Ackerman-type intelligent precision variable air-assisted spraying robot structure for orchards, its path planning, and variable spraying method

Abstract

The present invention discloses an Ackerman-type intelligent and precise variable air-assisted spraying robot structure for orchards, as well as its path planning and variable spraying method. It includes an Ackerman-type mobile chassis 1, on which a robot bracket 2 is fixed. On the robot bracket 2, a battery 3, a central processor 4, a variable spraying system 5, a data acquisition module 6, a water pump 7, a hardware circuit drive control board 8 and a water tank 9 are fixed; the variable spraying system 5 is composed of a centrifugal nozzle 51, a duct 52 and a fan 53; the data acquisition module 6 is composed of a navigation module 61 and a 3D lidar 62. The present invention uses the central processor 4 to process the data obtained by the data acquisition module 6, controls the forward speed and forward direction of the Ackerman-type mobile chassis 1, and controls the spraying amount, spraying time, air volume and wind speed of the variable spraying system 5, and finally realizes the path planning of the intelligent and precise variable air-assisted spraying robot for orchards and precise variable flow and variable air volume spraying.

Description

Technical Field

[0001] The present invention relates to an orchard spraying robot, and more specifically to an Ackerman orchard intelligent precise variable air-assisted spraying robot structure, its path planning, and variable spraying method. Background Art

[0002] As a major agricultural country, China attaches great importance to the development of agriculture. In recent years, the fruit industry has continued to develop, having an important impact on the national economy. Orchard plant protection is an important link in orchard management, accounting for about one-fourth of the total workload of orchard management. At the same time, orchard plant protection is also a key link in improving the quality and yield of fruits. To achieve the control effect, the current orchard plant protection method in China generally mainly uses manual backpack sprayers or driving air-assisted sprayers to continuously spray at a fixed dose. The utilization rate of the liquid medicine is low and the labor intensity is high. At the same time, excessive application of pesticides easily leads to excessive pesticide residues, which in turn affects the quality of fruits, pollutes the environment, and even causes poisoning of operators. To address the above problems, a robot that can perform variable spraying operations while navigating autonomously is designed. Summary of the Invention

[0003] The purpose of the present invention is to design an Ackerman orchard intelligent precise variable air-assisted spraying robot, which is suitable for performing variable spraying operations while navigating autonomously, realizing the separation of humans and machines, avoiding direct contact between operators and liquid medicine, greatly reducing the labor intensity and improving the operation efficiency; at the same time, it can accurately detect and extract the target characteristics of plants, perform precise variable spraying, effectively save the usage amount of pesticides, improve the utilization rate of pesticides, and reduce environmental pollution. The Ackerman mobile chassis and the inclined swing spraying structure broaden the operation environment of the robot, and the intelligent precise variable air-assisted spraying system improves the accuracy and efficiency of spraying.

[0004] To achieve the above object, the technical solution of the present invention is as follows:

[0005] An Ackermann-type intelligent and precise variable air-assisted spraying robot structure for orchards, comprising an Ackermann-type mobile chassis (1), a robot bracket (2), a battery (3), a central processor (4), a variable spraying system (5), a data acquisition module (6), a water pump (7), a hardware circuit drive control board (8), and a water tank (9); the Ackermann-type mobile chassis (1) is fixed with the robot bracket (2), and the robot bracket (2) is fixed with the battery (3), the central processor (4), the variable spraying system (5), the data acquisition module (6), the water pump (7), the hardware circuit drive control board (8), and the water tank (9); the robot bracket (2) is fixed on the Ackermann-type mobile chassis (1) to provide a supporting role for the entire robot platform; the battery (3) is fixed on the robot bracket (2) to supply power to the central processor (4), the variable spraying system (5), the data acquisition module (6), the water pump (7), and the hardware circuit drive control board (8); the central processor (4) is used to process the data obtained by the data acquisition module (6) and send it to the hardware circuit drive control board (8); the variable spraying system (5) consists of a centrifugal nozzle (51), a duct (52), and a fan (53), the centrifugal nozzle (51) and the fan (53) are fixed on a carbon fiber tube (534) and are located inside the duct (52), and can rotate synchronously with the carbon fiber tube (534). By adjusting the rotation speed of the fan (53), the wind speed and air volume are adjusted in real time, thereby adjusting the penetration and spraying range of the liquid medicine; the data acquisition module (6) consists of a navigation module (61) and a 3D lidar (62); the liquid medicine is stored in the water tank (9), is connected to the centrifugal nozzle (51) through the water pump (7) in sequence, and sprays the liquid medicine precisely and evenly onto the plant canopy; after receiving the data sent by the central processor (4), the hardware circuit drive control board (8) performs further processing to control the forward speed and forward direction of the Ackermann-type mobile chassis (1), and control the spraying amount, spraying time, air volume, and wind speed of the variable spraying system (5), so as to realize the path planning of the intelligent and precise variable air-assisted spraying robot and the precise variable flow rate and variable air volume spraying.

[0006] Further, the Ackermann-type mobile chassis (1) provides a mounting platform for the intelligent and precise variable air-assisted spraying system to realize the position adjustment of the intelligent and precise variable air-assisted spraying system.

[0007] Further, the robot bracket (2) consists of upper and lower layers. The lower layer is fixed on the Ackerman mobile chassis (1) using screws and nuts, and is equipped with a battery (3), a central processor (4), a water pump (7), a hardware circuit drive control board (8), and a water tank (9). To achieve the waterproof function, the upper and lower layers are separated by a board. The upper layer is equipped with a variable spray system (5) and a data acquisition module (6). The robot bracket (2) is connected by aluminum profiles and aluminum plates through metal corner fittings, which provides support for the entire intelligent and precise variable air-assisted spray robot platform while reducing its own weight.

[0008] Further, the water tank (9) is fixed on the robot bracket (2); the water pump (7) is fixed to the lower layer of the robot bracket (2) using screws and nuts; the water pump (7) is used to control the flow rate of the centrifugal nozzle (51) in the duct (52); the centrifugal nozzle (51) and the fan (53) are fixed on the upper layer of the robot bracket (2) through a carbon fiber tube (534) passing through the duct (52); the centrifugal nozzle (51) and the fan (53) are located in the duct (52), and the wind speed and air volume are controlled by controlling the rotation speed of the fan (53).

[0009] Further, the battery (3) fixed to the lower layer of the robot bracket (2) is a lithium battery, which provides power for the central processor (4), the variable spray system (5), the data acquisition module (6), the water pump (7), and the hardware circuit drive control board (8).

[0010] Further, the data acquisition module (6) consists of a navigation module (61) and a 3D lidar (62); the navigation module (61) is fixed at the top of the robot bracket (2) and is used to obtain information such as the position and attitude of the robot in real time; the 3D lidar (62) is fixed at the front end of the robot bracket (2) and is used to detect obstacles around the robot and scan the volume of the plant canopy around the robot; the central processor (4) is used to process the data obtained by the data acquisition module (6) for path planning and target plant volume calculation.

[0011] Further, the hardware circuit drive control board (8) is fixed to the lower layer of the robot bracket (2) and is an STM32F4 control board, which is electrically connected through the Ackerman mobile chassis (1), the central processor (4), the centrifugal nozzle (51), the fan (53), and the water pump (7) to achieve path planning and precise variable flow rate and variable air volume spraying of the intelligent and precise variable air-assisted spray robot.

[0012] The path planning method for the structure of an Ackerman orchard intelligent and precise variable air-assisted spray robot of the present invention is as follows:

[0013] Step 1: According to the motion model of the Ackermann mobile chassis (1), the inner and outer tire angles of the front wheels of the spray robot chassis are: , ;

[0014] where δ o is the outer tire angle of the front wheel, δ i is the inner tire angle of the front wheel, L is the wheelbase, l ω is the track width, R is the turning radius.

[0015] Step 2: Assume that the spray robot is traveling in a low-speed and non-slip scenario. Taking the center point of the rear axle of the spray robot chassis as the rotation reference point, the average values of the left front wheel tire angle and the right front wheel tire angle of the spray robot are taken to obtain the simplified motion model of the Ackermann mobile chassis (1), and its average front wheel angle is: ;

[0016] where δ is the average front wheel angle.

[0017] Step 3: According to the pure tracking algorithm, the motion trajectory of the spray robot from the current position O t to the path tracking point P t can be regarded as an arc trajectory with R as the radius. According to the sine theorem, it can be obtained that: ;

[0018] After conversion, it is obtained that: ;

[0019] where l d is the look-ahead distance, α is the heading deviation, that is, the angle between the current vehicle body attitude and the target point P t .

[0020] Step 4: According to the curvature formula, the curvature can be expressed as: ;

[0021] where K is the curvature of the path.

[0022] Step 5: The average angle of the front wheels of the spray robot can be expressed as: ;

[0023] Combining the above formula, the final expression of the control variable δ of the pure tracking algorithm can be obtained as: ;

[0024] In the formula, the time domain is introduced, and α( t ) is t the included angle between the spray robot and the target point at P t moment. Define l l as the lateral error between the current posture of the spray robot and the target point P t . Then, we can get: ;

[0025] Meanwhile, the curvature can be expressed as: .

[0026] Step 6: Express the forward viewing distance l d as a linear function of the linear velocity of the spray robot, and we can get: ;

[0027] Combining the above formula, we can get: ;

[0028] In the formula, k is the proportional coefficient, v is the linear velocity of the spray robot, and e is the initial forward viewing distance.

[0029] Step 7: According to the angular velocity calculation method, we can get: ;

[0030] In the formula, ω is the angular velocity of the spray robot.

[0031] Step 8: By adjusting the proportional coefficient k to adjust the pure tracking algorithm, the angular velocity ω of the spray robot is inversely proportional to the forward viewing distance l d , that is, the larger the forward viewing distance, the smaller the angular velocity, and the smoother the tracking trajectory; the smaller the forward viewing distance, the larger the angular velocity, and the more oscillating the tracking trajectory.

[0032] Step 9: Use the navigation module (61) to obtain the position and attitude information of the spray robot. According to the planned path, use the pure tracking algorithm. By comparing the lateral deviation between the current pose of the spray robot and the planned navigation line, adjust the motion angle and displacement of the spray robot in real time.

[0033] The variable spraying method of an Ackerman-type orchard intelligent precise variable air-assisted spray robot structure of the present invention is as follows:

[0034] Step 1: Use a 3D lidar (62) to scan the plant canopy to obtain the original lidar point cloud data. Preprocess the original lidar point cloud data to filter out the point cloud data generated by distant obstacles and focus on the target to be detected.

[0035] Step 2: Perform clustering on the preprocessed lidar point cloud data to fit the plant trunk. The point cloud data returned by the 3D lidar can be used to calculate the volume of the target plant.

[0036] Step 3: The distance between the lidar and the spray robot chassis is a , define the origin of the coordinate system at the point directly below the lidar a , x The axis points directly in front of the spray robot, y The axis is parallel to the chassis wheel axis of the spray robot and points to the left, z The axis points vertically upward.

[0037] Step 4: For ease of calculation, convert the 3D lidar data points P t from polar coordinates ([[]] r t , ω t , α t ) to Cartesian coordinates ([[]] x t , y t , z t ), and its conversion formula can be expressed as: ;

[0038] In the formula, r t is the distance between the data point P t and the lidar; ω is the vertical scanning angle resolution of the lidar; α is the horizontal scanning angle resolution of the lidar.

[0039] Step 5: According to the fitted plant trunk, after data analysis, the point with the largest z axis coordinate value on the plant is denoted as A , and its coordinates are denoted as ([[]] X a , Y a , Z max ); the point with the smallest z axis coordinate value on the plant is denoted as B , and its coordinates are denoted as ([[]] X b , Yb , Z min ); The point on the plant with the largest x axis coordinate value is denoted as C , and its coordinates are denoted as ([[]] X max , Y c , Z c ); The point on the plant with the smallest x axis coordinate value is denoted as D , and its coordinates are denoted as ([[]] X min , Y d , Z d ); The point on the plant with the largest y axis coordinate value is denoted as E , and its coordinates are denoted as ([[]] X e , Y max , Z e ); The point on the plant with the smallest y axis coordinate value is denoted as F , and its coordinates are denoted as ([[]] X f , Y min , Z f ) 。

[0040] Step 6: To improve the accuracy of calculating the volume of the target plant, the plant is longitudinally divided into xoy equal parts along the n plane. Except for the top layer, each part can be approximated as a frustum of a cone. Denote the radius of the upper base of the frustum as r i , the radius of the lower base as R i . It can be obtained that r i = R i+1 . Denote the height of the target plant as H , then . Denote the height of each part after longitudinal division as h , then . Let the coordinates of a point on the plant be ([[]] X , Y , Z ). The i -th part of the section z axis coordinate value of the point satisfies ([[]] i -1) h ≤ Z≤ ih ,distance x The point with the largest axis coordinate value is recorded as C i , whose coordinates are ,distance x The point with the smallest axis coordinate value is recorded as D i , whose coordinates are , No. i +1 point in the zone z The axis coordinates satisfy h ≤ Z ≤( i +1) h ,distance x The point with the largest axis coordinate value is recorded as C i+1 , whose coordinates are ,distance x The point with the smallest axis coordinate value is recorded as D i+1 , whose coordinates are ,Pick , , No. n The part is approximately a cone, that is r n =0, then i The volume of the plant canopy is calculated as follows: ;

[0041] In the formula, v i For the i The volume of the plant canopy.

[0042] Step 7: n By summing up the volumes of the target plants, we can get the canopy volume of the entire target plant. The calculation method is: ;

[0043] In the formula, V is the canopy volume of the entire target plant.

[0044] Step 8: Control the flow rate and air volume in the variable spray system (5) according to the volume of the target plant canopy. To ensure that the robot spray operation can completely cover the entire target plant canopy, the time compensation of advance and delay spraying is performed, and finally accurate variable flow rate and variable air volume spraying is achieved.

[0045] Traditional orchard plant protection generally uses manual backpack sprayers or driving air-assisted sprayers to continuously spray operations at a fixed dose. The labor intensity is relatively high, and during the spraying process, the amount of liquid medicine that can accurately reach the parts of the fruit trees to be sprayed is extremely limited. Most of the pesticides are lost in the surrounding environment, resulting in poor spraying uniformity, double spraying, missed spraying and other phenomena, directly leading to low drug utilization rate, causing waste of pesticides and environmental pollution, and being unable to be applied to various crops. In addition, due to the low atomization degree of the liquid medicine by traditional operating equipment, the sedimentation loss of the liquid medicine is relatively high after spraying on the canopy of the sprayed plants. Moreover, in orchard operation environments such as vineyards and kiwifruit orchards, not only the vertical plants on both sides of the ridges need to be sprayed with liquid medicine, but also the horizontal vines at their tops. When the spraying device sprays pesticides vertically upward to the horizontal vines at its top, with the dripping of the liquid medicine, it is easy to cause poisoning of the operators and corrosion of the spraying device.

[0046] To address the above problems, the present invention uses an Ackermann mobile chassis and an inclined swing spray structure to broaden the operating environment of the spray robot. By adopting the inclined swing spray structure, not only the height of the spray robot is reduced, enabling it to operate flexibly, but also when spraying, the inclined swing variable spray system sprays the liquid medicine obliquely backward and upward as well as to the left and right sides, achieving three-sided spraying simultaneously and improving the operating efficiency. In addition, when the liquid medicine is sprayed obliquely backward and upward, even if some of the liquid medicine drips, it will not fall on the surface of the spray robot, resulting in the spray robot being corroded by the liquid medicine. At the same time, a three-dimensional lidar is used to scan the canopy of the plants, calculate the volume of the target plant canopy, and based on the volume size, achieve precise variable flow rate and variable air volume spraying for the plants. In addition, the aviation centrifugal nozzles used in the spray system can atomize the liquid medicine to a great extent by high-speed rotation, reduce the sedimentation waste of the liquid medicine, increase the spraying range, and at the same time, with the huge wind field generated by the fan, the liquid medicine can be accurately and quickly sprayed onto the canopy of the plants to be sprayed, and the liquid medicine can effectively penetrate the leaf layer, improving the utilization rate of the liquid medicine. Brief Description of the Drawings

[0047] Figure 1 Left front view of the present invention;

[0048] Figure 2 Left rear view of the present invention;

[0049] Figure 3 Schematic structural diagram of the inclined swing variable spray system of the present invention;

[0050] Figure 4 Schematic structural diagram of the data acquisition module of the present invention;

[0051] Figure 5 Schematic diagram of the kinematic model of the Ackermann structure;

[0052] Figure 6It is a schematic diagram of path planning;

[0053] Figure 7 It is a schematic diagram for calculating the canopy volume of the target plant;

[0054] Wherein: 1 - Ackermann mobile chassis; 2 - robot bracket; 3 - battery; 4 - central processor; 5 - variable spraying system, 51 - centrifugal nozzle, 52 - duct, 53 - fan, 531 - joint motor, 532 - flange, 533 - coupling, 534 - carbon fiber tube, 535 - vertical bearing; 6 - data acquisition module, 61 - navigation module, 62 - 3D lidar; 7 - water pump; 8 - hardware circuit drive control board; 9 - water tank. Specific implementation manners

[0055] The following elaborates on the invention solution in detail with reference to the drawings in the invention examples.

[0056] Such as Figure 1 、 Figure 2Shown are the left front and left rear views of the structure of an Ackermann-type intelligent precision variable air-assisted spray robot for orchards. It mainly consists of an Ackermann-type mobile chassis (1), a robot bracket (2), a battery (3), a central processor (4), a variable spray system (5), a data acquisition module (6), a water pump (7), a hardware circuit drive control board (8), and a water tank (9). And it includes the following parts: 51 - centrifugal nozzle, 52 - duct, 53 - fan, 531 - joint motor, 532 - flange, 533 - coupling, 534 - carbon fiber tube, 535 - vertical bearing; 61 - navigation module, 62 - 3D lidar. The Ackermann-type mobile chassis (1) uses PID control to control the motion trajectory, and realizes the control of the angular velocity and speed of the robot by controlling the steering angle of the front wheels and the rotational speed of the rear wheels of the Ackermann-type mobile chassis, and further controls the motion trajectory of the robot. The robot bracket (2) is connected by aluminum profiles and aluminum plates through metal angle pieces, which reduces its own weight while providing support for the entire intelligent precision variable air-assisted spray robot platform. The battery (3) provides power for the central processor (4), the variable spray system (5), the data acquisition module (6), the water pump (7), and the hardware circuit drive control board (8). The navigation module (61) fixed at the top of the robot bracket (2) is used to obtain information such as the position and attitude of the robot in real time; the 3D lidar (62) fixed at the front end of the robot bracket (2) is used to detect obstacles around the robot and scan the volume of the plant canopy around the robot. The central processor (4) is used to process the data obtained by the data acquisition module (6) for path planning and target plant volume calculation, so as to realize the path planning of the intelligent precision variable air-assisted spray robot and the accurate variable flow rate and variable air volume spraying. The liquid medicine in the water tank (9) is pressurized and transported to the centrifugal nozzle (51) by the water pump (7). The centrifugal nozzle (51) rotates at a high speed to atomize and spray the liquid medicine. The joint motor (531) and the fan (53) start operating synchronously. The hardware circuit drive control board (8) controls the water pump (7), the variable spray system (5), and the Ackermann-type mobile chassis (1) according to different control requirements issued by the central processor (4).

[0057] Such as Figure 3The figure shows a schematic diagram of the structure of an inclined swing variable spray system. Since the top of the orchard operation environment such as vineyards and kiwifruit orchards has horizontal vines, it is generally relatively short. And in this orchard operation environment, not only the vertical plants on both sides of the ridge need to be sprayed with liquid medicine, but also the horizontal vines at the top need to be sprayed. The inclined swing spray structure not only reduces the height of the spray robot, enabling it to operate flexibly, but also when spraying, the inclined swing variable spray system sprays the liquid medicine obliquely backward and upward as well as to the left and right sides, which can achieve three-sided spraying at the same time and improve the operation efficiency. In addition, when the liquid medicine is sprayed obliquely backward and upward, even if some liquid medicine drips, it will not fall on the surface of the spray robot, resulting in the spray robot being corroded by the liquid medicine, thus broadening the operation environment of the spray robot. The joint motor (531) is connected to the carbon fiber tube (534) through a flange (532), and a coupling (533) connects two carbon fiber tubes (534) with different diameters. The flange (532) tightly fixes the joint motor (531), the duct (52) and the carbon fiber tube (534) so that the three can rotate synchronously. The two ends of the carbon fiber tube (534) passing through the duct (52) are fixed in the vertical bearing (535) and can rotate arbitrarily within the vertical bearing (535). The vertical bearing (535) is fixed to the upper layer of the robot bracket (2) by screws and nuts. The centrifugal nozzle (51) is fixed on the carbon fiber tube (534) through a nozzle fixing part. By controlling the pressure of the water pump (7), the flow rate of the centrifugal nozzle (51) is controlled. At the same time, by controlling the rotation speed of the centrifugal nozzle (51), the atomization degree of the liquid medicine is further controlled to achieve spraying with variable droplet diameters. The fan (53) is also fixed on the carbon fiber tube (534) through a fan fixing part. The electronic speed controller is controlled by the hardware circuit driving control board (8) to further control the rotation speed of the fan. The centrifugal nozzle (51) and the fan (53) can rotate with the rotation of the carbon fiber tube (534) under the fixation of their respective fixing parts, realizing the reciprocating swing of the variable spray system (5).

[0058] As Figure 4The figure shows a schematic diagram of the data acquisition module structure. The navigation module (61) and the 3D lidar (62) are respectively fixed to the top and the front end of the robot bracket (2) through fixing parts. The robot uses the mainstream robot software framework ROS (Robot Operation System) as the current development environment. The point cloud volume processing algorithm and path planning for the plant canopy developed based on ROS combine the advantages of the ROS system and have the characteristics of modularity. The spraying robot obtains information such as the position and attitude of the robot according to the data acquired by the navigation module (61) and the 3D lidar (62), scans the volume of the plant canopy around the robot, and adjusts the driving route and its own attitude in real time, so that when the spraying robot encounters a plant, it can maintain an appropriate safety distance and spraying distance, and at the same time can realize the obstacle avoidance function during the operation. The path planning of the intelligent and precise variable air-assisted spraying robot and the precise variable flow rate and variable air volume spraying are realized.

[0059] As Figure 5 shown is a schematic diagram of the kinematic model of the Ackermann structure. In the figure, R is the turning radius, δ o is the outer front wheel tire angle, δ i is the inner front wheel tire angle, δ is the average front wheel angle, L is the wheelbase, l ω is the track width. According to this figure, the inner and outer tire angles of the front wheels of the spraying robot chassis can be obtained as: 、 ;

[0060] In the formula, δ o is the outer front wheel tire angle, δ i is the inner front wheel tire angle, L is the wheelbase, l ω is the track width, R is the turning radius.

[0061] Assume that the spraying robot is driving in a low-speed and non-slip scenario. Taking the center point of the rear axle of the spraying robot chassis as the rotation reference point, the average values of the left front wheel tire angle and the right front wheel tire angle of the spraying robot are taken, and the simplified motion model of the Ackermann mobile chassis (1) is as shown by the dotted line in Figure 5 , and its average front wheel angle is: ;

[0062] In the formula, δ is the average front wheel angle.

[0063] As Figure 6 shown is a schematic diagram of path planning. In the figure, O t is the current position, P tis a target point on the path, R is the turning radius, l d is the look-ahead distance, α is the heading deviation, l l is the lateral error. According to the pure pursuit algorithm, the motion trajectory of the spray robot from the current position to the path tracking point can be regarded as an arc trajectory with R as the radius. According to the sine theorem, we can get: ;

[0064] After conversion, we get: ;

[0065] In the formula, l d is the look-ahead distance, α is the heading deviation, that is, the angle between the current vehicle body attitude and the target point P t between.

[0066] According to the curvature formula, the curvature can be expressed as: ;

[0067] In the formula, K is the curvature of the path.

[0068] The average steering angle of the front wheels of the spray robot can be expressed as: ;

[0069] Combining the above formula, the final expression of the control variable δ of the pure pursuit algorithm can be obtained as: ;

[0070] In the formula, the time domain is introduced, α( t ) is t the angle between the spray robot and the target point P t at time, define l l as the lateral error between the current attitude of the spray robot and the target point P t , we can get: ;

[0071] At the same time, the curvature can be expressed as: ;

[0072] Express the look-ahead distance l d as a linear function of the linear velocity of the spray robot, we can get: ;

[0073] Combining the above formula, we can get: ;

[0074] Wherein, k is the proportionality coefficient, v is the linear velocity of the spraying robot, e is the initial forward viewing distance;

[0075] According to the angular velocity calculation method, it can be obtained that: ;

[0076] Wherein, ω is the angular velocity of the spraying robot.

[0077] By adjusting the proportionality coefficient k and then adjusting the pure tracking algorithm, the angular velocity ω of the spraying robot is inversely proportional to the forward viewing distance l d That is, the larger the forward viewing distance, the smaller the angular velocity, and the smoother the tracking trajectory; the smaller the forward viewing distance, the larger the angular velocity, and the more oscillating the tracking trajectory. The navigation module (61) is used to obtain the position and attitude information of the spraying robot. According to the planned path, the pure tracking algorithm is used to compare the lateral deviation between the current pose of the spraying robot and the planned navigation line, and the motion angle and displacement of the spraying robot are adjusted in real time.

[0078] Such as Figure 7 shown in the schematic diagram of the target plant canopy volume calculation. The plant canopy is scanned by a 3D lidar (62) to obtain the original lidar point cloud data. The original lidar point cloud data is preprocessed to filter out the point cloud data generated by distant obstacles and focus on the target to be detected. The preprocessed lidar point cloud data is clustered to fit the plant trunk. The point cloud data returned by the 3D lidar can be used for the volume calculation of the target plant. According to the fitted plant trunk, after data analysis, the point on the plant z with the largest coordinate value on the A axis is denoted as X a , Y a , Z max ); The point on the plant z with the smallest coordinate value on the B axis is denoted as X b , Y b , Z min ); The point on the plant x with the largest coordinate value on the C axis is denoted as X max , Y c , Z c ); The point on the plant xThe point with the minimum axial coordinate value is denoted as D , and its coordinates are denoted as ([[]] X min , Y d , Z d ); On the plant y The point with the maximum axial coordinate value is denoted as E , and its coordinates are denoted as ([[]] X e , Y max , Z e ); On the plant y The point with the minimum axial coordinate value is denoted as F , and its coordinates are denoted as ([[]] X f , Y min , Z f ). In order to improve the accuracy of target plant volume calculation, the plant is longitudinally divided into xoy equal parts along the n plane, as shown in Figure 7 . Except for the top layer, each part can be approximated as a frustum of a cone. Denote the radius of the upper base of the frustum as r i , and the radius of the lower base as R i . We can obtain r i = R i+1 . Denote the height of the target plant as H , then . Denote the height of each part after longitudinal division as h , then . Let the coordinates of a point on the plant be ([[]] X , Y , Z ). The i axial coordinate value of the points in the z th section satisfies ([[]] i -1) h ≤ Z ≤ ih . The point with the maximum distance from the x axis is denoted as C i , and its coordinates are denoted as . The point with the minimum distance from the x axis is denoted as D i , and its coordinates are denoted as . The i +1 axial coordinate value of the points in the z th section satisfiesh ≤ Z ≤( i +1) h , from the point with the largest x axis coordinate value, it is denoted as C i+1 , and its coordinates are denoted as , from the point with the smallest x axis coordinate value, it is denoted as D i+1 , and its coordinates are denoted as . Take , . The n th part is approximately a conical body, that is r n =0. Then the calculation method for the volume of the i th part of the plant canopy is: ;

[0079] In the formula, v i is the volume of the i th part of the plant canopy.

[0080] By summing the volumes of n parts, the volume of the entire target plant canopy can be obtained, and the calculation method is: ;

[0081] Control the flow rate and air volume in the variable spray system (5) according to the volume of the plant canopy. To ensure that the robot spray operation can completely cover the entire target plant canopy, perform time compensation for early and delayed spraying, and finally achieve precise variable flow rate and variable air volume spraying.

Claims

1. A path planning method for the structure of an Ackerman-type intelligent precision variable air-assisted spraying robot in an orchard, characterized in that, The robot structure includes an Ackermann mobile chassis (1), a robot bracket (2), a battery (3), a central processor (4), a variable spray system (5), a data acquisition module (6), a water pump (7), a hardware circuit drive control board (8), and a water tank (9); the Ackermann mobile chassis (1) is fixed with a robot bracket (2), and the robot bracket (2) is fixed with a battery (3), a central processor (4), a variable spray system (5), a data acquisition module (6), a water pump (7), a hardware circuit drive control board (8), and a water tank (9); the robot bracket (2) is fixed on the Ackermann mobile chassis (1) to provide a support function for the entire robot platform; the battery (3) supplies power to the central processor (4), the variable spray system (5), the data acquisition module (6), the water pump (7), and the hardware circuit drive control board (8); the central processor (4) is used to process the data obtained by the data acquisition module (6) and send it to the hardware circuit drive control board (8); the variable spray system (5) consists of a centrifugal nozzle (51), a duct (52), and a fan (53), the centrifugal nozzle (51) and the fan (53) are fixed on a carbon fiber tube (534), located inside the duct (52), and can swing reciprocally synchronously with the carbon fiber tube (534). By adjusting the rotation speed of the fan (53), the wind speed and air volume are adjusted in real time, thereby adjusting the penetration and spraying range of the liquid medicine; the liquid medicine is stored in the water tank (9), connected to the centrifugal nozzle (51) through the water pump (7) in sequence, and the liquid medicine is sprayed precisely and evenly onto the plant canopy; after receiving the data sent by the central processor (4), the hardware circuit drive control board (8) performs further processing, electrically connects and controls the forward speed and forward direction of the Ackermann mobile chassis (1), as well as the spray volume, spray time, air volume, and wind speed of the variable spray system (5), to achieve the path planning and precise variable flow and variable air volume spraying of the orchard intelligent and precise variable air-assisted spray robot; The steps of the path planning method are as follows: Step 1: According to the motion model of the Ackerman mobile chassis (1), the inner and outer tire angles of the front wheels of the spray robot chassis are as follows: , ; where δ o is the outer front wheel steering angle, δ i is the inner front wheel steering angle, L is the wheelbase, l ω is the track width, R is the turning radius; Step 2: Assume that the spraying robot travels in a low-speed and non-slip scenario. Taking the center point of the rear axle of the spraying robot chassis as the rotation reference point, the average values of the tire angles of the left front wheel and the right front wheel of the spraying robot are taken to obtain a simplified motion model of the Ackermann mobile chassis (1), and its average front wheel angle is: ; In the formula, δ is the average front wheel angle; Step 3: According to the pure tracking algorithm, the spraying robot moves from the current position O t to the path tracking point P t The motion trajectory can be regarded as an arc trajectory with R as the radius. According to the sine theorem, we can get: ; After conversion, we get: ; In the formula, l d is the forward viewing distance, and α is the course deviation, that is, the angle between the current vehicle body attitude and the target point P t ; Step 4: According to the curvature formula, the curvature can be expressed as: ; In the formula, K is the curvature of the path; Step 5: The average steering angle of the front wheels of the spraying robot can be expressed as: ; Combined with the above formula, the final expression of the control variable δ of the pure tracking algorithm can be obtained as follows: ; In the formula, the time domain is introduced, and α( t ) is t the included angle between the spraying robot and the target point at P t the moment, and it is defined that l l is the lateral error between the current posture of the spraying robot and the target point P t . Then, we can get: ; Meanwhile, the curvature can be expressed as: ; Step 6: Express the forward viewing distance l d as a linear function of the linear velocity of the spray robot, and we can obtain: ; Combining the above formula, we can obtain: ; Wherein, k is a proportionality coefficient, v is the linear velocity of the spraying robot, e is the initial forward viewing distance; Step 7: According to the angular velocity calculation method, it can be obtained that: ; In the formula, ω is the angular velocity of the spray robot; Step 8: By adjusting the scale factor k further adjust the pure tracking algorithm. The angular velocity ω of the spray robot is inversely proportional to the forward viewing distance l d That is, the larger the forward viewing distance, the smaller the angular velocity, and the smoother the tracking trajectory; the smaller the forward viewing distance, the larger the angular velocity, and the more oscillating the tracking trajectory. Step 9: Use the navigation module (61) to obtain the position and attitude information of the spray robot. According to the planned path, use the pure tracking algorithm. By comparing the lateral deviation between the current pose of the spray robot and the planned navigation line, the motion angle and displacement of the spray robot are adjusted in real time.

2. The method according to claim 1, characterized in that The Ackermann mobile chassis (1) provides a mounting platform to realize the position adjustment of the intelligent and precise variable air-assisted spray robot.

3. The method according to claim 1, wherein The robot bracket (2) consists of upper and lower layers. The lower layer is fixed on the Ackerman mobile chassis (1) using screws and nuts, and is equipped with a battery (3), a central processor (4), a water pump (7), a hardware circuit drive control board (8), and a water tank (9). To achieve the waterproof function, the upper and lower layers are separated by a board. The upper layer is equipped with a variable spray system (5) and a data acquisition module (6). The robot bracket (2) is connected by aluminum profiles and aluminum plates through metal corner fittings, which provides support for the entire orchard intelligent precision variable air-assisted spray robot platform while reducing its own weight.

4. The method according to claim 1, wherein The water pump (7) is used to control the flow rate of the centrifugal nozzle (51) in the duct (52). The centrifugal nozzle (51) and the fan (53) are fixed on the upper layer of the robot bracket (2) through a carbon fiber tube (534) passing through the duct (52). The centrifugal nozzle (51) and the fan (53) are located in the duct (52), and the wind speed and air volume are controlled by controlling the rotation speed of the fan (53).

5. The method according to claim 1, characterized in that The battery (3) is a lithium battery.

6. The method according to claim 1, wherein The data acquisition module (6) consists of a navigation module (61) and a 3D lidar (62). The navigation module (61) is fixed at the top of the robot bracket (2) and is used to obtain the position and attitude information of the robot in real time. The 3D lidar (62) is fixed at the front end of the robot bracket (2) and is used to detect obstacles around the robot and scan the volume of the plant canopy around the robot. The central processor (4) is used to process the data obtained by the data acquisition module (6) for path planning and target plant volume calculation.

7. The method according to claim 1, wherein The hardware circuit drive control board (8) is an STM32F4 control board, which is electrically connected through the Ackerman mobile chassis (1), the central processor (4), the centrifugal nozzle (51), the fan (53), and the water pump (7) to achieve variable flow rate and variable air volume spraying.

8. The method according to claim 1, characterized in that The steps are as follows: Step 1: Use the 3D lidar (62) to scan the plant canopy to obtain the original laser point cloud data. Preprocess the original laser point cloud data to filter out the point cloud data generated by distant obstacles and focus on the target to be detected. Step 2: Perform clustering processing on the preprocessed laser point cloud data, fit the plant trunk, and the point cloud data returned by the 3D lidar can be used for the volume calculation of the target plant. Step 3: The distance between the lidar and the chassis of the spraying robot is a , and the point directly below the lidar is defined as the origin of the coordinate system a . The x axis points directly in front of the spraying robot, y axis is parallel to the chassis wheel axis of the spraying robot and points to the left, z axis points vertically upward; Step 4: For convenience of calculation, the 3D lidar data points P t from polar coordinates ( r t , ω t , α t ) are converted to Cartesian coordinates ( x t , y t , z t ), and the conversion formula can be expressed as: ; In the formula, r t is the data point P t is the distance from the lidar; ω is the vertical scanning angular resolution of the lidar; α is the horizontal scanning angular resolution of the lidar; Step 5: Based on the fitted plant trunk, after data analysis, the point with the maximum z axis coordinate value on the plant is denoted as A , and its coordinates are denoted as ( X a , Y a , Z max ); The point with the minimum z axis coordinate value on the plant is denoted as B , and its coordinates are denoted as ( X b , Y b , Z min ); The point with the maximum x axis coordinate value on the plant is denoted as C , and its coordinates are denoted as ( X max , Y c , Z c ); The point with the minimum x axis coordinate value on the plant is denoted as D , and its coordinates are denoted as ( X min , Y d , Z d ); The point with the maximum y axis coordinate value on the plant is denoted as E , and its coordinates are denoted as ( X e , Y max , Z e ); The point with the minimum y axis coordinate value on the plant is denoted as F , and its coordinates are denoted as ( X f , Y min , Z f ); Step 6: To improve the accuracy of target plant volume calculation, the plant is longitudinally divided along the xoy plane into n equal parts. Except for the top layer, each part can be approximated as a frustum of a cone. Denote the radius of the upper base of the frustum as r i , the radius of the lower base as R i , and we can get r i = R i+1 . Denote the height of the target plant as H , then . Denote the height of each part after longitudinal division as h , then . Set the coordinates of a point on the plant as ([[]] X , Y , Z ). The i -th part of the z -axis coordinate value of the points in the section satisfies ( i -1) h ≤ Z ≤ ih . Denote the point with the maximum x -axis coordinate value as C i , and its coordinates are denoted as . Denote the point with the minimum x -axis coordinate value as D i , and its coordinates are denoted as . The i +1-th part of the z -axis coordinate value of the points in the section satisfies h ≤ Z ≤( i +1) h . Denote the point with the maximum x -axis coordinate value as C i+1 , and its coordinates are denoted as . Denote the point with the minimum x -axis coordinate value as D i+1 , and its coordinates are denoted as . Take , . The n -th part is approximated as a cone, that is, r n =0. Then the volume calculation method for the i -th part of the plant canopy is: ; In the formula, v i is the i volume of the plant canopy of the th portion; Step 7: Summing the n volume portions, the volume of the canopy of the entire target plant can be obtained, and the calculation method is: ; In the formula, V is the volume of the canopy of the entire target plant; Step 8: Control the flow rate and air volume in the variable spray system (5) according to the volume of the target plant canopy. To ensure that the robot spraying operation can completely cover the entire target plant canopy, perform time compensation for early and delayed spraying, and finally achieve precise variable flow rate and variable air volume spraying.

Images