Multi-pesticide tank unmanned aerial vehicle centroid balance liquid supply control method based on flight attitude

Abstract

The application discloses a kind of based on flight attitude multi pesticide tank unmanned aerial vehicle centroid balance liquid supply control method, obtain unmanned aerial vehicle pitch angle data in flight process, each pesticide tank geometric parameter and liquid state information, according to the change of flight attitude, pesticide tank in liquid free surface form is classified, calculates each pesticide tank liquid centroid coordinate and unmanned aerial vehicle whole machine centroid coordinate. 0-1 mixed linear model is established, under the premise of meeting continuous liquid supply time length constraint, pesticide tank mutual exclusion liquid supply constraint, joint liquid supply amount constraint, liquid supply rate constraint and the liquid amount constraint required by liquid spray system, with the minimum unmanned aerial vehicle whole machine centroid offset of each pesticide tank liquid supply operation as target, dynamically determine the liquid supply state and liquid supply rate of each pesticide tank, realize the collaborative liquid supply control of multi pesticide tank.The application with initiative control and prediction model will instantaneous centroid always be constrained in ideal small range, effectively reduce the centroid deviation of multi pesticide tank unmanned aerial vehicle, improve the flight stability of plant protection unmanned aerial vehicle.

Description

Technical Field

[0001] This invention belongs to the field of UAV plant protection and flight control technology, specifically involving a centroid balance liquid supply control method for multi-pesticide tank UAVs based on flight attitude. During plant protection operations, the liquid supply of the multi-pesticide tanks of the UAV is dynamically adjusted based on changes in flight attitude to achieve centroid balance control of the UAV. Background Technology

[0002] Agricultural drones typically employ a multi-tank structure to meet the requirements of flight range, pesticide loading, structural layout, and safety redundancy in plant protection operations. During flight operations, as pesticide solution is continuously consumed and the spraying sequence changes, the distribution of pesticide solution in each tank constantly alters, causing a shift in the overall centroid coordinates of the drone.

[0003] Especially in actual plant protection operations, the pitch angle of the drone changes in real time with the operating environment (such as hills, terraces, or orchard canopies) and flight altitude. The free surface morphology of the pesticide liquid in the tank also changes accordingly, causing the centroid coordinates of the pesticide liquid in the tank to be in a dynamic state. If a fixed spraying sequence or experience-based spraying control method is still used, it is difficult to accurately reflect the actual situation of the pesticide tank's centroid changing with attitude. This can easily cause excessive centroid offset of the drone, thus affecting flight stability and plant protection operation accuracy, and even leading to problems such as missed spraying, double spraying, and drone tipping.

[0004] In addition, the process of supplying pesticides to multiple pesticide tanks by drones is usually subject to various constraints such as the supply rate, the duration of continuous supply, the structural relationship of the pesticide tanks, and the spraying requirements of the spraying system. There is a lack of a supply method in the existing technology that can achieve dynamic and precise control of the drone's center of mass while meeting the above constraints.

[0005] Therefore, it is necessary to provide a liquid supply control method that can comprehensively consider the changes in the flight attitude of the UAV, the structural characteristics of multiple pesticide tanks, and the liquid supply constraints, so as to improve the centroid control accuracy of the UAV and the pesticide utilization efficiency, and ensure the stability and accuracy of plant protection operations. Summary of the Invention

[0006] Purpose of the invention: In view of the problems pointed out in the background art, the present invention provides a centroid balance liquid supply control method for multi-pesticide tank UAVs based on flight attitude. By calculating the centroid coordinates of the pesticide liquid in the pesticide tanks in real time during the plant protection operation flight, and dynamically adjusting the liquid supply strategy of each pesticide tank under multiple constraints, the centroid offset of the UAV is minimized during the entire plant protection operation flight mission, thereby improving the flight stability of the UAV, reducing pesticide waste, and improving the accuracy of plant protection operations.

[0007] Technical solution: This invention discloses a centroid balance liquid supply control method for multi-pesticide tank UAVs based on flight attitude, comprising:

[0008] Obtain the geometric parameters, installation position parameters, initial liquid volume, drone pitch angle data, and spray system requirements of each medicine tank on the drone;

[0009] The shape of the free liquid surface of the medicine in the medicine container is determined based on the pitch angle data, and the centroid coordinates of the medicine in the container at the corresponding time are calculated.

[0010] Based on the centroid coordinates of the pesticide liquid in each pesticide tank and the mass of the pesticide liquid, calculate the centroid coordinates of the entire drone at the corresponding moment;

[0011] Establish a 0-1 hybrid linear model to plan the liquid supply constraints. The liquid supply constraints include at least the continuous liquid supply duration constraint, the mutual exclusion liquid supply constraint of the medicine tank, the joint liquid supply quantity constraint, the liquid supply rate constraint, and the liquid quantity required by the spraying system.

[0012] Under the premise of satisfying the liquid supply constraints, with the goal of minimizing the offset of the whole machine's center of gravity caused by the liquid supply of pesticide tanks from the drone, the liquid supply status and liquid supply rate of each pesticide tank at the corresponding time are determined.

[0013] Based on the determined liquid supply status and rate, the liquid supply to each medicine tank is controlled, and the drone flight status is updated for the next moment, entering the control cycle of the next period.

[0014] Furthermore, based on the pitch angle data, the pitch state of the UAV at the current moment is determined, and the free surface morphology of the medicine liquid in the medicine tank is classified according to the preset pitch angle range;

[0015] Within different pitch angle ranges, based on the geometric dimensions of the medicine container and the current volume of the medicine liquid, the cross-section of the free liquid surface of the medicine liquid inside the medicine container is divided into at least one combined geometric shape composed of rectangles and / or triangles;

[0016] Calculate the area of ​​the combined geometric shape and its corresponding geometric centroid coordinates, and determine the centroid coordinates of the medicine container at the current moment based on the area weighting method.

[0017] Furthermore, the free surface morphology of the pesticide solution inside the pesticide tank is classified, and a is defined as follows: n b n c n Let represent the length, width, and height of the nth pesticide tank, respectively; let θ be the pitch angle of the drone; and let S be the cross-sectional area of ​​the pesticide solution in the yz coordinate system. The specific classifications are as follows:

[0018] The first scenario: When the drone is flying upwards, y1 ≤ b n And z1≤c n At this time, the free surface of the pesticide solution forms a triangle with the pesticide tank. The length of the triangle on the y-axis is y1, the length on the z-axis is z1, and the area is S.

[0019] The second scenario: When the drone flies upwards, it satisfies... At that time, the pesticide solution in the pesticide tank was divided into a rectangle S2 and a triangle S1. The length of the rectangle on the y-axis is y1, and the length of the triangle on the y-axis is y2.

[0020] The third type: When the drone flies upwards, it satisfies... At that time, the pesticide solution in the pesticide tank is divided into a rectangle S2 and a triangle S1. The length of the rectangle on the z-axis is z1, and the length of the triangle on the z-axis is z2.

[0021] The fourth type: When the drone flies upwards, it satisfies... At that time, the length of the empty part inside the medicine container in the y-axis direction is set to y3, and the length in the z-axis direction is set to z3. The area S of the medicine liquid is divided into two rectangles S1 and S2 and a triangle S3.

[0022] The fifth type: When the drone flies downwards, it meets the requirements. At that time, the free surface of the pesticide solution forms a triangle with the pesticide tank. Let the length of the pesticide tank in the y-axis direction be y1, the length in the z-axis direction be z1, and the area of ​​the pesticide solution inside the pesticide tank be S.

[0023] The sixth type: When the drone flies downwards, it meets the requirements. At that time, the liquid medicine in the medicine pot is divided into a rectangle S2 and a triangle S1. Let the length of the cross-sectional triangle in the y-axis direction be y1, and the length of the cross-sectional rectangle in the y-axis direction be y2.

[0024] The seventh type: When the drone flies downwards, it meets the requirements. At that time, the liquid medicine in the medicine pot is divided into a rectangle S2 and a triangle S1. Let the length of the cross-sectional triangle in the z-axis direction be z2, and the length of the cross-sectional rectangle in the z-axis direction be z1.

[0025] The eighth type: When the drone flies downwards, it meets the requirements. At that time, the length of the empty part inside the medicine jar in the y-axis direction is set to y3, and the length in the z-axis direction is set to z3. The medicine liquid inside the medicine jar is divided into two rectangles S1 and S2 and a triangle S3.

[0026] The ninth type: When the drone is flying horizontally, it meets the following requirements. At that time, the remaining liquid in the medicine tank is a rectangle S with a length of z1 in the z-axis direction, and the center of mass of the UAV depends on the height z1 of the remaining liquid.

[0027] Furthermore, when calculating the overall centroid coordinates of the UAV at a given moment, it is necessary to consider the mass of the medicine in each medicine container at the current moment, the corresponding centroid coordinates of the medicine, and the mass of the UAV itself. Specifically:

[0028] Based on the coordinates of the centroid of the medicine liquid in the nth tank at the rth second. Formula for the centroid coordinates of the entire drone Find the actual center of mass of the entire UAV at the r-th second. ,Right now:

[0029] ;

[0030] Where M represents the weight of the drone itself, and the coordinates of the centroid of the pesticide solution in the nth pesticide tank at second r are... for T0(X0,Y0,Z0) represents the centroid coordinates of the drone when the pesticide tank is unloaded. This represents the weight of the pesticide solution in the nth pesticide tank at second r, where p represents the number of pesticide tanks. In practical work... The item can be ignored.

[0031] Furthermore, a 0-1 hybrid linear programming model is established. Under the premise of satisfying the constraints of continuous liquid supply duration, mutual exclusion of liquid supply to medicine tanks, joint liquid supply volume, liquid supply rate, and the liquid volume required by the spraying system, the liquid supply status and liquid supply rate of each medicine tank are dynamically determined. The specific liquid supply constraints are as follows:

[0032] The duration of liquid supply must be no less than 60 seconds. At time r, a variable is set. The value of the medicine container is 1 if it supplies liquid at time r, and 0 if it does not supply liquid. The determination of the liquid supply status of the medicine container at a given time is as follows:

[0033] ;

[0034] in, This indicates the number of times fluid was supplied within the past 60 seconds. When r ≤ 60, the number of times fluid was supplied starts from the initial time.

[0035] During the liquid supply process, a maximum of two pesticide tanks can supply liquid simultaneously and cannot be located on the same side. The amount of liquid supplied in the r-th second is less than the amount of liquid remaining in the previous second. The liquid supply speed cannot exceed the maximum liquid supply rate allowed by the pesticide tank. An interval constraint is set for the total liquid supply in the r-th second, with the lower limit being the amount of liquid required by the spraying system and the upper limit being 1.1 times the amount of liquid required by the spraying system.

[0036] Furthermore, given that all constraints are met, there are multiple liquid supply combinations for single or dual tanks. The least squares method is used to allocate the liquid supply rate of each pesticide tank, minimizing the overall centroid shift of the drone caused by liquid supply. Specifically:

[0037] Calculate the centroid influence coefficient of all non-empty medicine containers at the current moment. , = This indicates that for every unit of medicine consumed by the nth tank, the center of gravity of the entire machine shifts by an amount. Let represent the centroid coordinates of the pesticide solution in the nth pesticide tank at second r. This indicates the coordinates of the drone's center of mass. Let M represent the weight of the pesticide solution in the nth pesticide tank at second r, M represent the weight of the drone itself, and p represent the number of pesticide tanks.

[0038] Introducing dummy variables The optimization objective is to minimize the sum of squared Euclidean distances of the overall machine's center of gravity shift caused by liquid supply, i.e.: ;

[0039] For each feasible liquid supply combination, the optimal rate is obtained by solving the least squares method.

[0040] Furthermore, for each feasible liquid supply combination, the optimal rate is obtained by solving the least squares method:

[0041] If it is a single tank i, then = ;

[0042] If there are two tanks i and j, then the following conditions are met. Under the premise that:

[0043] ;

[0044] - ;

[0045] in, This represents the liquid dispensing rate of the i-th pesticide tank at the r-th second in the screening combination. This represents the liquid supply rate of the j-th pesticide tank at the r-th second in the screening combination. This represents the centroid influence coefficient of the i-th pesticide tank at second r in the screening combination. This represents the centroid influence coefficient of the j-th pesticide tank at second r in the screening combination. This indicates the amount of liquid required by the spraying system in the r-th second.

[0046] Furthermore, the liquid supply is controlled according to the optimal liquid supply strategy. After one control cycle ends, the remaining liquid quality and real-time status of each medicine tank are updated, and the control cycle for the next cycle begins. Specifically:

[0047] After a control cycle ends, the remaining liquid mass of each medicine tank is calculated for the next moment.

[0048] Substitute the values ​​into the machine's centroid formula to calculate the machine's centroid at the next moment, and calculate the corresponding centroid influence coefficient.

[0049] Calculate the minimum value of the sum of squared Euclidean distances of the total centroid shift caused by the liquid supply at the next moment. The combination with the smallest value is selected as the optimal liquid supply strategy for the next cycle.

[0050] Furthermore, the system controls each tank to supply liquid to the spraying system according to the liquid supply strategy, and updates the remaining liquid volume in the tank and the centroid coordinates of the UAV at each preset time interval, entering the next spraying control cycle until the plant protection operation flight mission ends.

[0051] Beneficial effects:

[0052] 1. This invention introduces a dynamic calculation method for the centroid of pesticide liquid in pesticide tanks based on flight attitude, which aligns with the actual characteristics of pesticide liquid loading on agricultural drones, thus improving the accuracy of drone centroid calculation. Dynamic problems that cannot be solved by static design are addressed by dynamic active control and prediction models. Dynamic optimization of spraying strategies under multiple pesticide tanks and multiple constraints effectively reduces the maximum centroid offset of the drone, avoiding flight tilting and overturning problems caused by centroid deviation, and significantly improving flight stability during agricultural operations.

[0053] 2. This invention determines the free surface morphology of the medicine liquid inside the medicine container based on pitch angle data, and calculates the centroid coordinates of the medicine liquid according to different free surface morphologies. This results in more accurate centroid coordinates, and the medicine liquid is classified into a combination of geometric shapes composed of rectangles and / or triangles, making the calculation of the free surface area and centroid process simpler.

[0054] 3. By constraining the liquid supply rate and continuous liquid supply duration, this invention reduces the impact of frequent start-stop liquid supply on the drone's spraying pipeline, atomizing nozzle, and other systems, while also avoiding missed spraying and over-spraying caused by disordered liquid supply rhythm, thereby improving pesticide utilization efficiency.

[0055] 4. This invention is applicable to complex plant protection flight missions where the pitch angle dynamically changes with the operating environment (such as plant protection in hilly and mountainous areas, fruit tree canopy protection, terraced field protection, etc.), and is compatible with various plant protection drones, including multi-rotor and fixed-wing drones, possessing good versatility and engineering application value. The liquid supply control strategy, centered on the drone's center of gravity balance, ensures the drone maintains a stable flight attitude, further improving the uniformity and accuracy of liquid spraying in plant protection operations. Attached Figure Description

[0056] Figure 1 This is a flight attitude diagram of the drone;

[0057] Figure 2 A schematic diagram for calculating the centroid of pesticide liquid in a drone's pesticide tank;

[0058] Figure 3A schematic diagram illustrating the calculation of the optimal liquid supply rate for pesticide tanks on drones under multiple constraints;

[0059] Figure 4 This is a schematic diagram of the liquid supply control method of the present invention;

[0060] Figure 5 The plot shows the pitch angle of the UAV as a function of time (s).

[0061] Figure 6 The motion curve of the center of mass of the UAV during flight (xyz coordinate system);

[0062] Figure 7 A comparison diagram of theoretical and experimental results regarding the center-of-mass balancing strategy for unmanned aerial vehicles (UAVs).

[0063] Figure 8 This is a schematic cross-sectional view of the free surface morphology of the pesticide liquid inside the pesticide tank of the present invention. Figure 1 ;

[0064] Figure 9 This is a schematic cross-sectional view of the free surface morphology of the pesticide liquid inside the pesticide tank of the present invention. Figure 2 . Detailed Implementation

[0065] The following description, in conjunction with the accompanying drawings, illustrates a specific embodiment of the present invention, but the invention is not limited thereto.

[0066] This invention discloses a centroid balance liquid supply control method for multi-pesticide tank UAVs based on flight attitude, see [link to relevant documentation]. Figure 4 It includes the following steps:

[0067] S1: Obtain the following parameters of the drone during the current agricultural spraying operation flight mission:

[0068] • Geometric dimensions of each pesticide tank and its installation location in the drone;

[0069] • Initial pesticide solution volume in each pesticide tank;

[0070] • The centroid coordinates of the UAV in an unloaded state;

[0071] • Pitch angle data of the drone during flight;

[0072] • Planned liquid injection volume data of the liquid injection system during flight operations;

[0073] • Maximum allowable liquid supply rate and minimum continuous liquid supply duration limits for each pesticide tank.

[0074] S2: Steps for calculating the centroid of pesticide liquid in a pesticide tank

[0075] Based on the pitch angle data, the pitch state of the UAV at the current moment is determined, and the free surface morphology of the pesticide liquid in the pesticide tank is classified according to the preset pitch angle range.

[0076] Within different pitch angle ranges, based on the geometric dimensions of the pesticide tank and the current volume of pesticide liquid, the cross-section of the free liquid surface inside the pesticide tank is divided into at least one combined geometric shape consisting of rectangles and / or triangles.

[0077] Calculate the area of ​​the combined geometric shape and its corresponding geometric centroid coordinates, and determine the centroid coordinates of the pesticide liquid in the pesticide tank at the current moment based on the area weighting method.

[0078] S3: Steps for calculating the overall centroid of a UAV

[0079] Based on the current mass of pesticide solution in each pesticide tank and its corresponding centroid coordinates, and combined with the drone's own mass and reference centroid coordinates, the centroid coordinates of the drone at the current moment are calculated.

[0080] S4: Steps for Constructing a Liquid Supply Constraint Strategy

[0081] Based on the structural characteristics of UAVs and the safety control requirements for plant protection operations, liquid supply constraints are established. These constraints should include at least the following:

[0082] • Each pesticide tank must meet the minimum continuous supply time requirement during the liquid supply process. The continuous supply time constraint is achieved by cumulatively judging the supply status.

[0083] • A preset number of pesticide tanks allows for simultaneous supply of pesticides to the spraying system;

[0084] • Pesticide tanks located on the same side or with structural relationships must satisfy mutual exclusion liquid supply constraints;

[0085] • The instantaneous liquid supply rate of each pesticide tank shall not exceed the corresponding maximum liquid supply rate;

[0086] • The combined liquid supply of all pesticide supply tanks meets the pesticide spraying requirements of the spraying system at the corresponding time.

[0087] S5: Steps for Determining the Spraying Strategy

[0088] Under the premise of meeting the liquid supply constraints, the liquid supply status and liquid supply rate of each pesticide tank at each moment during the entire flight operation of the UAV are determined by minimizing the overall center of gravity offset caused by liquid supply at each moment, thus forming a liquid supply control strategy.

[0089] S6: Spraying Execution and Status Update Procedures

[0090] The system controls each pesticide tank to supply liquid to the spraying system according to the liquid supply strategy, and updates the remaining liquid volume of the pesticide tank and the actual centroid coordinates of the drone at each preset time interval, and enters the next spraying control cycle until the plant protection operation flight mission ends.

[0091] In this embodiment of the invention, during a plant protection operation flight mission, the plant protection drone is equipped with six pesticide tanks, distributed on both sides and the middle of the drone's fuselage, with different geometric dimensions and installation positions. During the flight operation, the drone's flight control system acquires pitch angle data in real time and classifies the free surface morphology of the pesticide solution in the tanks based on the pitch angle changes. When the drone is operating, its pitch attitude only involves two flight modes: forward and ascent. Therefore, the overall X-axis centroid changes only slightly linearly with pesticide consumption. This change can be adaptively adjusted by the flight control system without requiring active control through a liquid supply strategy. This invention only performs centroidal balance liquid supply control on the Y-axis (left and right) and Z-axis (vertical); the X-axis can be calculated synchronously but does not participate in the liquid supply decision.

[0092] The implementation process of this invention is described below:

[0093] Step 1: First, set the center of the medicine pot as the origin of the coordinate system. To facilitate geometric calculations, first establish the corner coordinate system with the lower left corner of the pot as the origin. Finally, convert the center-of-mass coordinates to the center-of-mass coordinate system. See [link to relevant documentation]. Figure 1 Ideally, by varying the angle θ and the UAV's pitch flight attitude, nine different scenarios can be derived. See [link to relevant documentation]. Figures 8-9 Based on the centroid coordinates of each segmented figure in each case, the centroid coordinate formula is applied. Find the coordinates of the center of mass inside the tank. n Let S represent the centroid coordinates of the pesticide solution inside the nth pesticide tank, and let S represent the total area of ​​the pesticide solution pattern. i This represents the area of ​​the shape formed by the division of the medicinal liquid. a represents the centroid coordinates of the image after the drug solution is segmented. n b n c n These represent the dimensions (length, width, height) of the nth pesticide container.

[0094] (1) When the UAV is flying upwards, y1≤b n And z1≤c n At that time, assume that the length of the triangle (the cross-section of the medicine liquid) on the y-axis is y1, the length on the z-axis is z1, and the area is S.

[0095] From this, we can deduce that: .

[0096] From this, we can deduce that: .

[0097] The coordinates of the three points ABC of the triangle are A(0,y1,0), B(0,0,z1), and C(0,0,0).

[0098] The centroid coordinates t of the pesticide solution inside the pesticide tank in this case can be obtained using the centroid coordinate formula. n .

[0099] (2) When the drone flies upwards, it satisfies At that time, the pesticide solution in the pesticide tank was divided into a rectangle S2 and a triangle S1. The length of the rectangle on the y-axis is y1, and the length of the triangle on the y-axis is y2.

[0100] From this, we can deduce that: .

[0101] Their respective areas are: , .

[0102] So The coordinates of the three points A, B, and C of the triangle are A(0, y1, c). n ), B(0,y1,0), C(0,y1+y2,0).

[0103] The centroid coordinates t of the pesticide solution inside the pesticide tank in this case can be obtained using the centroid coordinate formula. n .

[0104] (3) When the drone flies upwards, it satisfies At that time, the pesticide solution in the pesticide tank was divided into a rectangle S2 and a triangle S1. The length of the rectangle on the z-axis is z1, and the length of the triangle on the z-axis is z2.

[0105] From this, we can deduce that:

[0106] The coordinates of the three points ABC of the triangle are A(0,0,z1), B(0,b) and C(z2,z3). n ,z1)C(0,0,z1+z2)

[0107] The centroid of the segmented figure is then: , .

[0108] The centroid coordinates t of the pesticide solution inside the pesticide tank in this case can be obtained using the centroid coordinate formula. n .

[0109] (4) When the drone flies upwards, it satisfies At that time, the length of the empty part inside the medicine container in the y-axis direction is set to y3, and the length in the z-axis direction is set to z3. The area S of the medicine liquid is divided into two rectangles S1 and S2 and a triangle S3.

[0110] From this, we can deduce that:

[0111]

[0112]

[0113]

[0114]

[0115]

[0116] The coordinates of the three points DEF of the triangle are D(0, ..., ... -y3, E(0, -y3, -z3), F(0, , -z3)

[0117] Then the centroid of the triangle is: .

[0118] (5) When the drone is flying downwards, when Let the length of the medicine container in the y-axis direction be y1, the length in the z-axis direction be z1, and the area of ​​the liquid inside the medicine container be S.

[0119] From this, we can deduce that:

[0120]

[0121]

[0122]

[0123] The coordinates of the three points A, B, and C of the triangle are A(0, ..., ...). -y1,0), B(0, ,0)C(0, , z1)

[0124] The centroid coordinates t of the pesticide solution inside the pesticide tank in this case can be obtained using the centroid coordinate formula. n .

[0125] (6) When the drone flies downwards, it meets the following requirements. At that time, the liquid medicine in the medicine pot is divided into a rectangle S2 and a triangle S1. Let the length of the cross-sectional triangle in the y-axis direction be y1, and the length of the cross-sectional rectangle in the y-axis direction be y2.

[0126] From this, we can deduce that:

[0127]

[0128]

[0129]

[0130]

[0131] The coordinates of the three points A, B, and C of the triangle are A(0, ..., ...). -y2,0),B(0, -y2, ), C(0, -y1-y2,0).

[0132] The centroids of the rectangle and triangle are: , .

[0133] The centroid coordinates t of the pesticide solution inside the pesticide tank in this case can be obtained using the centroid coordinate formula. n .

[0134] (7) When the drone flies downwards, it meets the following conditions. At that time, the liquid medicine in the medicine pot is divided into a rectangle S2 and a triangle S1. Let the length of the cross-sectional triangle in the z-axis direction be z2, and the length of the cross-sectional rectangle in the z-axis direction be z1.

[0135] From this, we can deduce that:

[0136]

[0137]

[0138]

[0139] The coordinates of the three points ABC of the triangle are A(0,0,z1), B(0,z2), and C(z3). ,z1)C(0, (z1+z2)

[0140] The centroids of the rectangle and triangle are: , .

[0141] The centroid coordinates t of the pesticide solution inside the pesticide tank in this case can be obtained using the centroid coordinate formula. n .

[0142] (8) When the drone flies downwards, it meets the following conditions. At that time, the length of the empty part inside the medicine jar in the y-axis direction is set to y3, and the length in the z-axis direction is set to z3. The medicine liquid inside the medicine jar is divided into two rectangles S1 and S2 and a triangle S3.

[0143] From this, we can deduce that:

[0144]

[0145]

[0146]

[0147]

[0148] The coordinates of the three points DEF of the triangle are D(0, y3, ..., y4, ...). E(0,y3, -z3), F(0,0, -z3)

[0149] Then the area and centroid of the triangle are: , .

[0150] The centroid coordinates t of the pesticide solution inside the pesticide tank in this case can be obtained using the centroid coordinate formula. n .

[0151] (9) When the drone is flying horizontally, it meets the following requirements. At that time, the remaining liquid in the medicine tank is a rectangle S with a length of z1 in the z-axis direction, and the center of mass of the UAV depends on the height z1 of the remaining liquid.

[0152] From this, we can deduce that: .

[0153] The centroid coordinates t of the pesticide solution inside the pesticide tank in this case can be obtained using the centroid coordinate formula. n .

[0154] See Figure 5 In this embodiment, the pitch angle of the UAV changes over time (s). Based on this pitch angle, the center of mass of the UAV is calculated. See [link / reference]. Figures 6-7 The diagram shows the motion curve of the UAV's center of mass during flight (xyz coordinate system) and a comparison diagram of the theoretical and experimental aspects of the UAV's center of mass balance strategy.

[0155] Step 2: Based on the included angle θ and the UAV's pitch flight attitude, the coordinates of the centroid of the pesticide liquid in the nth pesticide tank at second r can be obtained. for .

[0156] Step 3: Based on the coordinates of the centroid of the liquid in the nth container at the r-th second. Formula for the centroid coordinates of the entire drone Find the actual center of mass of the entire UAV at the r-th second. ,Right now:

[0157]

[0158] Where M represents the weight of the drone itself, and T0(X0,Y0,Z0) represents the centroid coordinates of the drone when the pesticide tank is unloaded. This represents the weight of the pesticide solution in the nth pesticide tank at second r. In discussing practical applications... This item can be ignored.

[0159] Step 4: The quality of pesticide liquid in the pesticide tank is limited by several engineering constraints, such as continuous supply time, mutual exclusion of supply between tanks, combined supply volume, supply rate, and the amount of pesticide liquid required by the spraying system. A 0-1 mixed linear programming model is established to dynamically determine the supply status and supply volume of each tank.

[0160] Step 4.1: To ensure the liquid supply duration is no less than 60 seconds, and to address the issue of determining if the liquid supply time meets the 60-second requirement, at time r, a variable is set... ,Right now:

[0161] .

[0162] To determine the state of the medicine supply at a given moment, i.e.:

[0163]

[0164] Regarding the judgment method described above, if there is no liquid supply from the medicine tank for one second, that is... It can be turned on or kept off at any time; the moment the medicine is dispensed from the tank, it is ready to start. If the continuous fluid supply in the past period was less than 60 seconds, then =1, fluid supply must continue; if continuous fluid supply has already reached 60 seconds in the past period, then =1 or 0, you can choose to stop or continue the fluid supply. This indicates the number of times liquid was supplied within the past 60 seconds. When r ≤ 60, it starts from the initial time.

[0165] Step 4.2: To ensure control stability and prevent tipping, the drone can only supply liquid to a maximum of two tanks simultaneously during the liquid supply process, and they cannot be located on the same side.

[0166]

[0167] Step 4.3, to ensure the rationality of the liquid supply, the weight of the liquid in the nth pesticide tank at second r is set to be... To ensure that the amount of liquid supplied in the r-th second is less than the amount of liquid remaining in the previous second, that is:

[0168] ;

[0169] This represents the liquid supply rate of the nth pesticide tank at second r. This represents the weight of the medicine in the nth medicine container at the (r-1)th second. Indicates the time step.

[0170] Step 4.4, to ensure that the liquid supply rate does not exceed the upper limit, that is:

[0171] v n ;

[0172] v n This represents the maximum allowable liquid supply rate for the nth pesticide tank.

[0173] Step 4.5: To ensure the stability of the liquid supply of the spraying system and avoid insufficient or excessive supply of liquid, an interval constraint is set for the total liquid supply at second r, with the lower limit being the amount of liquid required by the spraying system. To prevent insufficient spraying, the upper limit is the amount of liquid required by the spraying system. 1.1 times, that is:

[0174]

[0175] Step 5: Under the condition of satisfying all constraints, there are 15 possible liquid supply combinations for single or double tanks (6 single tank liquid supply combinations and 9 double tank liquid supply combinations). The least squares method is used to allocate the liquid supply rate of each pesticide tank so as to minimize the overall centroid offset of the UAV caused by liquid supply.

[0176] Step 5.1: Calculate the centroid influence coefficient of all non-empty medicine containers at the current moment. ,Right now:

[0177] =

[0178] The influence coefficient This indicates the amount of displacement of the entire machine's center of gravity caused by the consumption of 1 unit of medicine in the nth tank.

[0179] Step 5.2, introduce dummy variables The optimization objective is to minimize the sum of squared Euclidean distances of the overall machine's center of gravity offset caused by liquid supply:

[0180]

[0181] in, This represents the liquid supply rate of the nth pesticide tank at second r. This represents the centroid influence coefficient of the nth pesticide tank at second r.

[0182] Step 5.3: For each feasible liquid supply combination, the optimal rate is obtained by solving the least squares method.

[0183] If it is a single tank i, then = .

[0184] If there are two tanks i and j, then the following conditions are met. Under the premise that:

[0185] ;

[0186] - ;

[0187] in, This represents the liquid dispensing rate of the i-th pesticide tank at the r-th second in the screening combination. This represents the liquid supply rate of the j-th pesticide tank at the r-th second in the screening combination. This represents the centroid influence coefficient of the i-th pesticide tank at second r in the screening combination. This represents the centroid influence coefficient of the j-th pesticide tank at second r in the screening combination. This indicates the amount of liquid required by the spraying system in the r-th second.

[0188] Step 6: Control the liquid supply according to the optimal liquid supply strategy. After one control cycle ends, update the remaining liquid quality and real-time status of each medicine tank, and enter the control cycle of the next cycle.

[0189] Step 6.1: After a control cycle ends, begin calculating the remaining liquid mass of each medicine container at the next moment.

[0190] Step 6.2: Substitute the values ​​into the system's center of mass formula to calculate the system's center of mass at the next moment. And calculate the corresponding centroid influence coefficient. ;

[0191] Step 6.3: Calculate the minimum value of the sum of squared Euclidean distances of the overall machine's center of gravity shift caused by the liquid supply at the next moment. The combination with the smallest value is selected as the optimal liquid supply strategy for the next cycle.

[0192] The above embodiments are only for illustrating the technical concept and features of the present invention, and are intended to enable those skilled in the art to understand the content of the present invention and implement it accordingly. They should not be construed as limiting the scope of protection of the present invention. All equivalent transformations or modifications made in accordance with the spirit and essence of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A centroid-balanced liquid supply control method for multi-pesticide tank UAVs based on flight attitude, characterized in that, include: Obtain the geometric parameters, installation position parameters, initial liquid volume, drone pitch angle data, and spray system requirements of each medicine tank on the drone; The shape of the free liquid surface of the medicine in the medicine container is determined based on the pitch angle data, and the centroid coordinates of the medicine in the container at the corresponding time are calculated. Based on the centroid coordinates of the pesticide liquid in each pesticide tank and the mass of the pesticide liquid, calculate the centroid coordinates of the entire drone at the corresponding moment; Establish a 0-1 hybrid linear model to plan the liquid supply constraints. The liquid supply constraints include at least the continuous liquid supply duration constraint, the mutual exclusion liquid supply constraint of the medicine tank, the joint liquid supply quantity constraint, the liquid supply rate constraint, and the liquid quantity required by the spraying system. Under the premise of satisfying the liquid supply constraints, with the goal of minimizing the offset of the whole machine's center of gravity caused by the liquid supply of pesticide tanks from the drone, the liquid supply status and liquid supply rate of each pesticide tank at the corresponding time are determined. Based on the determined liquid supply status and rate, the liquid supply to each medicine tank is controlled, and the drone flight status is updated for the next moment, entering the control cycle of the next period.

2. The method for centroid balance liquid supply control of a multi-pesticide tank UAV based on flight attitude according to claim 1, characterized in that, Based on the pitch angle data, the pitch state of the UAV at the current moment is determined, and the free surface morphology of the medicine liquid in the medicine tank is classified according to the preset pitch angle range. Within different pitch angle ranges, based on the geometric dimensions of the medicine container and the current volume of the medicine liquid, the cross-section of the free liquid surface of the medicine liquid inside the medicine container is divided into at least one combined geometric shape composed of rectangles and / or triangles; Calculate the area of ​​the combined geometric shape and its corresponding geometric centroid coordinates, and determine the centroid coordinates of the medicine container at the current moment based on the area weighting method.

3. The method for centroid balance liquid supply control of a multi-pesticide tank UAV based on flight attitude according to claim 2, characterized in that, The free surface morphology of the pesticide solution inside the pesticide tank is classified, and a is defined as follows: n b n c n Let represent the length, width, and height of the nth pesticide tank, respectively; let θ be the pitch angle of the drone; and let S be the cross-sectional area of ​​the pesticide solution in the yz coordinate system. The specific classifications are as follows: The first scenario: When the drone is flying upwards, y1 ≤ b n And z1≤c n At this time, the free surface of the pesticide solution forms a triangle with the pesticide tank. The length of the triangle on the y-axis is y1, the length on the z-axis is z1, and the area is S. The second scenario: When the drone flies upwards, it satisfies... At that time, the pesticide solution in the pesticide tank was divided into a rectangle S2 and a triangle S1. The length of the rectangle on the y-axis is y1, and the length of the triangle on the y-axis is y2. The third type: When the drone flies upwards, it satisfies... At that time, the pesticide solution in the pesticide tank is divided into a rectangle S2 and a triangle S1. The length of the rectangle on the z-axis is z1, and the length of the triangle on the z-axis is z2. The fourth type: When the drone flies upwards, it satisfies... At that time, the length of the empty part inside the medicine container in the y-axis direction is set to y3, and the length in the z-axis direction is set to z3. The area S of the medicine liquid is divided into two rectangles S1 and S2 and a triangle S3. The fifth type: When the drone flies downwards, it meets the requirements. At that time, the free surface of the pesticide solution forms a triangle with the pesticide tank. Let the length of the pesticide tank in the y-axis direction be y1, the length in the z-axis direction be z1, and the area of ​​the pesticide solution inside the pesticide tank be S. The sixth type: When the drone flies downwards, it meets the requirements. At that time, the liquid medicine in the medicine pot is divided into a rectangle S2 and a triangle S1. Let the length of the cross-sectional triangle in the y-axis direction be y1, and the length of the cross-sectional rectangle in the y-axis direction be y2. The seventh type: When the drone flies downwards, it meets the requirements. At that time, the liquid medicine in the medicine pot is divided into a rectangle S2 and a triangle S1. Let the length of the cross-sectional triangle in the z-axis direction be z2, and the length of the cross-sectional rectangle in the z-axis direction be z1. The eighth type: When the drone flies downwards, it meets the requirements. At that time, the length of the empty part inside the medicine jar in the y-axis direction is set to y3, and the length in the z-axis direction is set to z3. The medicine liquid inside the medicine jar is divided into two rectangles S1 and S2 and a triangle S3. The ninth type: When the drone is flying horizontally, it meets the following requirements. At that time, the remaining liquid in the medicine tank is a rectangle S with a length of z1 in the z-axis direction, and the center of mass of the UAV depends on the height z1 of the remaining liquid.

4. The method for centroid balance liquid supply control of a multi-pesticide tank UAV based on flight attitude according to claim 1, characterized in that, Calculating the overall centroid coordinates of the drone at a given moment requires considering the mass of the medicine in each medicine container at that moment, the corresponding centroid coordinates of the medicine, and the mass of the drone itself. Specifically: Based on the coordinates of the centroid of the medicine liquid in the nth tank at the rth second. Formula for the centroid coordinates of the entire drone Find the actual center of mass of the entire UAV at the r-th second. ,Right now: ； Where M represents the weight of the drone itself, and the coordinates of the centroid of the pesticide solution in the nth pesticide tank at second r are... for T0(X0,Y0,Z0) represents the centroid coordinates of the drone when the pesticide tank is unloaded. This represents the weight of the pesticide solution in the nth pesticide tank at second r, where p represents the number of pesticide tanks. In practical work... The item can be ignored.

5. The method for centroid balance liquid supply control of a multi-pesticide tank UAV based on flight attitude according to claim 1, characterized in that, A 0-1 hybrid linear programming model is established to dynamically determine the liquid supply status and rate of each tank, under the premise of satisfying the following constraints: continuous liquid supply duration, mutual exclusion of tanks, combined liquid supply volume, liquid supply rate, and the required liquid volume of the spraying system. The specific liquid supply constraints are as follows: The duration of liquid supply must be no less than 60 seconds. At time r, a variable is set. The value of the medicine container is 1 if it supplies liquid at time r, and 0 if it does not supply liquid. The determination of the liquid supply status of the medicine container at a given time is as follows: ； in, This indicates the number of times fluid was supplied within the past 60 seconds. When r ≤ 60, the number of times fluid was supplied starts from the initial time. During the liquid supply process, a maximum of two pesticide tanks can supply liquid simultaneously and cannot be located on the same side. The amount of liquid supplied in the r-th second is less than the amount of liquid remaining in the previous second. The liquid supply speed cannot exceed the maximum liquid supply rate allowed by the pesticide tank. An interval constraint is set for the total liquid supply in the r-th second, with the lower limit being the amount of liquid required by the spraying system and the upper limit being 1.1 times the amount of liquid required by the spraying system.

6. The method for centroid balance liquid supply control of a multi-pesticide tank UAV based on flight attitude according to claim 1, characterized in that, Under the condition of satisfying all constraints, there are multiple liquid supply combinations for single-tank or dual-tank systems. The least squares method is used to allocate the liquid supply rate of each pesticide tank to minimize the overall centroid offset of the drone caused by liquid supply. Specifically: Calculate the centroid influence coefficient of all non-empty medicine containers at the current moment. , = This indicates that for every unit of medicine consumed by the nth tank, the center of gravity of the entire machine shifts by an amount. Let represent the centroid coordinates of the pesticide solution in the nth pesticide tank at second r. This indicates the coordinates of the drone's center of mass. Let M represent the weight of the pesticide solution in the nth pesticide tank at second r, M represent the weight of the drone itself, and p represent the number of pesticide tanks. Introducing dummy variables The optimization objective is to minimize the sum of squared Euclidean distances of the overall machine's center of gravity shift caused by liquid supply, i.e.: ; For each feasible liquid supply combination, the optimal rate is obtained by solving the least squares method.

7. The method for centroid balance liquid supply control of a multi-pesticide tank UAV based on flight attitude according to claim 6, characterized in that, For each feasible liquid supply combination, the optimal rate is obtained by solving the least squares method: If it is a single tank i, then = ; If there are two tanks i and j, then the following conditions are met. Under the premise that: ； - ； in, This represents the liquid dispensing rate of the i-th pesticide tank at the r-th second in the screening combination. This represents the liquid supply rate of the j-th pesticide tank at the r-th second in the screening combination. This represents the centroid influence coefficient of the i-th pesticide tank at second r in the screening combination. This represents the centroid influence coefficient of the j-th pesticide tank at second r in the screening combination. This indicates the amount of liquid required by the spraying system in the r-th second.

8. The method for centroid balance liquid supply control of a multi-pesticide tank UAV based on flight attitude according to claim 6, characterized in that, Liquid supply is controlled according to the optimal liquid supply strategy. After one control cycle ends, the remaining liquid quality and real-time status of each tank are updated, and the control cycle for the next cycle begins. Specifically: After a control cycle ends, the remaining liquid mass of each medicine tank is calculated for the next moment. Substitute the values ​​into the machine's centroid formula to calculate the machine's centroid at the next moment, and calculate the corresponding centroid influence coefficient. Calculate the minimum value of the sum of squared Euclidean distances of the total centroid shift caused by the liquid supply at the next moment. The combination with the smallest value is selected as the optimal liquid supply strategy for the next cycle.

9. The method for centroid balance liquid supply control of a multi-pesticide tank UAV based on flight attitude according to claim 1, characterized in that, The system controls each pesticide tank to supply liquid to the spraying system according to the liquid supply strategy, and updates the remaining liquid volume of the pesticide tank and the centroid coordinates of the UAV at each preset time interval, and enters the next spraying control cycle until the plant protection operation flight mission ends.

Images