can be calculated:

[0077] z1=-D+H

[0078]

[0079] z3=-D+H+3ΔH

[0080] After coordinate transformation, the 2D lidar 5 detects a set threshold range [y] based on the row spacing of the fruit trees. 2D_min ,y 2D_max ], Filter out laser data points outside the threshold range. When y i When the value is greater than 0.65m, Z1, Z2 and Z3 are -0.65m, -0.11m and 0.43m respectively.

[0081] like Figure 4 The diagram shows a longitudinal segmentation of the plant canopy based on lidar scanning. Let l be the distance from the center of the fruit tree to the center of the Hokuyo 2D lidar 5, and p be the adjacent data points of the lidar. i and p i+1 The height difference is D h The average thickness of the canopy surface to the trunk of adjacent laser data points is D. w Therefore, the area of ​​half of the plant canopy in a single frame of laser scan data can be longitudinally divided into discrete rectangles. The formula for calculating the area of ​​a single frame in each spraying area is as follows:

[0082] D h =|z i+1 -z i |

[0083]

[0084] S j =ΣD h ×D w

[0085] Among them, S j The total area of ​​each spraying zone is represented by a single frame scan, where j = 1, 2, ..., 8; l is the horizontal distance between the Hokuyo 2D lidar 5 and the center of the fruit tree; (y i ,z i Let be the rectangular coordinate value corresponding to the i-th laser data point. By dividing the spraying area, based on the height value z of the laser point... i Determine the spraying area to which the current laser point belongs, calculate the area of ​​that point and its neighboring points, and sum the areas within the same spraying area to obtain the total area S of the current data frame for all spraying areas. j .

[0086] like Figure 5The diagram shows a schematic of horizontal segmentation of the plant canopy based on lidar scanning. The Hokuyo 2D lidar 5 is set to continuous operation mode, continuously detecting targets at a fixed scanning frequency. After accumulating a certain number of frames, the forward speed v of the orchard sprayer, the scanning time Δt of a single lidar frame, and the accumulated number of frames n are combined. scan The lateral segmentation distance W during this time period can be calculated. The selected Hokuyo 2D lidar 5 has a scanning period Δt of 25ms, and the response time of the variable spray solenoid valve 11 is 100ms. To match the response time of the solenoid valve, every 4 frames of laser data are selected as a spray unit. The maximum area of ​​each sprayed area in the 4 frames of data is taken as the area of ​​the corresponding area within the spray unit, and then the volume corresponding to each sprayed area in the spray unit is calculated.

[0087] W = 4 × Δt × v

[0088] V j =W×S j_max

[0089] In the formula, V j The volume of the plant canopy in the j-th spraying area is expressed in m³. 3 W represents the width of the spray unit, in meters (m); v represents the forward speed of the orchard sprayer, in meters per second (m / s); Δt represents the scanning time of a single frame of data from the lidar, in seconds (s); S j_max The maximum area of ​​the j-th spraying region in a spray unit across four consecutive frames of data, expressed in meters (m). 2 .

[0090] like Figure 6 The diagram shows the working principle of a variable displacement sprayer for orchards based on a vertical spraying structure. During operation, the Hokuyo 2D lidar 5 installed at the front continuously scans the plants on both sides of the front row. The microcomputer 4 calculates the required output flow rate Q of each nozzle 12 in each spraying unit based on the received Hokuyo lidar point cloud data and the established vertical structure variable displacement spraying model. j The required duty cycle (DUC) for the corresponding variable spray solenoid valve 11 is calculated based on the fitting equation of duty cycle and output flow rate. j This is stored in a delayed spray dynamic array. It combines the two-dimensional lidar scanning period Δt and the hardware response time T. hardware_delay Algorithm processing time T soft_delay The distance L between the two-dimensional lidar and the spray mechanism c Given the forward speed v of the orchard variable sprayer, the software delay T of the variable spraying system can be calculated. delay :

[0091]

[0092] After a time delay, when the spraying mechanism at the tail of the orchard sprayer reaches the corresponding spraying area, it sends the stored data to the variable spray controller 7, which controls the opening and closing of the variable spray solenoid valve 11 to achieve variable spraying operation with independent control of nozzle flow.

[0093] like Figure 7 The flowchart shown illustrates the independent control method for nozzle flow rate in a vertical spraying structure variable spraying system. Communication between the various modules of the orchard variable sprayer is established to ensure normal communication between the variable spraying operation mechanism 3, microcomputer 4, 2D LiDAR 5, Beidou navigation system 6, and variable spray controller 7. Next, the distance *l* between the center of the fruit tree and the center of the Hokuyo 2D LiDAR 5, the required spray volume per unit volume *▽*, and the calibration constants *a* and *b* of the duty cycle fitting formula are set. After setting the relevant spraying parameters, the travel speed *v* of the orchard variable sprayer is obtained through the Beidou navigation system 6. During the sprayer's movement, the Hokuyo 2D LiDAR 5 detects the canopy characteristics of the plants on both sides in real time according to the scanning cycle, divides the spraying area, reconstructs and segments the plant canopy, calculates the required output flow rate of each nozzle 12 in each spraying unit, obtains the required duty cycle of the variable spray solenoid valve 11 corresponding to each nozzle 12 based on the duty cycle fitting formula, and stores this data in the delayed spray array. Combining the forward speed v of the orchard sprayer, the relative position L of the Hokuyo 2D lidar 5 and the spraying mechanism installation... c The spray delay is calculated based on the scanning cycle of the Hokuyo lidar 5, the algorithm processing time, and the hardware response time during spraying. After delay control, the microcomputer 4 sends duty cycle data to the variable spray controller 7. The controller independently drives the corresponding variable spray solenoid valve 11 according to the duty cycle data to achieve variable spray operation with independent flow control of each nozzle.

[0094] Matters not covered in this invention are common knowledge.

[0095] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or instance of the invention. In this document, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0096] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these examples without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A variable spraying method based on independent flow control of nozzles using a vertical application structure, characterized in that: The method includes the following steps: Step 1: Establish communication between modules, set variable spraying parameters, and acquire data from the 2D LiDAR and BeiDou navigation system in real time based on the robot operating system ROS: The 2D LiDAR is installed at the front of the orchard variable sprayer in a vertical scanning manner, and detects the canopy features of the plants on both sides in real time according to the scanning cycle; the BeiDou navigation system acquires the travel speed v of the orchard sprayer in real time. Step 2: Convert the polar coordinates of the 2D LiDAR to rectangular coordinates, and divide the spraying area based on the structure of the vertical spraying support, the spatial distribution of the nozzles, the spraying angle of the nozzles, and the relative position between the LiDAR and the nozzles. Step 3: Set the horizontal scanning threshold, and calculate the canopy volume of the plants in each spraying area, the required output flow rate of each nozzle, and the duty cycle of the corresponding variable spray solenoid valve based on the laser point cloud data within the threshold range. Step 4: Combine the LiDAR scanning cycle, hardware response time, algorithm processing time, and the distance between the 2D LiDAR and the spraying mechanism for delay control. The main controller receives the variable spraying command sent by the microcomputer and controls the opening and closing of the variable spray solenoid valve corresponding to each nozzle to realize the target variable spraying operation with independent control of nozzle flow rate. Step 1 specifically includes: Establish communication between the various modules of the orchard variable sprayer to ensure normal communication between the microcomputer, 2D lidar, Beidou navigation system, controller, and variable spraying actuator; set relevant parameters for variable spraying: the distance l between the 2D lidar and the center of the fruit trees on both sides, the required spray volume per unit volume of the plant canopy ▽, and the calibration constants a and b of the duty cycle fitting formula; acquire plant canopy feature information detected by 2D lidar and the travel speed v of the orchard variable sprayer collected by the Beidou navigation system in real time through the robot operating system ROS; Step 2 specifically includes: A two-dimensional lidar is installed vertically at the front end of the orchard variable sprayer. The vertical spraying structure consists of 8 electronic speed controllers, 8 fans, 8 variable spray solenoid valves, 8 nozzle assemblies, a DC motor driver, a water pump, and a pressure sensor. The 8 nozzle assemblies are symmetrically installed on both sides of the sprayer support, with the nozzles on both sides at the same installation height. The nozzles are numbered 1 to 8, starting from the lowest nozzle on the right side and arranged counter-clockwise. The height distance between adjacent nozzles on the same side is ΔH; D is the distance between the center of the two-dimensional lidar and the ground; H is the distance between the lowest nozzle in the orchard sprayer and the ground; θ spray ΔW represents the theoretical spray angle of the nozzle; ΔW is the horizontal distance between the center of the Hokuyo lidar and the nozzle. Coordinate transformation is performed on the two-dimensional lidar: Among them, (ρ i ,θ i (y) represents the polar coordinates of the two-dimensional lidar point cloud data; i ,z i () represents the rectangular coordinates of the two-dimensional lidar point cloud data; The spray patterns of adjacent nozzles on the same side will converge at a distance d from the nozzle. This is determined by the installation positions of the two-dimensional lidar and the nozzles, and the horizontal distance y between the two-dimensional lidar and the surface of the fruit tree canopy. i The height thresholds Z1, Z2, and Z3 for the spraying area can be obtained; the intersection distance d of the theoretical spray width output of the nozzle: d=ΔH×tanβ Where β is the theoretical spray angle θ of the nozzle. spray Half of, that is: Combining the above two equations, the intersection distance d of the theoretical spray width of the nozzle can be obtained as: Based on the horizontal distance ΔW and d between the two-dimensional lidar and the nozzle, the lateral threshold Δy between the two-dimensional lidar and the plant canopy surface can be calculated: Δy=ΔW+d When the distance y from the outer layer of the fruit tree canopy i When the value is greater than Δy, the height threshold for dividing the spraying area can be calculated: z1=-D+H z3=-D+H+3ΔH After coordinate transformation, the two-dimensional lidar detection threshold range [y] is set based on the row spacing of the fruit trees. 2D_min ,y 2D_max ], filter out laser data points outside the threshold range, based on the horizontal distance y between the two-dimensional lidar and the fruit tree canopy surface within the threshold range. i The positive and negative values ​​are used to divide the spraying area into left and right zones. When y i When |y>0, it represents the data for the right spraying area; otherwise, it represents the data for the left spraying area. i When |>Δy, the height division thresholds Z1, Z2 and Z3 of the spraying area are calculated according to the formula.

2. The variable spraying method for independent flow control of nozzles based on a vertical application structure according to claim 1, characterized in that: Step 3 specifically includes: Step 2 yields the lateral intersection distance d of the theoretical spray width output from the nozzle of the vertical application structure, the surface threshold Δy of the two-dimensional lidar, and the height division thresholds Z1, Z2, and Z3 of the spraying area. Using these height division thresholds, the detection area of ​​the two-dimensional lidar is divided into four areas on each side, for a total of eight spraying areas. The area S of each spraying area is calculated individually. j The volume V of a spray unit is calculated using 4 frames of lidar data. j The output flow rate Q of the nozzles corresponding to each spraying area j Duty cycle (DUC) of the spray solenoid valve corresponding to the nozzle j .

3. The variable spraying method for independent flow control of nozzles based on a vertical application structure according to claim 2, characterized in that: The area S of each spraying zone j : Among them, S j The area of ​​the j-th sprayed region is expressed in meters. 2 ; l represents the distance from the origin of the two-dimensional lidar coordinate system to the center of the fruit tree, in meters; (y i ,z i ) represents the coordinates of the i-th laser point of the two-dimensional lidar; (y i+1 ,z i+1 (i) represents the coordinates of the (i+1)th laser point of the two-dimensional lidar; the area of ​​each spraying area in a single frame can be calculated separately according to the above formula for calculating the area of ​​the spraying area. The volume V of the spraying area within the spray unit j : In j =4×Δt×v×S j_max Among them, V j The volume corresponding to the j-th spraying area within the spray unit, in meters. 3 S j_max This represents the maximum area of ​​the j-th sprayed region in 4 frames of data, in meters. 2 Δt is the scanning cycle of the two-dimensional lidar, in seconds; v is the forward speed of the orchard variable speed sprayer, in m / s; the volume of each spraying area within the spray unit can be calculated using the volume calculation formula above. The output flow rate Q of the nozzles corresponding to each spraying area within the spray unit j : Among them, Q j ∠ is the required output flow rate of the nozzle corresponding to the j-th spraying area within the spray unit, in L / s; ▽ is the spray volume required per unit volume of the plant canopy, in L / m². 3 The required spray volume per unit volume of the plant canopy ranges from 0.05L to 0.13L per cubic meter; for sparsely pruned plant canopies, a spray volume of 0.05L / m³ is recommended. 3 For plants with undried branches and leaves and dense canopies, use 0.13L / m². 3 ; Duty cycle (DUC) of the variable spray solenoid valve corresponding to the nozzle in each spraying area within the spray unit. j : Among them, DUC j , where is the duty cycle of the variable spray solenoid valve corresponding to the nozzle in the j-th spraying area within the spray unit, in %; a and b are the calibration constants of the fitting formula between the output flow rate and the solenoid valve duty cycle. The duty cycle of the variable spray solenoid valve corresponding to each nozzle can be obtained from the above formula.

4. The variable spraying method for independent flow control of nozzles based on a vertical application structure according to claim 1, characterized in that: Step 4 specifically includes: Based on step 3, the area S of each sprayed zone in the plant canopy can be calculated online. j The volume V of each spraying area within the spray unit j The output flow rate Q of the nozzles corresponding to each spraying area j Duty cycle (DUC) of the spray solenoid valve corresponding to the nozzle j To achieve independent control of nozzle flow rate in target-variable spraying operations, the main factors causing spray delay were analyzed, and delay control of target-variable spraying was implemented. The main factors causing spray delay are: the two-dimensional lidar scanning period Δt, and the hardware response time T. hardware_delay Algorithm processing time T soft_delay The distance L between the two-dimensional lidar and the spray mechanism c Combining the above delay factors and the forward speed v of the orchard variable sprayer, the delay T of the variable spraying software can be calculated. delay : Among them, T delay The time from generating the duty cycle data of the variable spray solenoid valve to sending the data to the main controller is measured in seconds (T). hardware_delay T represents the time it takes for the liquid medicine to travel from the tank to the nozzle, including the time for the liquid medicine to be transported and the time required for the variable spray solenoid valve to process and respond to the control signal. The unit is seconds (s). soft_delay The time taken by the microcomputer to process the algorithm from receiving lidar data to generating variable spray solenoid valve duty cycle data is expressed in seconds. Step 3 calculates the duty cycle data of the variable spray solenoid valve corresponding to each nozzle in the spray unit online, and then... delay After a delay control, data is sent to the main controller. Each variable spray solenoid valve receives the duty cycle array and independently controls the opening and closing of the solenoid valve, realizing target variable spraying operation with independent control of nozzle flow.

5. A variable spray control system based on a vertical application structure, the system being used to implement a variable spray method for independent control of nozzle flow rate based on a vertical application structure as described in any one of claims 1-4, characterized in that: It includes a power supply system (1), a walking control system (2), a variable spraying mechanism (3), a microcomputer (4), a two-dimensional lidar (5), and a Beidou navigation system (6); The power supply system (1) consists of a 24V 30AH first lithium battery (15) and a 24V 16AH second lithium battery (16), which are used to supply power to the airborne microcomputer (4), two-dimensional lidar (5), Beidou navigation system (6), variable spray controller (7), chassis controller (20) and variable spray operation mechanism (3). The stable first DC / DC power module (17) and second DC / DC power module (18) convert the 24V 30AH and 24V 16AH lithium batteries into 12V and 19V voltages respectively. The 12V power supply supplies power to the two-dimensional lidar (5), Beidou navigation system (6), variable spray controller (7), chassis controller (20) and variable spray operation mechanism (3), and the 19V power supply supplies power to the airborne microcomputer (4). The walking control system (2) includes a remote controller (19), a chassis controller (20), a motor drive (21), a travel motor (22), and a steering motor (23). The operator can directly control the movement of the orchard variable sprayer through the remote controller (19). The remote controller (19) uses the Serial Bus protocol to communicate with the chassis controller (20) through a 2.4G module. The travel motor (22) and steering motor (23) are controlled by remote control commands to realize the walking control of the orchard sprayer and realize the forward, backward, turning, and emergency braking functions of the orchard variable sprayer. The microcomputer (4) communicates with the chassis controller (20) through the CAN port. The variable spraying mechanism (3) of the vertical spraying structure includes a DC motor driver (8), a water pump (9), a pressure sensor (10), a variable spray solenoid valve (11), a nozzle (12), an electronic controller (13), a fan (14), and a distributor (26); one output of the variable spray controller (7) is connected to the electronic controller (13) and the fan (14) in sequence; the second output of the variable spray controller (7) is connected to the DC motor driver (8), the water pump (9), and the pressure sensor (10) in sequence; the third output of the variable spray controller (7) is connected to the variable spray solenoid valve (11) and the nozzle (12) in sequence. In the variable spraying mechanism (3) of the vertical spraying structure, before the variable spraying operation, the medicine tank switch (25) is opened in advance, and the liquid medicine flows out from the medicine tank (24) through the medicine tank switch (25). When the system is working, the microcomputer (4) obtains data from the two-dimensional laser radar (5) and the Beidou navigation system (6) in real time through the robot operating system ROS. Based on the vertical structure variable spraying model, it calculates the required output flow of each nozzle (12), obtains the duty cycle of the variable spraying solenoid valve (11) corresponding to each nozzle through the fitting formula, and sends the generated duty cycle array to the variable spraying controller (7) through the serial port. The variable spraying controller (7) receives the microcomputer After the computer (4) issues a variable spray command, it drives the water pump (9) to pressurize the liquid and send it to the distributor (26). The pressure sensor (10) mounted on the distributor (26) detects the water pressure inside the distributor (26) in real time and feeds it back to the variable spray controller (7) to achieve constant pressure operation. The variable spray controller (7) controls the opening and closing of the variable spray solenoid valve (11) according to the duty cycle data. At the same time, it adjusts the speed of the fan (14) through the electric speed controller (13). The liquid flows out from each variable spray solenoid valve (11), is atomized into fine droplets by the nozzle (12), and then blown towards the target by the fan (14), so as to achieve variable spray operation with independent control of nozzle flow.

6. A variable spray control system based on a vertical application structure according to claim 5, characterized in that: The two-dimensional lidar (5) uses the Hokuyo UTM-300TX model. This lidar is based on the time-of-flight ranging principle. According to the flight speed and time of the laser beam, it realizes non-contact detection of plants. The laser point cloud data is returned in polar coordinate form. The Hokuyo two-dimensional lidar (5) has an IP64 protection level, which can adapt to the detection requirements of the orchard environment. The 25ms scanning cycle, 0.1~30m detection distance, 0.25° angular resolution and 270° effective scanning angle range enable the lidar to generate 1080 effective laser data points per frame, realizing the detection of plant canopy features. The orchard variable sprayer is extended on the spray support rod of the vertical spraying structure. The two-dimensional lidar is fixed to the extended support rod in a rigid connection manner. The two-dimensional lidar is installed on the extended support rod in a vertical scanning manner. Its center position is 1.3m above the ground and the distance from the spraying mechanism is L. c The speed of the orchard variable sprayer is 0.6m; the Beidou navigation system (6) can obtain the speed v of the orchard variable sprayer in real time.

Images